News#
Release 13.0.1 - 2023-03-24#
Improvements#
[highlight_html] Added support for prefix search.
We can now use prefix search in
highlight_html.Note that highlight keyword is also highlighted not only at the first but also in the middle or at the end.
table_create Tags TABLE_NO_KEY column_create Tags name COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --normalizer 'NormalizerNFKC150' column_create Terms tags_name COLUMN_INDEX Tags name load --table Tags [ {"name": "Groonga"} ] select Tags \ --query "name:^g" \ --output_columns "highlight_html(name)" #[ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "highlight_html", # null # ] # ], # [ # "<span class=\"keyword\">G</span>roon<span class=\"keyword\">g</span>a" # ] # ] # ] #]
[Normalizers] Added new options for NormalizerNFKC*.
unify_kana_prolonged_sound_markWe can now normalize prolonged_sound_mark with this option as below.
ァー -> ァア, アー -> アア, ヵー -> ヵア, カー -> カア, ガー -> ガア, サー -> サア, ザー -> ザア, ター -> タア, ダー -> ダア, ナー -> ナア, ハー -> ハア, バー -> バア, パー -> パア, マー -> マア, ャー -> ャア, ヤー -> ヤア, ラー -> ラア, ヮー -> ヮア, ワー -> ワア, ヷー -> ヷア, ィー -> ィイ, イー -> イイ, キー -> キイ, ギー -> ギイ, シー -> シイ, ジー -> ジイ, チー -> チイ, ヂー -> ヂイ, ニー -> ニイ, ヒー -> ヒイ, ビー -> ビイ, ピー -> ピイ, ミー -> ミイ, リー -> リイ, ヰー -> ヰイ, ヸー -> ヸイ, ゥー -> ゥウ, ウー -> ウウ, クー -> クウ, グー -> グウ, スー -> スウ, ズー -> ズウ, ツー -> ツウ, ヅー -> ヅウ, ヌー -> ヌウ, フー -> フウ, ブー -> ブウ, プー -> プウ, ムー -> ムウ, ュー -> ュウ, ユー -> ユウ, ルー -> ルウ, ヱー -> ヱウ, ヴー -> ヴウ, ェー -> ェエ, エー -> エエ, ヶー -> ヶエ, ケー -> ケエ, ゲー -> ゲエ, セー -> セエ, ゼー -> ゼエ, テー -> テエ, デー -> デエ, ネー -> ネエ, ヘー -> ヘエ, ベー -> ベエ, ペー -> ペエ, メー -> メエ, レー -> レエ, ヹー -> ヹエ, ォー -> ォオ, オー -> オオ, コー -> コオ, ゴー -> ゴオ, ソー -> ソオ, ゾー -> ゾオ, トー -> トオ, ドー -> ドオ, ノー -> ノオ, ホー -> ホオ, ボー -> ボオ, ポー -> ポオ, モー -> モオ, ョー -> ョオ, ヨー -> ヨオ, ロー -> ロオ, ヲー -> ヲオ, ヺー -> ヺオ, ンー -> ンン ぁー -> ぁあ, あー -> ああ, ゕー -> ゕあ, かー -> かあ, がー -> があ, さー -> さあ, ざー -> ざあ, たー -> たあ, だー -> だあ, なー -> なあ, はー -> はあ, ばー -> ばあ, ぱー -> ぱあ, まー -> まあ, ゃー -> ゃあ, やー -> やあ, らー -> らあ, ゎー -> ゎあ, わー -> わあ ぃー -> ぃい, いー -> いい, きー -> きい, ぎー -> ぎい, しー -> しい, じー -> じい, ちー -> ちい, ぢー -> ぢい, にー -> にい, ひー -> ひい, びー -> びい, ぴー -> ぴい, みー -> みい, りー -> りい, ゐー -> ゐい ぅー -> ぅう, うー -> うう, くー -> くう, ぐー -> ぐう, すー -> すう, ずー -> ずう, つー -> つう, づー -> づう, ぬー -> ぬう, ふー -> ふう, ぶー -> ぶう, ぷー -> ぷう, むー -> むう, ゅー -> ゅう, ゆー -> ゆう, るー -> るう, ゑー -> ゑう, ゔー -> ゔう ぇー -> ぇえ, えー -> ええ, ゖー -> ゖえ, けー -> けえ, げー -> げえ, せー -> せえ, ぜー -> ぜえ, てー -> てえ, でー -> でえ, ねー -> ねえ, へー -> へえ, べー -> べえ, ぺー -> ぺえ, めー -> めえ, れー -> れえ ぉー -> ぉお, おー -> おお, こー -> こお, ごー -> ごお, そー -> そお, ぞー -> ぞお, とー -> とお, どー -> どお, のー -> のお, ほー -> ほお, ぼー -> ぼお, ぽー -> ぽお, もー -> もお, ょー -> ょお, よー -> よお, ろー -> ろお, をー -> をお んー -> んん
Here is an example of
unify_kana_prolonged_sound_mark.table_create --name Animals --flags TABLE_HASH_KEY --key_type ShortText column_create --table Animals --name name --type ShortText column_create --table Animals --name sound --type ShortText load --table Animals [ {"_key":"1","name":"羊", "sound":"メーメー"}, ] table_create \ --name idx_animals_sound \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_kana_prolonged_sound_mark", true)' column_create --table idx_animals_sound --name animals_sound --flags COLUMN_INDEX|WITH_POSITION --type Animals --source sound select --table Animals --query sound:@メエメエ #[ # [ # 0, # 1677829950.652696, # 0.01971983909606934 # ], # [ # [ # [ # 1 # ], # [ # [ # "_id", # "UInt32" # ], # [ # "_key", # "ShortText" # ], # [ # "name", # "ShortText" # ], # [ # "sound", # "ShortText" # ] # ], # [ # 1, # "1", # "羊", # "メーメー" # ] # ] # ] #]
unify_kana_hyphenWe can now normalize hyphen with this option as below.
ァ- -> ァア, ア- -> アア, ヵ- -> ヵア, カ- -> カア, ガ- -> ガア, サ- -> サア, ザ- -> ザア, タ- -> タア, ダ- -> ダア, ナ- -> ナア, ハ- -> ハア, バ- -> バア, パ- -> パア, マ- -> マア, ャ- -> ャア, ヤ- -> ヤア, ラ- -> ラア, ヮ- -> ヮア, ワ- -> ワア, ヷ- -> ヷア, ィ- -> ィイ, イ- -> イイ, キ- -> キイ, ギ- -> ギイ, シ- -> シイ, ジ- -> ジイ, チ- -> チイ, ヂ- -> ヂイ, ニ- -> ニイ, ヒ- -> ヒイ, ビ- -> ビイ, ピ- -> ピイ, ミ- -> ミイ, リ- -> リイ, ヰ- -> ヰイ, ヸ- -> ヸイ, ゥ- -> ゥウ, ウ- -> ウウ, ク- -> クウ, グ- -> グウ, ス- -> スウ, ズ- -> ズウ, ツ- -> ツウ, ヅ- -> ヅウ, ヌ- -> ヌウ, フ- -> フウ, ブ- -> ブウ, プ- -> プウ, ム- -> ムウ, ュ- -> ュウ, ユ- -> ユウ, ル- -> ルウ, ヱ- -> ヱウ, ヴ- -> ヴウ, ェ- -> ェエ, エ- -> エエ, ヶ- -> ヶエ, ケ- -> ケエ, ゲ- -> ゲエ, セ- -> セエ, ゼ- -> ゼエ, テ- -> テエ, デ- -> デエ, ネ- -> ネエ, ヘ- -> ヘエ, ベ- -> ベエ, ペ- -> ペエ, メ- -> メエ, レ- -> レエ, ヹ- -> ヹエ, ォ- -> ォオ, オ- -> オオ, コ- -> コオ, ゴ- -> ゴオ, ソ- -> ソオ, ゾ- -> ゾオ, ト- -> トオ, ド- -> ドオ, ノ- -> ノオ, ホ- -> ホオ, ボ- -> ボオ, ポ- -> ポオ, モ- -> モオ, ョ- -> ョオ, ヨ- -> ヨオ, ロ- -> ロオ, ヲ- -> ヲオ, ヺ- -> ヺオ, ン- -> ンン ぁ- -> ぁあ, あ- -> ああ, ゕ- -> ゕあ, か- -> かあ, が- -> があ, さ- -> さあ, ざ- -> ざあ, た- -> たあ, だ- -> だあ, な- -> なあ, は- -> はあ, ば- -> ばあ, ぱ- -> ぱあ, ま- -> まあ, ゃ- -> ゃあ, や- -> やあ, ら- -> らあ, ゎ- -> ゎあ, わ- -> わあ ぃ- -> ぃい, い- -> いい, き- -> きい, ぎ- -> ぎい, し- -> しい, じ- -> じい, ち- -> ちい, ぢ- -> ぢい, に- -> にい, ひ- -> ひい, び- -> びい, ぴ- -> ぴい, み- -> みい, り- -> りい, ゐ- -> ゐい ぅ- -> ぅう, う- -> うう, く- -> くう, ぐ- -> ぐう, す- -> すう, ず- -> ずう, つ- -> つう, づ- -> づう, ぬ- -> ぬう, ふ- -> ふう, ぶ- -> ぶう, ぷ- -> ぷう, む- -> むう, ゅ- -> ゅう, ゆ- -> ゆう, る- -> るう, ゑ- -> ゑう, ゔ- -> ゔう ぇ- -> ぇえ, え- -> ええ, ゖ- -> ゖえ, け- -> けえ, げ- -> げえ, せ- -> せえ, ぜ- -> ぜえ, て- -> てえ, で- -> でえ, ね- -> ねえ, へ- -> へえ, べ- -> べえ, ぺ- -> ぺえ, め- -> めえ, れ- -> れえ ぉ- -> ぉお, お- -> おお, こ- -> こお, ご- -> ごお, そ- -> そお, ぞ- -> ぞお, と- -> とお, ど- -> どお, の- -> のお, ほ- -> ほお, ぼ- -> ぼお, ぽ- -> ぽお, も- -> もお, ょ- -> ょお, よ- -> よお, ろ- -> ろお, を- -> をお ん- -> んん
Here is an example of
unify_kana_hyphen.table_create --name Animals --flags TABLE_HASH_KEY --key_type ShortText column_create --table Animals --name name --type ShortText column_create --table Animals --name sound --type ShortText load --table Animals [ {"_key":"1","name":"羊", "sound":"メ-メ-"}, ] table_create \ --name idx_animals_sound \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_kana_hyphen", true)' column_create --table idx_animals_sound --name animals_sound --flags COLUMN_INDEX|WITH_POSITION --type Animals --source sound select --table Animals --query sound:@メエメエ #[ # [ # 0,1677829950.652696, # 0.01971983909606934 # ], # [ # [ # [ # 1 # ], # [ # [ # "_id", # "UInt32" # ], # [ # "_key", # "ShortText" # ], # [ # "name", # "ShortText" # ], # [ # "sound", # "ShortText" # ] # ], # [ # 1, # "1", # "羊", # "メ-メ-" # ] # ] # ] #]
[Near search condition][Near search operator] Added a new option
${MIN_INTERVAL}for near-search family.We can now specifiy the minimum interval between phrases (words) with
${MIN_INTERVAL}. The interval between phrases (words) must be at least this value.Here are new syntax:
*N${MAX_INTERVAL},${MAX_TOKEN_INTERVAL_1}|${MAX_TOKEN_INTERVAL_2}|...,${MIN_INTERVAL} "word1 word2 ..." *NP${MAX_INTERVAL},${ADDITIONAL_LAST_INTERVAL},${MAX_PHRASE_INTERVAL_1}|${MAX_PHRASE_INTERVAL_2}|...,${MIN_INTERVAL} "phrase1 phrase2 ..." *NPP${MAX_INTERVAL},${ADDITIONAL_LAST_INTERVAL},${MAX_PHRASE_INTERVAL_1}|${MAX_PHRASE_INTERVAL_2}|...,${MIN_INTERVAL} "(phrase1-1 phrase1-2 ...) (phrase2-1 phrase2-2 ...) ..." *ONP${MAX_INTERVAL},${ADDITIONAL_LAST_INTERVAL},${MAX_PHRASE_INTERVAL_1}|${MAX_PHRASE_INTERVAL_2}|...,${MIN_INTERVAL} "phrase1 phrase2 ..." *ONPP${MAX_INTERVAL},${ADDITIONAL_LAST_INTERVAL},${MAX_PHRASE_INTERVAL_1}|${MAX_PHRASE_INTERVAL_2}|...,${MIN_INTERVAL} "(phrase1-1 phrase1-2 ...) (phrase2-1 phrase2-2 ...) ..."The default value of
${MIN_INTERVAL}isINT32_MIN(-2147483648). We use the default value when${MIN_INTERVAL}is omitted.This option is useful when we want to ignore overlapped phrases.
The interval for
*NPis culculated asinterval between the top tokens of phrases - tokens in the left phrase + 1.When a tokenizer is Bigram, for exmaple,
東京has one token東京, also京都has one token京都.Considering
東京都as a target value of*NP "東京 京都":interval between the top tokens of phrases:1(interval between東京and京都)tokens in the left phrase:1(東京)
The interval for
*NPof東京都is1 - 1 + 1=1.As a result, the interval for
*NPis greater than1when東京and京都are not overlapped.Here is an example for ignoring overlapped phrases.
table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer 'TokenNgram("unify_alphabet", false, \ "unify_digit", false)' \ --normalizer NormalizerNFKC150 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "東京都"}, {"content": "東京京都"} ] select Entries \ --match_columns content \ --query '*NP-1,0,,2"東京 京都"' \ --output_columns '_score, content' #[ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "_score", # "Int32" # ], # [ # "content", # "Text" # ] # ], # [ # 1, # "東京京都" # ] # ] # ] #]
In the example above,
東京都is not matched as the interval is1but東京京都is matched as the interval is2.[Normalizers] Added support for new values in the
unify_katakana_trailing_ooption.We added support for normalizing the following new values in the
unify_katakana_trailing_ooption because a vowel of those left letters isO.ォオ->ォウョオ->ョウヺオ->ヺウ
Add support for MessagePack v6.0.0. [GitHub#1536][Reported by Carlo Cabrera]
Groonga can not found MessagePack v6.0.0 or later when we execute configure or cmake until now. Groonga can found MessagePack since this release even if the version of MessagePack is v6.0.0 or later.
Fixes#
[Normalizers] Fixed a bug that NormalizerNFKC* did incorrect normalization.
This bug occured when
unify_kana_caseandunify_katakana_v_soundsused at the same time.For example,
ヴァwas normalized toバアwithunify_kana_caseandunify_katakana_v_sounds, butヴァshould be normalized toバ.This was because
ヴァwas normalized toヴアwithunify_kana_case, and after that,ヴアwas normalized toバアwithunify_katakana_v_sounds. We fixed to normalize characters withunify_katakana_v_soundsbeforeunify_kana_case.Here is an example of the bug in the previous version.
normalize \ 'NormalizerNFKC150("unify_katakana_v_sounds", true, \ "unify_kana_case", true)' \ "ヴァーチャル" #[ # [ # 0, # 1678097412.913053, # 0.00019073486328125 # ], # { # "normalized":"ブアーチヤル", # "types":[], # "checks":[] # }
From this version,
ヴァーチャルis normalized toバーチヤル.[Ordered near phrase search condition][Ordered near phrase search operator] Fixed a bug that
${MAX_PHRASE_INTERVALS}doesn’t work correctly.When this bug occured, intervals are regarded as
0. Therefore, if this bug occurs, records may hit too many.This bug occured when:
Use
*ONPwith${MAX_PHRASE_INTERVALS}The number of tokens in the matched left phrase is greater than or equal to the number of
${MAX_PHRASE_INTERVALS}elements.
Here is an example of this bug.
table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer 'TokenNgram("unify_alphabet", false, \ "unify_digit", false)' \ --normalizer NormalizerNFKC150 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "abcXYZdef"}, {"content": "abcdef"}, {"content": "abc123456789def"}, {"content": "abc12345678def"}, {"content": "abc1de2def"} ] select Entries --filter 'content *ONP-1,0,1 "abc def"' --output_columns '_score, content' #[ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 5 # ], # [ # [ # "_score", # "Int32" # ], # [ # "content", # "Text" # ] # ], # [ # 1, # "abcXYZdef" # ], # [ # 1, # "abcdef" # ], # [ # 1, # "abc123456789def" # ], # [ # 1, # "abc12345678def" # ], # [ # 1, # "abc1de2def" # ] # ] # ] #]
In the example above, the first element interval is specified as
1with*ONP-1,0,1 "abc def", but allcontentis matched, including those farther than1.This is because the example satisfies the condition for the bug and the interval is regarded as
0.Use
*ONPwith${MAX_PHRASE_INTERVALS}*ONP-1,0,1 "abc def"specifies${MAX_PHRASE_INTERVALS}.The number of tokens in the matched left phrase is greater than or equal to the number of
${MAX_PHRASE_INTERVALS}elements.The matched left phrase:
abcIncluded tokens:
ab,bcTokenized with
TokenNgram("unify_alphabet", false, "unify_digit", false)
The number of elements specified with
max_element_intervals:11of*ONP-1,0,1 "abc def"
The number of tokens in the left phrase (
2) > the number of elements specified withmax_element_intervals(1)
[Near phrase search condition][Near phrase search operator] Fixed a bug that phrases specified as last didn’t used as last in near-phrase family.
When this bug occured,
${ADDITIONAL_LAST_INTERVAL}was ignored and only${MAX_INTERVAL}was used.This bug occured when:
A phrase specified as last contains multiple tokens.
The size of the last token of the phrase is smaller than or equals to the sizes of other tokens in the phrase.
The token size is the number of tokens appeared in the all records.
Here is an example of this bug.
table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer 'TokenNgram("unify_alphabet", false, \ "unify_digit", false)' \ --normalizer NormalizerNFKC150 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "abc123456789defg"}, {"content": "dededede"} ] select Entries \ --filter 'content *NP10,1"abc defg$"' \ --output_columns '_score, content' #[ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 0 # ], # [ # [ # "_score", # "Int32" # ], # [ # "content", # "Text" # ] # ] # ] # ] #]
In the example above, for
abc123456789defg, the intervalabctodefgis11.${MAX_INTERVAL}is10and${ADDITIONAL_LAST_INTERVAL}is1, so a threshold for matching last phrase is11. So it should be matched, but isn’t.This is because the example satisfies the condition for the bug as below, and only
`${MAX_INTERVAL}is used.A phrase specified as last contains multiple tokens.
defg$is specified as last because the suffix is$.defg$is tokenized tode,ef,fgwithTokenNgram("unify_alphabet", false, "unify_digit", false)The size of the last token of the phrase is smaller than or equals to the sizes of other tokens in the phrase.
fgis the last token ofdefg$.abc123456789defgcontains onefgandde, anddedededecontains 4de.So, the size of
fgis 1 anddeis 5.
[Near phrase search condition] Fixed interval calculation.
If we use near phrase search, records may hit too many by this bug.
[highlight_html] Fixed a bug that highlight position may move over when we use loose_symbol=true.
Thanks#
Carlo Cabrera
Release 13.0.0 - 2023-02-09#
This is a major version up! But It keeps backward compatibility. We can upgrade to 13.0.0 without rebuilding database.
First of all, we introduce the main changes in 13.0.0. Then, we introduce the hilight and summary of changes from Groonga 12.0.0 to 12.1.2.
New Features and Improvements in 13.0.0#
Improvements#
[Normalizers] Added a new Normalizer
NormalizerNFKC150based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 15.0.[Token filters] Added a new TokenFilter
TokenFilterNFKC150based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 15.0.[NormalizerNFKC150] Added new options for NormalizerNFKC* as below.
unify_katakana_gu_small_soundsWe can normalize “グァ -> ガ”, “グィ -> ギ”, “グェ -> ゲ”, and “グォ -> ゴ” with this option.
Here is an example of
unify_katakana_gu_small_soundsoption.table_create --name Countries --flags TABLE_HASH_KEY --key_type ShortText column_create --table Countries --name name --type ShortText load --table Countries [ {"_key":"JP","name":"日本"}, {"_key":"GT","name":"グァテマラ共和国"}, ] table_create \ --name idx_contry_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_gu_small_sounds", true)' column_create --table idx_contry_name --name contry_name --flags COLUMN_INDEX|WITH_POSITION --type Countries --source name select --table Countries --query name:@ガテマラ共和国 [ [0, 0, 0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 2, "GT", "グァテマラ共和国" ] ] ] ]
unify_katakana_di_soundWe can normalize “ヂ -> ジ” with this option.
Here is an example of
unify_katakana_di_soundoption.table_create --name Foods --flags TABLE_HASH_KEY --key_type ShortText column_create --table Foods --name name --type ShortText load --table Foods [ {"_key":"1","name":"チジミ"}, {"_key":"2","name":"パジョン"}, ] table_create \ --name idx_food_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_di_sound", true)' column_create --table idx_food_name --name food_name --flags COLUMN_INDEX|WITH_POSITION --type Foods --source name select --table Foods --query name:@チヂミ [ [ 0, 0, 0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 1, "1", "チジミ" ] ] ] ]
unify_katakana_wo_soundWe can normalize “ヲ -> オ” with this option.
Here is an example of
unify_katakana_wo_soundoption.table_create --name Foods --flags TABLE_HASH_KEY --key_type ShortText column_create --table Foods --name name --type ShortText load --table Foods [ {"_key":"1","name":"アヲハタ"}, {"_key":"2","name":"ヴェルデ"}, {"_key":"3","name":"ランプ"}, ] table_create \ --name idx_food_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_wo_sound", true)' column_create --table idx_food_name --name food_name --flags COLUMN_INDEX|WITH_POSITION --type Foods --source name select --table Foods --query name:@アオハタ [ [ 0, 0, 0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 1, "1", "アヲハタ" ] ] ] ]
unify_katakana_zu_small_soundsWe can normalize “ズァ -> ザ”, “ズィ -> ジ”, “ズェ -> ゼ”, and “ズォ -> ゾ” with this option.
Here is an example of
unify_katakana_zu_small_soundsoption.table_create --name Cities --flags TABLE_HASH_KEY --key_type ShortText column_create --table Cities --name name --type ShortText load --table Cities [ {"_key":"1","name":"ガージヤーバード"}, {"_key":"2","name":"デリー"}, ] table_create \ --name idx_city_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_zu_small_sounds", true)' column_create --table idx_city_name --name city_name --flags COLUMN_INDEX|WITH_POSITION --type Cities --source name select --table Cities --query name:@ガーズィヤーバード [ [ 0, 0, 0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 1, "1", "ガージヤーバード" ] ] ] ]
unify_katakana_du_soundWe can normalize “ヅ -> ズ” with this option.
Here is an example of
unify_katakana_du_soundoption.table_create --name Plants --flags TABLE_HASH_KEY --key_type ShortText column_create --table Plants --name name --type ShortText load --table Plants [ {"_key":"1","name":"ハスノカヅラ"}, {"_key":"2","name":"オオツヅラフジ"}, {"_key":"3","name":"アオツヅラフジ"}, ] table_create \ --name idx_plant_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_du_sound", true)' column_create --table idx_plant_name --name plant_name --flags COLUMN_INDEX|WITH_POSITION --type Plants --source name select --table Plants --query name:@ツズラ [ [ 0, 0, 0 ], [ [ [ 2 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 2, "2", "オオツヅラフジ" ], [ 3, "3", "アオツヅラフジ" ] ] ] ]
unify_katakana_trailing_oWe can normalize following characters with this option.
“オオ -> オウ”
“コオ -> コウ”
“ソオ -> ソウ”
“トオ -> トウ”
“ノオ -> ノウ”
“ホオ -> ホウ”
“モオ -> モウ”
“ヨオ -> ヨウ”
“ロオ -> ロウ”
“ゴオ -> ゴウ”
“ゾオ -> ゾウ”
“ドオ -> ドウ”
“ボオ -> ボウ”
“ポオ -> ポウ”
Here is an example of
unify_katakana_trailing_ooption.table_create --name Sharks --flags TABLE_HASH_KEY --key_type ShortText column_create --table Sharks --name name --type ShortText load --table Sharks [ {"_key":"1","name":"ホオジロザメ"}, {"_key":"2","name":"ジンベイザメ"}, ] table_create \ --name idx_shark_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_trailing_o", true)' column_create --table idx_shark_name --name shark_name --flags COLUMN_INDEX|WITH_POSITION --type Sharks --source name select --table Sharks --query name:@ホウジロザメ [ [ 0, 0, 0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 1, "1", "ホオジロザメ" ] ] ] ]
unify_katakana_du_small_soundsWe can normalize “ヅァ -> ザ”, “ヅィ -> ジ”, “ヅェ -> ゼ”, and “ヅォ -> ゾ” with this option.
Here is an example of
unify_katakana_du_small_soundsoption.table_create --name Airports --flags TABLE_HASH_KEY --key_type ShortText column_create --table Airports --name name --type ShortText load --table Airports [ {"_key":"HER","name":"イラクリオ・ニコスカザンヅァキス国際空港"}, {"_key":"ATH","name":"アテネ国際空港"}, ] table_create \ --name idx_airport_name \ --flags TABLE_PAT_KEY \ --key_type ShortText \ --default_tokenizer TokenBigram \ --normalizer 'NormalizerNFKC150("unify_katakana_du_small_sounds", true)' column_create --table idx_airport_name --name airport_name --flags COLUMN_INDEX|WITH_POSITION --type Airports --source name select --table Airports --query name:@ニコスカザンザキス [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "name", "ShortText" ] ], [ 1, "HER", "イラクリオ・ニコスカザンヅァキス国際空港" ] ] ]
[Oracle Linux] Added newly support for Oracle Linux 8 and 9.
Higlight and Summary of changes from 12.0.0 to 12.1.2#
Higlight#
[Normalizers] Added
NormalizerHTML. (Experimental)
NormalizerHTMLis a normalizer for HTML.Currently
NormalizerHTMLsupports removing tags like<span>or</span>and expanding character references like&or&.Here are sample queries for
NormalizerHTML.normalize NormalizerHTML "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666923364.883798,0.0005481243133544922],{"normalized":" Groonga & Mroonga & Rroonga ","types":[],"checks":[]}]In this sample
<span>and</span>are removed, and&and&are expanded to&.We can specify whether removing the tags with the
remove_tagoption. (The default value of theremove_tagoption istrue.)normalize 'NormalizerHTML("remove_tag", false)' "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666924069.278549,0.0001978874206542969],{"normalized":"<span> Groonga & Mroonga & Rroonga </span>","types":[],"checks":[]}]In this sample,
<span>and</span>are not removed.We can specify whether expanding the character references with the
expand_character_referenceoption. (The default value of theexpand_character_referenceoption istrue.)normalize 'NormalizerHTML("expand_character_reference", false)' "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666924357.099782,0.0002346038818359375],{"normalized":" Groonga & Mroonga & Rroonga ","types":[],"checks":[]}]In this sample,
&and&are not expanded.
[snippet],[snippet_html] Added support for text vector as input. [groonga-dev,04956][Reported by shinonon]
For example, we can extract snippets of target text around search keywords against vector in JSON data as below.
table_create Entries TABLE_NO_KEY column_create Entries title COLUMN_SCALAR ShortText column_create Entries contents COLUMN_VECTOR ShortText table_create Tokens TABLE_PAT_KEY ShortText --default_tokenizer TokenNgram --normalizer NormalizerNFKC130 column_create Tokens entries_title COLUMN_INDEX|WITH_POSITION Entries title column_create Tokens entries_contents COLUMN_INDEX|WITH_SECTION|WITH_POSITION Entries contents load --table Entries [ { "title": "Groonga and MySQL", "contents": [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] } ] select Entries\ --output_columns 'snippet_html(contents), contents'\ --match_columns 'title'\ --query Groonga [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "snippet_html", null ], [ "contents", "ShortText" ] ], [ [ "<span class=\"keyword\">Groonga</span> is a full text search engine", "Mroonga is a MySQL storage engine based on <span class=\"keyword\">Groonga</span>" ], [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] ] ] ] ]Until now, if we specified
snippet*like--output_columns 'snippet_html(contents[1]), we could extract snippets of target text around search keywords against the vector as below. However, we didn’t know which we should output elements. Because we didn’t know which element was hit on search.select Entries\ --output_columns 'snippet_html(contents[0]), contents'\ --match_columns 'title'\ --query Groonga [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "snippet_html", null ], [ "contents", "ShortText" ] ], [ [ "<span class=\"keyword\">Groonga</span> is a full text search engine" ], [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] ] ] ] ]
[query_expand] Added a support for synonym group.
Until now, We had to each defined a keyword and synonyms of the keyword as below when we use the synonym search.
table_create Thesaurus TABLE_PAT_KEY ShortText --normalizer NormalizerAuto # [[0, 1337566253.89858, 0.000355720520019531], true] column_create Thesaurus synonym COLUMN_VECTOR ShortText # [[0, 1337566253.89858, 0.000355720520019531], true] load --table Thesaurus [ {"_key": "mroonga", "synonym": ["mroonga", "tritonn", "groonga mysql"]}, {"_key": "groonga", "synonym": ["groonga", "senna"]} ]In the above case, if we search
mroonga, Groonga searchmroonga OR tritonn OR "groonga mysql"as we intended. However, if we searchtritonn, Groonga search onlytritonn. If we want to searchtritonn OR mroonga OR "groonga mysql"even if we searchtritonn, we need had added a definition as below.load --table Thesaurus [ {"_key": "tritonn", "synonym": ["tritonn", "mroonga", "groonga mysql"]}, ]In many cases, if we expand
mroongatomroonga OR tritonn OR "groonga mysql", we feel we want to expandtritonnand"groonga mysql"tomroonga OR tritonn OR "groonga mysql". However, until now, we had needed additional definitions in such a case. Therefore, if target keywords for synonyms are many, we are troublesome to define synonyms. Because we need to define many similar definitions.In addition, when we remove synonyms, we are troublesome because we need to execute remove against many records.
We can make a group by deciding on a representative synonym record since this release. For example, the all following keywords are the “mroonga” group.
load --table Synonyms [ {"_key": "mroonga": "representative": "mroonga"} ] load --table Synonyms [ {"_key": "tritonn": "representative": "mroonga"}, {"_key": "groonga mysql": "representative": "mroonga"} ]In this case,
mroongais expanded tomroonga OR tritonn OR "groonga mysql". In addition,tritonnand"groonga mysql"are also expanded tomroonga OR tritonn OR "groonga mysql".When we want to remove synonyms, we execute just remove against a target record. For example, if we want to remove
"groonga mysql"from synonyms, we just remove{"_key": "groonga mysql": "representative": "mroonga"}.
[index_column_have_source_record] Added a new function
index_column_have_source_record().We can confirm whether a token that is existing in the index is included in any of the records that are registered in Groonga or not.
Groonga does not remove a token even if the token become never used from records in Groonga by updating records. Therefore, for example, when we use the feature of autocomplete, Groonga may return a token that is not included in any of the records as candidates for search words. However, we can become that we don’t return the needless token by using this function.
Because this function can detect a token that is not included in any of the records.
[select] Added new arguments
drilldown_max_n_target_recordsanddrilldown[${LABEL}].max_n_target_records.We can specify the max number of records of the drilldown target table (filtered result) to use drilldown. If the number of filtered result is larger than the specified value, some records in filtered result aren’t used for drilldown. The default value of this arguments are
-1. If these arguments are set-1, Groonga uses all records for drilldown.This argument is useful when filtered result may be very large. Because a drilldown against large filtered result may be slow. We can limit the max number of records to be used for drilldown by this feature.
Here is an example to limit the max number of records to be used for drilldown. The last 2 records,
{\"_id\": 4, \"tag\": \"Senna\"}and{\"_id\": 5, \"tag\": \"Senna\"}, aren’t used.table_create Entries TABLE_HASH_KEY ShortText column_create Entries content COLUMN_SCALAR Text column_create Entries n_likes COLUMN_SCALAR UInt32 column_create Entries tag COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto column_create Terms entries_key_index COLUMN_INDEX|WITH_POSITION Entries _key column_create Terms entries_content_index COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"_key": "The first post!", "content": "Welcome! This is my first post!", "n_likes": 5, "tag": "Hello"}, {"_key": "Groonga", "content": "I started to use Groonga. It's very fast!", "n_likes": 10, "tag": "Groonga"}, {"_key": "Mroonga", "content": "I also started to use Mroonga. It's also very fast! Really fast!", "n_likes": 15, "tag": "Groonga"}, {"_key": "Good-bye Senna", "content": "I migrated all Senna system!", "n_likes": 3, "tag": "Senna"}, {"_key": "Good-bye Tritonn", "content": "I also migrated all Tritonn system!", "n_likes": 3, "tag": "Senna"} ] select Entries \ --limit -1 \ --output_columns _id,tag \ --drilldown tag \ --drilldown_max_n_target_records 3 [ [ 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "tag", "ShortText" ] ], [ 1, "Hello" ], [ 2, "Groonga" ], [ 3, "Groonga" ], [ 4, "Senna" ], [ 5, "Senna" ] ], [ [ 2 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ] ], [ "Hello", 1 ], [ "Groonga", 2 ] ] ] ]
Summary#
Improvements#
[httpd] Updated bundled nginx to 1.23.3.
[AlmaLinux] Added support for AlmaLinux 9.
[escalate] Added a document for the
escalate()function.[Normalizers] Added
NormalizerHTML. (Experimental)[httpd] Updated bundled nginx to 1.23.2.
Suppressed logging a lot of same messages when no memory is available.
Changed specification of the
escalate()function (Experimental) to make it easier to use.[How to build Groonga with CMake] Added a document about how to build Groonga with CMake.
[Others] Added descriptions about how to enable/disable Apache Arrow support when building with GNU Autotools.
[select] Added a document about drilldowns[${LABEL}].table.
[I18N] Updated the translation procedure.
Added a new function
escalate(). (experimental)[httpd] Updated bundled nginx to 1.23.1.
[select] Add a document for the
--n_workersoption.
Added new Munin plugins for groonga-delta.
[column_copy] Added support for weight vector.
[Ubuntu] Dropped support for Ubuntu 21.10 (Impish Indri).
[Debian GNU/Linux] Dropped Debian 10 (buster) support.
[select] Improved a little bit of performance for prefix search by search escalation.
[select] Added support for specifying a reference vector column with weight in
drilldowns[LABEL]._key.[select] Added support for doing drilldown with a reference vector with weight even if we use
queryorfilter, orpost_filter.
[logical_count] Improved memory usage while
logical_countexecuted.[dump] Added support for
MISSING_IGNORE/MISSING_NIL.[snippet],[snippet_html] Added support for text vector as input.
[
vector_join] Added a new functionvector_join().[Indexing] Ignore too large a token like online index construction.
[logical_range_filter] Added support for reducing reference immediately after processing a shard.
We increased the stability of the feature of recovering on crashes.
Improved performance for mmap if anonymous mmap available.
[Indexing] Added support for the static index construction against the following types of columns.
[column_create] Added new flags
MISSING_*andINVALID_*.[dump][column_list] Added support for
MISSING_*andINVALID_*flags.[schema] Added support for
MISSING_*andINVALID_*flags.We provided the package of Amazon Linux 2.
[Windows] Dropped support for building with Visual Studio 2017.
[query_expand] Added a support for synonym group.
[query_expand] Added a support for text vector and index.
Added support for disabling a backtrace by the environment variable.
[select] Improved performance for
--slices.[Windows] Added support for Visual Studio 2022.
[select] Added support for specifing max intervals for each elements in near search.
[Groonga HTTP server] We could use
groonga-server-httpeven if Groonga of RPM packages.
[sub_filter] Added a new option
pre_filter_threshold.[index_column_have_source_record] Added a new function
index_column_have_source_record().[NormalizerNFKC130] Added a new option
strip[select] Added new arguments
drilldown_max_n_target_recordsanddrilldown[${LABEL}].max_n_target_records.[httpd] Updated bundled nginx to 1.21.6.
Fixes#
[select] Fix a bug displaying a wrong label in
drilldownresults whencommand_versionis3.[NormalizerTable] Fix a bug for Groonga to crush with specific definition setting in
NormalizerTable.
[select][Vector column] Fixed a bug displaying integer in the results when a weight vector column specifies WEIGHT FLOAT32.
Fixed a bug that Groonga could return incorrect results when we use NormalizerTable and it contains a non-idempotent (results can be changed when executed repeatedly) definition.
Fixed a bug Groonga’s response may be slow when we execute the
request_cancelwhile executing a OR search.
Fixed a bug that Groonga may crash when we execute drilldown in a parallel by
n_workersoption.[select] Fixed a bug that the syntax error occurred when we specify a very long expression in
--filter.
Fixed a bug Groonga’s response may be slow when we execute
request_cancelwhile executing a search.Fixed a bug that string list can’t be casted to int32 vector.
Fixed a bug that Groonga Munin Plugins do not work on AlmaLinux 8 and CentOS 7.
Fixed a bug that we may be not able to add a key to a table of patricia trie.
Thanks#
Atsushi Shinoda
i10a
naoa
shinonon
Zhanzhao (Deo) Liang
David CARLIER
Release 12.1.2 - 2023-01-29#
Improvements#
[httpd] Updated bundled nginx to 1.23.3.
Release 12.1.1 - 2023-01-06#
Improvements#
[select][POWER_SET] Vector’s power set is now able to aggregate with the drilldowns.
A new option
key_vector_expansionis added to drilldowns. Currently,NONEorPOWER_SETcan be specified forkey_vector_expansion.Specifying
POWER_SETtokey_vector_expansionallows to aggregate for power set case. This method of aggregation is useful to aggregate total number of individual and combination tag occurrence at once.Following example is to see aggregating total number of individual and combination occurrence for following 3 tags,
Groonga,Mroonga, andPGroonga.Sample case:
table_create PowersetDrilldownMemos TABLE_HASH_KEY ShortText # [[0, 1337566253.89858, 0.000355720520019531], true] column_create PowersetDrilldownMemos tags COLUMN_VECTOR ShortText # [[0, 1337566253.89858, 0.000355720520019531], true] load --table PowersetDrilldownMemos [ {"_key": "Groonga is fast!", "tags": ["Groonga"]}, {"_key": "Mroonga uses Groonga!", "tags": ["Groonga", "Mroonga"]}, {"_key": "PGroonga uses Groonga!", "tags": ["Groonga", "PGroonga"]}, {"_key": "Mroonga and PGroonga are Groonga family", "tags": ["Groonga", "Mroonga", "PGroonga"]} ] # [[0, 1337566253.89858, 0.000355720520019531], 4] select PowersetDrilldownMemos \ --drilldowns[tags].keys tags \ --drilldowns[tags].key_vector_expansion POWER_SET \ --drilldowns[tags].columns[power_set].stage initial \ --drilldowns[tags].columns[power_set].value _key \ --drilldowns[tags].columns[power_set].flags COLUMN_VECTOR \ --drilldowns[tags].sort_keys 'power_set' \ --drilldowns[tags].output_columns 'power_set, _nsubrecs' \ --limit 0 # [ # [ # 0, # 1337566253.89858, # 0.000355720520019531 # ], # [ # [ # [ # 4 # ], # [ # [ # "_id", # "UInt32" # ], # [ # "_key", # "ShortText" # ], # [ # "tags", # "ShortText" # ] # ] # ], # { # "tags": [ # [ # 7 # ], # [ # [ # "power_set", # "Text" # ], # [ # "_nsubrecs", # "Int32" # ] # ], # [ # [ # "Groonga" # ], # 4 # ], # [ # [ # "Mroonga" # ], # 2 # ], # [ # [ # "PGroonga" # ], # 2 # ], # [ # [ # "Groonga", # "Mroonga" # ], # 2 # ], # [ # [ # "Groonga", # "PGroonga" # ], # 2 # ], # [ # [ # "Mroonga", # "PGroonga" # ], # 1 # ], # [ # [ # "Groonga", # "Mroonga", # "PGroonga" # ], # 1 # ] # ] # } # ] # ]
This result shows following.
tag
number of occurrence
Groonga4
Mroonga2
PGroonga2
GroongaandMroonga2
GroongaandPGroonga2
MroongaandPGroonga1
GroongaandMroongaandPGroonga1
This feature is complex. For more information, please refer to POWER_SET.
[select] Specific element of vector column is now able to be search target.
It allows specific elements of vector column to be search targets that specifying the specific elements to
match_columnswith index number.Following is a sample case.
table_create Memos TABLE_NO_KEY column_create Memos contents COLUMN_VECTOR ShortText table_create Lexicon TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram column_create Lexicon memo_index COLUMN_INDEX|WITH_POSITION|WITH_SECTION Memos contents load --table Memos [ ["contents"], [["I like Groonga", "Use Groonga with Ruby"]], [["I like Ruby", "Use Groonga"]] ] select Memos \ --match_columns "contents[1]" \ --query Ruby \ --output_columns "contents, _score" # [ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "contents", # "ShortText" # ], # [ # "_score", # "Int32" # ] # ], # [ # [ # "I like Groonga", # "Use Groonga with Ruby" # ], # 1 # ] # ] # ] # ]
--match_columns "contents[1]"specifies only 2nd vector elements ofcontentsas the search target. In this sample,["I like Groonga", "Use Groonga with Ruby"]is shown in the results becauseRubyis in 2nd elementUse Groonga with Ruby. However,["I like Ruby", "Use Groonga"]is not shown in results becauseRubyis not in 2nd elementUse Groonga.[load] Added support for
YYYY-MM-DDtime format.YYYY-MM-DDis a general time format. Supporting this time format madeloadmore useful.The time of the loaded value is set to
00:00:00on the local time.plugin_register functions/time # [[0,0.0,0.0],true] table_create Logs TABLE_NO_KEY # [[0,0.0,0.0],true] column_create Logs created_at COLUMN_SCALAR Time # [[0,0.0,0.0],true] column_create Logs created_at_text COLUMN_SCALAR ShortText # [[0,0.0,0.0],true] load --table Logs [ {"created_at": "2000-01-01", "created_at_text": "2000-01-01"} ] # [[0,0.0,0.0],1] select Logs --output_columns "time_format_iso8601(created_at), created_at_text" # [ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "time_format_iso8601", # null # ], # [ # "created_at_text", # "ShortText" # ] # ], # [ # "2000-01-01T00:00:00.000000+09:00", # "2000-01-01" # ] # ] # ] # ]
Fixes#
[select] Fix a bug displaying a wrong label in
drilldownresults whencommand_versionis3. [groonga-dev,05005][Reported by Atsushi Shinoda]Following is a sample case.
table_create Documents TABLE_NO_KEY column_create Documents tag1 COLUMN_SCALAR ShortText column_create Documents tag2 COLUMN_SCALAR ShortText load --table Documents [ {"tag1": "1", "tag2": "2"} ] select Documents --drilldown tag1,tag2 --command_version 3 # { # "header": { # "return_code": 0, # "start_time": 1672123380.653039, # "elapsed_time": 0.0005846023559570312 # }, # "body": { # "n_hits": 1, # "columns": [ # { # "name": "_id", # "type": "UInt32" # }, # { # "name": "tag1", # "type": "ShortText" # }, # { # "name": "tag2", # "type": "ShortText" # } # ], # "records": [ # [ # 1, # "1", # "2" # ] # ], # "drilldowns": { # "ctor": { # "n_hits": 1, # "columns": [ # { # "name": "_key", # "type": "ShortText" # }, # { # "name": "_nsubrecs", # "type": "Int32" # } # ], # "records": [ # [ # "1", # 1 # ] # ] # }, # "tag2": { # "n_hits": 1, # "columns": [ # { # "name": "_key", # "type": "ShortText" # }, # { # "name": "_nsubrecs", # "type": "Int32" # } # ], # "records": [ # [ # "2", # 1 # ] # ] # } # } # } # }
ctor, displaying right afterdrilldownsas result ofselect, should betag1in correct case. In this sample,ctoris shown instead oftag1. However, what kind of value to be shown is unknown.[NormalizerTable] Fix a bug for Groonga to crush with specific definition setting in
NormalizerTable. [GitHub:pgroonga/pgroonga#279][Reported by i10a]Following case as sample.
table_create Normalizations TABLE_PAT_KEY ShortText --normalizer NormalizerNFKC130 column_create Normalizations normalized COLUMN_SCALAR ShortText load --table Normalizations [ {"_key": "Ⅰ", "normalized": "1"}, {"_key": "Ⅱ", "normalized": "2"}, {"_key": "Ⅲ", "normalized": "3"} ] normalize 'NormalizerTable("normalized", "Normalizations.normalized")' "ⅡⅡ"
This bug is reported to occur when condition meet following 1., 2., and 3..
Keys are normalized in the target table.
In this sample, it meets condition specifying
--normalizer NormalizerNFKC130inNormalizations. Original keys,Ⅰ,Ⅱ,andⅢ, are normalized each intoi,ii,iiiwithNormalizerNFKC130.Same characters in the normalized key are included in the other normalized key.
In this sample, it meets condition because normalized key
iiiincludes the charactersiiandi, same with other normalized keys which are original keyⅡandⅠ.Same characters of 2nd condition are used multiple times.
In this sample, it meets condition because normalized key
iiii, original keyⅡⅡwithNormalizerNFKC130, is considered as same with normalized key forⅢandⅠwithNormalizerNFKC130.Normalizing
iiiiwithNormalizationstakes following steps and it meets the condition.First
iii( applied forⅢ)iioriare not used at first because NormalizerTable works with the Longest-Common-Prefix search.Last
i( applied forⅠ)
Thanks#
i10a
Atsushi Shinoda
Release 12.1.0 - 2022-11-29#
Improvements#
[load] Added support for slow log output of
load.This feature is for Groonga’s performance tuning. For example, you can use this feature to detect records that are taking time longer than average when
loadis slow.Slow log output would be enabled with specifying
GRN SLOWLOG THRESHOLDas the Environment variable.Here is about specifying
GRN_SLOW_LOG_THRESHOLD.GRN_SLOW_LOG_THRESHOLDrequires specifying time in seconds as a threshold. The time of the threshold can shorter than a second with specifying decimal number.A log with debug level would be output if the processing time takes longer than specifyed time with
GRN_SLOW_LOG_THRESHOLD.
Setting for log level would be controled with
log-levelor log_level.What value to specify
GRN_SLOW_LOG_THRESHOLDwould depend on its environment and checking purpose. For an example, we can use following setting to check which records are taking longer time forload. For this, we specify the value based on necesarry time per 1 record caliculated with total number and time ofload.Necessary time to process load would be checked in Query log.
2022-11-16 16:41:27.139000|000000CE63AFF640|>load --table Memo 2022-11-16 16:43:40.841000|000000CE63AFF640|:000133702000000 load(100000): [0][0][100000] 2022-11-16 16:43:40.842000|000000CE63AFF640|:000133703000000 send(46) 2022-11-16 16:43:40.842000|000000CE63AFF640|<000133703000000 rc=0
In this example, the time would be as following;
Number of records: 100000
Time to process: 2 minutes 13 seconds = 133 seconds ( Based on Time stamp for beginning
load: 16:43:27 and time stamp for end ofload(rc=0): 16:43:40 )Time to process 1 record: 0.00133 seconds (133 divided with 100000)
Therefore, we specify
0.00133as a threshold inGRN_SLOW_LOG_THRESHOLDto check which records are taking longer time forload.Note: Enabling slow log may cause following bad effects.
Performance degradation
Larger log size
Thus, the slow log is recommended to be enabled only necessary occasion.
[API] Added new API
grn_is_reference_count_enable().This new API would return boolean weather reference count mode is enabled or not.
[API] Added new API
grn_set_reference_count_enable(bool enable).This new API would enable or disable reference count mode.
For secure usage, this API can’t switch reference count mode if there are multiple open database.
[API] Added new API
grn_is_back_trace_enable().This new API would return boolean weather logging back trace is enabled or not.
[API] Added new API
grn_set_back_trace_enable(bool enable).This new API would enable or disable logging back trace.
In some environments, Groonga crashes when logging back trace, logging back trace should be disabled in such envoronments.
[status] Added new items:
back_traceandreference_count.reference_countindicates weather logging back trace is enabled or not as boolean.back_traceindicates weather logging back trace is enabled or not as boolean.status [ [ 0, 1654237168.304533, 0.0001480579376220703 ], { (omitted) "back_trace": true, "reference_count": false, } ]
Fixes#
[select][Vector column] Fixed a bug displaying integer in the results when a weight vector column specifies WEIGHT FLOAT32.
This bug was only appeared in use of a weight vector column without reference type. A reference type weight vector column does not have this bug.
The bug only affected on the final result display even though internal processes was in floating-point number.
An example for this bug as follows;
table_create Memos TABLE_HASH_KEY ShortText # [[0,0.0,0.0],true] column_create Memos tags COLUMN_VECTOR|WITH_WEIGHT|WEIGHT_FLOAT32 ShortText # [[0,0.0,0.0],true] load --table Memos [ { "_key": "Groonga is fast", "tags": { "groonga": 2.8, "full text search": 1.2 } } ] # [[0,0.0,0.0],1] select Memos # [ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "_id", # "UInt32" # ], # [ # "_key", # "ShortText" # ], # [ # "tags", # "ShortText" # ] # ], # [ # 1, # "Groonga is fast", # { # "groonga": 2, # "full text search": 1 # } # ] # ] # ] # ]
tagscolumn is aShortTexttype weight vector column, sample of non-reference type weight vector column.The results in this sample, the value 2 and 1 are returned as below, evne though the correct value should be 2.8 and 1.2.
{ "groonga": 2, "full text search": 1 }
Applying this fix, the results would be returned as follows;
select Memos # [ # [ # 0, # 0.0, # 0.0 # ], # [ # [ # [ # 1 # ], # [ # [ # "_id", # "UInt32" # ], # [ # "_key", # "ShortText" # ], # [ # "tags", # "ShortText" # ] # ], # [ # 1, # "Groonga is fast", # { # "groonga": 2.8, # "full text search": 1.2 # } # ] # ] # ] # ]
Release 12.0.9 - 2022-10-28#
Improvements#
[AlmaLinux] Added support for AlmaLinux 9.
We had added this support at Release 12.0.8 - 2022-10-03 but haven’t announced it.
[escalate] Added a document for the
escalate()function.[Normalizers] Added
NormalizerHTML. (Experimental)NormalizerHTMLis a normalizer for HTML.Currently
NormalizerHTMLsupports removing tags like<span>or</span>and expanding character references like&or&.Here are sample queries for
NormalizerHTML.normalize NormalizerHTML "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666923364.883798,0.0005481243133544922],{"normalized":" Groonga & Mroonga & Rroonga ","types":[],"checks":[]}]
In this sample
<span>and</span>are removed, and&and&are expanded to&.We can specify whether removing the tags with the
remove_tagoption. (The default value of theremove_tagoption istrue.)normalize 'NormalizerHTML("remove_tag", false)' "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666924069.278549,0.0001978874206542969],{"normalized":"<span> Groonga & Mroonga & Rroonga </span>","types":[],"checks":[]}]
In this sample,
<span>and</span>are not removed.We can specify whether expanding the character references with the
expand_character_referenceoption. (The default value of theexpand_character_referenceoption istrue.)normalize 'NormalizerHTML("expand_character_reference", false)' "<span> Groonga & Mroonga & Rroonga </span>" [[0,1666924357.099782,0.0002346038818359375],{"normalized":" Groonga & Mroonga & Rroonga ","types":[],"checks":[]}]
In this sample,
&and&are not expanded.[httpd] Updated bundled nginx to 1.23.2.
Contains security fixes of CVE-2022-41741 and CVE-2022-41742. Please refer to https://nginx.org/en/CHANGES about these security fixes.
Suppressed logging a lot of same messages when no memory is available.
Groonga could log a lot of
mmap failed!!!!when no memory is available. We improved to log the above message as less duplicates as possible.
Fixes#
[select] Fixed a bug that Groonga could crash or return incorrect results when specifying n_workers.
This bug had occurred when using n_workers with a value greater than
1anddrilldowns[{LABEL}].filterat the same time.The reason why this bug occurred was because Groonga referenced incorrect values (objects) when performing internal parallel processing. So if the condition above was satisfied, Groonga sometimes crashed or returned incorrect results depending on the timing of the parallel processing.
Release 12.0.8 - 2022-10-03#
Improvements#
Changed specification of the
escalate()function (Experimental) to make it easier to use.We changed to not use results out of
escalate().In the previous specification, users had to guess how many results would be passed to
escalate()to determin the first threshold, which was incovenient.Here is a example for the previous
escalate().number_column > 10 && escalate(THRESHOLD_1, CONDITION_1, ..., THRESHOLD_N, CONDITION_N)
CONDITION1was executed when the results ofnumber_column > 10was less or equal toTHRESHOLD_1. Users had to guess how many results would they get fromnumber_column > 10to determineTHRESHOLD_1.From this release, the users don’t need to guess how many results will they get from
number_column > 10, making it easier to set the thresholds.With this change, the syntax of
escalate()changed as follow.The previous syntax
escalate(THRESHOLD_1, CONDITION_1,THRESHOLD_2, CONDITION_2, ..., THRESHOLD_N, CONDITION_N)
The new syntax
escalate(CONDITION_1, THRESHOLD_2, CONDITION_2, ..., THRESHOLD_N, CONDITION_N)
Here are details of the syntax changes.
Don’t require the threshold for the first condition.
Don’t allow empty arguments call. The first condition is required.
Always execute the first condition.
This function is experimental. These behaviors may be changed in the future.
[How to build Groonga with CMake] Added a document about how to build Groonga with CMake.
[Others] Added descriptions about how to enable/disable Apache Arrow support when building with GNU Autotools.
[select] Added a document about drilldowns[${LABEL}].table.
[I18N] Updated the translation procedure.
Fixes#
Fixed a bug that Groonga could return incorrect results when we use NormalizerTable and it contains a non-idempotent (results can be changed when executed repeatedly) definition.
This was caused by that we normalized a search value multiple times: after the value was input and after the value was tokenized.
Groonga tokenizes and normalizes the data to be registered using the tokenizer and normalizer set in the index table when adding a record. The search value is also tokenized and normalized using the tokenizer and normalizer set in the index table, and then the search value and the index are matched. If the search value is the same as the data registered in the index, it will be in the same state as stored in the index because both use the same tokenizer and normalizer.
However, Groonga had normalized extra only search keywords.
Built-in normalizers like NormalizerAuto did’t cause this bug because they are idempotent (results aren’t changed if they are executed repeatedly). On the other hand, NormalizerTable allows the users specify their own normalization definitions, so they can specify non-idempotent (results can be changed when executed repeatedly) definitions.
If there were non-idempotent definitions in NormalizerTable, the indexed data and the search value did not match in some cases because the search value was normalized extra.
In such cases, the data that should hit was not hit or the data that should not hit was hit.
Here is a example.
table_create ColumnNormalizations TABLE_NO_KEY column_create ColumnNormalizations target_column COLUMN_SCALAR ShortText column_create ColumnNormalizations normalized COLUMN_SCALAR ShortText load --table ColumnNormalizations [ {"target_column": "a", "normalized": "b"}, {"target_column": "b", "normalized": "c"} ] table_create Targets TABLE_PAT_KEY ShortText column_create Targets column_normalizations_target_column COLUMN_INDEX \ ColumnNormalizations target_column table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText load --table Memos [ {"content":"a"}, {"content":"c"}, ] table_create \ Terms \ TABLE_PAT_KEY \ ShortText \ --default_tokenizer 'TokenNgram' \ --normalizers 'NormalizerTable("normalized", \ "ColumnNormalizations.normalized", \ "target", \ "target_column")' column_create Terms memos_content COLUMN_INDEX|WITH_POSITION Memos content select Memos --query content:@a [[0,1664781132.892326,0.03527212142944336],[[[1],[["_id","UInt32"],["content","ShortText"]],[2,"c"]]]]
The expected result of
select Memos --query content:@aisa, but Groonga returnedcas a result. This was because we normalized the inputatobby definitions ofColumnNormalizations, and after that, we normalized the normalizedbagain and it was normalized toc. As a result, the inputawas converted tocand matched to{"content":"c"}of theMemostable.
Release 12.0.7 - 2022-08-29#
Improvements#
Added a new function
escalate(). (experimental)The
escalate()function is similar to the existing match escalation ( 検索 ). We can use this function for any conditions. (The existing match escalation is just for one full text search by invert index.)The
escalate()function is useful when we want to limit the number of results of a search. Even if we use--limit, we can limit the number of results of a search. However,--limitis evaluated after evaluating all conditions in our query. Theescalate()function finish the evaluation of conditions at that point when the result set has greater thanTHRESHOLDrecords. In other words, Theescalate()function may reduce the number of evaluating conditions.The syntax of the
escalate()function as below:escalate(THRESHOLD_1, CONDITION_1, THRESHOLD_2, CONDITION_2, ..., THRESHOLD_N, CONDITION_N)
THRESHOLD_Nis a positive number such as 0 and 29.CONDITION_Nis a string that uses Script syntax such asnumber_column > 29.If the current result set has less than or equal to
THRESHOLD_1records, the correspondingCONDITION_1is executed. Similarly, if the next result set has less than or equal toTHRESHOLD_2records, the correspondingCONDITION_2is executed. If the next result set has greater thanTHRESHOLD_3records, theescalate()function is finished.If all
CONDITIONs are executed,escalate(THRESHOLD_1, CONDITION_1, ..., THRESHOLD_N, CONDITION_N)is same asCONDITION_1 || ... || CONDITION_N.The
escalate()function can be worked with logical operators such as&&and&!number_column > 10 && escalate(THRESHOLD_1, CONDITION_1, ..., THRESHOLD_N, CONDITION_N) number_column > 10 &! escalate(THRESHOLD_1, CONDITION_1, ..., THRESHOLD_N, CONDITION_N)They are same as
number_column > 10 && (CONDITION_1 || ... || CONDITION_N)andnumber_column > 10 &! (CONDITION_1 || ... || CONDITION_N).However, these behaviors may be changed because they may not be useful.
[httpd] Updated bundled nginx to 1.23.1.
[select] Add a document for the
--n_workersoption.
Fixes#
Fixed a bug Groonga’s response may be slow when we execute the
request_cancelwhile executing a OR search.When the number of results of the OR search is many and a query has many OR conditions, Groonga may response slow with the “request_cancel” command.
Release 12.0.6 - 2022-08-04#
Improvements#
Added new Munin plugins for groonga-delta.
We can monitoring the following items by plugins for groonga-delta.
Whether
groonga-delta-importcan import or not.grnfile on local storage.Whether
groonga-delta-importcan import or not difference data of MySQL.Whether
groonga-delta-applycan apply imported data or not.The total size of applying data.
[column_copy] Added support for weight vector.
We can copy the value of weight vector by
column_copyas below.table_create Tags TABLE_HASH_KEY ShortText [[0,0.0,0.0],true] table_create CopyFloat32Value TABLE_HASH_KEY ShortText [[0,0.0,0.0],true] column_create CopyFloat32Value source_tags COLUMN_VECTOR|WITH_WEIGHT|WEIGHT_FLOAT32 Tags [[0,0.0,0.0],true] column_create CopyFloat32Value destination_tags COLUMN_VECTOR|WITH_WEIGHT|WEIGHT_FLOAT32 Tags [[0,0.0,0.0],true] load --table CopyFloat32Value [ { "_key": "Groonga is fast!!!", "source_tags": { "Groonga": 2.8, "full text search": 1.5 } } ] [[0,0.0,0.0],1] column_copy CopyFloat32Value source_tags CopyFloat32Value destination_tags [[0,0.0,0.0],true] select CopyFloat32Value [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "destination_tags", "Tags" ], [ "source_tags", "Tags" ] ], [ 1, "Groonga is fast!!!", { "Groonga": 2.8, "full text search": 1.5 }, { "Groonga": 2.8, "full text search": 1.5 } ] ] ] ]
[Ubuntu] Dropped support for Ubuntu 21.10 (Impish Indri).
Because Ubuntu 21.10 reached EOL in July 2022.
[Debian GNU/Linux] Dropped Debian 10 (buster) support.
Because Debian 10 reaches EOL in August 2022.
Fixes#
Fixed a bug that Groonga may crash when we execute drilldown in a parallel by
n_workersoption.[select] Fixed a bug that the syntax error occurred when we specify a very long expression in
--filter.Because the max stack size for the expression of
--filterwas 100 until now.
Release 12.0.5 - 2022-06-29#
Improvements#
[select] Improved a little bit of performance for prefix search by search escalation.
[select] Added support for specifying a reference vector column with weight in
drilldowns[LABEL]._key. [GitHub#1366][Patched by naoa]If we specified a reference vector column with weight in drilldown’s key, Groonga had returned incorrect results until now.
For example, the following tag search had returned incorrect results until now.
table_create Tags TABLE_PAT_KEY ShortText table_create Memos TABLE_HASH_KEY ShortText column_create Memos tags COLUMN_VECTOR|WITH_WEIGHT Tags column_create Memos date COLUMN_SCALAR Time load --table Memos [ {"_key": "Groonga is fast!", "tags": {"full-text-search": 100}, "date": "2014-11-16 00:00:00"}, {"_key": "Mroonga is fast!", "tags": {"mysql": 100, "full-text-search": 80}, "date": "2014-11-16 00:00:00"}, {"_key": "Groonga sticker!", "tags": {"full-text-search": 100, "sticker": 10}, "date": "2014-11-16 00:00:00"}, {"_key": "Rroonga is fast!", "tags": {"full-text-search": 100, "ruby": 20}, "date": "2014-11-17 00:00:00"}, {"_key": "Groonga is good!", "tags": {"full-text-search": 100}, "date": "2014-11-17 00:00:00"} ] select Memos \ --drilldowns[tags].keys tags \ --drilldowns[tags].output_columns _key,_nsubrecs [ [ 0, 1656480220.591901, 0.0005342960357666016 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "date", "Time" ], [ "tags", "Tags" ] ], [ 1, "Groonga is fast!", 1416063600.0, {"full-text-search":100} ], [ 2, "Mroonga is fast!", 1416063600.0, {"mysql":100,"full-text-search":80} ], [ 3, "Groonga sticker!", 1416063600.0, {"full-text-search":100,"sticker":10} ], [ 4, "Rroonga is fast!", 1416150000.0, {"full-text-search":100,"ruby":20} ], [ 5, "Groonga is good!", 1416150000.0, {"full-text-search":100} ] ], { "tags": [ [ 8 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ] ], [ "full-text-search", 5 ], [ "f", 5 ], [ "mysql", 1 ], [ "f", 1 ], [ "sticker", 1 ], [ "f", 1 ], [ "ruby", 1 ], [ "f", 1 ] ] } ]
The above query returns correct results as below since this release.
select Memos --drilldowns[tags].keys tags --drilldowns[tags].output_columns _key,_nsubrecs [ [ 0, 0.0, 0.0 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "date", "Time" ], [ "tags", "Tags" ] ], [ 1, "Groonga is fast!", 1416063600.0, { "full-text-search": 100 } ], [ 2, "Mroonga is fast!", 1416063600.0, { "mysql": 100, "full-text-search": 80 } ], [ 3, "Groonga sticker!", 1416063600.0, { "full-text-search": 100, "sticker": 10 } ], [ 4, "Rroonga is fast!", 1416150000.0, { "full-text-search": 100, "ruby": 20 } ], [ 5, "Groonga is good!", 1416150000.0, { "full-text-search": 100 } ] ], { "tags": [ [ 4 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ] ], [ "full-text-search", 5 ], [ "mysql", 1 ], [ "sticker", 1 ], [ "ruby", 1 ] ] } ] ]
[select] Added support for doing drilldown with a reference vector with weight even if we use
queryorfilter, orpost_filter. [GitHub#1367][Patched by naoa]If we specified a reference vector column with weight in drilldown’s key when we use
queryorfilter, orpost_filter, Groonga had returned incorrect results or errors until now.For example, the following query had been erred until now.
table_create Tags TABLE_PAT_KEY ShortText table_create Memos TABLE_HASH_KEY ShortText column_create Memos tags COLUMN_VECTOR|WITH_WEIGHT Tags column_create Memos date COLUMN_SCALAR Time load --table Memos [ {"_key": "Groonga is fast!", "tags": {"full-text-search": 100}, "date": "2014-11-16 00:00:00"}, {"_key": "Mroonga is fast!", "tags": {"mysql": 100, "full-text-search": 80}, "date": "2014-11-16 00:00:00"}, {"_key": "Groonga sticker!", "tags": {"full-text-search": 100, "sticker": 10}, "date": "2014-11-16 00:00:00"}, {"_key": "Rroonga is fast!", "tags": {"full-text-search": 100, "ruby": 20}, "date": "2014-11-17 00:00:00"}, {"_key": "Groonga is good!", "tags": {"full-text-search": 100}, "date": "2014-11-17 00:00:00"} ] select Memos \ --filter true \ --post_filter true \ --drilldowns[tags].keys tags \ --drilldowns[tags].output_columns _key,_nsubrecs [ [ -22, 1656473820.734894, 0.03771400451660156, "[hash][add][ ] key size unmatch", [ [ "grn_hash_add", "hash.c", 3405 ] ] ], [ [ ] ] ]
The above query returns correct results as below since this release.
select Memos \ --filter true \ --post_filter true \ --drilldowns[tags].keys tags \ --drilldowns[tags].output_columns _key,_nsubrecs [ [ 0, 0.0, 0.0 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "date", "Time" ], [ "tags", "Tags" ] ], [ 1, "Groonga is fast!", 1416063600.0, { "full-text-search": 100 } ], [ 2, "Mroonga is fast!", 1416063600.0, { "mysql": 100, "full-text-search": 80 } ], [ 3, "Groonga sticker!", 1416063600.0, { "full-text-search": 100, "sticker": 10 } ], [ 4, "Rroonga is fast!", 1416150000.0, { "full-text-search": 100, "ruby": 20 } ], [ 5, "Groonga is good!", 1416150000.0, { "full-text-search": 100 } ] ], { "tags": [ [ 4 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ] ], [ "full-text-search", 5 ], [ "mysql", 1 ], [ "sticker", 1 ], [ "ruby", 1 ] ] } ] ]
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
naoa
Release 12.0.4 - 2022-06-06#
Improvements#
[Ubuntu] Added support for Ubuntu 22.04 (Jammy Jellyfish).
We don’t provide groonga-benchmark.
Because nobody will not use it and we can’t maintain it.
[status] Added a new item
memory_map_size.We can get the total memory map size in bytes of Groonga with
statuscommand.status [ [ 0, 1654237168.304533, 0.0001480579376220703 ], { (omitted) "memory_map_size": 70098944 } ]
For example, in Windows, if Groonga uses up physical memory and swap area, Groonga can’t more mapping memory than that. Therefore, we can control properly memory map size by monitoring this value even if the environment does have not enough memory.
Fixes#
Fixed a bug Groonga’s response may be slow when we execute
request_cancelwhile executing a search.Especially, when the number of records of the search target is many, Groonga’s response may be very slow.
Fixed a bug that string list can’t be casted to int32 vector.
For example, the following cast had failed.
[“10”, “100”] -> [10, 100]
This bug only occurs when we specify
apache-arrowintoinput_typeas the argument ofload. This bug occurs in Groonga 12.0.2 or later.Fixed a bug that Groonga Munin Plugins do not work on AlmaLinux 8 and CentOS 7.
Release 12.0.3 - 2022-04-29#
Improvements#
[logical_count] Improved memory usage while
logical_countexecuted.Up to now, Groonga had been keeping objects(objects are tables and columns and indexes, and so on) and temporary tables that were allocated while
logical_countexecuted until the execution oflogical_countfinished.Groonga reduces reference immediately after processing a shard by this feature. Therefore, Groonga can release memory while
logical_countexecuted. The usage of memory of Groonga may reduce because of these reasons.This improvement is only valid for the reference count mode. We can valid the reference count mode by setting
GRN_ENABLE_REFERENCE_COUNT=yes.In addition, Groonga releases temporary tables dynamically while
logical_countis executed by this feature. Therefore, the usage of memory of Groonga reduces. This improvement is valid even if we don’t set the reference count mode.[dump] Added support for
MISSING_IGNORE/MISSING_NIL.If columns had
MISSING_IGNORE/MISSING_NIL, the dump of these columns had failed until now.dumpcommand supports these columns since this release.[snippet],[snippet_html] Added support for text vector as input. [groonga-dev,04956][Reported by shinonon]
For example, we can extract snippets of target text around search keywords against vector in JSON data as below.
table_create Entries TABLE_NO_KEY column_create Entries title COLUMN_SCALAR ShortText column_create Entries contents COLUMN_VECTOR ShortText table_create Tokens TABLE_PAT_KEY ShortText --default_tokenizer TokenNgram --normalizer NormalizerNFKC130 column_create Tokens entries_title COLUMN_INDEX|WITH_POSITION Entries title column_create Tokens entries_contents COLUMN_INDEX|WITH_SECTION|WITH_POSITION Entries contents load --table Entries [ { "title": "Groonga and MySQL", "contents": [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] } ] select Entries\ --output_columns 'snippet_html(contents), contents'\ --match_columns 'title'\ --query Groonga [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "snippet_html", null ], [ "contents", "ShortText" ] ], [ [ "<span class=\"keyword\">Groonga</span> is a full text search engine", "Mroonga is a MySQL storage engine based on <span class=\"keyword\">Groonga</span>" ], [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] ] ] ] ]
Until now, if we specified
snippet*like--output_columns 'snippet_html(contents[1]), we could extract snippets of target text around search keywords against the vector as below. However, we didn’t know which we should output elements. Because we didn’t know which element was hit on search.select Entries\ --output_columns 'snippet_html(contents[0]), contents'\ --match_columns 'title'\ --query Groonga [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "snippet_html", null ], [ "contents", "ShortText" ] ], [ [ "<span class=\"keyword\">Groonga</span> is a full text search engine" ], [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] ] ] ] ]
[
vector_join] Added a new functionvector_join().[groonga-dev,04956][Reported by shinonon]This function can concatenate each elements. We can specify delimiter in the second argument in this function.
For example, we could execute
snippet()andsnippet_html()against vector that concatenate each elements as below.plugin_register functions/vector table_create Entries TABLE_NO_KEY column_create Entries title COLUMN_SCALAR ShortText column_create Entries contents COLUMN_VECTOR ShortText table_create Tokens TABLE_PAT_KEY ShortText --default_tokenizer TokenNgram --normalizer NormalizerNFKC130 column_create Tokens entries_title COLUMN_INDEX|WITH_POSITION Entries title column_create Tokens entries_contents COLUMN_INDEX|WITH_SECTION|WITH_POSITION Entries contents load --table Entries [ { "title": "Groonga and MySQL", "contents": [ "Groonga is a full text search engine", "MySQL is a RDBMS", "Mroonga is a MySQL storage engine based on Groonga" ] } ] select Entries\ --output_columns 'snippet_html(vector_join(contents, "\n")), contents'\ --match_columns 'title'\ --query Groonga --output_pretty yes [ [ 0, 1650849001.524027, 0.0003361701965332031 ], [ [ [ 1 ], [ [ "snippet_html", null ], [ "contents", "ShortText" ] ], [ [ "<span class=\"keyword\">Groonga</span> is a full text search engine\nMySQL is a RDBMS\nMroonga is a MySQL storage engine based on <span class=\"keyword\">Groonga</span>" ], [ "Groonga is a full text search engine","MySQL is a RDBMS","Mroonga is a MySQL storage engine based on Groonga" ] ] ] ] ]
[Indexing] Ignore too large a token like online index construction. [GitHub:pgroonga/pgroonga#209][Reported by Zhanzhao (Deo) Liang]
Groonga doesn’t occur error, but Groonga ignores too large a token when we execute static index construction. However, Groonga output warning in this case.
Fixes#
Fixed a bug that we may be not able to add a key to a table of patricia trie.
This bug occurs in the following conditon.
If a table of patricia trie already has a key.
If the additional key is 4096 bytes.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
shinonon
Zhanzhao (Deo) Liang
Release 12.0.2 - 2022-03-29#
Improvements#
[logical_range_filter] Added support for reducing reference immediately after processing a shard.
Groonga had reduced reference all shards when the finish of
logical_range_filteruntil now. Groonga reduces reference immediately after processing a shard by this feature. The usage of memory may reduce whilelogical_range_filterexecutes by this feature.This feature is only valid for the reference count mode. We can valid the reference count mode by setting
GRN_ENABLE_REFERENCE_COUNT=yes.Normally, Groonga keep objects(tables and column and index, and so on) that Groonga opened even once on memory. However, if we open many objects, Groonga uses much memory. In the reference count mode release objects that are not referenced anywhere from memory. The usage of memory of Groonga may reduce by this.
We increased the stability of the feature of recovering on crashes.
This feature is experimental and it is disabled by default. Therefore, the following improvements are no influence on ordinary users.
We fixed a bug that the index was broken when Groonga crashed.
We fixed a bug that might remain a lock.
We fixed a bug that Groonga crashed while it was recovering the crash.
Improved performance for mmap if anonymous mmap available.[GitHub:MariaDB/server#1999][Suggested by David CARLIER]
The performance of Groonga is improved a bit by this improvement.
[Indexing] Added support for the static index construction against the following types of columns.
The non-reference vector column with weight
The reference vector column with weight
The reference scalar column
These columns have not supported the static index construction until now. Therefore, the time of making the index has longed even if we set the index against these columns after we loaded data into them. By this improvement, the time of making the index is short in this case.
[column_create] Added new flags
MISSING_*andINVALID_*.We added the following new flags for
column_create.MISSING_ADDMISSING_IGNOREMISSING_NILINVALID_ERRORINVALID_WARNINVALID_IGNORE
Normally, if the data column is a reference data column and the nonexistent key is specified, a new record for the nonexistent key is newly created automatically.
The behavior that Groonga adds the key automatically into the column of reference destination is useful in the search like tag search. Because Groonga adds data automatically when we load data.
However, this behavior is inconvenient if we need the other data except for the key. Because a record that only has the key exists.
We can change this behavior by using flags that are added in this release.
MISSING_ADD: This is the default value. This is the same behavior as the current.If the data column is a reference data column and the nonexistent key is specified, a new record for the nonexistent key is newly created automatically.
MISSING_IGNORE:If the data column is a reference data column and the nonexistent key is specified, the nonexistent key is ignored. If the reference data column is a scalar column, the value is
GRN_ID_NIL. If the reference data column is a vector column, the element is just ignored as below["existent1", "nonexistent", "existent2"] -> ["existent1" "existent2"]
MISSING_NIL:If the data column is a reference data column and the nonexistent key is specified, the nonexistent key in a scalar column and a vector column is treated as
GRN_ID_NIL["existent1", "nonexistent", "existent2"] -> ["existent1", "" (GRN_ID_NIL), "existent2"]
INVALID_ERROR: This is the default value. This is the same behavior as the current except an error response of a vector column case.If we set the invalid value (e.g.:
XXXforUInt8scalar column), the set operation is treated as an error. If the data column is a scalar column,loadreports an error in log and response. If the data column is a vector column,loadreports an error in log but doesn’t report an error in response. This is an incompatible change.INVALID_WARN:If we set the invalid value (e.g.:
XXXforUInt8scalar column), a warning message is logged and the set operation is ignored. If the target data column is a reference vector data column,MISSING_IGNOREandMISSING_NILare used to determine the behavior.INVALID_IGNORE:If we set the invalid value (e.g.:
XXXforUInt8scalar column), the set operation is ignored. If the target data column is a reference vector data column,MISSING_IGNOREandMISSING_NILare used to determine the behavior.
[dump][column_list] Added support for
MISSING_*andINVALID_*flags.These commands doesn’t show
MISSING_ADDandINVALID_ERRORflags to keep backward compatibility. Because these flags show the default behavior.[schema] Added support for
MISSING_*andINVALID_*flags.MISSING_ANDandINVALID_ERRORflags aren’t shown inflagsto keep backward compatibility. However, newmissingandinvalidkeys are added to each column.We provided the package of Amazon Linux 2.
[Windows] Dropped support for building with Visual Studio 2017.
Because we could not use windows-2016 image on GitHub Actions.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
David CARLIER
Release 12.0.1 - 2022-02-28#
Improvements#
[query_expand] Added a support for synonym group.
Until now, We had to each defined a keyword and synonyms of the keyword as below when we use the synonym search.
table_create Thesaurus TABLE_PAT_KEY ShortText --normalizer NormalizerAuto # [[0, 1337566253.89858, 0.000355720520019531], true] column_create Thesaurus synonym COLUMN_VECTOR ShortText # [[0, 1337566253.89858, 0.000355720520019531], true] load --table Thesaurus [ {"_key": "mroonga", "synonym": ["mroonga", "tritonn", "groonga mysql"]}, {"_key": "groonga", "synonym": ["groonga", "senna"]} ]
In the above case, if we search
mroonga, Groonga searchmroonga OR tritonn OR "groonga mysql"as we intended. However, if we searchtritonn, Groonga search onlytritonn. If we want to searchtritonn OR mroonga OR "groonga mysql"even if we searchtritonn, we need had added a definition as below.load --table Thesaurus [ {"_key": "tritonn", "synonym": ["tritonn", "mroonga", "groonga mysql"]}, ]
In many cases, if we expand
mroongatomroonga OR tritonn OR "groonga mysql", we feel we want to expandtritonnand"groonga mysql"tomroonga OR tritonn OR "groonga mysql". However, until now, we had needed additional definitions in such a case. Therefore, if target keywords for synonyms are many, we are troublesome to define synonyms. Because we need to define many similar definitions.In addition, when we remove synonyms, we are troublesome because we need to execute remove against many records.
We can make a group by deciding on a representative synonym record since this release. For example, the all following keywords are the “mroonga” group.
load --table Synonyms [ {"_key": "mroonga": "representative": "mroonga"} ] load --table Synonyms [ {"_key": "tritonn": "representative": "mroonga"}, {"_key": "groonga mysql": "representative": "mroonga"} ]
In this case,
mroongais expanded tomroonga OR tritonn OR "groonga mysql". In addition,tritonnand"groonga mysql"are also expanded tomroonga OR tritonn OR "groonga mysql".When we want to remove synonyms, we execute just remove against a target record. For example, if we want to remove
"groonga mysql"from synonyms, we just remove{"_key": "groonga mysql": "representative": "mroonga"}.[query_expand] Added a support for text vector and index.
We can use text vector in a synonym group as below.
table_create SynonymGroups TABLE_NO_KEY [[0,0.0,0.0],true] column_create SynonymGroups synonyms COLUMN_VECTOR ShortText [[0,0.0,0.0],true] table_create Synonyms TABLE_PAT_KEY ShortText [[0,0.0,0.0],true] column_create Synonyms group COLUMN_INDEX SynonymGroups synonyms [[0,0.0,0.0],true] load --table SynonymGroups [ ["synonyms"], [["rroonga", "Ruby groonga"]], [["groonga", "rroonga", "mroonga"]] ] [[0,0.0,0.0],2] query_expand Synonyms.group "rroonga" [ [ 0, 0.0, 0.0 ], "((rroonga) OR (Ruby groonga) OR (groonga) OR (rroonga) OR (mroonga))" ]
Added support for disabling a backtrace by the environment variable.
We can disable output a backtrace by using
GRN_BACK_TRACE_ENABLE. If we setGRN_BACK_TRACE_ENABLE=no, Groonga doesn’t output a backtrace.Groonga output backtrace to a stack area. Therefore, Groonga may crash because Groonga uses up stack area depending on the OS. In such cases, we can avoid crashes by using
GRN_BACK_TRACE_ENABLE=no.[select] Improved performance for
--slices.[Windows] Added support for Visual Studio 2022.
[select] Added support for specifing max intervals for each elements in near search.
For example, we can specify max intervals for each phrase in a near phrase search. We make documentation for this feature in the future. Therefore, we will make more details later.
[Groonga HTTP server] We could use
groonga-server-httpeven if Groonga of RPM packages.
Fixes#
[Windows] Fixed a crash bug when Groonga output backtrace.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Release 12.0.0 - 2022-02-09#
This is a major version up! But It keeps backward compatibility. We can upgrade to 12.0.0 without rebuilding database.
First of all, we introduce the Summary of changes from Groonga 11.0.0 to 11.1.3. Then, we introduce the main changes in 12.0.0.
Summary of changes from Groonga 11.0.0 to 11.1.3#
New Features and Improvements#
[snippet] Added support for using the keyword of 32 or more.
We could not specify the keyword of 32 or more with snippet until now. However, we can specify the keyword of 32 or more by this improvement.
We don’t specify the keyword of 32 or more with snippet in normal use. However, if the keyword increments automatically by using such as
query_expand, the number of target keywords may be 32 or more.In this case, Groonga occurs an error until now. However, Groonga doesn’t occur an error by this improvement.
See release 11.1.3 for details.
[NormalizerNFKC130] Added a new option
remove_symbol.This option removes symbols (e.g. #, !, “, &, %, …) from the string that the target of normalizing. For example, this option useful when we prevent orthographical variants such as a title of song and name of artist, and a name of store.
See release 11.1.3 for details.
[load] Added support for ISO 8601 time format.
ISO 8601 format is the format generally. Therefore
loadbecomes easy to use by Groonga support the more standard format.See release 11.1.0 for details.
[snippet] Added a new option
delimiter_regexpfor detecting snippet delimiter with regular expression.This feature is useful in that we want to show by the sentence the result of the search.
snippet extracts text around search keywords. We call the text that is extracted by snippet snippet.
Normally, snippet () returns the text of 200 bytes around search keywords. However, snippet () gives no thought to a delimiter of sentences. The snippet may be composed of multi sentences.
delimiter_regexpoption is useful if we want to only extract the text of the same sentence as search keywords. For example, we can use\.\s*to extract only text in the target sentence. Note that you need to escape\in string.See release 11.0.9 for details.
[cache_limit] Groonga remove query cache when we execute
cache_limit 0.Groonga stores query cache to internally table. The maximum total size of keys of this table is 4GiB. Because this table is hash table. Therefore, If we execute many huge queries, Groonga may be unable to store query cache, because the maximum total size of keys may be over 4GiB. In such cases, We can clear the table for query cache by using
cache_limit 0, and Groonga can store query cacheWe needed to restart Groonga to resolve this problem until now. However, We can resolve this problem if we just execute
cache_limit 0by this improvement.See release 11.0.6 for details.
[between] Added support for optimizing the order of evaluation of a conditional expression.
We can use the optimization of the order of evaluation of a conditional expression in
between()by settingGRN_EXPR_OPTIMIZE=yes. This optimization is effective with respect ifbetween()narrow down records enough orbetween()can’t narrow down few records.[Log] Added support for outputting to stdout and stderr.
This feature is useful when we execute Groonga on Docker. Docker has the feature that records stdout and stderr in standard. Therefore, we don’t need to login into the environment of Docker to get Groonga’s log.
See release 11.0.4 for details.
[query] Added support for ignoring
TokenFilterStemandTokenFilterStopWordby the query.We are able to search without
TokenFilterStemandTokenFilterStopWordin only a specific query.This feature is useful when we want to search for the same words exactly as a search keyword. Normally, Groonga gets better results with enable stemming and stopwords excepting. However, if we want to search words the same as a keyword of search exactly, These features are needless.
Until now, If we want to search words the same as a keyword of search exactly, We needed to make the index of the exclusive use. By this improvement, we can search words the same as a keyword of search exactly without making the index of the exclusive use.
See release 11.0.3 for details.
[string_slice] Added a new function
string_slice().string_slice()extracts a substring of a string of search results by position or regular expression. This function is useful if we want to edit search results.For example, this feature is useful in that we exclude tags from search results.
[query_parallel_or] Added a new function for processing queries in parallel.
query_parallel_oris similar to query butquery_parallel_orprocesses query that has multiple OR conditions in parallel. We can increase in speed of the execution of many OR conditions by using this function.However,
query_parallel_oralone uses multiple CPUs. Therefore, queries that are executing at the same time as the query that executesquery_parallel_ormay be slow.[Token filters] Added support for multiple token filters with options.
We can use multiple token filters with options as the following example.
--token_filters 'TokenFilterStopWord("column", "ignore"), TokenFilterNFKC130("unify_kana", true)'
[select] Added support for
--post_filterand--slices[].post_filter.We can filter again after we execute
--filterby using--post_filterand--slices[].post_filter. The difference between--post_filterand--slices[].post_filteris response format.The response format of
--post_filtersame as the response of--filter. The response format of--slices[].post_filtershow the result of before and after--slices[].post_filterexecuted.Note that if we use
--slices[].post_filter, the format of the response is different from the normalselect‘s response.
Fixes#
Fixed a bug that the version up of Groonga failed Because the version up of arrow-libs on which Groonga depends.
However, if arrow-libs update a major version, this problem reproduces. In this case, we will handle that by rebuilding the Groonga package.
This bug only occurs AlmaLinux 8 and CentOS 7.
[Windows] Fixed a resource leak when Groonga fail open a new file caused by out of memory.
Fixed a bug that Groonga may not have returned a result of a search query if we sent many search queries when tokenizer, normalizer, or token_filters that support options were used.
Fixed a bug that there is possible that index is corrupt when Groonga executes many additions, delete, and update information in it.
This bug occurs when we only execute many delete information from index. However, it doesn’t occur when we only execute many additions information into index.
See release 11.0.0 for details.
Newly supported OSes#
[AlmaLinux] Added support for AlmaLinux 8.
[AlmaLinux] Added support for AlmaLinux 8 for ARM64.
[Debian GNU/Linux] Added support for Debian 11 (Bullseye).
[Debian GNU/Linux] Added support for Debian 11 (Bullseye) for ARM64 and Debian 10 (buster) for ARM64.
[Ubuntu] Added support for Ubuntu 21.10 (Impish Indri).
Dropped support OSes#
[CentOS] Dropped support for CentOS 8.
[Ubuntu] Dropped support for Ubuntu 21.04 (Hirsute Hippo).
[Ubuntu] Dropped support for Ubuntu 20.10 (Groovy Gorilla).
[Ubuntu] Dropped support for Ubuntu 16.04 LTS (Xenial Xerus).
[Windows] Dropped support for the following packages of Windows version that we had cross-compiled by using MinGW on Linux.
groonga-x.x.x-x86.exe
groonga-x.x.x-x86.zip
groonga-x.x.x-x64.exe
groonga-x.x.x-x86.zip
Thanks#
naoa
Anthony M. Cook
MASUDA Kazuhiro
poti
Takashi Hashida
higchi
wi24rd
Josep Sanz
Keitaro YOSHIMURA
shibanao4870
The main changes in 12.0.0 are as follows.
Improvements#
[sub_filter] Added a new option
pre_filter_threshold.We can change the value of
GRN_SUB_FILTER_PRE_FILTER_THRESHOLDby this option. If the number of records is less thanGRN_SUB_FILTER_PRE_FILTER_THRESHOLDwhen Groonga executessub_filter, Groonga executesub_filteragainst records that have been already narrowed down.We can use -1 to always use this optimization.
[index_column_have_source_record] Added a new function
index_column_have_source_record().We can confirm whether a token that is existing in the index is included in any of the records that are registered in Groonga or not.
Groonga does not remove a token even if the token become never used from records in Groonga by updating records. Therefore, for example, when we use the feature of autocomplete, Groonga may return a token that is not included in any of the records as candidates for search words. However, we can become that we don’t return the needless token by using this function.
Because this function can detect a token that is not included in any of the records.
[NormalizerNFKC130] Added a new option
stripThis option removes spaces from the start and the end as below.
normalize \ 'NormalizerNFKC121("strip", true, \ "report_source_offset", true)' \ " hello world\t! \t " \ WITH_CHECKS|WITH_TYPES [ [ 0, 0.0, 0.0 ], { "normalized": "hello world!", "types": [ "alpha", "alpha", "alpha", "alpha", "alpha", "others", "alpha", "alpha", "alpha", "alpha", "alpha|blank", "symbol|blank" ], "checks": [ 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2 ], "offsets": [ 0, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14 ] } ]
[select] Added new arguments
drilldown_max_n_target_recordsanddrilldown[${LABEL}].max_n_target_records.We can specify the max number of records of the drilldown target table (filtered result) to use drilldown. If the number of filtered result is larger than the specified value, some records in filtered result aren’t used for drilldown. The default value of this arguments are
-1. If these arguments are set-1, Groonga uses all records for drilldown.This argument is useful when filtered result may be very large. Because a drilldown against large filtered result may be slow. We can limit the max number of records to be used for drilldown by this feature.
Here is an example to limit the max number of records to be used for drilldown. The last 2 records,
{\"_id\": 4, \"tag\": \"Senna\"}and{\"_id\": 5, \"tag\": \"Senna\"}, aren’t used.table_create Entries TABLE_HASH_KEY ShortText column_create Entries content COLUMN_SCALAR Text column_create Entries n_likes COLUMN_SCALAR UInt32 column_create Entries tag COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto column_create Terms entries_key_index COLUMN_INDEX|WITH_POSITION Entries _key column_create Terms entries_content_index COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"_key": "The first post!", "content": "Welcome! This is my first post!", "n_likes": 5, "tag": "Hello"}, {"_key": "Groonga", "content": "I started to use Groonga. It's very fast!", "n_likes": 10, "tag": "Groonga"}, {"_key": "Mroonga", "content": "I also started to use Mroonga. It's also very fast! Really fast!", "n_likes": 15, "tag": "Groonga"}, {"_key": "Good-bye Senna", "content": "I migrated all Senna system!", "n_likes": 3, "tag": "Senna"}, {"_key": "Good-bye Tritonn", "content": "I also migrated all Tritonn system!", "n_likes": 3, "tag": "Senna"} ] select Entries \ --limit -1 \ --output_columns _id,tag \ --drilldown tag \ --drilldown_max_n_target_records 3 [ [ 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "tag", "ShortText" ] ], [ 1, "Hello" ], [ 2, "Groonga" ], [ 3, "Groonga" ], [ 4, "Senna" ], [ 5, "Senna" ] ], [ [ 2 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ] ], [ "Hello", 1 ], [ "Groonga", 2 ] ] ] ]
[httpd] Updated bundled nginx to 1.21.6.
Release 11.1.3 - 2022-01-29#
Improvements#
[snippet] Added support for using the keyword of 32 or more. [GitHub#1313][Pathched by Takashi Hashida]
We could not specify the keyword of 32 or more with
snippetuntil now. However, we can specify the keyword of 32 or more by this improvement as below.table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR ShortText load --table Entries [ {"content": "Groonga is a fast and accurate full text search engine based on inverted index. One of the characteristics of Groonga is that a newly registered document instantly appears in search results. Also, Groonga allows updates without read locks. These characteristics result in superior performance on real-time applications.\nGroonga is also a column-oriented database management system (DBMS). Compared with well-known row-oriented systems, such as MySQL and PostgreSQL, column-oriented systems are more suited for aggregate queries. Due to this advantage, Groonga can cover weakness of row-oriented systems.\nThe basic functions of Groonga are provided in a C library. Also, libraries for using Groonga in other languages, such as Ruby, are provided by related projects. In addition, groonga-based storage engines are provided for MySQL and PostgreSQL. These libraries and storage engines allow any application to use Groonga. See usage examples."}, {"content": "In widely used DBMSs, updates are immediately processed, for example, a newly registered record appears in the result of the next query. In contrast, some full text search engines do not support instant updates, because it is difficult to dynamically update inverted indexes, the underlying data structure.\nGroonga also uses inverted indexes but supports instant updates. In addition, Groonga allows you to search documents even when updating the document collection. Due to these superior characteristics, Groonga is very flexible as a full text search engine. Also, Groonga always shows good performance because it divides a large task, inverted index merging, into smaller tasks."} ] select Entries \ --output_columns ' \ snippet(content, \ "groonga", "inverted", "index", "fast", "full", "text", "search", "engine", "registered", "document", \ "results", "appears", "also", "system", "libraries", "for", "mysql", "postgresql", "column-oriented", "dbms", \ "basic", "ruby", "projects", "storage", "allow", "application", "usage", "sql", "well-known", "real-time", \ "weakness", "merging", "performance", "superior", "large", "dynamically", "difficult", "query", "examples", "divides", \ { \ "default_open_tag": "[", \ "default_close_tag": "]", \ "width" : 2048 \ })' [ [ 0, 1643165838.691991, 0.0003311634063720703 ], [ [ [ 2 ], [ [ "snippet", null ] ], [ [ "[Groonga] is a [fast] and accurate [full] [text] [search] [engine] based on [inverted] [index]. One of the characteristics of [Groonga] is that a newly [registered] [document] instantly [appears] in [search] [results]. [Also], [Groonga] [allow]s updates without read locks. These characteristics result in [superior] [performance] on [real-time] [application]s.\n[Groonga] is [also] a [column-oriented] database management [system] ([DBMS]). Compared with [well-known] row-oriented [system]s, such as [MySQL] and [PostgreSQL], [column-oriented] [system]s are more suited [for] aggregate queries. Due to this advantage, [Groonga] can cover [weakness] of row-oriented [system]s.\nThe [basic] functions of [Groonga] are provided in a C library. [Also], [libraries] [for] using [Groonga] in other languages, such as [Ruby], are provided by related [projects]. In addition, [groonga]-based [storage] [engine]s are provided [for] [MySQL] and [PostgreSQL]. These [libraries] and [storage] [engine]s [allow] any [application] to use [Groonga]. See [usage] [examples]." ] ], [ [ "In widely used [DBMS]s, updates are immediately processed, [for] example, a newly [registered] record [appears] in the result of the next [query]. In contrast, some [full] [text] [search] [engine]s do not support instant updates, because it is [difficult] to [dynamically] update [inverted] [index]es, the underlying data structure.\n[Groonga] [also] uses [inverted] [index]es but supports instant updates. In addition, [Groonga] [allow]s you to [search] [document]s even when updating the [document] collection. Due to these [superior] characteristics, [Groonga] is very flexible as a [full] [text] [search] [engine]. [Also], [Groonga] always shows good [performance] because it [divides] a [large] task, [inverted] [index] [merging], into smaller tasks." ] ] ] ] ]
[NormalizerNFKC130] Added a new option
remove_symbol.This option removes symbols (e.g. #, !, “, &, %, …) from the string that the target of normalizing as below.
normalize 'NormalizerNFKC130("remove_symbol", true)' "#This & is %% a pen." WITH_TYPES [ [ 0, 1643595008.729597, 0.0005540847778320312 ], { "normalized": "this is a pen", "types": [ "alpha", "alpha", "alpha", "alpha", "others", "others", "alpha", "alpha", "others", "others", "alpha", "others", "alpha", "alpha", "alpha" ], "checks": [ ] } ]
[AlmaLinux] Added support for AlmaLinux 8 on ARM64.
[httpd] Updated bundled nginx to 1.21.5.
[Documentation] Fixed a typo in ruby_eval. [GitHub#1317][Pathched by wi24rd]
[Ubuntu] Dropped Ubuntu 21.04 (Hirsute Hippo) support.
Because Ubuntu 21.04 reached EOL January 20, 2022.
Fixes#
[load] Fixed a crash bug when we load data with specifying a nonexistent column.
This bug only occurs when we specify
apache-arrowintoinput_typeas the argument ofload.Fixed a bug that the version up of Groonga failed Because the version up of arrow-libs on which Groonga depends. [groonga-talk,540][Reported by Josep Sanz][Gitter,61eaaa306d9ba23328d23ce1][Reported by shibanao4870][GitHub#1316][Reported by Keitaro YOSHIMURA]
However, if arrow-libs update a major version, this problem reproduces. In this case, we will handle that by rebuilding the Groonga package.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
Takashi Hashida
wi24rd
Josep Sanz
Keitaro YOSHIMURA
shibanao4870
Release 11.1.1 - 2021-12-29#
Improvements#
[select] Added support for near phrase product search.
This feature is a shortcut of
'*NP"..." OR *NP"..." OR ...'. For example, we can use*NPPinstead of the expression that execute mulitiple*NPwithqueryas below.query ("title * 10 || content", "*NP"a 1 x" OR *NP"a 1 y" OR *NP"a 1 z" OR *NP"a 2 x" OR *NP"a 2 y" OR *NP"a 2 z" OR *NP"a 3 x" OR *NP"a 3 y" OR *NP"a 3 z" OR *NP"b 1 x" OR *NP"b 1 y" OR *NP"b 1 z" OR *NP"b 2 x" OR *NP"b 2 y" OR *NP"b 2 z" OR *NP"b 3 x" OR *NP"b 3 y" OR *NP"b 3 z"")
We can be written as
*NPP"(a b) (1 2 3) (x y z)"the above expression by this feature. In addition,*NPP"(a b) (1 2 3) (x y z)"is faster than'*NP"..." OR *NP"..." OR ...'.query ("title * 10 || content", "*NPP"(a b) (1 2 3) (x y z)"")
We implements this feature for improving performance near phrase search like
'*NP"..." OR *NP"..." OR ...'.[select] Added support for order near phrase product search.
This feature is a shortcut of
'*ONP"..." OR *ONP"..." OR ...'. For example, we can use*ONPPinstead of the expression that execute multiple*ONPwithqueryas below.query ("title * 10 || content", "*ONP"a 1 x" OR *ONP"a 1 y" OR *ONP"a 1 z" OR *ONP"a 2 x" OR *ONP"a 2 y" OR *ONP"a 2 z" OR *ONP"a 3 x" OR *ONP"a 3 y" OR *ONP"a 3 z" OR *ONP"b 1 x" OR *ONP"b 1 y" OR *ONP"b 1 z" OR *ONP"b 2 x" OR *ONP"b 2 y" OR *ONP"b 2 z" OR *ONP"b 3 x" OR *ONP"b 3 y" OR *ONP"b 3 z"")
We can be written as
*ONPP"(a b) (1 2 3) (x y z)"the above expression by this feature. In addition,*ONPP"(a b) (1 2 3) (x y z)"is faster than'*ONP"..." OR *ONP"..." OR ...'.query ("title * 10 || content", "*ONPP"(a b) (1 2 3) (x y z)"")
We implements this feature for improving performance near phrase search like
'*ONP"..." OR *ONP"..." OR ...'.[request_cancel] Groonga became easily detects
request_cancelwhile executing a search.Because we added more checks of return code to detect
request_cancel.[thread_dump] Added a new command
thread_dumpCurrently, this command works only on Windows.
We can put a backtrace of all threads into a log as logs of NOTICE level at the time of running this command.
This feature is useful when we solve a problem such as Groonga doesn’t return a response.
[CentOS] Dropped support for CentOS 8.
Because CentOS 8 will reach EOL at 2021-12-31.
Fixes#
Fixed a bug that we can’t remove a index column with invalid parameter. [GitHub#1301][Patched by Takashi Hashida]
For example, we can’t remove a table when we create an invalid index column with
column_createas below.table_create Statuses TABLE_NO_KEY column_create Statuses start_time COLUMN_SCALAR UInt16 column_create Statuses end_time COLUMN_SCALAR UInt16 table_create Times TABLE_PAT_KEY UInt16 column_create Times statuses COLUMN_INDEX Statuses start_time,end_time [ [ -22, 1639037503.16114, 0.003981828689575195, "grn_obj_set_info(): GRN_INFO_SOURCE: multi column index must be created with WITH_SECTION flag: <Times.statuses>", [ [ "grn_obj_set_info_source_validate", "../../groonga/lib/db.c", 9605 ], [ "/tmp/d.grn", 6, "column_create Times statuses COLUMN_INDEX Statuses start_time,end_time" ] ] ], false ] table_remove Times [ [ -22, 1639037503.16515, 0.0005414485931396484, "[object][remove] column is broken: <Times.statuses>", [ [ "remove_columns", "../../groonga/lib/db.c", 10649 ], [ "/tmp/d.grn", 8, "table_remove Times" ] ] ], false ]
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
Takashi Hashida
Release 11.1.0 - 2021-11-29#
Improvements#
[load] Added support for ISO 8601 time format.[GitHub#1228][Patched by Takashi Hashida]
loadsupport the following format by this modification.YYYY-MM-ddThh:mm:ss.sZ
YYYY-MM-ddThh:mm:ss.s+10:00
YYYY-MM-ddThh:mm:ss.s-10:00
We can also use
tandzcharacters instead ofTandZin this syntax. We can also use/character instead of-in this syntax. However, note that this is not an ISO 8601 format. This format is present for compatibility.plugin_register functions/time table_create Logs TABLE_NO_KEY column_create Logs case COLUMN_SCALAR ShortText column_create Logs created_at COLUMN_SCALAR Time column_create Logs created_at_text COLUMN_SCALAR ShortText load --table Logs [ {"case": "timezone: Z", "created_at": "2000-01-01T10:00:00Z", "created_at_text": "2000-01-01T10:00:00Z"}, {"case": "timezone: z", "created_at": "2000-01-01t10:00:00z", "created_at_text": "2000-01-01T10:00:00z"}, {"case": "timezone: 00:00", "created_at": "2000-01-01T10:00:00+00:00", "created_at_text": "2000-01-01T10:00:00+00:00"}, {"case": "timezone: +01:01", "created_at": "2000-01-01T11:01:00+01:01", "created_at_text": "2000-01-01T11:01:00+01:01"}, {"case": "timezone: +11:11", "created_at": "2000-01-01T21:11:00+11:11", "created_at_text": "2000-01-01T21:11:00+11:11"}, {"case": "timezone: -01:01", "created_at": "2000-01-01T08:59:00-01:01", "created_at_text": "2000-01-01T08:59:00-01:01"}, {"case": "timezone: -11:11", "created_at": "1999-12-31T22:49:00-11:11", "created_at_text": "1999-12-31T22:49:00-11:11"}, {"case": "timezone hour threshold: +23:00", "created_at": "2000-01-02T09:00:00+23:00", "created_at_text": "2000-01-02T09:00:00+23:00"}, {"case": "timezone minute threshold: +00:59", "created_at": "2000-01-01T10:59:00+00:59", "created_at_text": "2000-01-01T10:59:00+00:59"}, {"case": "timezone omitting minute: +01", "created_at": "2000-01-01T11:00:00+01", "created_at_text": "2000-01-01T11:00:00+01"}, {"case": "timezone omitting minute: -01", "created_at": "2000-01-01T09:00:00-01", "created_at_text": "2000-01-01T09:00:00-01"}, {"case": "timezone: localtime", "created_at": "2000-01-01T19:00:00", "created_at_text": "2000-01-01T19:00:00"}, {"case": "compatible: date delimiter: /", "created_at": "2000/01/01T10:00:00Z", "created_at_text": "2000/01/01T10:00:00Z"}, {"case": "decimal", "created_at": "2000-01-01T11:01:00.123+01:01", "created_at_text": "2000-01-01T11:01:00.123+01:01"} ] select Logs \ --limit -1 \ --output_columns "case, time_format_iso8601(created_at), created_at_text" [ [ 0, 0.0, 0.0 ], [ [ [ 14 ], [ [ "case", "ShortText" ], [ "time_format_iso8601", null ], [ "created_at_text", "ShortText" ] ], [ "timezone: Z", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T10:00:00Z" ], [ "timezone: z", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T10:00:00z" ], [ "timezone: 00:00", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T10:00:00+00:00" ], [ "timezone: +01:01", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T11:01:00+01:01" ], [ "timezone: +11:11", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T21:11:00+11:11" ], [ "timezone: -01:01", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T08:59:00-01:01" ], [ "timezone: -11:11", "2000-01-01T19:00:00.000000+09:00", "1999-12-31T22:49:00-11:11" ], [ "timezone hour threshold: +23:00", "2000-01-01T19:00:00.000000+09:00", "2000-01-02T09:00:00+23:00" ], [ "timezone minute threshold: +00:59", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T10:59:00+00:59" ], [ "timezone omitting minute: +01", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T11:00:00+01" ], [ "timezone omitting minute: -01", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T09:00:00-01" ], [ "timezone: localtime", "2000-01-01T19:00:00.000000+09:00", "2000-01-01T19:00:00" ], [ "compatible: date delimiter: /", "2000-01-01T19:00:00.000000+09:00", "2000/01/01T10:00:00Z" ], [ "decimal", "2000-01-01T19:00:00.123000+09:00", "2000-01-01T11:01:00.123+01:01" ] ] ] ]
[select] Added a new
query_flagsDISABLE_PREFIX_SEARCH.We can use the prefix search operators
^and*as search keywords byDISABLE_PREFIX_SEARCHas below.This feature is useful if we want to search documents including
^and*.table_create Users TABLE_PAT_KEY ShortText load --table Users [ {"_key": "alice"}, {"_key": "alan"}, {"_key": "ba*"} ] select Users \ --match_columns "_key" \ --query "a*" \ --query_flags "DISABLE_PREFIX_SEARCH" [[0,0.0,0.0],[[[1],[["_id","UInt32"],["_key","ShortText"]],[3,"ba*"]]]]
table_create Users TABLE_PAT_KEY ShortText load --table Users [ {"_key": "alice"}, {"_key": "alan"}, {"_key": "^a"} ] select Users \ --query "_key:^a" \ --query_flags "ALLOW_COLUMN|DISABLE_PREFIX_SEARCH" [[0,0.0,0.0],[[[1],[["_id","UInt32"],["_key","ShortText"]],[3,"^a"]]]]
[select] Added a new
query_flagsDISABLE_AND_NOT.We can use
AND NOToperators-as search keywords byDISABLE_AND_NOTas below.This feature is useful if we want to search documents including
-.table_create Users TABLE_PAT_KEY ShortText load --table Users [ {"_key": "alice"}, {"_key": "bob"}, {"_key": "cab-"} ] select Users --match_columns "_key" --query "b - a" --query_flags "DISABLE_AND_NOT" [[0,0.0,0.0],[[[1],[["_id","UInt32"],["_key","ShortText"]],[3,"cab-"]]]]
Fixes#
[The browser based administration tool] Fixed a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list. [GitHub#1186][Patched by Takashi Hashida]
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
Takashi Hashida
Release 11.0.9 - 2021-11-04#
Improvements#
[snippet] Added a new option
delimiter_regexpfor detecting snippet delimiter with regular expression.snippet extracts text around search keywords. We call the text that is extracted by snippet snippet.
Normally, snippet () returns the text of 200 bytes around search keywords. However, snippet () gives no thought to a delimiter of sentences. The snippet may be composed of multi sentences.
delimiter_regexpoption is useful if we want to only extract the text of the same sentence as search keywords. For example, we can use\.\s*to extract only text in the target sentence as below. Note that you need to escape\in string.table_create Documents TABLE_NO_KEY column_create Documents content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto column_create Terms documents_content_index COLUMN_INDEX|WITH_POSITION Documents content load --table Documents [ ["content"], ["Groonga is a fast and accurate full text search engine based on inverted index. One of the characteristics of groonga is that a newly registered document instantly appears in search results. Also, groonga allows updates without read locks. These characteristics result in superior performance on real-time applications."], ["Groonga is also a column-oriented database management system (DBMS). Compared with well-known row-oriented systems, such as MySQL and PostgreSQL, column-oriented systems are more suited for aggregate queries. Due to this advantage, groonga can cover weakness of row-oriented systems."] ] select Documents \ --output_columns 'snippet(content, \ { \ "default_open_tag": "[", \ "default_close_tag": "]", \ "delimiter_regexp": "\\\\.\\\\s*" \ })' \ --match_columns content \ --query "fast performance" [ [ 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 1 ], [ [ "snippet", null ] ], [ [ "Groonga is a [fast] and accurate full text search engine based on inverted index", "These characteristics result in superior [performance] on real-time applications" ] ] ] ] ]
[window_rank] Added a new function
window_rank().We can calculate a rank that includes a gap of each record. Normally, the rank isn’t incremented when multiple records that are the same order. For example, if values of sort keys are 100, 100, 200 then the ranks of them are 1, 1, 3. The rank of the last record is 3 not 2 because there are two 1 rank records.
This is similar to window_record_number. However, window_record_number gives no thought to gap.
table_create Points TABLE_NO_KEY column_create Points game COLUMN_SCALAR ShortText column_create Points score COLUMN_SCALAR UInt32 load --table Points [ ["game", "score"], ["game1", 100], ["game1", 200], ["game1", 100], ["game1", 400], ["game2", 150], ["game2", 200], ["game2", 200], ["game2", 200] ] select Points \ --columns[rank].stage filtered \ --columns[rank].value 'window_rank()' \ --columns[rank].type UInt32 \ --columns[rank].window.sort_keys score \ --output_columns 'game, score, rank' \ --sort_keys score [ [ 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 8 ], [ [ "game", "ShortText" ], [ "score", "UInt32" ], [ "rank", "UInt32" ] ], [ "game1", 100, 1 ], [ "game1", 100, 1 ], [ "game2", 150, 3 ], [ "game2", 200, 4 ], [ "game2", 200, 4 ], [ "game1", 200, 4 ], [ "game2", 200, 4 ], [ "game1", 400, 8 ] ] ] ]
[in_values] Added support for auto cast when we search tables.
For example, if we load values of
UInt32into a table that a key type isUInt64, Groonga cast the values toUInt64automatically when we search the table within_values(). However,in_values(_key, 10)doesn’t work withUInt64key table. Because 10 is parsed asInt32.table_create Numbers TABLE_HASH_KEY UInt64 load --table Numbers [ {"_key": 100}, {"_key": 200}, {"_key": 300} ] select Numbers --output_columns _key --filter 'in_values(_key, 200, 100)' --sortby _id [[0,0.0,0.0],[[[2],[["_key","UInt64"]],[100],[200]]]]
[httpd] Updated bundled nginx to 1.21.3.
[AlmaLinux] Added support for AlmaLinux 8.
[Ubuntu] Added support for Ubuntu 21.10 (Impish Indri).
Fixes#
Fixed a bug that Groonga doesn’t return a response when an error occurred in command (e.g. sytax error in filter).
This bug only occurs when we use
--output_type apache-arrow.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Release 11.0.7 - 2021-09-29#
Improvements#
[load] Added support for casting a string like as “[int, int,…]” to a vector of integer like as [int, int,…].
For example, Groonga handle as a vector of integer like as [1, -2] even if we load vector of string like as “[1, -2]” as below.
table_create Data TABLE_NO_KEY column_create Data numbers COLUMN_VECTOR Int16 table_create Numbers TABLE_PAT_KEY Int16 column_create Numbers data_numbers COLUMN_INDEX Data numbers load --table Data [ {"numbers": "[1, -2]"}, {"numbers": "[-3, 4]"} ] dump --dump_plugins no --dump_schema no load --table Data [ ["_id","numbers"], [1,[1,-2]], [2,[-3,4]] ] column_create Numbers data_numbers COLUMN_INDEX Data numbers select Data --filter 'numbers @ -2' [[0,0.0,0.0],[[[1],[["_id","UInt32"],["numbers","Int16"]],[1,[1,-2]]]]]
This feature supports for the floowings types.
Int8
UInt8
Int16
UInt16
Int32
UInt32
Int64
UInt64
[load] Added support for loading a JSON array expressed as a text string as a vector of string.
For example, Groonga handle as a vector that has two elements like as [“hello”, “world”] if we load JSON array expressed as a text string like as “["hello", "world"]” as below.
table_create Data TABLE_NO_KEY [[0,0.0,0.0],true] column_create Data strings COLUMN_VECTOR ShortText [[0,0.0,0.0],true] table_create Terms TABLE_PAT_KEY ShortText --normalizer NormalizerNFKC130 --default_tokenizer TokenNgram [[0,0.0,0.0],true] column_create Terms data_strings COLUMN_INDEX Data strings [[0,0.0,0.0],true] load --table Data [ {"strings": "[\"Hello\", \"World\"]"}, {"strings": "[\"Good-bye\", \"World\"]"} ] [[0,0.0,0.0],2] dump --dump_plugins no --dump_schema no load --table Data [ ["_id","strings"], [1,["Hello","World"]], [2,["Good-bye","World"]] ] column_create Terms data_strings COLUMN_INDEX Data strings select Data --filter 'strings @ "bye"' [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "strings", "ShortText" ] ], [ 2, [ "Good-bye", "World" ] ] ] ] ]
In before version, Groonga handled as a vector that had one element like as [“["hello", "world"]”] if we loaded JSON array expressed as a text string like as “["hello", "world"]”.
[Documentation] Added a documentation about the following items.
[column_create] Added a documentation about
WEIGHT_FLOAT32flag.[NormalizerNFKC121] Added a documentation about
NormalizerNFKC121.[NormalizerNFKC130] Added a documentation about
NormalizerNFKC130.[NormalizerTable] Added a documentation about
NormalizerTable.
Updated to 3.0.0 that the version of Apache Arrow that Groonga requires. [GitHub#1265][Patched by Takashi Hashida]
Fixes#
Fixed a memory leak when we created a table with a tokenizer with invalid option.
Fixed a bug that may not add a new entry in Hash table.
This bug only occurs in Groonga 11.0.6, and it may occur if we quite a lot of add and delete data. If this bug occurs in your environment, you can resolve this problem by executing the following steps.
We upgrade Groonga to 11.0.7 or later from 11.0.6.
We make a new table that has the same schema as the original table.
We copy data to the new table from the original table.
[Windows] Fixed a resource leak when Groonga fail open a new file caused by out of memory.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
Takashi Hashida
Release 11.0.6 - 2021-08-29#
Warning
Groonga 11.0.6 has had a bug that may not add a new entry in Hash table.
We fixed this bug on Groonga 11.0.7. This bug only occurs in Groonga 11.0.6. Therefore, if you were using Groonga 11.0.6, we highly recommended that you use Groonga 11.0.7 or later.
Improvements#
Added support for recovering on crash. (experimental)
This is a experimental feature. Currently, this feature is still not stable.
If Groonga crashes, it recovers the database automatically when it opens a database for the first time since the crash. However, This feature can’t recover the database automatically in all crash cases. We need to recover the database manually depending on timing even if this feature enables.
Groonga execute WAL (write ahead log) when this feature is enable. We can dump WAL by the following tools, but currently, users doesn’t need to use them.
[grndb]
dump-walcommand.dump-wal.rbscritp.
[cache_limit] Groonga remove cache when we execute
cache_limit 0. [GitHub#1224][Reported by higchi]Groonga stores query cache to internally table. The maximum total size of keys of this table is 4GiB. Because this table is hash table. Therefore, If we execute many huge queries, Groonga may be unable to store query cache, because the maximum total size of keys may be over 4GiB. In such cases, We can clear the table for query cache by using
cache_limit 0, and Groonga can store query cache
Fixes#
Fixed a bug that Groonga doesn’t clear lock when some threads open the same object around the same time.
If some threads open the same object around the same time, threads except for a thread that executes the opening object at first are waiting for opening the target object. At this time, threads that wait for an opening object take locks, but these locks are not released. Therefore, these locks remain until Groonga’s process is restarted in the above case, and a new thread can’t also open the object all the time until Groonga’s process is restarted.
However, this bug rarely happens. Because a time of a thread open the object is a very short time.
[query_parallel_or] Fixed a bug that result may be different from the
query().For example, If we used
query("tags || tags2", "beginner man"), the following record was a match, but if we usedquery_parallel_or("tags || tags2", "beginner man"), the following record wasn’t a match until now.{"_key": "Bob", "comment": "Hey!", "tags": ["expert", "man"], "tags2": ["beginner"]}
Even if we use
query_parallel_or("tags || tags2", "beginner man"), the above record is match by this modification.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Groonga may not return records that should match caused by
GRN_II_CURSOR_SET_MIN_ENABLE.
Thanks#
higchi
Release 11.0.5 - 2021-07-29#
Improvements#
[Normalizers] Added support for multiple normalizers.
We can specify multiple normalizers by
--notmalizersoption when we create a table since this release. If we can also specify them by--normalizerexisting option because of compatibility.We added
NormalizerTablefor customizing a normalizer in Groonga 11.0.4. We can more flexibly behavior of the normalizer by combiningNormalizerTablewith existing normalizer.For example, this feature is useful in the following case.
Search for a telephone number. However, we import data handwritten by OCR. If data is handwritten, OCR may misunderstand a number and string(e.g. 5 and S).
The details are as follows.
table_create Normalizations TABLE_PAT_KEY ShortText column_create Normalizations normalized COLUMN_SCALAR ShortText load --table Normalizations [ {"_key": "s", "normalized": "5"} ] table_create Tels TABLE_NO_KEY column_create Tels tel COLUMN_SCALAR ShortText table_create TelsIndex TABLE_PAT_KEY ShortText \ --normalizers 'NormalizerNFKC130("unify_hyphen_and_prolonged_sound_mark", true), \ NormalizerTable("column", "Normalizations.normalized")' \ --default_tokenizer 'TokenNgram("loose_symbol", true, "loose_blank", true)' column_create TelsIndex tel_index COLUMN_INDEX|WITH_SECTION Tels tel load --table Tels [ {"tel": "03-4S-1234"} {"tel": "03-45-9876"} ] select --table Tels \ --filter 'tel @ "03-45-1234"' [ [ 0, 1625227424.560146, 0.0001730918884277344 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "tel", "ShortText" ] ], [ 1, "03-4S-1234" ] ] ] ]
Existing normalizers can’t meet in such case, but we can meet it by combining
NormalizerTablewith existing normalizer since this release.[query_parallel_or][query] Added support for customizing thresholds for sequential search.
We can customize thresholds in each queries whether to use sequential search by the following options.
{"max_n_enough_filtered_records": xx}max_n_enough_filtered_recordsspecify the number of records.queryorquery_parallel_oruse sequential search when they seems to narrow down until under this number.{"enough_filtered_ratio": x.x}enough_filtered_ratiospecify percentage of total.queryorquery_parallel_oruse sequential search when they seems to narrow down until under this percentage. For example, if we specify{"enough_filtered_ratio": 0.5},queryorquery_parallel_oruse sequential search when they seems to narrow down until half of the whole.
The details are as follows.
table_create Products TABLE_NO_KEY column_create Products name COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText --normalizer NormalizerAuto column_create Terms products_name COLUMN_INDEX Products name load --table Products [ ["name"], ["Groonga"], ["Mroonga"], ["Rroonga"], ["PGroonga"], ["Ruby"], ["PostgreSQL"] ] select \ --table Products \ --filter 'query("name", "r name:Ruby", {"enough_filtered_ratio": 0.5})'
table_create Products TABLE_NO_KEY column_create Products name COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText --normalizer NormalizerAuto column_create Terms products_name COLUMN_INDEX Products name load --table Products [ ["name"], ["Groonga"], ["Mroonga"], ["Rroonga"], ["PGroonga"], ["Ruby"], ["PostgreSQL"] ] select \ --table Products \ --filter 'query("name", "r name:Ruby", {"max_n_enough_filtered_records": 10})'
[between][in_values] Added support for customizing thresholds for sequential search.
[between] and [in_values] have a feature that they switch to sequential search when the target of search records is narrowed down enough.
The value of
GRN_IN_VALUES_TOO_MANY_INDEX_MATCH_RATIO/GRN_BETWEEN_TOO_MANY_INDEX_MATCH_RATIOis used as threshold whether Groonga execute sequential search or search with indexes in such a case.This behavior is customized by only the following environment variable until now.
in_values():# Don't use auto sequential search GRN_IN_VALUES_TOO_MANY_INDEX_MATCH_RATIO=-1 # Set threshold to 0.02 GRN_IN_VALUES_TOO_MANY_INDEX_MATCH_RATIO=0.02
between():# Don't use auto sequential search GRN_BETWEEN_TOO_MANY_INDEX_MATCH_RATIO=-1 # Set threshold to 0.02 GRN_BETWEEN_TOO_MANY_INDEX_MATCH_RATIO=0.02
if customize by the environment variable, this threshold applies to all queries, but we can specify it in each query by using this feature.
The details are as follows. We can specify the threshold by using
{"too_many_index_match_ratio": x.xx}option. The value type of this option isdouble.table_create Memos TABLE_HASH_KEY ShortText column_create Memos timestamp COLUMN_SCALAR Time table_create Times TABLE_PAT_KEY Time column_create Times memos_timestamp COLUMN_INDEX Memos timestamp load --table Memos [ {"_key": "001", "timestamp": "2014-11-10 07:25:23"}, {"_key": "002", "timestamp": "2014-11-10 07:25:24"}, {"_key": "003", "timestamp": "2014-11-10 07:25:25"}, {"_key": "004", "timestamp": "2014-11-10 07:25:26"}, {"_key": "005", "timestamp": "2014-11-10 07:25:27"}, {"_key": "006", "timestamp": "2014-11-10 07:25:28"}, {"_key": "007", "timestamp": "2014-11-10 07:25:29"}, {"_key": "008", "timestamp": "2014-11-10 07:25:30"}, {"_key": "009", "timestamp": "2014-11-10 07:25:31"}, {"_key": "010", "timestamp": "2014-11-10 07:25:32"}, {"_key": "011", "timestamp": "2014-11-10 07:25:33"}, {"_key": "012", "timestamp": "2014-11-10 07:25:34"}, {"_key": "013", "timestamp": "2014-11-10 07:25:35"}, {"_key": "014", "timestamp": "2014-11-10 07:25:36"}, {"_key": "015", "timestamp": "2014-11-10 07:25:37"}, {"_key": "016", "timestamp": "2014-11-10 07:25:38"}, {"_key": "017", "timestamp": "2014-11-10 07:25:39"}, {"_key": "018", "timestamp": "2014-11-10 07:25:40"}, {"_key": "019", "timestamp": "2014-11-10 07:25:41"}, {"_key": "020", "timestamp": "2014-11-10 07:25:42"}, {"_key": "021", "timestamp": "2014-11-10 07:25:43"}, {"_key": "022", "timestamp": "2014-11-10 07:25:44"}, {"_key": "023", "timestamp": "2014-11-10 07:25:45"}, {"_key": "024", "timestamp": "2014-11-10 07:25:46"}, {"_key": "025", "timestamp": "2014-11-10 07:25:47"}, {"_key": "026", "timestamp": "2014-11-10 07:25:48"}, {"_key": "027", "timestamp": "2014-11-10 07:25:49"}, {"_key": "028", "timestamp": "2014-11-10 07:25:50"}, {"_key": "029", "timestamp": "2014-11-10 07:25:51"}, {"_key": "030", "timestamp": "2014-11-10 07:25:52"}, {"_key": "031", "timestamp": "2014-11-10 07:25:53"}, {"_key": "032", "timestamp": "2014-11-10 07:25:54"}, {"_key": "033", "timestamp": "2014-11-10 07:25:55"}, {"_key": "034", "timestamp": "2014-11-10 07:25:56"}, {"_key": "035", "timestamp": "2014-11-10 07:25:57"}, {"_key": "036", "timestamp": "2014-11-10 07:25:58"}, {"_key": "037", "timestamp": "2014-11-10 07:25:59"}, {"_key": "038", "timestamp": "2014-11-10 07:26:00"}, {"_key": "039", "timestamp": "2014-11-10 07:26:01"}, {"_key": "040", "timestamp": "2014-11-10 07:26:02"}, {"_key": "041", "timestamp": "2014-11-10 07:26:03"}, {"_key": "042", "timestamp": "2014-11-10 07:26:04"}, {"_key": "043", "timestamp": "2014-11-10 07:26:05"}, {"_key": "044", "timestamp": "2014-11-10 07:26:06"}, {"_key": "045", "timestamp": "2014-11-10 07:26:07"}, {"_key": "046", "timestamp": "2014-11-10 07:26:08"}, {"_key": "047", "timestamp": "2014-11-10 07:26:09"}, {"_key": "048", "timestamp": "2014-11-10 07:26:10"}, {"_key": "049", "timestamp": "2014-11-10 07:26:11"}, {"_key": "050", "timestamp": "2014-11-10 07:26:12"} ] select Memos \ --filter '_key == "003" && \ between(timestamp, \ "2014-11-10 07:25:24", \ "include", \ "2014-11-10 07:27:26", \ "exclude", \ {"too_many_index_match_ratio": 0.03})'
table_create Tags TABLE_HASH_KEY ShortText table_create Memos TABLE_HASH_KEY ShortText column_create Memos tag COLUMN_SCALAR Tags load --table Memos [ {"_key": "Rroonga is fast!", "tag": "Rroonga"}, {"_key": "Groonga is fast!", "tag": "Groonga"}, {"_key": "Mroonga is fast!", "tag": "Mroonga"}, {"_key": "Groonga sticker!", "tag": "Groonga"}, {"_key": "Groonga is good!", "tag": "Groonga"} ] column_create Tags memos_tag COLUMN_INDEX Memos tag select \ Memos \ --filter '_id >= 3 && \ in_values(tag, \ "Groonga", \ {"too_many_index_match_ratio": 0.7})' \ --output_columns _id,_score,_key,tag
[between] Added support for
GRN_EXPR_OPTIMIZE=yes.between()supported for optimizing the order of evaluation of a conditional expression.[query_parallel_or][query] Added support for specifying group of match_columns as vector. [GitHub#1238][Patched by naoa]
We can use vector in
match_columnsofqueryandquery_parallel_oras below.table_create Users TABLE_NO_KEY column_create Users name COLUMN_SCALAR ShortText column_create Users memo COLUMN_SCALAR ShortText column_create Users tag COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenNgram \ --normalizer NormalizerNFKC130 column_create Terms name COLUMN_INDEX|WITH_POSITION Users name column_create Terms memo COLUMN_INDEX|WITH_POSITION Users memo column_create Terms tag COLUMN_INDEX|WITH_POSITION Users tag load --table Users [ {"name": "Alice", "memo": "Groonga user", "tag": "Groonga"}, {"name": "Bob", "memo": "Rroonga user", "tag": "Rroonga"} ] select Users \ --output_columns _score,name \ --filter 'query(["name * 100", "memo", "tag * 10"], \ "Alice OR Groonga")'
[select] Added support for section and weight in prefix search. [GitHub#1240][Patched by naoa]
We can use multi column index and adjusting score in prefix search.
table_create Memos TABLE_NO_KEY column_create Memos title COLUMN_SCALAR ShortText column_create Memos tags COLUMN_VECTOR ShortText table_create Terms TABLE_PAT_KEY ShortText column_create Terms index COLUMN_INDEX|WITH_SECTION Memos title,tags load --table Memos [ {"title": "Groonga", "tags": ["Groonga"]}, {"title": "Rroonga", "tags": ["Groonga", "Rroonga", "Ruby"]}, {"title": "Mroonga", "tags": ["Groonga", "Mroonga", "MySQL"]} ] select Memos \ --match_columns "Terms.index.title * 2" \ --query 'G*' \ --output_columns title,tags,_score [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "title", "ShortText" ], [ "tags", "ShortText" ], [ "_score", "Int32" ] ], [ "Groonga", [ "Groonga" ], 2 ] ] ] ]
[grndb] Added support for closing used object immediately in
grndb recover.We can reduce memory usage by this. This may decrease performance but it will be acceptable.
Note that
grndb checkdoesn’t close used objects immediately yet.[query_parallel_or][query] Added support for specifying
scorer_tf_idfinmatch_columnsas below.table_create Tags TABLE_HASH_KEY ShortText table_create Users TABLE_HASH_KEY ShortText column_create Users tags COLUMN_VECTOR Tags load --table Users [ {"_key": "Alice", "tags": ["beginner", "active"]}, {"_key": "Bob", "tags": ["expert", "passive"]}, {"_key": "Chris", "tags": ["beginner", "passive"]} ] column_create Tags users COLUMN_INDEX Users tags select Users \ --output_columns _key,_score \ --sort_keys _id \ --command_version 3 \ --filter 'query_parallel_or("scorer_tf_idf(tags)", \ "beginner active")' { "header": { "return_code": 0, "start_time": 0.0, "elapsed_time": 0.0 }, "body": { "n_hits": 1, "columns": [ { "name": "_key", "type": "ShortText" }, { "name": "_score", "type": "Float" } ], "records": [ [ "Alice", 2.098612308502197 ] ] } }
[query_expand] Added support for weighted increment, decrement, and negative.
We can specify weight against expanded words.
If we want to increment score, we use
>. If we want to decrement score, we use<.We can specify the quantity of scores as a number. We can also use a negative numbers in it.
table_create TermExpansions TABLE_NO_KEY column_create TermExpansions term COLUMN_SCALAR ShortText column_create TermExpansions expansions COLUMN_VECTOR ShortText load --table TermExpansions [ {"term": "Rroonga", "expansions": ["Rroonga", "Ruby Groonga"]} ] query_expand TermExpansions "Groonga <-0.2Rroonga Mroonga" \ --term_column term \ --expanded_term_column expansions [[0,0.0,0.0],"Groonga <-0.2((Rroonga) OR (Ruby Groonga)) Mroonga"]
[httpd] Updated bundled nginx to 1.21.1.
Updated bundled Apache Arrow to 5.0.0.
[Ubuntu] Dropped Ubuntu 20.10 (Groovy Gorilla) support.
Because Ubuntu 20.10 reached EOL July 22, 2021.
Fixes#
[query_parallel_or][query] Fixed a bug that if we specify
query_optionsand the other options, the other options are ignored.For example,
"default_operator": "OR"option had been ignored in the following case.plugin_register token_filters/stop_word table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto \ --token_filters TokenFilterStopWord column_create Terms memos_content COLUMN_INDEX|WITH_POSITION Memos content column_create Terms is_stop_word COLUMN_SCALAR Bool load --table Terms [ {"_key": "and", "is_stop_word": true} ] load --table Memos [ {"content": "Hello"}, {"content": "Hello and Good-bye"}, {"content": "and"}, {"content": "Good-bye"} ] select Memos \ --filter 'query_parallel_or( \ "content", \ "Hello and", \ {"default_operator": "OR", \ "options": {"TokenFilterStopWord.enable": false}})' \ --match_escalation_threshold -1 \ --sort_keys -_score [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "content", "ShortText" ] ], [ 2, "Hello and Good-bye" ] ] ] ]
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
If we repeat that we remove any data and load them again, Groonga may not return records that should match.
Thanks#
naoa
Release 11.0.4 - 2021-06-29#
Improvements#
[Normalizer] Added support for customized normalizer.
We define a table for normalize to use this feature. We can normalize with use that table. In other words, we can use customized normalizer.
For example, we define that “S” normalize to “5” in the following example. The
Substitutionstable is for nromalize.table_create Substitutions TABLE_PAT_KEY ShortText column_create Substitutions substituted COLUMN_SCALAR ShortText load --table Substitutions [ {"_key": "S", "substituted": "5"} ] table_create TelLists TABLE_NO_KEY column_create TelLists tel COLUMN_SCALAR ShortText table_create Terms TABLE_HASH_KEY ShortText \ --default_tokenizer TokenNgram \ --normalizer 'NormalizerTable("column", "Substitutions.substituted", \ "report_source_offset", true)' column_create Terms tel_index COLUMN_INDEX|WITH_POSITION TelLists tel load --table TelLists [ {"tel": "03-4S-1234"} ] select TelLists --filter 'tel @ "03-45-1234"' [ [ 0, 1624686303.538532, 0.001319169998168945 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "tel", "ShortText" ] ], [ 1, "03-4S-1234" ] ] ] ]
For example, we can define to the table easy to false recognize words when we input a handwritten data. By this, we can normalize incorrect data to correct data.
Note that we need to reconstruct the index if we updated the table for normalize.
Added a new command
object_warm.This commnad ship Groonga’s DB to OS’s page cache.
If we never startup Groonga after OS startup, Groonga’s DB doesn’t exist on OS’s page cache When Groonga on the first run. Therefore, the first operation to Groonga is slow.
If we execute this command in advance, the first operation to Groonga is fast. In Linux, we can do the same by also executing
cat *.db > dev/null. However, we could not do the same thing in Windows until now.By using this command, we can ship Groonga’s DB to OS’s page cache in both Linux and Windows. Then, we can also do that in units of table, column, and index. Therefore, we can ship only table, column, and index that we often use to OS’s page cache.
We can execute this command against various targets as below.
If we specify
object_warm --name index_name, the index is shipped to OS’s page cache.If we specify
object_warm --name column_name, the column is shipped to OS’s page cache.If we specify
object_warm --name table_nameis shipped to OS’s page cache.If we specify
object_warm, whole Groonga’s database is shipped to OS’s page cache.
However, note that if OS has not empty space on memory, this command has no effect.
[select] Added support for adjusting the score of a specific record in
--filter.We can adjust the score of a specific record by using a oprtator named
*~.*~is logical operator same as&&and||. Therefore, we can use*~like as&&ans||. Default weight of*~is -1.Therefore, for example,
'content @ "Groonga" *~ content @ "Mroonga"'mean the following operations.Extract records that match
'content @ "Groonga"andcontent @ "Mroonga"'.Add a score as below.
Calculate the score of
'content @ "Groonga".Calculate the score of
'content @ "Mroonga"'.b’s score multiplied by -1 by
*~.The socre of this record is a + b Therefore, if a’s socre is 1 and b’s score is 1, the score of this record is 1 + (1 * -1) = 0.
Then, we can specify score quantity by
*~${score_quantity}.In particular, the following query adjust the score of match records by the following condition(
'content @ "Groonga" *~2.5 content @ "Mroonga")').table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto column_create Terms index COLUMN_INDEX|WITH_POSITION Memos content load --table Memos [ {"content": "Groonga is a full text search engine."}, {"content": "Rroonga is the Ruby bindings of Groonga."}, {"content": "Mroonga is a MySQL storage engine based of Groonga."} ] select Memos \ --command_version 3 \ --filter 'content @ "Groonga" *~2.5 content @ "Mroonga"' \ --output_columns 'content, _score' \ --sort_keys -_score,_id { "header": { "return_code": 0, "start_time": 1624605205.641078, "elapsed_time": 0.002965450286865234 }, "body": { "n_hits": 3, "columns": [ { "name": "content", "type": "ShortText" }, { "name": "_score", "type": "Float" } ], "records": [ [ "Groonga is a full text search engine.", 1.0 ], [ "Rroonga is the Ruby bindings of Groonga.", 1.0 ], [ "Mroonga is a MySQL storage engine based of Groonga.", -1.5 ] ] } }
We can do the same by also useing adjuster . If we use adjuster , we need to make
--filtercondition and--adjusterconditon on our application, but we make only--filtercondition on it by this improvement.We can also describe filter condition as below by using
query().--filter 'content @ "Groonga" *~2.5 content @ "Mroonga"'
[select] Added support for
&&with weight.We can use
&&with weight by using*<or*>. Default weight of*<is 0.5. Default weight of*>is 2.0.We can specify score quantity by
*<${score_quantity}and*>${score_quantity}. Then, if we specify*<${score_quantity}, a plus or minus sign of${score_quantity}is reverse.For example,
'content @ "Groonga" *<2.5 query("content", "MySQL")'is as below.Extract records that match
'content @ "Groonga"andcontent @ "Mroonga"'.Add a score as below.
Calculate the score of
'content @ "Groonga".Calculate the score of
query("content", "MySQL").b’s score multiplied by -2.5 by
*<.The socre of this record is a + b Therefore, if a’s socre is 1 and b’s score is 1, the score of this record is 1 + (1 * -2.5) = -1.5.
In particular, the following query adjust the score of match records by the following condition(
'content @ "Groonga" *<2.5 query("content", "Mroonga")').table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto column_create Terms index COLUMN_INDEX|WITH_POSITION Memos content load --table Memos [ {"content": "Groonga is a full text search engine."}, {"content": "Rroonga is the Ruby bindings of Groonga."}, {"content": "Mroonga is a MySQL storage engine based of Groonga."} ] select Memos \ --command_version 3 \ --filter 'content @ "Groonga" *<2.5 query("content", "Mroonga")' \ --output_columns 'content, _score' \ --sort_keys -_score,_id { "header": { "return_code": 0, "start_time": 1624605205.641078, "elapsed_time": 0.002965450286865234 }, "body": { "n_hits": 3, "columns": [ { "name": "content", "type": "ShortText" }, { "name": "_score", "type": "Float" } ], "records": [ [ "Groonga is a full text search engine.", 1.0 ], [ "Rroonga is the Ruby bindings of Groonga.", 1.0 ], [ "Mroonga is a MySQL storage engine based of Groonga.", -1.5 ] ] } }
[Log] Added support for outputting to stdout and stderr.
[Process log] and [Query log] supported output to stdout and stderr.
If we specify as
--log-path -,--query-log-path -, Groonga output log to stdout.If we specify as
--log-path +,--query-log-path +, Groonga output log to stderr.
[Process log] is for all of Groonga works. [Query log] is just for query processing.
This feature is useful when we execute Groonga on Docker. Docker has the feature that records stdout and stderr in standard. Therefore, we don’t need to login into the environment of Docker to get Groonga’s log.
For example, this feature is useful as he following case.
If we want to analyze slow queries of Groonga on Docker.
If we specify
--query-log-path -when startup Groonga, we can analyze slow queries by only execution the following commands.docker logs ${container_name} | groonga-query-log-analyze
By this, we can analyze slow query with the query log that output from Groonga on Docker simply.
[Documentation] Filled missing documentation of
string_substring. [GitHub#1209][Patched by Takashi Hashida]
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
*<and*>only valid when we usequery()the right side of filter condition. If we specify as below,*<and*>work as&&.'content @ "Groonga" *< content @ "Mroonga"'
Thanks#
Takashi Hashida
Release 11.0.3 - 2021-05-29#
Improvements#
[query] Added support for ignoring
TokenFilterStemby the query.TokenFilterStemcan search by using a stem. For example, all ofdevelop,developing,developedanddevelopstokens are stemmed asdevelop. So we can finddevelop,developinganddevelopedbydevelopsquery.In this release, we are able to search without
TokenFilterStemin only a specific query as below.plugin_register token_filters/stem table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto \ --token_filters 'TokenFilterStem("keep_original", true)' column_create Terms memos_content COLUMN_INDEX|WITH_POSITION Memos content load --table Memos [ {"content": "I develop Groonga"}, {"content": "I'm developing Groonga"}, {"content": "I developed Groonga"} ] select Memos \ --match_columns content \ --query '"developed groonga"' \ --query_options '{"TokenFilterStem.enable": false}' [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "content", "ShortText" ] ], [ 3, "I developed Groonga" ] ] ] ]
This feature is useful when users want to search by a stemmed word generally but users sometimes want to search by a exact (not stemmed) word as below.
If Groonga returns many results when searching by a stemmed word.
If
TokenFilterStemreturns the wrong result of stemming.If we want to find only records that have an exact (not stemmed) word.
[query] Added support for ignoring
TokenFilterStopWordby the query.TokenFilterStopWordsearched without stop word that we registered beforehand. It uses for reducing noise of search by ignoring frequently word (e.g.,and,is, and so on.).However, we sometimes want to search include these words only a specific query. In this release, we are able to search without
TokenFilterStopWordin only a specific query as below.plugin_register token_filters/stop_word table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto \ --token_filters TokenFilterStopWord column_create Terms memos_content COLUMN_INDEX|WITH_POSITION Memos content column_create Terms is_stop_word COLUMN_SCALAR Bool load --table Terms [ {"_key": "and", "is_stop_word": true} ] load --table Memos [ {"content": "Hello"}, {"content": "Hello and Good-bye"}, {"content": "Good-bye"} ] select Memos \ --match_columns content \ --query "Hello and" \ --query_options '{"TokenFilterStopWord.enable": false}' \ --match_escalation_threshold -1 \ --sort_keys -_score [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "content", "ShortText" ] ], [ 2, "Hello and Good-bye" ] ] ] ]
In the above example, we specify
TokenFilterStopWord.enableby using--query-options, but we also specify it by using{"options": {"TokenFilterStopWord.enable": false}}as below.plugin_register token_filters/stop_word table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerAuto \ --token_filters TokenFilterStopWord column_create Terms memos_content COLUMN_INDEX|WITH_POSITION Memos content column_create Terms is_stop_word COLUMN_SCALAR Bool load --table Terms [ {"_key": "and", "is_stop_word": true} ] load --table Memos [ {"content": "Hello"}, {"content": "Hello and Good-bye"}, {"content": "Good-bye"} ] select Memos \ --filter 'query("content", \ "Hello and", \ {"options": {"TokenFilterStopWord.enable": false}})' \ --match_escalation_threshold -1 \ --sort_keys -_score [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_id", "UInt32" ], [ "content", "ShortText" ] ], [ 2, "Hello and Good-bye" ] ] ] ]
This feature is useful if that Groonga can’t return results correctly if we don’t search by keywords include commonly used words (e.g., if a search for a song title, a shop name, and so on.).
[Normalizers][NormalizerNFKC] Added a new option
remove_new_line.If we want to normalize the key of a table that stores data, we set a normalizer to it. However, normally, normalizers remove a new line.
Groonga can’t handle a key that is only a new line.
We can register data that is only a new line as key by this option.
[string_slice] Added a new function
string_slice(). [Github#1177][Patched by Takashi Hashida]string_slice()extracts a substring of a string.To enable this function, we need to register
functions/stringplugin.We can use two different extraction methods depending on the arguments as below.
Extraction by position:
plugin_register functions/string table_create Memos TABLE_HASH_KEY ShortText load --table Memos [ {"_key": "Groonga"} ] select Memos --output_columns '_key, string_slice(_key, 2, 3)' [ [ 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 1 ], [ [ "_key", "ShortText" ], [ "string_slice", null ] ], [ "Groonga", "oon" ] ] ] ]
Extraction by regular expression:
plugin_register functions/string table_create Memos TABLE_HASH_KEY ShortText load --table Memos [ {"_key": "Groonga"} ] select Memos --output_columns '_key, string_slice(_key, "(Gro+)(.*)", 2)' [ [p 0, 1337566253.89858, 0.000355720520019531 ], [ [ [ 1 ], [ [ "_key", "ShortText" ], [ "string_slice", null ] ], [ "Groonga", "nga" ] ] ] ]
[Ubuntu] Dropped support for Ubuntu 16.04 LTS (Xenial Xerus).
Added EditorConfig for Visual Studio. [GitHub#1191][Patched by Takashi Hashida]
Most settings are for Visual Studio only.
[httpd] Updated bundled nginx to 1.20.1.
Contains security fix of CVE-2021-23017.
Fixes#
Fixed a bug that Groonga may not have returned a result of a search query if we sent many search queries when tokenizer, normalizer, or token_filters that support options were used.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list.
Thanks#
Takashi Hashida
Release 11.0.2 - 2021-05-10#
Improvements#
[Documentation] Removed a reference about
ruby_loadcommand. [GitHub#1172][Patched by Anthony M. Cook]Because this command has already deleted.
[Debian GNU/Linux] Added support for Debian 11(Bullseye).
[select] Added support for
--post_filter.We can use
post_filterto filter byfilteredstage dynamic columns as below.table_create Items TABLE_NO_KEY column_create Items price COLUMN_SCALAR UInt32 load --table Items [ {"price": 100}, {"price": 150}, {"price": 200}, {"price": 250}, {"price": 300} ] select Items \ --filter "price >= 150" \ --columns[price_with_tax].stage filtered \ --columns[price_with_tax].type UInt32 \ --columns[price_with_tax].flags COLUMN_SCALAR \ --columns[price_with_tax].value "price * 1.1" \ --post_filter "price_with_tax <= 250" [ [ 0, 0.0, 0.0 ], [ [ [ 2 ], [ [ "_id", "UInt32" ], [ "price_with_tax", "UInt32" ], [ "price", "UInt32" ] ], [ 2, 165, 150 ], [ 3, 220, 200 ] ] ] ]
[select] Added support for
--slices[].post_filter.We can use
post_filterto filter by--slices[].filteras below.table_create Items TABLE_NO_KEY column_create Items price COLUMN_SCALAR UInt32 load --table Items [ {"price": 100}, {"price": 200}, {"price": 300}, {"price": 1000}, {"price": 2000}, {"price": 3000} ] select Items \ --slices[expensive].filter 'price >= 1000' \ --slices[expensive].post_filter 'price < 3000' [ [ 0, 0.0, 0.0 ], [ [ [ 6 ], [ [ "_id", "UInt32" ], [ "price", "UInt32" ] ], [ 1, 100 ], [ 2, 200 ], [ 3, 300 ], [ 4, 1000 ], [ 5, 2000 ], [ 6, 3000 ] ], { "expensive": [ [ 2 ], [ [ "_id", "UInt32" ], [ "price", "UInt32" ] ], [ 4, 1000 ], [ 5, 2000 ] ] } ] ]
[select] Added support for describing expression into
--sort_keys.We can describe the expression into
--sort_keys.If nonexistent keys into expression as a
--sort_keys, they are ignored and outputted warns into a log.
By this, for example, we can specify a value of an element of
VECTOR COLUMNto--sort_keys. And we can sort a result with it.We can sort a result with an element of
VECTOR COLUMNeven if the before version by using dynamic column. However, we can sort a result with an element ofVECTOR COLUMNwithout using dynamic column by this feature.table_create Values TABLE_NO_KEY column_create Values numbers COLUMN_VECTOR Int32 load --table Values [ {"numbers": [127, 128, 129]}, {"numbers": [126, 255]}, {"numbers": [128, -254]} ] select Values --sort_keys 'numbers[1]' --output_columns numbers [ [ 0, 0.0, 0.0 ], [ [ [ 3 ], [ [ "numbers", "Int32" ] ], [ [ 128, -254 ] ], [ [ 127, 128, 129 ] ], [ [ 126, 255 ] ] ] ] ]
[Token filters] Added support for multiple token filters with options.
We can specify multiple token filters with options like
--token_filters 'TokenFilterStopWord("column", "ignore"), TokenFilterNFKC130("unify_kana", true)'. [Github#mroonga/mroonga#399][Reported by MASUDA Kazuhiro]
[query] Added support a dynamic column of
result_setstage with complex expression.Complex expression is that it needs temporary result sets internally like a following expression.
'(true && query("name * 10", "ali", {"score_column": "ali_score"})) || \ (true && query("name * 2", "li", {"score_column": "li_score"}))'
In the above expressions, the temporary result sets are used to store the result of evaluating the
true.Therefore, for example, in the following expression, we can use a value of dynamic column of
result_setstage in expression. Because temporary result sets internally are needless as below expression.'(query("name * 10", "ali", {"score_column": "ali_score"})) || \ (query("name * 2", "li", {"score_column": "li_score"}))'
In this release, for example, we can set a value to
li_scoreas below. (The value ofli_scorehad been0in before version. Because the second expression could not get dynamic column.)table_create Users TABLE_NO_KEY column_create Users name COLUMN_SCALAR ShortText table_create Lexicon TABLE_HASH_KEY ShortText \ --default_tokenizer TokenBigramSplitSymbolAlphaDigit \ --normalizer NormalizerAuto column_create Lexicon users_name COLUMN_INDEX|WITH_POSITION Users name load --table Users [ {"name": "Alice"}, {"name": "Alisa"}, {"name": "Bob"} ] select Users \ --columns[ali_score].stage result_set \ --columns[ali_score].type Float \ --columns[ali_score].flags COLUMN_SCALAR \ --columns[li_score].stage result_set \ --columns[li_score].type Float \ --columns[li_score].flags COLUMN_SCALAR \ --output_columns name,_score,ali_score,li_score \ --filter '(true && query("name * 10", "ali", {"score_column": "ali_score"})) || \ (true && query("name * 2", "li", {"score_column": "li_score"}))' [ [ 0, 0.0, 0.0 ], [ [ [ 2 ], [ [ "name", "ShortText" ], [ "_score", "Int32" ], [ "ali_score", "Float" ], [ "li_score", "Float" ] ], [ "Alice", 14, 10.0, 2.0 ], [ "Alisa", 14, 10.0, 2.0 ] ] ] ]
We also supported a dynamic vector column of
result_setstage as below.table_create Users TABLE_NO_KEY column_create Users name COLUMN_SCALAR ShortText table_create Lexicon TABLE_HASH_KEY ShortText \ --default_tokenizer TokenBigramSplitSymbolAlphaDigit \ --normalizer NormalizerAuto column_create Lexicon users_name COLUMN_INDEX|WITH_POSITION Users name load --table Users [ {"name": "Alice"}, {"name": "Alisa"}, {"name": "Bob"} ] select Users \ --columns[tags].stage result_set \ --columns[tags].type ShortText \ --columns[tags].flags COLUMN_VECTOR \ --output_columns name,tags \ --filter '(true && query("name", "al", {"tags": ["al"], "tags_column": "tags"})) || \ (true && query("name", "sa", {"tags": ["sa"], "tags_column": "tags"}))' [ [ 0, 0.0, 0.0 ], [ [ [ 2 ], [ [ "name", "ShortText" ], [ "tags", "ShortText" ] ], [ "Alice", [ "al" ] ], [ "Alisa", [ "al", "sa" ] ] ] ] ]
If we use a dynamic vector column, the storing values are appended values of each element.
[Ubuntu] Added support for Ubuntu 21.04 (Hirsute Hippo).
[httpd] Updated bundled nginx to 1.19.10.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
[The browser based administration tool] Currently, Groonga has a bug that a search query that is inputted to non-administration mode is sent even if we input checks to the checkbox for the administration mode of a record list. [Github#1186][Reported by poti]
Thanks#
Anthony M. Cook
MASUDA Kazuhiro
poti
Release 11.0.1 - 2021-03-31#
Improvements#
[Debian GNU/Linux] Added support for a ARM64 package.
[select] Added support for customizing adjust weight every key word.
We need to specify
<or>to all keywords to adjust scores until now. Because the default adjustment of weight (6 or 4) is larger than the default score (1).Therefore, for example, “A“‘s weight is 1 and “B“‘s weight is 4 in
A <B. Decremented “B“‘s weight (4) is larger than not decremented “A“‘s weight (1). This is not works as expected. we need to specify>A <Bto use smaller weight than “A” for “B”. “A“‘s weight is 6 and “B“‘s weight is 4 in>A <B.
We can customize adjustment of weight every key word by only specifying
<${WEIGHT}or>${WEIGHT}to target keywords since this release. For example, “A“‘s weight is 1 and “B“‘s weight is 0.9 inA <0.1B(“B“‘s weight decrement 0.1).However, note that these forms (
>${WEIGHT}...,<${WEIGHT}..., and~${WEIGHT}...) are incompatible.
[select] Added support for outputting
FloatandFloat32value in Apache Arrow format.For example, Groonga output as below.
table_create Data TABLE_NO_KEY column_create Data float COLUMN_SCALAR Float load --table Data [ {"float": 1.1} ] select Data \ --command_version 3 \ --output_type apache-arrow return_code: int32 start_time: timestamp[ns] elapsed_time: double -- metadata -- GROONGA:data_type: metadata return_code start_time elapsed_time 0 0 1970-01-01T09:00:00+09:00 0.000000 ======================================== _id: uint32 float: double -- metadata -- GROONGA:n_hits: 1 _id float 0 1 1.100000
[select] Added support for getting a reference destination data via index column when we output a result.
Until now, Groonga had returned involuntary value when we specified output value like
index_column.xxx. For example, A value of--columns[tags].value purchases.tagwas["apple",["many"]],["banana",["man"]],["cacao",["man"]]in the following example. In this case, the expected values was["apple",["man","many"]],["banana",["man"]],["cacao",["woman"]]. In this release, we can get a correct reference destination data via index column as below.table_create Products TABLE_PAT_KEY ShortText table_create Purchases TABLE_NO_KEY column_create Purchases product COLUMN_SCALAR Products column_create Purchases tag COLUMN_SCALAR ShortText column_create Products purchases COLUMN_INDEX Purchases product load --table Products [ {"_key": "apple"}, {"_key": "banana"}, {"_key": "cacao"} ] load --table Purchases [ {"product": "apple", "tag": "man"}, {"product": "banana", "tag": "man"}, {"product": "cacao", "tag": "woman"}, {"product": "apple", "tag": "many"} ] select Products \ --columns[tags].stage output \ --columns[tags].flags COLUMN_VECTOR \ --columns[tags].type ShortText \ --columns[tags].value purchases.tag \ --output_columns _key,tags [ [ 0, 0.0, 0.0 ], [ [ [ 3 ], [ [ "_key", "ShortText" ], [ "tags", "ShortText" ] ], [ "apple", [ "man", "many" ] ], [ "banana", [ "man" ] ], [ "cacao", [ "woman" ] ] ] ] ]
[select] Added support for specifying index column directly as a part of nested index.
We can search source table after filtering by using
index_column.except_source_column. For example, we specifycomments.contentwhen searching in the following example. In this case, at first, this query execute full text search fromcontentcolumn ofCommenttstable, then fetch the records of Articles table which refers to already searched records of Comments table.table_create Articles TABLE_HASH_KEY ShortText table_create Comments TABLE_NO_KEY column_create Comments article COLUMN_SCALAR Articles column_create Comments content COLUMN_SCALAR ShortText column_create Articles content COLUMN_SCALAR Text column_create Articles comments COLUMN_INDEX Comments article table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenBigram \ --normalizer NormalizerNFKC130 column_create Terms articles_content COLUMN_INDEX|WITH_POSITION \ Articles content column_create Terms comments_content COLUMN_INDEX|WITH_POSITION \ Comments content load --table Articles [ {"_key": "article-1", "content": "Groonga is fast!"}, {"_key": "article-2", "content": "Groonga is useful!"}, {"_key": "article-3", "content": "Mroonga is fast!"} ] load --table Comments [ {"article": "article-1", "content": "I'm using Groonga too!"}, {"article": "article-3", "content": "I'm using Mroonga!"}, {"article": "article-1", "content": "I'm using PGroonga!"} ] select Articles --match_columns comments.content --query groonga \ --output_columns "_key, _score, comments.content [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "_key", "ShortText" ], [ "_score", "Int32" ], [ "comments.content", "ShortText" ] ], [ "article-1", 1, [ "I'm using Groonga too!", "I'm using PGroonga!" ] ] ] ] ]
[load] Added support for loading reference vector with inline object literal.
For example, we can load data like
"key" : "[ { "key" : "value", ..., "key" : "value" } ]"as below.table_create Purchases TABLE_NO_KEY column_create Purchases item COLUMN_SCALAR ShortText column_create Purchases price COLUMN_SCALAR UInt32 table_create Settlements TABLE_HASH_KEY ShortText column_create Settlements purchases COLUMN_VECTOR Purchases column_create Purchases settlements_purchases COLUMN_INDEX Settlements purchases load --table Settlements [ { "_key": "super market", "purchases": [ {"item": "apple", "price": 100}, {"item": "milk", "price": 200} ] }, { "_key": "shoes shop", "purchases": [ {"item": "sneakers", "price": 3000} ] } ]
It makes easier to add JSON data into reference columns by this feature.
Currently, this feature only support with JSON input.
[load] Added support for loading reference vector from JSON text.
We can load data to reference vector from source table with JSON text as below.
table_create Purchases TABLE_HASH_KEY ShortText column_create Purchases item COLUMN_SCALAR ShortText column_create Purchases price COLUMN_SCALAR UInt32 table_create Settlements TABLE_HASH_KEY ShortText column_create Settlements purchases COLUMN_VECTOR Purchases column_create Purchases settlements_purchases COLUMN_INDEX Settlements purchases load --table Settlements [ { "_key": "super market", "purchases": "[{\"_key\": \"super market-1\", \"item\": \"apple\", \"price\": 100}, {\"_key\": \"super market-2\", \"item\": \"milk\", \"price\": 200}]" }, { "_key": "shoes shop", "purchases": "[{\"_key\": \"shoes shop-1\", \"item\": \"sneakers\", \"price\": 3000}]" } ] dump \ --dump_plugins no \ --dump_schema no load --table Purchases [ ["_key","item","price"], ["super market-1","apple",100], ["super market-2","milk",200], ["shoes shop-1","sneakers",3000] ] load --table Settlements [ ["_key","purchases"], ["super market",["super market-1","super market-2"]], ["shoes shop",["shoes shop-1"]] ] column_create Purchases settlements_purchases COLUMN_INDEX Settlements purchases
Currently, this feature doesn’t support nested reference record.
[Windows] Added support for UNIX epoch for
time_classify_*functions.Groonga handles timestamps on local time. Therefore, for example, if we input the UNIX epoch in Japan, inputting time is 9 hours behind the UNIX epoch.
The Windows API outputs an error when we input the time before the UNIX epoch.
We can use the UNIX epoch in
time_classify_*functions as below in this release.plugin_register functions/time table_create Timestamps TABLE_PAT_KEY Time load --table Timestamps [ {"_key": 0}, {"_key": "2016-05-06 00:00:00.000001"}, {"_key": "2016-05-06 23:59:59.999999"}, {"_key": "2016-05-07 00:00:00.000000"}, {"_key": "2016-05-07 00:00:00.000001"}, {"_key": "2016-05-08 23:59:59.999999"}, {"_key": "2016-05-08 00:00:00.000000"} ] select Timestamps \ --sortby _id \ --limit -1 \ --output_columns '_key, time_classify_day_of_week(_key)' [ [ 0, 0.0, 0.0 ], [ [ [ 7 ], [ [ "_key", "Time" ], [ "time_classify_day_of_week", null ] ], [ 0.0, 4 ], [ 1462460400.000001, 5 ], [ 1462546799.999999, 5 ], [ 1462546800.0, 6 ], [ 1462546800.000001, 6 ], [ 1462719599.999999, 0 ], [ 1462633200.0, 0 ] ] ] ]
[query_parallel_or] Added a new function for processing queries in parallel.
query_parallel_orrequires Apache Arrow for processing queries in parallel. If it does not enable,query_parallel_orprocesses queries in sequence.query_parallel_orprocesses combination ofmatch_columnsandquery_stringin parallel.Syntax of
query_parallel_oris as follow:query_parallel_or(match_columns, query_string1, query_string2, . . . query_stringN, {"option": "value", ...})
[select] Added support for ignoring nonexistent sort keys.
Groonga had been outputted error when we specified nonexistent sort keys until now. However, Groonga ignore nonexistent sort keys since this release. (Groonga doesn’t output error.)
This feature implements for consistency. Because we just ignore invalid values in
output_columnsand most of invalid values insort_keys.
[select] Added support for ignoring nonexistent tables in
drilldowns[].table. [GitHub#1169][Reported by naoa]Groonga had been outputted error when we specified nonexistent tables in
drilldowns[].tableuntil now. However, Groonga ignore nonexistent tables indrilldowns[].tablesince this release. (Groonga doesn’t output error.)This feature implements for consistency. Because we just ignore invalid values in
output_columnsand most of invalid values insort_keys.
[httpd] Updated bundled nginx to 1.19.8.
Fixes#
[reference_acquire] Fixed a bug that Groonga crash when a table’s reference is acquired and a column is added to the table before auto release is happened.
Because the added column’s reference isn’t acquired but it’s released on auto release.
[Windows] Fixed a bug that one or more processes fail an output backtrace on SEGV when a new backtrace logging process starts when another backtrace logging process is running in another thread.
Known Issues#
Currently, Groonga has a bug that there is possible that data is corrupt when we execute many additions, delete, and update data to vector column.
Thanks#
naoa
Release 11.0.0 - 2021-02-09#
This is a major version up! But It keeps backward compatibility. We can upgrade to 11.0.0 without rebuilding database.
Improvements#
[select] Added support for outputting values of scalar column and vector column via nested index.
The nested index is that has structure as below.
table_create Products TABLE_PAT_KEY ShortText table_create Purchases TABLE_NO_KEY column_create Purchases product COLUMN_SCALAR Products column_create Purchases tag COLUMN_SCALAR ShortText column_create Products purchases COLUMN_INDEX Purchases product
The
Products.purchasescolumn is a index ofPurchases.productcolumn in the above example. Also,Purchases.productis a reference toProductstable.We had not got the correct search result when we search via nested index until now.
The result had been as follows until now. We can see that
{"product": "apple", "tag": "man"}is not output.table_create Products TABLE_PAT_KEY ShortText table_create Purchases TABLE_NO_KEY column_create Purchases product COLUMN_SCALAR Products column_create Purchases tag COLUMN_SCALAR ShortText column_create Products purchases COLUMN_INDEX Purchases product load --table Products [ {"_key": "apple"}, {"_key": "banana"}, {"_key": "cacao"} ] load --table Purchases [ {"product": "apple", "tag": "man"}, {"product": "banana", "tag": "man"}, {"product": "cacao", "tag": "woman"}, {"product": "apple", "tag": "many"} ] select Products \ --output_columns _key,purchases.tag [ [ 0, 1612504193.380738, 0.0002026557922363281 ], [ [ [ 3 ], [ [ "_key", "ShortText" ], [ "purchases.tag", "ShortText" ] ], [ "apple", "many" ], [ "banana", "man" ], [ "cacao", "man" ] ] ] ]
The result will be as follows from this release. We can see that
{"product": "apple", "tag": "man"}is output.select Products \ --output_columns _key,purchases.tag [ [ 0, 0.0, 0.0 ], [ [ [ 3 ], [ [ "_key", "ShortText" ], [ "purchases.tags", "Tags" ] ], [ "apple", [ [ "man", "one" ], [ "child", "many" ] ] ], [ "banana", [ [ "man", "many" ] ] ], [ "cacao", [ [ "woman" ] ] ] ] ] ]
[Windows] Dropped support for packages of Windows version that we had cross compiled by using MinGW on Linux.
Because there aren’t probably many people who use that.
These above packages are that We had provided as below name until now.
groonga-x.x.x-x86.exegroonga-x.x.x-x86.zipgroonga-x.x.x-x64.exegroonga-x.x.x-x86.zip
From now on, we use the following packages for Windows.
groonga-latest-x86-vs2019-with-vcruntime.zipgroonga-latest-x64-vs2019-with-vcruntime.zip
If a system already has installed Microsoft Visual C++ Runtime Library, we suggest that we use the following packages.
groonga-latest-x86-vs2019.zipgroonga-latest-x64-vs2019.zip
Fixes#
Fixed a bug that there is possible that index is corrupt when Groonga executes many additions, delete, and update information in it.
This bug occurs when we only execute many delete information from index. However, it doesn’t occur when we only execute many additions information into index.
We can repair the index that is corrupt by this bug using reconstruction of it.
This bug doesn’t detect unless we reference the broken index. Therefore, the index in our indexes may has already broken.
We can use [index_column_diff] command to confirm whether the index has already been broken or not.
Release 10.1.1 - 2021-01-25#
Improvements#
[select] Added support for outputting UInt64 value in Apache Arrow format.
[select] Added support for outputting a number of hits in Apache Arrow format as below.
-- metadata -- GROONGA:n-hits: 10
[query] Added support for optimization of “order by estimated size”.
Normally, we can search at high speed when we first execute a condition which number of hits is a little.
The “B (There are few) && A (There are many)” faster than “A (There are many) && B (There are few)”.
This is a known optimization. However we need reorder conditions by ourselves.
Groonga executes this reorder automatically by using optimization of “order by estimated size”.
This optimization is valid by
GRN_ORDER_BY_ESTIMATED_SIZE_ENABLE=yes.
[between] Improved performance by the following improvements.
Improved the accuracy of a decision whether the
between()use sequential search or not.Improved that we set a result of
between()into a result set in bulk.
[select] Improved performance for a prefix search.
For example, the performance of the following prefix search by using “*” improved.
table_create Memos TABLE_PAT_KEY ShortText table_create Contents TABLE_PAT_KEY ShortText --normalizer NormalizerAuto column_create Contents entries_key_index COLUMN_INDEX Memos _key load --table Memos [ {"_key": "(groonga) Start to try!"}, {"_key": "(mroonga) Installed"}, {"_key": "(groonga) Upgraded!"} ] select \ --table Memos \ --match_columns _key \ --query '\\(groonga\\)*'
[Tokenizers][TokenMecab] Improved performance for parallel construction fo token column. [GitHub#1158][Patched by naoa]
Fixes#
[sub_filter] Fixed a bug that
sub_filterdoesn’t work inslices[].filter.For example, the result of
sub_filterwas 0 records by the following query by this bug.table_create Users TABLE_HASH_KEY ShortText column_create Users age COLUMN_SCALAR UInt8 table_create Memos TABLE_NO_KEY column_create Memos user COLUMN_SCALAR Users column_create Memos content COLUMN_SCALAR Text load --table Users [ {"_key": "alice", "age": 9}, {"_key": "bob", "age": 29} ] load --table Memos [ {"user": "alice", "content": "hello"}, {"user": "bob", "content": "world"}, {"user": "alice", "content": "bye"}, {"user": "bob", "content": "yay"} ] select \ --table Memos \ --slices[adult].filter '_id > 1 && sub_filter(user, "age >= 18")'
Fixed a bug that it is possible that we can’t add data and Groonga crash when we repeat much addition of data and deletion of data against a hash table.
Thanks#
naoa
Release 10.1.0 - 2020-12-29#
Improvements#
[highlight_html] Added support for removing leading full width spaces from highlight target. [PGroonga#GitHub#155][Reported by Hung Nguyen V]
Until now, leading full width spaces had also included in the target of highlight as below.
table_create Entries TABLE_NO_KEY column_create Entries body COLUMN_SCALAR ShortText table_create Terms TABLE_PAT_KEY ShortText --default_tokenizer TokenBigram --normalizer NormalizerAuto column_create Terms document_index COLUMN_INDEX|WITH_POSITION Entries body load --table Entries [ {"body": "Groonga 高速！"} ] select Entries --output_columns \ --match_columns body --query '高速' \ --output_columns 'highlight_html(body)' [ [ 0, 0.0, 0.0 ], [ [ [ 1 ], [ [ "highlight_html", null ] ], [ "Groonga<span class=\"keyword\"> 高速</span>！" ] ] ] ]
However, this is needless as the target of highlight. Therefore, in this release,
highlight_html()removes leading full width spaces.
[status] Added a new item
features.We can display which Groonga’s features are enabled as below.
status --output_pretty yes [ [ 0, 0.0, 0.0 ], { "alloc_count": 361, "starttime": 1608087311, "start_time": 1608087311, "uptime": 35, "version": "10.1.0", "n_queries": 0, "cache_hit_rate": 0.0, "command_version": 1, "default_command_version": 1, "max_command_version": 3, "n_jobs": 0, "features": { "nfkc": true, "mecab": true, "message_pack": true, "mruby": true, "onigmo": true, "zlib": true, "lz4": false, "zstandard": false, "kqueue": false, "epoll": true, "poll": false, "rapidjson": false, "apache_arrow": false, "xxhash": false } } ]
[status] Added a new item
apache_arrow.We can display Apache Arrow version that Groonga use as below.
[ [ 0, 1608088628.440753, 0.0006628036499023438 ], { "alloc_count": 360, "starttime": 1608088617, "start_time": 1608088617, "uptime": 11, "version": "10.0.9-39-g5a4c6f3", "n_queries": 0, "cache_hit_rate": 0.0, "command_version": 1, "default_command_version": 1, "max_command_version": 3, "n_jobs": 0, "features": { "nfkc": true, "mecab": true, "message_pack": true, "mruby": true, "onigmo": true, "zlib": true, "lz4": true, "zstandard": false, "kqueue": false, "epoll": true, "poll": false, "rapidjson": false, "apache_arrow": true, "xxhash": false }, "apache_arrow": { "version_major": 2, "version_minor": 0, "version_patch": 0, "version": "2.0.0" } } ]
This item only display when Apache Arrow is valid in Groonga.
[Window function] Added support for processing all tables at once even if target tables straddle a shard. (experimental)
If the target tables straddle a shard, the window function had processed each shard until now.
Therefore, if we used multiple group keys in windows functions, the value of the group keys from the second had to be one kind of value.
However, we can use multiple kind of values for it as below by this improvement.
plugin_register sharding plugin_register functions/time table_create Logs_20170315 TABLE_NO_KEY column_create Logs_20170315 timestamp COLUMN_SCALAR Time column_create Logs_20170315 price COLUMN_SCALAR UInt32 column_create Logs_20170315 item COLUMN_SCALAR ShortText table_create Logs_20170316 TABLE_NO_KEY column_create Logs_20170316 timestamp COLUMN_SCALAR Time column_create Logs_20170316 price COLUMN_SCALAR UInt32 column_create Logs_20170316 item COLUMN_SCALAR ShortText table_create Logs_20170317 TABLE_NO_KEY column_create Logs_20170317 timestamp COLUMN_SCALAR Time column_create Logs_20170317 price COLUMN_SCALAR UInt32 column_create Logs_20170317 item COLUMN_SCALAR ShortText load --table Logs_20170315 [ {"timestamp": "2017/03/15 10:00:00", "price": 1000, "item": "A"}, {"timestamp": "2017/03/15 11:00:00", "price": 900, "item": "A"}, {"timestamp": "2017/03/15 12:00:00", "price": 300, "item": "B"}, {"timestamp": "2017/03/15 13:00:00", "price": 200, "item": "B"} ] load --table Logs_20170316 [ {"timestamp": "2017/03/16 10:00:00", "price": 530, "item": "A"}, {"timestamp": "2017/03/16 11:00:00", "price": 520, "item": "B"}, {"timestamp": "2017/03/16 12:00:00", "price": 110, "item": "A"}, {"timestamp": "2017/03/16 13:00:00", "price": 410, "item": "A"}, {"timestamp": "2017/03/16 14:00:00", "price": 710, "item": "B"} ] load --table Logs_20170317 [ {"timestamp": "2017/03/17 10:00:00", "price": 800, "item": "A"}, {"timestamp": "2017/03/17 11:00:00", "price": 400, "item": "B"}, {"timestamp": "2017/03/17 12:00:00", "price": 500, "item": "B"}, {"timestamp": "2017/03/17 13:00:00", "price": 300, "item": "A"} ] table_create Times TABLE_PAT_KEY Time column_create Times logs_20170315 COLUMN_INDEX Logs_20170315 timestamp column_create Times logs_20170316 COLUMN_INDEX Logs_20170316 timestamp column_create Times logs_20170317 COLUMN_INDEX Logs_20170317 timestamp logical_range_filter Logs \ --shard_key timestamp \ --filter 'price >= 300' \ --limit -1 \ --columns[offsetted_timestamp].stage filtered \ --columns[offsetted_timestamp].type Time \ --columns[offsetted_timestamp].flags COLUMN_SCALAR \ --columns[offsetted_timestamp].value 'timestamp - 39600000000' \ --columns[offsetted_day].stage filtered \ --columns[offsetted_day].type Time \ --columns[offsetted_day].flags COLUMN_SCALAR \ --columns[offsetted_day].value 'time_classify_day(offsetted_timestamp)' \ --columns[n_records_per_day_and_item].stage filtered \ --columns[n_records_per_day_and_item].type UInt32 \ --columns[n_records_per_day_and_item].flags COLUMN_SCALAR \ --columns[n_records_per_day_and_item].value 'window_count()' \ --columns[n_records_per_day_and_item].window.group_keys 'offsetted_day,item' \ --output_columns "_id,time_format_iso8601(offsetted_day),item,n_records_per_day_and_item" [ [ 0, 0.0, 0.0 ], [ [ [ "_id", "UInt32" ], [ "time_format_iso8601", null ], [ "item", "ShortText" ], [ "n_records_per_day_and_item", "UInt32" ] ], [ 1, "2017-03-14T00:00:00.000000+09:00", "A", 1 ], [ 2, "2017-03-15T00:00:00.000000+09:00", "A", 2 ], [ 3, "2017-03-15T00:00:00.000000+09:00", "B", 1 ], [ 1, "2017-03-15T00:00:00.000000+09:00", "A", 2 ], [ 2, "2017-03-16T00:00:00.000000+09:00", "B", 2 ], [ 4, "2017-03-16T00:00:00.000000+09:00", "A", 2 ], [ 5, "2017-03-16T00:00:00.000000+09:00", "B", 2 ], [ 1, "2017-03-16T00:00:00.000000+09:00", "A", 2 ], [ 2, "2017-03-17T00:00:00.000000+09:00", "B", 2 ], [ 3, "2017-03-17T00:00:00.000000+09:00", "B", 2 ], [ 4, "2017-03-17T00:00:00.000000+09:00", "A", 1 ] ] ]
This feature requires Apache Arrow 3.0.0 that is not released yet.
Added support for sequential search against reference column.
This feature is only used if an index search will match many records and the current result set is enough small.
Because the sequential search is faster than the index search in the above case.
It is invalid by default.
It is valid if we set
GRN_II_SELECT_TOO_MANY_INDEX_MATCH_RATIO_REFERENCEenvironment variable.GRN_II_SELECT_TOO_MANY_INDEX_MATCH_RATIO_REFERENCEenvironment variable is a threshold to switch the sequential search from the index search.For example, if we set
GRN_II_SELECT_TOO_MANY_INDEX_MATCH_RATIO_REFERENCE=0.7as below, if the number of records for the result set less than 70 % of total records, a search is executed by a sequential search.$ export GRN_II_SELECT_TOO_MANY_INDEX_MATCH_RATIO_REFERENCE=0.7 table_create Tags TABLE_HASH_KEY ShortText table_create Memos TABLE_HASH_KEY ShortText column_create Memos tag COLUMN_SCALAR Tags load --table Memos [ {"_key": "Rroonga is fast!", "tag": "Rroonga"}, {"_key": "Groonga is fast!", "tag": "Groonga"}, {"_key": "Mroonga is fast!", "tag": "Mroonga"}, {"_key": "Groonga sticker!", "tag": "Groonga"}, {"_key": "Groonga is good!", "tag": "Groonga"} ] column_create Tags memos_tag COLUMN_INDEX Memos tag select Memos --query '_id:>=3 tag:@Groonga' --output_columns _id,_score,_key,tag [ [ 0, 0.0, 0.0 ], [ [ [ 2 ], [ [ "_id", "UInt32" ], [ "_score", "Int32" ], [ "_key", "ShortText" ], [ "tag", "Tags" ] ], [ 4, 2, "Groonga sticker!", "Groonga" ], [ 5, 2, "Groonga is good!", "Groonga" ] ] ] ]
[tokenizers] Added support for the token column into
TokenDocumentVectorTFIDFandTokenDocumentVectorBM25.If there is the token column that has the same source as the index column, these tokenizer use the token id of the token column.
The token column has already had data that has already finished tokenize.
Therefore, these tokenizer are improved performance by using a token column.
For example, we can use this feature by making a token column named
content_tokensas below.table_create Memos TABLE_NO_KEY column_create Memos content COLUMN_SCALAR Text load --table Memos [ {"content": "a b c a"}, {"content": "c b c"}, {"content": "b b a"}, {"content": "a c c"}, {"content": "a"} ] table_create Tokens TABLE_PAT_KEY ShortText \ --normalizer NormalizerNFKC121 \ --default_tokenizer TokenNgram column_create Tokens memos_content COLUMN_INDEX|WITH_POSITION Memos content column_create Memos content_tokens COLUMN_VECTOR Tokens content table_create DocumentVectorBM25 TABLE_HASH_KEY Tokens \ --default_tokenizer \ 'TokenDocumentVectorBM25("index_column", "memos_content", \ "df_column", "df")' column_create DocumentVectorBM25 df COLUMN_SCALAR UInt32 column_create Memos content_feature COLUMN_VECTOR|WITH_WEIGHT|WEIGHT_FLOAT32 \ DocumentVectorBM25 content select Memos [ [ 0, 0.0, 0.0 ], [ [ [ 5 ], [ [ "_id", "UInt32" ], [ "content", "Text" ], [ "content_feature", "DocumentVectorBM25" ], [ "content_tokens", "Tokens" ] ], [ 1, "a b c a", { "a": 0.5095787, "b": 0.6084117, "c": 0.6084117 }, [ "a", "b", "c", "a" ] ], [ 2, "c b c", { "c": 0.8342565, "b": 0.5513765 }, [ "c", "b", "c" ] ], [ 3, "b b a", { "b": 0.9430448, "a": 0.3326656 }, [ "b", "b", "a" ] ], [ 4, "a c c", { "a": 0.3326656, "c": 0.9430448 }, [ "a", "c", "c" ] ], [ 5, "a", { "a": 1.0 }, [ "a" ] ] ] ] ]
TokenDocumentVectorTFIDFandTokenDocumentVectorBM25give a weight against each tokens.TokenDocumentVectorTFIDFcalculate a weight for token by using TF-IDF.Please refer to https://radimrehurek.com/gensim/models/tfidfmodel.html about TF-IDF.
TokenDocumentVectorBM25calculate a weight for token by using Okapi BM25.Please refer to https://en.wikipedia.org/wiki/Okapi_BM25 about Okapi BM25.
Improved performance when below case.
(column @ "value") && (column @ "value")
[Ubuntu] Added support for Ubuntu 20.10 (Groovy Gorilla).
[Debian GNU/Linux] Dropped stretch support.
It reached EOL.
[CentOS] Dropped CentOS 6.
It reached EOL.
[httpd] Updated bundled nginx to 1.19.6.
Fixes#
Fixed a bug that Groonga crash when we use multiple keys drilldown and use multiple accessor as below. [GitHub#1153][Patched by naoa]
table_create Tags TABLE_PAT_KEY ShortText table_create Memos TABLE_HASH_KEY ShortText column_create Memos tags COLUMN_VECTOR Tags column_create Memos year COLUMN_SCALAR Int32 load --table Memos [ {"_key": "Groonga is fast!", "tags": ["full-text-search"], "year": 2019}, {"_key": "Mroonga is fast!", "tags": ["mysql", "full-text-search"], "year": 2019}, {"_key": "Groonga sticker!", "tags": ["full-text-search", "sticker"], "year": 2020}, {"_key": "Rroonga is fast!", "tags": ["full-text-search", "ruby"], "year": 2020}, {"_key": "Groonga is good!", "tags": ["full-text-search"], "year": 2020} ] select Memos \ --filter '_id > 0' \ --drilldowns[tags].keys 'tags,year >= 2020' \ --drilldowns[tags].output_columns _key[0],_key[1],_nsubrecs select Memos \ --filter '_id > 0' \ --drilldowns[tags].keys 'tags,year >= 2020' \ --drilldowns[tags].output_columns _key[1],_nsubrecs
Fixed a bug that the near phrase search did not match when the same phrase occurs multiple times as below.
table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer TokenNgram \ --normalizer NormalizerNFKC121 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "a x a x b x x"}, {"content": "a x x b x"} ] select Entries \ --match_columns content \ --query '*NP2"a b"' \ --output_columns '_score, content'
Thanks#
Hung Nguyen V
naoa
timgates42 [Provided the patch at GitHub#1155]
Release 10.0.9 - 2020-12-01#
Improvements#
Improved performance when we specified
-1tolimit.[reference_acquire] Added a new option
--auto_release_count.Groonga reduces a reference count automatically when a request reaching the number of value that is specified in
--auto_release_count.For example, the acquired reference of
Usersis released automatically after the secondstatusis processed as below.reference_acquire --target_name Users --auto_release_count 2 status # Users is still referred. status # Users' reference is released after this command is processed.
We can prevent a leak of the release of acquired reference by this option.
Modify behavior when Groonga evaluated empty
vectoranduvector.Empty
vectoranduvectorare evaluated tofalsein command version 3.This behavior is valid for only command version 3.
Note that the different behavior until now.
[Normalizers] Added a new Normalizer
NormalizerNFKC130based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 13.0.[Token filters] Added a new TokenFilter
TokenFilterNFKC130based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 13.0.[select] Improved performance for
"_score = column - X".[reference_acquire] Improved that
--reference_acquiredoesn’t get unnecessary reference of index column when we specify the--recursive dependentoption.From this release, the targets of
--recursive dependentare only the target table’s key and the index column that is set to data column.
[select] Add support for ordered near phrase search.
Until now, the near phrase search have only looked for records that the distance of between specified phrases near.
This feature look for satisfy the following conditions records.
If the distance of between specified phrases is near.
If the specified phrases are in line with specified order.
This feature use
*ONPas an operator. (Note that the operator isn’t*NP.)$is handled as itself in the query syntax. Note that it isn’t a special character in the query syntax.If we use script syntax, this feature use as below.
table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer 'TokenNgram("unify_alphabet", false, \ "unify_digit", false)' \ --normalizer NormalizerNFKC121 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "abcXYZdef"}, {"content": "abebcdXYZdef"}, {"content": "abcdef"}, {"content": "defXYZabc"}, {"content": "XYZabc"}, {"content": "abc123456789def"}, {"content": "abc12345678def"}, {"content": "abc1de2def"} ] select Entries --filter 'content *ONP "abc def"' --output_columns '_score, content' [ [ 0, 0.0, 0.0 ], [ [ [ 4 ], [ [ "_score", "Int32" ], [ "content", "Text" ] ], [ 1, "abcXYZdef" ], [ 1, "abcdef" ], [ 1, "abc12345678def" ], [ 1, "abc1de2def" ] ] ] ]
If we use query syntax, this feature use as below.
select Entries --query 'content:*ONP "abc def"' --output_columns '_score, content' [ [ 0, 0.0, 0.0 ], [ [ [ 4 ], [ [ "_score", "Int32" ], [ "content", "Text" ] ], [ 1, "abcXYZdef" ], [ 1, "abcdef" ], [ 1, "abc12345678def" ], [ 1, "abc1de2def" ] ] ] ]
[httpd] Updated bundled nginx to 1.19.5.
Fixes#
[Groonga HTTP server] Fixed that Groonga HTTP server finished without waiting all woker threads finished completely.
Until now, Groonga HTTP server had started the process of finish after worker threads finished itself process. However, worker threads may not finish completely at this timing. Therefore, Groonga HTTP server may crashed according to timing. Because Groonga HTTP server may free area of memory that worker threads are using them yet.
Release 10.0.8 - 2020-10-29#
Improvements#
[select] Added support for large drilldown keys.
The maximum on the key size of Groonga’s tables are 4KiB. However, if we specify multiple keys in drilldown, the size of drilldown keys may be larger than 4KiB.
For example, if the total size for
tagkey andn_likekey is lager than 4KiB in the following case, the drilldown had failed.
select Entries \ --limit -1 \ --output_columns tag,n_likes \ --drilldowns[tag.n_likes].keys tag,n_likes \ --drilldowns[tag.n_likes].output_columns _value.tag,_value.n_likes,_nsubrecs
Because the drilldown packes specifying all keys for drilldown. So, if each the size of drilldown key is large, the size of the packed drilldown keys is larger than 4KiB.
This feature requires xxHash .
However, if we install Groonga from package, we can use this feature without doing anything special. Because Groonga’s package already include xxHash .
[select] Added support for handling as the same dynamic column even if columns refer to different tables.
We can’t have handled the same dynamic column if columns refer to different tables until now. Because the type of columns is different.
However, from this version, we can handle the same dynamic column even if columns refer to different tables by casting to built-in types as below.
table_create Store_A TABLE_HASH_KEY ShortText table_create Store_B TABLE_HASH_KEY ShortText table_create Customers TABLE_HASH_KEY Int32 column_create Customers customers_A COLUMN_VECTOR Store_A column_create Customers customers_B COLUMN_VECTOR Store_B load --table Customers [ {"_key": 1604020694, "customers_A": ["A", "B", "C"]}, {"_key": 1602724694, "customers_B": ["Z", "V", "Y", "T"]}, ] select Customers \ --filter '_key == 1604020694' \ --columns[customers].stage output \ --columns[customers].flags COLUMN_VECTOR \ --columns[customers].type ShortText \ --columns[customers].value 'customers_A' \ --output_columns '_key, customers'
We have needed to set
Store_AorStore_Bin thetypeofcustomerscolumn until now.The type of
customers_Acolumn cast toShortTextin the above example.By this, we can also set the value of
customers_Bin the value ofcustomerscolumn. Because both the key ofcustomers_Aandcustomers_BareShortTexttype.
[select] Improved performance when the number of records for search result are huge.
This optimization works when below cases.
--filter 'column <= "value"'or--filter 'column >= "value"'--filter 'column == "value"'--filter 'between(...)'or--filter 'between(_key, ...)'--filter 'sub_filter(reference_column, ...)'Comparing against
_keysuch as--filter '_key > "value"'.--filter 'geo_in_circle(...)'
Updated bundled LZ4 to 1.9.2 from 1.8.2.
Added support xxHash 0.8
[httpd] Updated bundled nginx to 1.19.4.
Fixes#
Fixed the following bugs related the browser based administration tool. [GitHub#1139][Reported by sutamin]
The problem that Groonga’s logo had not been displayed.
The problem that the throughput chart had not been displayed on the index page for Japanese.
[between] Fixed a bug that
between(_key, ...)is always evaluated by sequential search.
Thanks#
sutamin
Release 10.0.7 - 2020-09-29#
Improvements#
[highlight], [highlight_full] Added support for normalizer options.
[Return code] Added a new return code
GRN_CONNECTION_RESETfor resetting connection.it is returned when an existing connection was forcibly close by the remote host.
Dropped Ubuntu 19.10 (Eoan Ermine).
Because this version has been EOL.
[httpd] Updated bundled nginx to 1.19.2.
[grndb] Added support for detecting duplicate keys.
grndb checkis also able to detect duplicate keys since this release.This check valid except a table of
TABLE_NO_KEY.If the table that was detected duplicate keys by
grndb checkhas only index columns, we can recover bygrndb recover.
[table_create], [column_create] Added a new option
--pathWe can store specified a table or a column to any path using this option.
This option is useful if we want to store a table or a column that we often use to fast storage (e.g. SSD) and store them that we don’t often use to slow storage (e.g. HDD).
We can specify both relative path and absolute path in this option.
If we specify relative path in this option, the path is resolved the path of
groongaprocess as the origin.
However, if we specify
--path, the result ofdumpcommand includes--pathinformations.Therefore, if we specify
--path, we can’t restore to host in different enviroment.Because the directory configuration and the location of
groongaprocess in different by each enviroment.
If we don’t want include
--pathinformations to a dump, we need specify--dump_paths noindumpcommand.
[dump] Added a new option
--dump_paths.--dump_pathsoption control whether--pathis dumped or not.The default value of it is
yes.If we specify
--pathwhen we create tables or columns and we don’t want include--pathinformations to a dump, we specifynointo--dump_pathswhen we executedumpcommand.
[functions] Added a new function
string_toknize().It tokenizes the column value that is specified in the second argument with the tokenizer that is specified in the first argument.
[tokenizers] Added a new tokenizer
TokenDocumentVectorTFIDF(experimental).It generates automatically document vector by TF-IDF.
[tokenizers] Added a new tokenizer
TokenDocumentVectorBM25(experimental).It generates automatically document vector by BM25.
[select] Added support for near search in same sentence.
Fixes#
[load] Fixed a bug that
loaddidn’t a return response when we executed it against 257 columns.This bug may occur from 10.0.4 or later.
This bug only occur when we load data by using
[a, b, c, ...]format.If we load data by using
[{...}], this bug doesn’t occur.
[MessagePack] Fixed a bug that float32 value wasn’t unpacked correctly.
Fixed the following bugs related multi column index.
_scoremay be broken with full text search.The records that couldn’t hit might hit.
Please refer to the following URL for the details about occurrence conditionin of this bug.
Release 10.0.6 - 2020-08-29#
Improvements#
[logical_range_filter] Improved search plan for large data.
Normally,
logical_range_filteris faster thanlogical_select. However, it had been slower thanlogical_selectin the below case.If Groonga can’t get the number of required records easily, it has the feature that switches index search from sequential search. (Normally,
logical_range_filteruses a sequential search when records of search target are many.)The search process for it is almost the same as
logical_selectif the above switching occurred. So,logical_range_filteris severalfold slower thanlogical_selectin the above case if the search target is large data. Becauselogical_range_filterexecutes sort after the search.
If we search for large data, Groonga easily use sequential search than until now since this release.
Therefore,
logical_range_filterwill improve performance. Because the case of the search process almost the same aslogical_selectdecreases.
[httpd] Updated bundled nginx to 1.19.1.
Modify how to install into Debian GNU/Linux.
We modify to use
groonga-apt-sourceinstead ofgroonga-archive-keyring. Because thelintiancommand recommends usingapt-sourceif a package that it puts files under the/etc/apt/sources.lists.d/.The
lintiancommand is the command which checks for many common packaging errors.Please also refer to the following URL for the details about installation procedures.
[logical_select] Added a support for
highlight_htmlandhighlight_full.Added support for recycling the IDs of records that are deleted when an array without value space delete.[GitHub#mroonga/mroonga#327][Reported by gaeeyo]
If an array that doesn’t have value space is deleted, deleted IDs are never recycled.
Groonga had used large storage space by large ID. Because it uses large storage space by itself.
For example, large ID is caused after many adds and deletes like Mroonga’s
mroonga_operations
[select] Improved performance of full-text-search without index.
[Function] Improved performance for calling of function that all arguments a variable reference or literal.
[Indexing] Improved performance of offline index construction by using token column. [GitHub#1126][Patched by naoa]
Improved performance for
"_score = func(...)".The performance when the
_scorevalue calculate by using only function like"_score = func(...)"improved.
Fixes#
Fixed a bug that garbage may be included in response after response send error.
It may occur if a client didn’t read all responses and closed the connection.
Thanks#
gaeeyo
naoa
Release 10.0.5 - 2020-07-30#
Improvements#
[select] Added support for storing reference in table that we specify with
--load_table.--load_tableis a feature that stores search results to the table in a prepared.If the searches are executed multiple times, we can cache the search results by storing them to this table.
We can shorten the search times that the search after the first time by using this table.
We can store a reference to other tables into the key of this table as below since this release.
We can make a smaller size of this table. Because we only store references without store column values.
If we search against this table, we can search by using indexes for reference destination.
table_create Logs TABLE_HASH_KEY ShortText column_create Logs timestamp COLUMN_SCALAR Time table_create Times TABLE_PAT_KEY Time column_create Times logs_timestamp COLUMN_INDEX Logs timestamp table_create LoadedLogs TABLE_HASH_KEY Logs load --table Logs [ { "_key": "2015-02-03:1", "timestamp": "2015-02-03 10:49:00" }, { "_key": "2015-02-03:2", "timestamp": "2015-02-03 12:49:00" }, { "_key": "2015-02-04:1", "timestamp": "2015-02-04 00:00:00" } ] select \ Logs \ --load_table LoadedLogs \ --load_columns "_key" \ --load_values "_key" \ --limit 0 select \ --table LoadedLogs \ --filter 'timestamp >= "2015-02-03 12:49:00"' [ [ 0, 0.0, 0.0 ], [ [ [ 2 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "timestamp", "Time" ] ], [ 2, "2015-02-03:2", 1422935340.0 ], [ 3, "2015-02-04:1", 1422975600.0 ] ] ] ]
[select] Improved sort performance on below cases.
When many sort keys need ID resolution.
For example, the following expression needs ID resolution.
--filter true --sort_keys column
For example, the following expression doesn’t need ID resolution. Because the
_scorepseudo column exists in the result table not the source table.--filter true --sort_keys _score
When a sort target table has a key.
Therefore,
TABLE_NO_KEYisn’t supported this improvement.
[select] Improved performance a bit on below cases.
A case of searching for many records matches.
A case of drilldown for many records.
[aggregator] Added support for score accessor for target. [GitHub#1120][Patched by naoa]
For example, we can
_scoresubject toaggregator_*as below.table_create Items TABLE_HASH_KEY ShortText column_create Items price COLUMN_SCALAR UInt32 column_create Items tag COLUMN_SCALAR ShortText load --table Items [ {"_key": "Book", "price": 1000, "tag": "A"}, {"_key": "Note", "price": 1000, "tag": "B"}, {"_key": "Box", "price": 500, "tag": "B"}, {"_key": "Pen", "price": 500, "tag": "A"}, {"_key": "Food", "price": 500, "tag": "C"}, {"_key": "Drink", "price": 300, "tag": "B"} ] select Items \ --filter true \ --drilldowns[tag].keys tag \ --drilldowns[tag].output_columns _key,_nsubrecs,score_mean \ --drilldowns[tag].columns[score_mean].stage group \ --drilldowns[tag].columns[score_mean].type Float \ --drilldowns[tag].columns[score_mean].flags COLUMN_SCALAR \ --drilldowns[tag].columns[score_mean].value 'aggregator_mean(_score)' [ [ 0, 0.0, 0.0 ], [ [ [ 6 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "price", "UInt32" ], [ "tag", "ShortText" ] ], [ 1, "Book", 1000, "A" ], [ 2, "Note", 1000, "B" ], [ 3, "Box", 500, "B" ], [ 4, "Pen", 500, "A" ], [ 5, "Food", 500, "C" ], [ 6, "Drink", 300, "B" ] ], { "tag": [ [ 3 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ], [ "score_mean", "Float" ] ], [ "A", 2, 1.0 ], [ "B", 3, 1.0 ], [ "C", 1, 1.0 ] ] } ] ]
[Indexing] Improved performance of offline index construction on VC++ version.
[select] Use
nullinsteadNaN,Infinity, and-Infinitywhen Groonga outputs result for JSON format.Because JSON doesn’t support them.
[select] Add support fot aggregating standard deviation value.
For example, we can calculate a standard deviation for every group as below.
table_create Items TABLE_HASH_KEY ShortText column_create Items price COLUMN_SCALAR UInt32 column_create Items tag COLUMN_SCALAR ShortText load --table Items [ {"_key": "Book", "price": 1000, "tag": "A"}, {"_key": "Note", "price": 1000, "tag": "B"}, {"_key": "Box", "price": 500, "tag": "B"}, {"_key": "Pen", "price": 500, "tag": "A"}, {"_key": "Food", "price": 500, "tag": "C"}, {"_key": "Drink", "price": 300, "tag": "B"} ] select Items \ --drilldowns[tag].keys tag \ --drilldowns[tag].output_columns _key,_nsubrecs,price_sd \ --drilldowns[tag].columns[price_sd].stage group \ --drilldowns[tag].columns[price_sd].type Float \ --drilldowns[tag].columns[price_sd].flags COLUMN_SCALAR \ --drilldowns[tag].columns[price_sd].value 'aggregator_sd(price)' \ --output_pretty yes [ [ 0, 1594339851.924836, 0.002813816070556641 ], [ [ [ 6 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "price", "UInt32" ], [ "tag", "ShortText" ] ], [ 1, "Book", 1000, "A" ], [ 2, "Note", 1000, "B" ], [ 3, "Box", 500, "B" ], [ 4, "Pen", 500, "A" ], [ 5, "Food", 500, "C" ], [ 6, "Drink", 300, "B" ] ], { "tag": [ [ 3 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ], [ "price_sd", "Float" ] ], [ "A", 2, 250.0 ], [ "B", 3, 294.3920288775949 ], [ "C", 1, 0.0 ] ] } ] ]
We can also calculate sample standard deviation to specifing
aggregate_sd(target, {"unbiased": true}).
[Windows] Dropped Visual Studio 2013 support.
Fixes#
[Groonga HTTP server] Fixed a bug that a request can’t halt even if we execute
shutdown?mode=immediatewhen the response was halted by error occurrence.Fixed a crash bug when an error occurs while a request.
It only occurs when we use Apache Arrow Format.
Groonga crashes when we send request to Groonga again after the previous request was halted by error occurrence.
[between] Fixed a crash bug when temporary table is used.
For example, if we specify a dynamic column in the first argument for
between, Groonga had crashed.
Fixed a bug that procedure created by plugin is freed unexpectedly.
It’s only occurred in reference count mode.
It’s not occurred if we don’t use
plugin_register.It’s not occurred in the process that executes
plugin_register.It’s occurred in the process that doesn’t execute
plugin_register.
Fixed a bug that normalization error occurred while static index construction by
token_column. [GitHub#1122][Reported by naoa]
Thanks#
naoa
Release 10.0.4 - 2020-06-29#
Improvements#
[Tables] Added support for registering 400M records into a hash table.
[select] Improve scorer performance when the
_scoredoesn’t get recursively values.Groonga get recursively value of
_scorewhen search result is search target.For example, the search targets of
slicesare search result. Therefore, if we useslicein a query, this improvement doesn’t ineffective.
[Log] Improved that we output drilldown keys in query-log.
[reference_acquire], [reference_release] Added new commands for reference count mode.
If we need to call multiple
loadin a short time, auto close by the reference count mode will degrade performance.We can avoid the performance degrading by calling
reference_acquirebefore multipleloadand callingreference_releaseafter multipleload. Betweenreference_acquireandreference_release, auto close is disabled.Because
reference_acquireacquires a reference of target objects.
We can must call
reference_releaseafter you finish performance impact operations.If we don’t call
reference_release, the reference count mode doesn’t work.
[select] Added support for aggregating multiple groups on one time
drilldown.We came to be able to calculate sum or arithmetic mean every different multiple groups on one time
drilldownas below.table_create Items TABLE_HASH_KEY ShortText column_create Items price COLUMN_SCALAR UInt32 column_create Items quantity COLUMN_SCALAR UInt32 column_create Items tag COLUMN_SCALAR ShortText load --table Items [ {"_key": "Book", "price": 1000, "quantity": 100, "tag": "A"}, {"_key": "Note", "price": 1000, "quantity": 10, "tag": "B"}, {"_key": "Box", "price": 500, "quantity": 15, "tag": "B"}, {"_key": "Pen", "price": 500, "quantity": 12, "tag": "A"}, {"_key": "Food", "price": 500, "quantity": 111, "tag": "C"}, {"_key": "Drink", "price": 300, "quantity": 22, "tag": "B"} ] select Items \ --drilldowns[tag].keys tag \ --drilldowns[tag].output_columns _key,_nsubrecs,price_sum,quantity_sum \ --drilldowns[tag].columns[price_sum].stage group \ --drilldowns[tag].columns[price_sum].type UInt32 \ --drilldowns[tag].columns[price_sum].flags COLUMN_SCALAR \ --drilldowns[tag].columns[price_sum].value 'aggregator_sum(price)' \ --drilldowns[tag].columns[quantity_sum].stage group \ --drilldowns[tag].columns[quantity_sum].type UInt32 \ --drilldowns[tag].columns[quantity_sum].flags COLUMN_SCALAR \ --drilldowns[tag].columns[quantity_sum].value 'aggregator_sum(quantity)' [ [ 0, 0.0, 0.0 ], [ [ [ 6 ], [ [ "_id", "UInt32" ], [ "_key", "ShortText" ], [ "price", "UInt32" ], [ "quantity", "UInt32" ], [ "tag", "ShortText" ] ], [ 1, "Book", 1000, 100, "A" ], [ 2, "Note", 1000, 10, "B" ], [ 3, "Box", 500, 15, "B" ], [ 4, "Pen", 500, 12, "A" ], [ 5, "Food", 500, 111, "C" ], [ 6, "Drink", 300, 22, "B" ] ], { "tag": [ [ 3 ], [ [ "_key", "ShortText" ], [ "_nsubrecs", "Int32" ], [ "price_sum", "UInt32" ], [ "quantity_sum", "UInt32" ] ], [ "A", 2, 1500, 112 ], [ "B", 3, 1800, 47 ], [ "C", 1, 500, 111 ] ] } ] ]
[groonga executable file] Added support for
--pid-pathin standalone mode.Because
--pid-pathhad been ignored in standalone mode in before version.
[io_flush] Added support for reference count mode.
[logical_range_filter], [logical_count] Added support for reference count mode.
[Groonga HTTP server] We didn’t add header after the last chunk.
Because there is a possibility to exist that the HTTP client ignores header after the last chunk.
[vector_slice] Added support for a vector that has the value of the
Float32type. [GitHub#1112 patched by naoa]Added support for parallel offline index construction using token column.
We came to be able to construct an offline index on parallel threads from data that are tokenized in advance.
We can tune parameters of parallel offline construction by the following environment variables
GRN_TOKEN_COLUMN_PARALLEL_CHUNK_SIZE: We specify how many records are processed per thread.The default value is
1024records.
GRN_TOKEN_COLUMN_PARALLEL_TABLE_SIZE_THRESHOLD: We specify how many source records are required for parallel offline construction.The default value is
102400records.
[select] Improved performance for
load_tableon the reference count mode.
Fixes#
Fixed a bug that the database of Groonga was broken when we search by using the dynamic columns that don’t specify a
--filterand stridden over shard.Fixed a bug that
Float32type had not displayed on a result ofschemacommand.Fixed a bug that we count in surplus to
_nsubrecswhen the referenceuvectorhasn’t element.
Thanks#
naoa
Release 10.0.3 - 2020-05-29#
Improvements#
We came to be able to construct an inverted index from data that are tokenized in advance.
The construct of an index is speeded up from this.
We need to prepare token column to use this improvement.
token column is an auto generated value column like an index column.
token column value is generated from source column value by tokenizing the source column value.
We can create a token column by setting the source column as below.
table_create Terms TABLE_PAT_KEY ShortText \ --normalizer NormalizerNFKC121 \ --default_tokenizer TokenNgram table_create Notes TABLE_NO_KEY column_create Notes title COLUMN_SCALAR Text # The last "title" is the source column. column_create Notes title_terms COLUMN_VECTOR Terms title
[select] We came to be able to specify a
vectorfor the argument of a function.For example,
flagsoptions ofquerycan describe by avectoras below.select \ --table Memos \ --filter 'query("content", "-content:@mroonga", \ { \ "expander": "QueryExpanderTSV", \ "flags": ["ALLOW_LEADING_NOT", "ALLOW_COLUMN"] \ })'
[select] Added a new stage
result_setfor dynamic columns.This stage generates a column into a result set table. Therefore, it is not generated if
queryorfilterdoesn’t existBecause if
queryorfilterdoesn’t exist, Groonga doesn’t make a result set table.
We can’t use
_valuefor the stage. Theresult_setstage is for storing value byscore_column.
[vector_slice] Added support for weight vector that has weight of
Float32type. [GitHub#1106 patched by naoa][select] Added support for
filteredstage andoutputstage of dynamic columns on drilldowns. [GitHub#1101 patched by naoa][GitHub#1100 patched by naoa]We can use
filteredandoutputstage of dynamic columns on drilldowns as withdrilldowns[Label].stage filteredanddrilldowns[Label].stage output.
[select] Added support for
Floattype value in aggregating on drilldown.We can aggregate max value, min value, and sum value for
Floattype value usingMAX,MIN, andSUM.
[query] [geo_in_rectangle] [geo_in_circle] Added a new option
score_columnforquery(),geo_in_rectangle(), andgeo_in_circle().We can store a score value by condition using
score_column.Normally, Groonga calculate a score by adding scores of all conditions. However, we sometimes want to get a score value by condition.
For example, if we want to only use how near central coordinate as score as below, we use
score_column.
table_create LandMarks TABLE_NO_KEY column_create LandMarks name COLUMN_SCALAR ShortText column_create LandMarks category COLUMN_SCALAR ShortText column_create LandMarks point COLUMN_SCALAR WGS84GeoPoint table_create Points TABLE_PAT_KEY WGS84GeoPoint column_create Points land_mark_index COLUMN_INDEX LandMarks point load --table LandMarks [ {"name": "Aries" , "category": "Tower" , "point": "11x11"}, {"name": "Taurus" , "category": "Lighthouse", "point": "9x10" }, {"name": "Gemini" , "category": "Lighthouse", "point": "8x8" }, {"name": "Cancer" , "category": "Tower" , "point": "12x12"}, {"name": "Leo" , "category": "Tower" , "point": "11x13"}, {"name": "Virgo" , "category": "Temple" , "point": "22x10"}, {"name": "Libra" , "category": "Tower" , "point": "14x14"}, {"name": "Scorpio" , "category": "Temple" , "point": "21x9" }, {"name": "Sagittarius", "category": "Temple" , "point": "43x12"}, {"name": "Capricorn" , "category": "Tower" , "point": "33x12"}, {"name": "Aquarius" , "category": "mountain" , "point": "55x11"}, {"name": "Pisces" , "category": "Tower" , "point": "9x9" }, {"name": "Ophiuchus" , "category": "mountain" , "point": "21x21"} ] select LandMarks \ --sort_keys 'distance' \ --columns[distance].stage initial \ --columns[distance].type Float \ --columns[distance].flags COLUMN_SCALAR \ --columns[distance].value 0.0 \ --output_columns 'name, category, point, distance, _score' \ --limit -1 \ --filter 'geo_in_circle(point, "11x11", "11x1", {"score_column": distance}) && category == "Tower"' [ [ 0, 1590647445.406149, 0.0002503395080566406 ], [ [ [ 5 ], [ [ "name", "ShortText" ], [ "category","ShortText" ], [ "point", "WGS84GeoPoint" ], [ "distance", "Float" ], [ "_score", "Int32" ] ], [ "Aries", "Tower", "11x11", 0.0, 1 ], [ "Cancer", "Tower", "12x12", 0.0435875803232193, 1 ], [ "Leo", "Tower", "11x13", 0.06164214760065079, 1 ], [ "Pisces", "Tower", "9x9", 0.0871751606464386, 1 ], [ "Libra", "Tower", "14x14", 0.1307627409696579, 1 ] ] ] ]
The sort by
_scoreis meaningless in the above example. Because the value of_scoreis all1bycategory == "Tower". However, we can sort distance from central coordinate usingscore_column.
[Windows] Groonga came to be able to output backtrace when it occurs error even if it doesn’t crash.
[Windows] Dropped support for old Windows.
Groonga for Windows come to require Windows 8 (Windows Server 2012) or later from 10.0.3.
[select] Improved sort performance when sort keys were mixed referable sort keys and the other sort keys.
We improved sort performance if mixed referable sort keys and the other and there are referable keys two or more.
Referable sort keys are sort keys that except below them.
Compressed columns
_valueagainst the result of drilldown that is specified multiple values to the key of drilldown._keyagainst patricia trie table that has not the key ofShortTexttype._score
The more sort keys that except string, a decrease in the usage of memory for sort.
[select] Improved sort performance when sort keys are all referable keys case.
[select] Improve scorer performance as a
_socre = column1*X + column2*Y + ...case.This optimization effective when there are many
+or*in_score.At the moment, it has only effective against
+and*.
[select] Added support for phrase near search.
We can search phrase by phrase by a near search.
Query syntax for near phrase search is
*NP"Phrase1 phrase2 ...".Script syntax for near phrase search is
column *NP "phrase1 phrase2 ...".If the search target phrase includes space, we can search for it by surrounding it with
"as below.table_create Entries TABLE_NO_KEY column_create Entries content COLUMN_SCALAR Text table_create Terms TABLE_PAT_KEY ShortText \ --default_tokenizer 'TokenNgram("unify_alphabet", false, \ "unify_digit", false)' \ --normalizer NormalizerNFKC121 column_create Terms entries_content COLUMN_INDEX|WITH_POSITION Entries content load --table Entries [ {"content": "I started to use Groonga. It's very fast!"}, {"content": "I also started to use Groonga. It's also very fast! Really fast!"} ] select Entries --filter 'content *NP "\\"I started\\" \\"use Groonga\\""' --output_columns 'content' [ [ 0, 1590469700.715882, 0.03997230529785156 ], [ [ [ 1 ], [ [ "content", "Text" ] ], [ "I started to use Groonga. It's very fast!" ] ] ] ]
[Vector column] Added support for
float32weight vector.We can store weight as
float32instead ofuint32.We need to add
WEIGHT_FLOAT32flag when executecolumn_createto use this feature.column_create Records tags COLUMN_VECTOR|WITH_WEIGHT|WEIGHT_FLOAT32 Tags
However,
WEIGHT_FLOAT32flag isn’t available withCOLUMN_INDEXflag for now.
Added following APIs
Added
grn_obj_is_xxxfunctions. For more information as below.grn_obj_is_weight_vector(grn_ctx *ctx, grn_obj *obj)It returns as a
boolwhether the object is a weight vector.
grn_obj_is_uvector(grn_ctx *ctx, grn_obj *obj)It returns as a
boolwhether the object is auvector.uvectoris avectorthat size of elements forvectorare fixed.
grn_obj_is_weight_uvector(grn_ctx *ctx, grn_obj *obj)It returns as a
boolwhether the object is a weight uvector.
Added
grn_type_id_size(grn_ctx *ctx, grn_id id).It returns the size of Groonga data type as a
size_t.
Added
grn_selector_data_get_xxxfunctions. For more information as below.These functions return selector related data.
These functions are supposed to call in selector. If they are called except in selector, they return
NULL.grn_selector_data_get(grn_ctx *ctx)It returns all information that relating calling selector as
grn_selector_data *structure.
grn_selector_data_get_selector(grn_ctx *ctx, grn_selector_data *data)It returns selector itself as
grn_obj *.
grn_selector_data_get_expr(grn_ctx *ctx, grn_selector_data *data)It returns selector is used
--filtercondition and--querycondition asgrn_obj *.
grn_selector_data_get_table(grn_ctx *ctx, grn_selector_data *data)It returns target table as
grn_obj *
grn_selector_data_get_index(grn_ctx *ctx, grn_selector_data *data)It returns index is used by selector as
grn_obj *.
grn_selector_data_get_args(grn_ctx *ctx, grn_selector_data *data, size_t *n_args)It returns arguments of function that called selector as
grn_obj **.
grn_selector_data_get_result_set(grn_ctx *ctx, grn_selector_data *data)It returns result table as
grn_obj *.
grn_selector_data_get_op(grn_ctx *ctx, grn_selector_data *data)It returns how to perform the set operation on existing result set as
grn_operator.
Added
grn_plugin_proc_xxxfunctions. For more information as below.grn_plugin_proc_get_value_operator(grn_ctx *ctx, grn_obj *value, grn_operator default_operator, const char *context)It returns the operator of a query as a
grn_operator.For example,
&&is returned as aGRN_OP_AND.
grn_plugin_proc_get_value_bool(grn_ctx *ctx, grn_obj *value, bool default_value, const char *tag)It returns the value that is specified
trueorfalselikewith_transpositionargument of the below function as abool(boolis the data type of C language).fuzzy_search(column, query, {"max_distance": 1, "prefix_length": 0, "max_expansion": 0, "with_transposition": true})
Added
grn_proc_options_xxxfunctions. For more information as below.query()only uses them for now.grn_proc_options_parsev(grn_ctx *ctx, grn_obj *options, const char *tag, const char *name, va_list args)This function execute parse options.
We had to implement parsing to options ourselves until now, however, we can parse them by just call this function from this version.
grn_proc_options_parse(grn_ctx *ctx, grn_obj *options, const char *tag, const char *name, ...)It calls
grn_proc_options_parsev(). Therefore, features of this function samegrn_proc_options_parsev().It only differs in the interface compare with
grn_proc_options_parsev().
Added
grn_text_printfv(grn_ctx *ctx, grn_obj *bulk, const char *format, va_list args)grn_text_vprintfis deprecated from 10.0.3. We usegrn_text_printfvinstead.
Added
grn_type_id_is_float_family(grn_ctx *ctx, grn_id id).It returns whether
grn_type_idisGRN_DB_FLOAT32orGRN_DB_FLOATor not as abool.
Added
grn_dat_cursor_get_max_n_records(grn_ctx *ctx, grn_dat_cursor *c).It returns the number of max records the cursor can have as a
size_t. (This API is for the DAT table)
Added
grn_table_cursor_get_max_n_records(grn_ctx *ctx, grn_table_cursor *cursor).It returns the number of max records the cursor can have as a
size_t.It can use against all table type (
TABLE_NO_KEY,TABLE_HASH_KEY,TABLE_DAT_KEY, andTABLE_PAT_KEY).
Added
grn_result_set_add_xxxfunctions. For more information as below.grn_result_set_add_record(grn_ctx *ctx, grn_hash *result_set, grn_posting *posting, grn_operator op)It adds a record into the table of result sets.
grn_ii_posting_add_floatis deprecated from 10.0.3. We usegrn_rset_add_records()instead.
grn_result_set_add_table(grn_ctx *ctx, grn_hash *result_set, grn_obj *table, double score, grn_operator op)It adds a table into the result sets.
grn_result_set_add_table_cursor(grn_ctx *ctx, grn_hash *result_set, grn_table_cursor *cursor, double score, grn_operator op)It adds records that a table cursor has into the result sets.
Added
grn_vector_copy(grn_ctx *ctx, grn_obj *src, grn_obj *dest).It copies a
vectorobject. It returns whether success copy avectorobject.
Added
grn_obj_have_source(grn_ctx *ctx, grn_obj *obj).It returns whether the column has a source column as a
bool.
Added
grn_obj_is_token_column(grn_ctx *ctx, grn_obj *obj).It returns whether the column is a token column as a
bool.
Added
grn_hash_add_table_cursor(grn_ctx *ctx, grn_hash *hash, grn_table_cursor *cursor, double score).It’s for bulk result set insert. It’s faster than inserting records by
grn_ii_posting_add().
Fixes#
Fixed a crash bug if the modules (tokenizers, normalizers, and token filters) are used at the same time from multiple threads.
Fixed precision of
Float32value when it outputted.The precision of it changes to 8-digit to 7-digit from 10.0.3.
Fixed a bug that Groonga used the wrong cache when the query that just the parameters of dynamic column different was executed. [GitHub#1102 patched by naoa]
Thanks#
naoa
Release 10.0.2 - 2020-04-29#
Improvements#
Added support for
uvectorfortime_classify_*functions. [GitHub#1089][Patched by naoa]uvectoris a vector that size of elements for vector are fixed.For example, a vector that has values of Time type as elements is a
uvector.
Improve sort performance if sort key that can’t refer value with zero-copy is mixed.
Some sort key (e.g.
_score) values can’t be referred with zero-copy.If there is at least one sort key that can’t be referable is included, all sort keys are copied before.
With this change, we just copy sort keys that can’t be referred. Referable sort keys are just referred without a copy.
However, this change may have performance regression when all sort keys are referable.
Added support for loading weight vector as a JSON string.
We can load weight vector as a JSON string as below example.
table_create Tags TABLE_PAT_KEY ShortText table_create Data TABLE_NO_KEY column_create Data tags COLUMN_VECTOR|WITH_WEIGHT Tags column_create Tags data_tags COLUMN_INDEX|WITH_WEIGHT Data tags load --table Data [ {"tags": "{\"fruit\": 10, \"apple\": 100}"}, {"tags": "{\"fruit\": 200}"} ]
Added support for
Float32type.Groonga already has a
Floattype. However, it is double precision floating point number. Therefore if we only use single precision floating point number, it is not efficient.We can select more suitable type by adding a
Float32type.
Added following APIs
grn_obj_unref(grn_ctx *ctx, grn_obj *obj)This API is only used on the reference count mode (The reference count mode is a state of
GRN_ENABLE_REFERENCE_COUNT=yes.).It calls
grn_obj_unlink()only on the reference count mode. It doesn’t do anything except when the reference count mode.We useful it when we need only call
grn_obj_unlink()on the referece count mode.Because as the following example, we don’t write condition that whether the reference count mode or not.
The example if we don’t use
grn_obj_unref().if (grn_enable_reference_count) { grn_obj_unlink(ctx, obj); }
The example if we use
grn_obj_unref().grn_obj_ubref(ctx, obj);
grn_get_version_major(void)grn_get_version_minor(void)grn_get_version_micro(void)They return Groonga’s major, minor, and micor version numbers as a
uint32_t.
grn_posting_get_record_id(grn_ctx *ctx, grn_posting *posting)grn_posting_get_section_id(grn_ctx *ctx, grn_posting *posting)grn_posting_get_position(grn_ctx *ctx, grn_posting *posting)grn_posting_get_tf(grn_ctx *ctx, grn_posting *posting)grn_posting_get_weight(grn_ctx *ctx, grn_posting *posting)grn_posting_get_weight_float(grn_ctx *ctx, grn_posting *posting)grn_posting_get_rest(grn_ctx *ctx, grn_posting *posting)They return information on the posting list.
These APIs return value as a
uint32_texceptgrn_posting_get_weight_float.grn_posting_get_weight_floatreturns value as afloat.grn_posting_get_section_id(grn_ctx *ctx, grn_posting *posting)Section id is the internal representation of the column name.
If column name store in posting list as a string, it is a large amount of information and it use waste capacity.
Therefore, Groonga compresses the amount of information and use storage capacity is small by storing column name in the posting list as a number called section id.
grn_posting_get_tf(grn_ctx *ctx, grn_posting *posting)tfofgrn_posting_get_tfis Term Frequency score.
grn_posting_get_weight_float(grn_ctx *ctx, grn_posting *posting)It returns weight of token as a
float.We suggest using this API when we get a weight of token after this.
Because we modify the internal representation of the weight from
uint32_ttofloatin the near future.
Fixes#
Fixed a bug that Groonga for 32bit on GNU/Linux may crash.
Fixed a bug that unrelated column value may be cleared. [GtiHub#1087][Reported by sutamin]
Fixed a memory leak when we dumped records with
dumpcommand.Fixed a memory leak when we specified invalid value into
output_columns.Fixed a memory leak when we executed
snippetfunction.Fixed a memory leak when we filled the below conditions.
If we used dynamic columns on the
initialstage.If we used
slicesargument withselectcommand.
Fixed a memory leak when we deleted tables with
logical_table_remove.Fixed a memory leak when we use the reference count mode.
The reference count mode is a
GRN_ENABLE_REFERENCE_COUNT=yesstate.This mode is experimental. Performance may degrade by this mode.
Fixed a bug that Groonga too much unlink
_keyaccessor when we load data for apache arrow format.
Thanks#
sutamin
naoa
Release 10.0.1 - 2020-03-30#
We have been released Groonga 10.0.1. Because Ubuntu and Windows(VC++ version) package in Groonga 10.0.0 were mistaken.
If we have already used Groonga 10.0.0 for CentOS, Debian, Windows(MinGW version), no problem with continued use it.
Fixes#
Added a missing runtime(vcruntime140_1.dll) in package for Windows VC++ version.
Release 10.0.0 - 2020-03-29#
Improvements#
[httpd] Updated bundled nginx to 1.17.9.
[httpd] Added support for specifying output type as an extension.
For example, we can write
load.jsoninstead ofload?output_type=json.
[Log] Outputted a path of opened or closed file into a log of dump level on Linux.
[Log] Outputted a path of closed file into a log of debug level on Windows.
Added following API and macros
grn_timeval_from_double(grn_ctx, double)This API converts
doubletype togrn_timevaltype.It returns value of
grn_timevaltype.
GRN_TIMEVAL_TO_NSEC(timeval)This macro converts value of
grn_timevaltype to nanosecond as the value ofuint64_ttype.
GRN_TIME_USEC_TO_SEC(usec)This macro converts microsecond to second.
Deprecated the following macro.
GRN_OBJ_FORMAT_FIN(grn_ctx, grn_obj_format)We
grn_obj_format_fin(grn_ctx, grn_obj_format)use instead since 10.0.0.
[logical_range_filter],[dump] Added support for stream output.
This feature requires
command_version 3or later. The header content is outputted after the body content.Currently, this feature support only
dumpandlogical_range_filter.logical_range_filteralways returns the output as a stream oncommand_version 3or later.This feature has the following limitations.
-1 is only allowed for negative
limitMessagePack output isn’t supported
We a little changed the response contents of JSON by this modify.
The key order differs from before versions as below.
The key order in before versions.
{ "header": {...}, "body": {...} }
The key order in this version(10.0.0).
{ "body": {...}, "header": {...} }
Disabled caches of
dumpandlogical_range_filterwhen they execute oncommand_version 3.Because of
dumpandlogical_range_filteroncommand_version 3returns stream since 10.0.0, Groonga can not cache the whole response.
[logical_range_filter] Added support for outputting response as Apache Arrow format.
Supported data type as below.
UInt8Int8UInt16Int16UInt32Int32UInt64Int64TimeShortTextTextLongTextVectorofInt32Reference vector
Supported Ubuntu 20.04 (Focal Fossa).
Dropped Ubuntu 19.04 (Disco Dingo).
Because this version has been EOL.
Release 9.1.2 - 2020-01-29#
Improvements#
[tools] Added a script for copying only files of specify tables or columns.
This script name is copy-related-files.rb.
This script is useful if we want to extract specifying tables or columns from a huge database.
Related files of specific tables or columns may need for reproducing fault.
If we difficult to offer a database whole, we can extract related files of target tables or columns by this tool.
[shutdown] Accept
/d/shutdown?mode=immediateimmediately even when all threads are used.This feature can only use on the Groonga HTTP server.
Unused objects free immediately by using
GRN_ENABLE_REFERENCE_COUNT=yes.This feature is experimental. Performance degrade by this feature.
If we load to span many tables, we can expect to keep in the usage of memory by this feature.
[CentOS] We prepare
groonga-releaseby version.Note that the little modification how to install.
[Debian GNU/Linux] We use
groonga-archive-keyringfor adding the Groonga apt repository.We can easy to add the Groonga apt repository to our system by this improvement.
groonga-archive-keyringincludes all information for using the Groonga apt repository. Thus, we need not be conscious of changing of repository information or PGP key by installing this package.groonga-archive-keyringis deb package. Thus, we can easy to update byapt update.
Release 9.1.1 - 2020-01-07#
Improvements#
[load] Added support for Apache Arrow format data.
If we use Apache Arrow format data, we may reduce parse cost. Therefore, data might be loading faster than other formats.
Groonga can also directly input data for Apache Arrow format from other data analysis systems by this improvement.
However, Apache Arrow format can use in the HTTP interface only. We can’t use it in the command line interface.
[load] Added how to load Apache Arrow format data in the document.
[load] Improve error message.
Response of
loadcommand includes error message also.If we faile data load, Groonga output error detail of
loadcommand by this Improvement.
[httpd] Updated bundled nginx to 1.17.7.
[Groonga HTTP server] Added support for sending command parameters by body of HTTP request.
We must set
application/x-www-form-urlencodedtoContent-Typefor this case.
[Groonga HTTP server] Added how to use HTTP POST in the document.
Release 9.1.0 - 2019-11-29#
Improvements#
Improved the performance of the “&&” operation.
For example, the performance of condition expression such as the following is increased.
( A || B ) && ( C || D ) && ( E || F) …
[TokenMecab] Added a new option
use_base_formWe can search using the base form of a token by this option.
For example, if we search “支えた” using this option, “支える” is hit also.
Fixes#
Fix a bug that when the accessor is index, performance decreases.
For example, it occurs with the query include the following conditions.
accessor @ queryaccessor == query
Fixed a bug the estimated size of a search result was overflow when the buffer is big enough. [PGroonga#GitHub#115][Reported by Albert Song]
Improved a test(1) portability. [GitHub#1065][Patched by OBATA Akio]
Added missing tools.
Because
index-column-diff-all.shandobject-inspect-all.shhad not bundled in before version.
Thanks#
Albert Song
OBATA Akio
Release 9.0.9 - 2019-10-30#
Note
Maybe performance decreases from this version. Therefore, If performance decreases than before, please report us with reproducible steps.
Improvements#
[Log] Improved that output the sending time of response into query-log.
[status] Added that the number of current jobs in the
statuscommand response.[groonga-httpd] Added support for
$request_timein log.In the previous version, even if we specified the
$request_timein thelog_formatdirective, it was always 0.00.If we specify the
$request_time, groonga-httpd output the correct time form this version.
[groonga-httpd] Added how to set the
$request_timein the document.Supported Ubuntu 19.10 (Eoan Ermine)
Supported CentOS 8 (experimental)
The package for CentOS 8 can’t use a part of features(e.g. we can’t use
TokenMecaband can’t cast to int32 vector from JSON string) for lacking some packages for development.
[tools] Added a script for executing the
index_column_diffcommand simply.This script name is index-column-diff-all.sh.
This script extracts index columns form Groonga’s database and execute the
index_column_diffcommand to them.
[tools] Added a script for executing
object_inspectagainst all objects.This script name is object-inspect-all.sh.
Fixes#
Fixed a bug that Groonga crash when we specify the value as the first argument of between.[GitHub#1045][Reported by yagisumi]
Thanks#
yagisumi
Release 9.0.8 - 2019-09-27#
Improvements#
[log_reopen] Added a supplementary explanation when we use
groonga-httpdwith 2 or more workers.Improved that Groonga ignores the index being built.
We can get correct search results even if the index is under construction.
However, the search is slow because of Groonga out of use the index to search in this case.
[sub_filter] Added a feature that
sub_filterexecutes a sequential search when Groonga is building indexes for the target column or the target column hasn’t indexed.sub_filterwas an error if the above situation in before version.From this version,
sub_filterreturns search results if the above situation.However if the above situation,
sub_filteris slow. Because it is executed as a sequential search.
[CentOS] Dropped 32-bit package support on CentOS 6.
Fixes#
[logical_range_filter] Fixed a bug that exception about closing the same object twice occurs when we have enough records and the number of records that unmatch filter search criteria is more than the estimated value of it.
Release 9.0.7 - 2019-08-29#
Improvements#
[httpd] Updated bundled nginx to 1.17.3.
Contains security fix for CVE-2019-9511, CVE-2019-9513, and CVE-2019-9516.
Fixes#
Fixed a bug that Groonga crash when posting lists were huge.
However, this bug almost doesn’t occur by general data. Because posting lists don’t grow bigger so much by them.
Fixed a bug that returns an empty result when we specify
initialinto a stage of a dynamic column and search for using index. [GitHub#683]Fixed a bug that the configure phase didn’t detect libedit despite installing it. [GitHub#1030][Patched by yu]
Fixed a bug that
--offsetand--limitoptions didn’t work with--sort_keysand--slicesoptions. [clear-code/redmine_full_text_search#70][Reported by a9zawa]Fixed a bug that search result is empty when the result of
selectcommand is huge. [groonga-dev,04770][Reported by Yutaro Shimamura]Fixed a bug that doesn’t use a suitable index when prefix search and suffix search. [GitHub#1007, PGroonga#GitHub#96][Reported by oknj]
Thanks#
oknj
Yutaro Shimamura
yu
a9zawa
Release 9.0.6 - 2019-08-05#
Improvements#
Added support for Debian 10 (buster).
Fixes#
[select] Fixed a bug that search is an error when occurring search escalation.
[select] Fixed a bug that may return wrong search results when we use nested equal condition.
[geo_distance_location_rectangle] Fixed an example that has wrong
loadformat. [GitHub#1023][Patched by yagisumi][Let’s create micro-blog] Fixed an example that has wrong search results. [GutHub#1024][Patched by yagisumi]
Thanks#
yagisumi
Release 9.0.5 - 2019-07-30#
Warning
There are some critical bugs are found in this release. select command returns wrong search results.
We will release the new version (9.0.6) which fixes the issues.
Please do not use Groonga 9.0.5, and recommends to upgrade to 9.0.6 in the future.
The detail of this issues are explained at https://groonga.org/en/blog/2019/07/30/groonga-9.0.5.html.
Improvements#
[logical_range_filter] Improved that only apply an optimization when the search target shard is large enough.
This feature reduces that duplicate search result between offset when we use same sort key.
Large enough threshold is 10000 records by default.
[Normalizers] Added new option
unify_to_katakanaforNormalizerNFKC100.This option normalize hiragana to katakana.
For example,
ゔぁゔぃゔゔぇゔぉis normalized toヴァヴィヴヴェヴォ.
[select] Added drilldowns support as a slices parameter.
[select] Added columns support as a slices parameter.
[select] Improved that we can refer
_scorein the initial stage for slices parameter.[highlight_html], [snippet_html] Improved that extract a keyword also from an expression of before executing a slices when we specify the slices parameter.
Improved that collect scores also from an expression of before executing a slices when we specify the slices parameter.
Stopped add 1 in score automatically when add posting to posting list.
grn_ii_posting_addis backward incompatible changed by this change. * Caller must increase the score to maintain compatibility.
Added support for index search for nested equal like
XXX.YYY.ZZZ == AAA.Reduce rehash interval when we use hash table.
This feature improve performance for output result.
Improved to we can add tag prefix in the query log.
We become easy to understand that it is filtered which the condition.
Added support for Apache Arrow 1.0.0.
However, It’s not released this version yet.
Added support for Amazon Linux 2.
Fixes#
Fixed a bug that vector values of JSON like
"[1, 2, 3]"are not indexed.Fixed wrong parameter name in
table_createtests. [GitHub#1000][Patch by yagisumi]Fixed a bug that drilldown label is empty when a drilldown command is executed by
command_version=3. [GitHub#1001][Reported by yagisumi]Fixed build error for Windows package on MinGW.
Fixed install missing COPYING for Windows package on MinGW.
Fixed a bug that don’t highlight when specifing non-text query as highlight target keyword.
Fixed a bug that broken output of MessagePack format of [object_inspect]. [GitHub#1009][Reported by yagisumi]
Fixed a bug that broken output of MessagePack format of
index_column_diff. [GitHub#1010][Reported by yagisumi]Fixed a bug that broken output of MessagePack format of [suggest]. [GitHub#1011][Reported by yagisumi]
Fixed a bug that allocate size by realloc isn’t enough when a search for a table of patricia trie and so on. [Reported by Shimadzu Corporation]
Groonga may be crashed by this bug.
Fix a bug that
groonga.repois removed when updating 1.5.0 fromgroonga-releaseversion before 1.5.0-1. [groonga-talk:429][Reported by Josep Sanz]
Thanks#
yagisumi
Shimadzu Corporation
Josep Sanz
Release 9.0.4 - 2019-06-29#
Improvements#
Added support for array literal with multiple elements.
Added support equivalence operation of a vector.
[logical_range_filter] Increase outputting logs into query log.
logical_range_filtercommand comes to output a log for below timing.After filtering by
logical_range_filter.After sorting by
logical_range_filter.After applying dynamic column.
After output results.
We can see how much has been finished this command by this feature.
[Tokenizers] Added document for
TokenPatterndescription.[Tokenizers] Added document for
TokenTabledescription.[Tokenizers] Added document for
TokenNgramdescription.[grndb] Added output operation log into groonga.log
grndbcommand comes to output execution result and execution process.
[grndb] Added support for checking empty files.
We can check if the empty files exist by this feature.
[grndb] Added support new option
--sinceWe can specify a scope of an inspection.
[grndb] Added document about new option
--sinceBundle RapidJSON
We can use RapidJson as Groonga’s JSON parser partly. (This feature is partly yet)
We can more exactly parse JSON by using this.
Added support for casting to int32 vector from JSON string.
This feature requires RapidJSON.
[query] Added
default_operator.We can customize operator when “keyword1 keyword2”.
“keyword1 Keyword2” is AND operation in default.
We can change “keyword1 keyword2“‘s operator except AND.
Fixes#
[optimizer] Fix a bug that execution error when specified multiple filter conditions and like
xxx.yyy=="keyword".Added missing LICENSE files in Groonga package for Windows(VC++ version).
Added UCRT runtime into Groonga package for Windows(VC++ version).
[Window function] Fix a memory leak.
This occurs when multiple windows with sort keys are used. [Patched by Takashi Hashida]
Thanks#
Takashi Hashida
Release 9.0.3 - 2019-05-29#
Improvements#
[select] Added more query logs.
selectcommand comes to output a log for below timing.After sorting by drilldown.
After filter by drilldown.
We can see how much has been finished this command by this feature.
[logical_select] Added more query logs.
logical_selectcommand comes to output a log for below timing.After making dynamic columns.
After grouping by drilldown.
After sorting by drilldown.
After filter by drilldown.
After sorting by
logical_select.
We can see how much has been finished this command by this feature.
[logical_select] Improved performance of sort a little when we use
limitoption.[index_column_diff] Improved performance.
We have greatly shortened the execution speed of this command.
[index_column_diff] Improved ignore invalid reference.
[index_column_diff] Added support for duplicated vector element case.
[Normalizers] Added a new Normalizer
NormalizerNFKC121based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 12.1.[TokenFilters] Added a new TokenFilter
TokenFilterNFKC121based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 12.1.[grndb] Added a new option
--log-flagsWe can specify output items of a log as with groonga executable file.
See [groonga executable file] to know about supported log flags.
[snippet_html] Added a new option for changing a return value when no match by search.
[plugin_unregister] Added support full path of Windows.
Added support for multiline log message.
The multiline log message is easy to read by this feature.
Output key in Groonga’s log when we search by index.
[match_columns parameter] Added a document for indexes with weight.
[logical_range_filter] Added a explanation for
orderparameter.[object_inspect] Added an explanation for new statistics
INDEX_COLUMN_VALUE_STATISTICS_NEXT_PHYSICAL_SEGMENT_IDandINDEX_COLUMN_VALUE_STATISTICS_N_PHYSICAL_SEGMENTS.Dropped Ubuntu 14.04 support.
Fixes#
[index_column_diff] Fixed a bug that too much
remainsare reported.Fixed a build error when we use
--without-onigmooption. [GitHub#951] [Reported by Tomohiro KATO]Fixed a vulnerability of “CVE: 2019-11675”. [Reported by Wolfgang Hotwagner]
Removed extended path prefix
\\?\at Windows version of Groonga. [GitHub#958] [Reported by yagisumi]This extended prefix causes a bug that plugin can’t be found correctly.
Thanks#
Tomohiro KATO
Wolfgang Hotwagner
yagisumi
Release 9.0.2 - 2019-04-29#
We provide a package for Windows made from VC++ from this release.
We also provide a package for Windows made form MinGW as in the past. However, we will provide it made from VC++ instead of making from MinGW sooner or later.
Improvements#
[column_create] Added a new flag
INDEX_LARGEfor index column.We can make an index column has space that two times of default by this flag.
However, note that it also uses two times of memory usage.
This flag useful when index target data are large.
Large data must have many records (normally at least 10 millions records) and at least one of the following features.
Index targets are multiple columns
Index table has tokenizer
[object_inspect] Added a new statistics
next_physical_segment_idandmax_n_physical_segmentsfor physical segment information.We can confirm usage of index column space and max value of index column space by this information.
[logical_select] Added support for window function over shard.
[logical_range_filter] Added support for window function over shard.
[logical_count] Added support for window function over shard.
We provided a package for Windows made from VC++.
[io_flush] Added a new option
--recursive dependentWe can all of the specified flush target object, child objects, corresponding table of an index column and corresponding index column are flush target objects.
Fixes#
Fixed “unknown type name ‘bool’” compilation error in some environments.
Fixed a bug that incorrect output number over Int32 by command of execute via mruby (e.g.
logical_select,logical_range_filter,logical_count, etc.). [GitHub#936] [Patch by HashidaTKS]
Thanks#
HashidaTKS
Release 9.0.1 - 2019-03-29#
Improvements#
Added support to acccept null for vector value.
You can use select … –columns[vector].flags COLUMN_VECTOR –columns[vector].value “null”
[dump] Translated document into English.
Added more checks and logging for invalid indexes. It helps to clarify the index related bugs.
Improved an explanation about
GRN_TABLE_SELECT_ENOUGH_FILTERED_RATIObehavior in news at Release 8.0.6 - 2018-08-29.[select] Added new argument
--load_table,--load_columnsand--load_values.You can store a result of
selectin a table that specifying--load_table.--load_valuesoption specifies columns of result ofselect.--load_columnsoptions specifies columns of table that specifying--load_table.In this way, you can store values of columns that specifying with
--load_valuesinto columns that specifying with--load_columns.
[select] Added documentation about
load_table,load_columnsandload_values.[load] Added supoort to display a table of load destination in a query log.
A name of table of load destination display as string in
[]as below.:000000000000000 load(3): [LoadedLogs][3]
Added a new API:
grn_ii_get_flags()grn_index_column_diff()grn_memory_get_usage()
Added
index_column_diffcommand to check broken index column. If you want to log progress of command execution, set log level to debug.
Fixes#
[snippet_html] Changed to return an empty vector for no match.
In such a case, an empty vector
[]is returned instead ofnull.
Fixed a warning about possibility of counting threads overflow. In real world, it doesn’t affect user because enourmous number of threads is not used. [GitHub#904]
Fixed build error on macOS [GitHub#909] [Reported by shiro615]
Fixed a stop word handling bug.
This bug occurs when we set the first token as a stop word in our query.
If this bug occurs, our search query isn’t hit.
[Global configurations] Fixed a typo about parameter name of
grn_lock_set_timeout.Fixed a bug that deleted records may be matched because of updating indexes incorrectly.
It may occure when large number of records is added or deleted.
Fixed a memory leak when
logical_range_filterreturns no records. [GitHub#911] [Patch by HashidaTKS]Fixed a bug that query will not match because of loading data is not normalized correctly. [PGroonga#GitHub#93, GitHub#912,GitHub#913] [Reported by kamicup and dodaisuke]
This bug occurs when load data contains whitespace after KATAKANA and
unify_kanaoption is used for normalizer.
Fixed a bug that an indexes is broken during updating indexes.
It may occurs when repeating to add large number of records or delete them for a long term.
Fixed a crash bug that allocated working area is not enough size when updating indexes.
Thanks#
shiro615
HashidaTKS
kamicup
dodaisuke
Release 9.0.0 - 2019-02-09#
This is a major version up! But It keeps backward compatibility. You can upgrade to 9.0.0 without rebuilding database.
Improvements#
[Tokenizers] Added a new tokenizer
TokenPattern.You can extract tokens by regular expression.
This tokenizer extracts only token that matches the regular expression.
You can also specify multiple patterns of regular expression.
[Tokenizers] Added a new tokenizer
TokenTable.You can extract tokens by a value of columns of existing a table.
[dump] Added support for dumping binary data.
[select] Added support for similer search against index column.
If you have used multi column index, you can similar search against all source columns by this feature.
[Normalizers] Added new option
remove_blankforNormalizerNFKC100.This option remove white spaces.
[groonga executable file] Improve display of thread id in log.
Because It was easy to confuse thread id and process id on Windows version, it made clear which is a thread id or a process id.
Release 8.1.1 - 2019-01-29#
Improvements#
[logical_select] Added new argument
--load_table,--load_columnsand--load_values.You can store a result of
logical_selectin a table that specifying--load_table.--load_valuesoption specifies columns of result oflogical_select.--load_columnsoptions specifies columns of table that specifying--load_table.In this way, you can store values of columns that specifying with
--load_valuesinto columns that specifying with--load_columns.
Improve error log when update error of index.
Added more information in the log.
For example, output source buffer and chunk when occur merge of posting lists error.
Also, outputting the log a free space size of a buffer and request size of a buffer when occurs error of allocating a buffer.
[groonga executable file] Added a new option
--log-flags.We can specify output items of a log of the Groonga.
We can output as below items.
Timestamp
Log message
Location(the location where the log was output)
Process id
Thread id
We can specify prefix as below.
+This prefix means that “add the flag”.
-This prefix means that “remove the flag”.
No prefix means that “replace existing flags”.
Specifically, we can specify flags as below.
noneOutput nothing into the log.
timeOutput a timestamp into the log.
messageOutput log messages into the log.
locationOutput the location where the log was output( a file name, a line and a function name) and process id.
process_idOutput a process id into the log.
pidThis flag is an alias of
process_id.
thread_idOutput thread id into the log.
allThis flag specifies all flags except
noneanddefaultflags.
defaultOutput a timestamp and log messages into the log.
We can also specify multiple log flags by separating flags with
|.
Fixes#
Fixed a memory leak when occurs index update error.
[Normalizers] Fixed a bug that stateless normalizers and stateful normalizers return wrong results when we use them at the same time.
Stateless normalizers are below.
unify_kanaunify_kana_caseunify_kana_voiced_sound_markunify_hyphenunify_prolonged_sound_markunify_hyphen_and_prolonged_sound_markunify_middle_dot
Stateful normalizers are below.
unify_katakana_v_soundsunify_katakana_bu_soundunify_to_romaji
Release 8.1.0 - 2018-12-29#
Improvements#
[httpd] Updated bundled nginx to 1.15.8.
Fixes#
Fixed a bug that unlock against DB is always executed after flush when after execute a
io_flushcommand.Fixed a bug that
reindexcommand doesn’t finish when execute areindexcommand against table that has record that has not references.
Release 8.0.9 - 2018-11-29#
Improvements#
[Tokenizers] Improved that output a tokenizer name in error message when create tokenizer fail.
[Tokenizers][TokenDelimit] Supported that customizing delimiter of a token.
You can use token other than whitespace as a token of delimiter.
[Tokenizers][TokenDelimit] Added new option
pattern.You can specify delimiter with regular expression by this option.
[Tokenizers] Added force_prefix_search value to each token information.
“force_prefix” is kept for backward compatibility.
[Token filters] Added built-in token filter
TokenFilterNFKC100.You can convert katakana to hiragana like NormalizerNFKC100 with a
unify_kanaoption.
[Token filters][TokenFilterStem] Added new option
algorithm.You can also stem language other than English(French, Spanish, Portuguese, Italian, Romanian, German, Dutch, Swedish, Norwegian, Danish, Russian, Finnish) by this option.
[Token filters][TokenFilterStopWord] Added new option
column.You can specify stop word in a column other than is_stop_word column by this option.
[dump] Supported output options of token filter options.
If you specify a tokenizer like
TokenNgramorTokenMecabetc that has options, you can output these options withtable_listcommand.
[truncate] Supported a table that it has token filter option.
You can
truncateeven a tabel that it has token filter likeTokenFilterStemorTokenStopWordthat has options.
[schema] Support output of options of token filter.
[Normalizers] Added new option for
NormalizerNFKC100thatunify_to_romajioption.You can normalize hiragana and katakana to romaji by this option.
[query-log][show-condition] Supported “func() > 0” case.
[Windows] Improved that ensure flushing on unmap.
Improved error message on opening input file error.
[httpd] Updated bundled nginx to 1.15.7.
contains security fix for CVE-2018-16843 and CVE-2018-16844.
Fixes#
Fixed a memory leak when evaluating window function.
[groonga-httpd] Fixed bug that log content may be mixed.
Fixed a bug that generates invalid JSON when occurs error of slice on output_columns.
Fixed a memory leak when getting nested reference vector column value.
Fixed a crash bug when outputting warning logs of index corruption.
Fix a crash bug when temporary vector is reused in expression evaluation.
For example, crash when evaluating an expression that uses a vector as below.
_score = _score + (vector_size(categories) > 0)Fix a bug that hits a value of vector columns deleted by a delete command.[GitHub PGroonga#85][Reported by dodaisuke]
Thanks#
dodaisuke
Release 8.0.8 - 2018-10-29#
Improvements#
[table_list] Supported output options of default tokenizer.
If you specify a tokenizer like
TokenNgramorTokenMecabetc that has options, you can output these options withtable_listcommand.
[select] Supported normalizer options in sequential match with
record @ 'query'.[truncate] Supported a table that it has tokenizer option.
You can
truncateeven a tabel that it has tokenizer likeTokenNgramorTokenMecabetc that has options.
[Tokenizers][TokenMecab] Added new option
target_classThis option searches a token of specifying a part-of-speech. For example, you can search only a noun.
This option can also specify subclasses and exclude or add specific part-of-speech of specific using
+or-. So, you can also search except a pronoun as below.'TokenMecab("target_class", "-名詞/代名詞", "target_class", "+")'
[io_flush] Supported locking of a database during a
io_flush.Because Groonga had a problem taht is a crash when deleteing a table of a target of a
io_flushduring execution of aio_flush.
[cast_loose] Added a new function
cast_loose.This function cast to a type to specify. If a value to specify can’t cast, it become to a default value to specify.
Added optimize the order of evaluation of a conditional expression.(experimental)
You can active this feature by setting environment value as below.
GRN_EXPR_OPTIMIZE=yes
Supported
(?-mix:XXX)form for index searchable regular expression. [groonga-dev,04683][Reported by Masatoshi SEKI](?-mix:XXX)form treats the same as XXX.
[httpd] Updated bundled nginx to 1.15.5.
Supported Ubuntu 18.10 (Cosmic Cuttlefish)
Fixes#
Fixed a bug that the Groonga GQTP server may fail to accept a new connection. [groonga-dev,04688][Reported by Yutaro Shimamura]
It’s caused when interruption client process without using quit.
Thanks#
Masatoshi SEKI
Yutaro Shimamura
Release 8.0.7 - 2018-09-29#
Improvements#
[Tokenizers][TokenMecab] support outputting metadata of Mecab.
Added new option
include_classforTokenMecab.This option outputs
classandsubclassin Mecab’s metadata.Added new option
include_readingforTokenMecab.This option outputs
readingin Mecab’s metadata.Added new option
include_formforTokenMecab.This option outputs
inflected_type,inflected_formandbase_formin Mecab’s metadata.Added new option
use_readingforTokenMecab.This option supports a search by kana.
This option is useful for countermeasure of orthographical variants because it searches with kana.
[plugin] Groonga now can grab plugins from multiple directories.
You can specify multiple directories to
GRN_PLUGINS_PATHseparated with “:” on non Windows, “;” on Windows.GRN_PLUGINS_PATHhas high priority than the existingGRN_PLUGINS_DIR. Currently, this option is not supported Windows.[Tokenizers][TokenNgram] Added new option
unify_alphabetforTokenNgram.If we use
unify_alphabetasfalse,TokenNgramuses bigram tokenize method for ASCII character.[Tokenizers][TokenNgram] Added new option
unify_symbolforTokenNgram.TokenNgram("unify_symbol", false)is same behavior ofTokenBigramSplitSymbol.[Tokenizers][TokenNgram] Added new option
unify_digitforTokenNgram.If we use
unify_digitasfalse, If we set false,TokenNgramuses bigram tokenize method for digits.[httpd] Updated bundled nginx to 1.15.4.
Fixes#
Fixed wrong score calculations on some cases.
It’s caused when adding, multiplication or division numeric to a bool value.
It’s caused when comparing a scalar and vector columns using
!=or==.
Release 8.0.6 - 2018-08-29#
Improvements#
[Tokenizers][TokenMecab] add
chunked_tokenizeandchunk_size_thresholdoptions.[optimizer] support estimation for query family expressions. It will generate more effective execution plan with query family expressions such as
column @ query,column @~ patternand so on.[optimizer] plug-in -> built-in It’s disabled by default for now. We can enable it by defining
GRN_EXPR_OPTIMIZE=yesenvironment variable or usingexpression_rewriterstable as before.Enable sequential search for enough filtered case by default. If the current result is enough filtered, sequential search is faster than index search. If the current result has only 1% records of all records in a table and less than 1000 records, sequential search is used even when index search is available.
Cullently, this optimization is applied when search by
==,>,<,>=, or<=.When a key of a table that has columns specified by the filter is
ShortText, you must setNormalizerAutoto normalizer of the table to apply this optimization.You can disable this feature by
GRN_TABLE_SELECT_ENOUGH_FILTERED_RATIO=0.0environment variable.[load] improve error message. Table name is included.
[load] add
lock_tableoption. If--lock_table yesis specified,loadlocks the target table while updating columns and applying--each. This option avoidsloadanddeleteconflicts but it’ll reduce load performance.[vector_find] avoid to crash with unsupported modes
Fixes#
[index] fix a bug that offline index construction for text vector with
HASH_KEY. It creates index with invalid section ID.Fix a bug that
--match_columns 'index[0] || index[9]'uses wrong section.[highlighter] fix a wrong highlight bug It’s caused when lexicon is hash table and keyword is less than N of N-gram.
[mruby] fix a bug that real error is hidden. mruby doesn’t support error propagation by no argument raise. mruby/mruby#290
[Tokenizers][TokenNgram loose]: fix a not found bug when query has only loose types.
highlight_html()with lexicon was also broken.Fix a bug that text->number cast ignores trailing garbage. “0garbage” should be cast error.
Fix an optimization bug for
reference_column >= 'key_value'case
Release 8.0.5 - 2018-07-29#
Improvements#
[Script syntax] Added complementary explain about similar search against Japanese documents. [GitHub#858][Patch by Yasuhiro Horimoto]
[time_classify_day_of_week] Added a new API:
time_classify_day_of_week().Suppressed a warning with
-fstack-protector. Suggested by OBATA Akio.Added a new API:
time_format_iso8601().Exported a struct
grn_raw_string.Added a new API:
grn_obj_clear_option_values(). It allows you to clear option values on remove (for persistent) / close (for temporary.)[log] Reported index column name for error message
[ii][update][one].[httpd] Updated bundled nginx to 1.15.2.
[Ubuntu] Dropped Ubuntu 17.10 (Artful Aardvark) support. It has reached EOL at July 19, 2018.
[Debian GNU/Linux] Dropped jessie support. Debian’s security and release team will no longer produce updates for jessie.
Fixes#
Fixed returning wrong result after unfinished
/d/loaddata by POST.Fixed wrong function call around KyTea.
[grndb] Added a missing label for the
--force-truncateoption.Fixed crash on closing of a database, when a normalizer provided by a plugin (ex.
groonga-normalizer-mysql) is used with any option.Fixed a bug that normalizer/tokenizer options may be ignored. It’s occurred when the same object ID is reused.
Release 8.0.4 - 2018-06-29#
Improvements#
[log] Add sub error for error message
[ii][update][one].Added a new API:
grn_highlighter_clear_keywords().Added a new predicate:
grn_obj_is_number_family_bulk().Added a new API:
grn_plugin_proc_get_value_mode().[vector_find] Added a new function
vector_find().Suppress memcpy warnings in msgpack.
Updated mruby from 1.0.0 to 1.4.1.
[doc][grn_obj] Added API reference for
grn_obj_is_index_column().[windows] Suppress printf format warnings.
[windows] Suppress warning by msgpack.
[grn_obj][Plugin] Added encoding converter. rules:
grn_ctx::errbuf: grn_encoding
grn_logger_put: grn_encoding
mruby: UTF-8
path: locale
[mrb] Added
LocaleOutput.[windows] Supported converting image path to grn_encoding.
[Tokenizers][TokenMecab] Convert error message encoding.
[window_sum] Supported dynamic column as a target column.
[doc][grn_obj] Added API reference for
grn_obj_is_vector_column().[column_create] Added more validations.
1: Full text search index for vector column must have
WITH_SECTIONflag. (Note that TokenDelmit withWITH_POSITIONwithoutWITH_SECTIONis permitted. It’s useful pattern for tag search.)2: Full text search index for vector column must not be multi column index. detail: groonga/groonga
[grndb] Disabled log check temporarily. Because it’s not completed yet.
Fixes#
[sub_filter] Fixed too much score with a too filtered case.
Fixed build error if KyTea is installed.
[grndb] Fixed output channel.
[query-log][show-condition] Maybe fixed a crash bug.
[highlighter][lexicon] Fixed a not highlighted bug. The keyword wasn’t highlighted if keyword length is less than N (“N”-gram. In many cases, it’s Bigram so “less than 2”).
[windows] Fixed a base path detection bug. If system locale DLL path includes 0x5c (
\in ASCII) such as “U+8868 CJK UNIFIED IDEOGRAPH-8868” in CP932, the base path detection is buggy.[Tokenizers][TokenNgram] Fixed wrong first character length. It’s caused for “PARENTHESIZED IDEOGRAPH” characters such as “U+3231 PARENTHESIZED IDEOGRAPH STOCK”.
Release 8.0.3 - 2018-05-29#
Improvements#
[highlight_html] Support highlight of results of the search by
NormalizerNFKC100orTokenNgram.[Tokenizers] Added new option for
TokenNgramthatreport_source_location option. This option used when highlighting withhighlight_htmluse a lexicon.[Normalizers] Added new option for
NormalizerNFKC100thatunify_middle_dot option. This option normalizes middle dot. You can search with or without・(middle dot) and regardless of・position.[Normalizers] Added new option for
NormalizerNFKC100thatunify_katakana_v_sounds option. This option normalizesヴァヴィヴヴェヴォ(katakana) toバビブベボ(katakana). For example, you can searchバイオリン(violin) inヴァイオリン(violin).[Normalizers] Added new option for
NormalizerNFKC100thatunify_katakana_bu_sound option. This option normalizesヴァヴィヴゥヴェヴォ(katakana) toブ(katakana). For example, you can searchセーブル(katakana) andセーヴルinセーヴェル(katakana).[sub_filter] Supported
sub_filteroptimization for the too filter case. this optimize is valid when records are enough narrowed down beforesub_filterexecution as below.[groonga-httpd] Made all workers context address to unique. context address is
#{ID}of below query log.#{TIME_STAMP}|#{MESSAGE}#{TIME_STAMP}|#{ID}|>#{QUERY}#{TIME_STAMP}|#{ID}|:#{ELAPSED_TIME} #{PROGRESS}#{TIME_STAMP}|#{ID}|<#{ELAPSED_TIME} #{RETURN_CODE}[delete] Added new options that
limit. You can limit the number of delete records as below example.delete --table Users --filter '_key @^ "b"' --limit 4[httpd] Updated bundled nginx to 1.14.0.
Fixes#
[logical_select] Fixed memory leak when an error occurs in filtered dynamic columns.
[logical_count] Fixed memory leak on initial dynamic column error.
[logical_range_filter] Fixed memory leak when an error occurs in dynamic column evaluation.
[Tokenizers] Fixed a bug that the wrong
source_offsetwhen a loose tokenizing such asloose_symboloption.[Normalizers] Fixed a bug that FULLWIDTH LATIN CAPITAL LETTERs such as
U+FF21 FULLWIDTH LATIN CAPITAL LETTER Aaren’t normalized to LATIN SMALL LETTERs such asU+0061 LATIN SMALL LETTER A. If you have been usedNormalizerNFKC100, you must recreate your indexes.
Release 8.0.2 - 2018-04-29#
Improvements#
[grndb][--force-truncate] Improved
grndb recover --force-truncateoption that it can be truncated even if locks are left on the table.[logical_range_filter] Added
sort_keysoption.Added a new function
time_format(). You can specify time format against a column ofTimetype. You can specify with use format ofstrftime.[Tokenizers] Support new tokenizer
TokenNgram. You can change its behavior dynamically via options. Here is a list of available options:n: “N” of Ngram. For example, “3” for trigram.loose_symbol: Tokenize keywords including symbols, to be searched by both queries with/without symbols. For example, a keyword “090-1111-2222” will be found by any of “09011112222”, “090”, “1111”, “2222” and “090-1111-2222”.loose_blank: Tokenize keywords including blanks, to be searched by both queries with/without blanks. For example, a keyword “090 1111 2222” will be found by any of “09011112222”, “090”, “1111”, “2222” and “090 1111 2222”.remove_blank: Tokenize keywords including blanks, to be searched by queries without blanks. For example, a keyword “090 1111 2222” will be found by any of “09011112222”, “090”, “1111” or “2222”. Note that the keyword won’t be found by a query including blanks like “090 1111 2222”.
[Normalizers] Support new normalizer “NormalizerNFKC100” based on Unicode NFKC (Normalization Form Compatibility Composition) for Unicode 10.0.
[Normalizers] Support options for “NormalizerNFKC51” and “NormalizerNFKC100” normalizers. You can change their behavior dynamically. Here is a list of available options:
unify_kana: Same pronounced characters in all of full-width Hiragana, full-width Katakana and half-width Katakana are regarded as the same character.unify_kana_case: Large and small versions of same letters in all of full-width Hiragana, full-width Katakana and half-width Katakana are regarded as the same character.unify_kana_voiced_sound_mark: Letters with/without voiced sound mark and semi voiced sound mark in all of full-width Hiragana, full-width Katakana and half-width Katakana are regarded as the same character.unify_hyphen: The characters like hyphen are regarded as the hyphen.unify_prolonged_sound_mark: The characters like prolonged sound mark are regarded as the prolonged sound mark.unify_hyphen_and_prolonged_sound_mark: The characters like hyphen and prolonged sound mark are regarded as the hyphen.
[dump] Support output of tokenizer’s options and normalizer’s options. Groonga 8.0.1 and earlier versions cannot import dump including options for tokenizers or normalizers generated by Groonga 8.0.2 or later, and it will occurs error due to unsupported information.
[schema] Support output of tokenizer’s options and normalizer’s options. Groonga 8.0.1 and earlier versions cannot import schema including options for tokenizers or normalizers generated by Groonga 8.0.2 or later, and it will occurs error due to unsupported information.
Supported Ubuntu 18.04 (Bionic Beaver)
Fixes#
Fixed a bug that unexpected record is matched with space only query. [groonga-dev,04609][Reported by satouyuzh]
Fixed a bug that wrong scorer may be used. It’s caused when multiple scorers are used as below.
--match_columns 'title || scorer_tf_at_most(content, 2.0)'.Fixed a bug that it may also take so much time to change “thread_limit”.
Thanks#
satouyuzh
Release 8.0.1 - 2018-03-29#
Improvements#
[Log] Show
filterconditions in query log. It’s disabled by default. To enable it, you need to set an environment variableGRN_QUERY_LOG_SHOW_CONDITION=yes.Install
*.pdbinto the directory where*.dlland*.exeare installed.[logical_count] Support
filteredstage dynamic columns.[logical_count] [post_filter] Added a new filter timing. It’s executed after
filteredstage columns are generated.[logical_select] [post_filter] Added a new filter timing. It’s executed after
filteredstage columns are generated.Support LZ4/Zstd/zlib compression for vector data.
Support alias to accessor such as
_key.[logical_range_filter] Optimize window function for large result set. If we find enough matched records, we don’t apply window function to the remaining windows.
TODO: Disable this optimization for small result set if its overhead is not negligible. The overhead is not evaluated yet.
[select] Added
match_escalationparameter. You can force to enable match escalation by--match_escalation yes. It’s stronger than--match_escalation_threshold 99999....999because--match_escalation yesalso works withSOME_CONDITIONS && column @ 'query'.--match_escalation_thresholdisn’t used in this case.The default is
--match_escalation auto. It doesn’t change the current behavior.You can disable match escalation by
--match_escalation no. It’s the same as--match_escalation_threshold -1.[httpd] Updated bundled nginx to 1.13.10.
Fixes#
Fixed memory leak that occurs when a prefix query doesn’t match any token. [GitHub#820][Patch by Naoya Murakami]
Fixed a bug that a cache for different databases is used when multiple databases are opened in the same process.
Fixed a bug that a wrong index is constructed. This occurs only when the source of a column is a vector column and
WITH_SECTIONisn’t specified.Fixed a bug that a constant value can overflow or underflow in comparison (>,>=,<,<=,==,!=).
Thanks#
Naoya Murakami
Release 8.0.0 - 2018-02-09#
This is a major version up! But It keeps backward compatibility. You can upgrade to 8.0.0 without rebuilding database.
Improvements#
[select] Added
--drilldown_adjusterand--drilldowns[LABEL].adjuster. You can adjust score against result of drilldown.[Online index construction] Changed environment variable name
GRN_II_REDUCE_EXPIRE_ENABLEtoGRN_II_REDUCE_EXPIRE_THRESHOLD.GRN_II_REDUCE_EXPIRE_THRESHOLD=0 == GRN_II_REDUCE_EXPIRE_ENABLE=no.GRN_II_REDUCE_EXPIRE_THRESHOLD=-1usesii->chunk->max_map_seg / 2as threshold.GRN_II_REDUCE_EXPIRE_THRESHOLD > 0usesMIN(ii->chunk->max_map_seg / 2, GRN_II_REDUCE_EXPIRE_THRESHOLD)as threshold.GRN_II_REDUCE_EXPIRE_THRESHOLD=32is the default.[between] Accept
between()without borders. If the number of arguments passed tobetween()is 3, the 2nd and 3rd arguments are handled as the inclusive edges. [GitHub#685]
Fixes#
Fixed a memory leak for normal hash table. [GitHub:mroonga/mroonga#190][Reported by fuku1]
Fix a memory leak for normal array.
[select] Stopped to cache when
output_columnsuses not stable function.[Windows] Fixed wrong value report on
WSASenderror.
Thanks#
fuku1
Release 7.1.1 - 2018-01-29#
Improvements#
[Ubuntu] Dropped Ubuntu 17.04 (Zesty Zapus) support. It has reached EOL at Jan 13, 2018.
Added quorum match support. You can use quorum match in both script syntax and query syntax. [groonga-talk,385][Suggested by 付超群]
TODO: Add documents for quorum match syntax and link to them.
Added custom similarity threshold support in script syntax. You can use custom similarity threshold in script syntax.
TODO: Add document for the syntax and link to it.
[grndb][--force-lock-clear] Added
--force-lock-clearoption. With this option,grndbforces to clear locks of database, tables and data columns. You can use your database again even if locks are remained in database, tables and data columns.But this option very risky. Normally, you should not use it. If your database is broken, your database is still broken. This option just ignores locks.
[load] Added surrogate pairs support in escape syntax. For example,
\uD83C\uDF7Ais processed as🍺.[Windows] Changed to use sparse file on Windows. It reduces disk space and there are no performance demerit.
[Online index construction] Added
GRN_II_REDUCE_EXPIRE_THRESHOLDenvironment variable to control when memory maps are expired in index column. It’s-1by default. It means that expire timing is depends on index column size. If index column is smaller, expire timing is more. If index column is larger, expire timing is less.You can use the previous behavior by
0. It means that Groonga always tries to expire.[logical_range_filter] [post_filter] Added a new filter timing. It’s executed after
filteredstage generated columns are generated.
Fixes#
Reduced resource usage for creating index for reference vector. [GitHub#806][Reported by Naoya Murakami]
[table_create] Fixed a bug that a table is created even when
token_filtersis invalid. [GitHub#266]
Thanks#
付超群
Naoya Murakami
Release 7.1.0 - 2017-12-29#
Improvements#
[load] Improved the
load’s query-log format. Added detail below items in theload’s query-log.outputs number of loaded records.
outputs number of error records and columns.
outputs number of total records.
[logical_count] Improved the
logical_count’s query-log format. Added detail below items in thelogical_count’s query-log.outputs number of count.
[logical_select] Improve the
logical_select’s query-log format. Added detail below items in thelogical_select’s query-log.log N outputs.
outputs plain drilldown.
outputs labeled drilldown.
outputs selected in each shard.
use “[…]” for target information.
[delete] Improved the
delete’s query-log format. Added detail below items in thedelete’s query-log.outputs number of deleted and error records.
outputs number of rest number of records.
[Groonga HTTP server] The server executed by
groonga -sensure stopping by C-c.Used
NaNandInfinity,-Infinityinstead of Lisp representations(#<nan>and#i1/0,#-i1/0).Supported vector for drilldown calc target.
Partially supported keyword extraction from regexp search. It enables
highlight_htmlandsnippet_htmlfor regexp search. [GitHub#787][Reported by takagi01][bulk] Reduced the number of
realloc().grn_bulk_*()API supports it.It improves performance for large output case on Windows. For example, it causes 100x faster for 100MB over output.
Because
realloc()is heavy on Windows.Enabled
GRN_II_OVERLAP_TOKEN_SKIP_ENABLEonly when its value is “yes”.Deprecated
GRN_NGRAM_TOKENIZER_REMOVE_BLANK_DISABLE. UseGRN_NGRAM_TOKENIZER_REMOVE_BLANK_ENABLE=noinstead.Added new function
index_column_source_records. It gets source records of index column.[Patch by Naoya Murakami][select] Supported negative “offset” for “offset + size - limit” >= 0
Added
grn_column_cache. It’ll improve performance for getter of fixed size column value.[groonga executable file] Added
--listen-backlog option. You can customizelisten(2)’s backlog by this option.[httpd] Updated bundled nginx to 1.13.8.
Fixes#
Fixed a memory leak in
highlight_fullFixed a crash bug by early unlink It’s not caused by instruction in
grn_expr_parse()but it’s caused when libgroonga user such as Mroonga uses the following instructions:grn_expr_append_const("_id")grn_expr_append_op(GRN_OP_GET_VALUE)
Thanks#
takagi01
Naoya Murakami
Release 7.0.9 - 2017-11-29#
Improvements#
Supported newer version of Apache Arrow. In this release, 0.8.0 or later is required for Apache Arrow support.
[sharding] Added new API for dynamic columns.
Groonga::LabeledArguments
[sharding] Added convenient
Table#select_allmethod.[logical_range_filter] Supported dynamic columns. Note that
initialandfilteredstage are only supported.[logical_range_filter] Added documentation about
cacheparameter and dynamic columns.[logical_count] Supported dynamic columns. Note that
initialstage is only supported.[logical_count] Added documentation about named parameters.
[select] Supported
--match_columns _keywithout index.[in_values] Supported to specify more than 126 values. [GitHub#760] [GitHub#781] [groonga-dev,04449] [Reported by Murata Satoshi]
[httpd] Updated bundled nginx to 1.13.7.
Fixes#
[httpd] Fixed build error when old Groonga is already installed. [GitHub#775] [Reported by myamanishi3]
[in_values] Fixed a bug that
in_valueswith too many arguments can cause a crash. This bug is found during supporting more than 126 values. [GitHub#780][cmake] Fixed LZ4 and MessagePack detection. [Reported by Sergei Golubchik]
[Offline index construction] Fixed a bug that offline index construction for vector column consumes unnecessary resources. If you have a log of elements in one vector column and many records, Groonga will crash. [groonga-dev,04533][Reported by Toshio Uchiyama]
Thanks#
Murata Satoshi
myamanishi3
Sergei Golubchik
Toshio Uchiyama
Release 7.0.8 - 2017-10-29#
Improvements#
[windows] Supported backtrace on crash. This feature not only function call history but also source filename and number of lines can be displayed as much as possible. This feature makes problem solving easier.
Supported
( )(empty block) only query (--query "( )") forQUERY_NO_SYNTAX_ERROR. In the previous version, it caused an error. [GitHub#767]Supported
(+)(only and block) only query (--query "(+)") forQUERY_NO_SYNTAX_ERROR. In the previous version, it caused an error. [GitHub#767]Supported
~foo(starting with “~”) query (--query "~y") forQUERY_NO_SYNTAX_ERROR. In the previous version, it caused an error. [GitHub#767]Modified log level of
expiredfrominfotodebug.2017-10-29 14:05:34.123456|i| <0000000012345678:0> expired i=000000000B123456 max=10 (2/2)This message is logged when memory mapped area for index is unmapped. Thus, this log message is useful information for debugging, in other words, as it is unnecessary information in normal operation, we changed log level frominfotodebug.Supported Ubuntu 17.10 (Artful Aardvark)
Fixes#
[dat] Fixed a bug that large file is created unexpectedly in the worst case during database expansion process. This bug may occurs when you create/delete index columns so frequently. In 7.0.7 release, a related bug was fixed - “
table_createcommand fails when there are many deleted keys”, but it turns out that it is not enough in the worst case.[logical_select] Fixed a bug that when
offsetandlimitwere applied to multiple shards at the same time, there is a case that it returns a fewer number of records unexpectedly.
Release 7.0.7 - 2017-09-29#
Improvements#
Supported
+only query (--query "+") forQUERY_NO_SYNTAX_ERROR. In the previous version, it caused an error.[httpd] Updated bundled nginx to 1.13.5.
[dump] Added the default argument values to the syntax section.
[Command version] Supported
--default-command-version 3.Supported caching select result with function call. Now, most of existing functions supports this feature. There are two exception, when
now()andrand()are used in query, select result will not cached. Because of this default behavior change, new APIs are introduced.grn_proc_set_is_stable()grn_proc_is_stable()
Note that if you add a new function that may return different result with the same argument, you must call
grn_proc_is_stable(ctx, proc, GRN_FALSE). If you don’t call it, select result with the function call is cached and is wrong result for multiple requests.
Fixes#
[windows] Fixed to clean up file handle correctly on failure when
database_unmapis executed. There is a case that critical section is not initialized when request is canceled before executingdatabase_unmap. In such a case, it caused a crach bug.[Tokenizers] Fixed document for wrong tokenizer names. It should be
TokenBigramIgnoreBlankSplitSymbolAlphaandTokenBigramIgnoreBlankSplitSymbolAlphaDigit.Changed not to keep created empty file on error.
In the previous versions, there is a case that empty file keeps remain on error.
Here is the senario to reproduce:
creating new file by grn_fileinfo_open succeeds
mapping file by DO_MAP() is failed
In such a case, it causes an another error such as “already file exists” because of the file which isn’t under control. so these file should be removed during cleanup process.
Fixed a bug that Groonga may be crashed when search process is executed during executing many updates in a short time.
[table_create] Fixed a bug that
table_createfailed when there are many deleted keys.
Release 7.0.6 - 2017-08-29#
Improvements#
Supported prefix match search using multiple indexes. (e.g.
--query "Foo*" --match_columns "TITLE_INDEX_COLUMN||BODY_INDEX_COLUMN").[window_count] Supported
window_countfunction to add count data to result set. It is useful to analyze or filter additionally.Added the following API
grn_obj_get_disk_usage():GRN_EXPR_QUERY_NO_SYNTAX_ERRORgrn_expr_syntax_expand_query_by_table()grn_table_find_reference_object()
[object_inspect] Supported to show disk usage about specified object.
Supported falling back query parse feature. It is enabled when
QUERY_NO_SYNTAX_ERRORflag is set toquery_flags. (this feature is disabled by default). If this flag is set, query never causes syntax error. For example, “A +” is parsed and escaped automatically into “A +”. This behavior is useful when application uses user input directly and doesn’t want to show syntax error to user and in log.Supported to adjust score for term in query. “>”, “<”, and “~” operators are supported. For example, “>Groonga” increments score of “Groonga”, “<Groonga” decrements score of “Groonga”. “~Groonga” decreases score of matched document in the current search result. “~” operator doesn’t change search result itself.
Improved performance to remove table.
thread_limit=1is not needed for it. The process about checking referenced table existence is done without opening objects. As a result, performance is improved.[httpd] Updated bundled nginx to 1.13.4.
Fixes#
[dump] Fixed a bug that the 7-th unnamed parameter for –sort_hash_table option is ignored.
[schema] Fixed a typo in command line parameter name. It should be source instead of sources. [groonga-dev,04449] [Reported by murata satoshi]
[ruby_eval] Fixed crash when ruby_eval returned syntax error. [GitHub#751] [Patch by ryo-pinus]
Thanks#
murata satoshi
ryo-pinus
Release 7.0.5 - 2017-07-29#
Improvements#
[httpd] Updated bundled nginx to 1.13.3. Note that this version contains security fix for CVE-2017-7529.
[load] Supported to load the value of max UInt64. In the previous versions, max UInt64 value is converted into 0 unexpectedlly.
Added the following API
grn_window_get_size()[GitHub#725] [Patch by Naoya Murakami]
[math_abs] Supported
math_abs()function to calculate absolute value. [GitHub#721]Supported to make
grn_default_logger_set_path()andgrn_default_query_logger_set_path()thread safe.[windows] Updated bundled pcre library to 8.41.
[normalize] Improved not to output redundant empty string
""on error. [GitHub#730][functions/time] Supported to show error message when division by zero was happened. [GitHub#733] [Patch by Naoya Murakami]
[windows] Changed to map
ERROR_NO_SYSTEM_RESOURCEStoGRN_RESOURCE_TEMPORARILY_UNAVAILABLE. In the previous versions, it returnsrc=-1as a result code. It is not helpful to investigate what actually happened. With this fix, it returnsrc=-12.[functions/min][functions/max] Supported vector column. Now you need not to care scalar column or vector column to use. [GitHub#735] [Patch by Naoya Murakami]
[dump] Supported
--sort_hash_tableoption to sort by_keyfor hash table. Specify--sort_hash_table yesto use it.[between] Supported to specify index column. [GitHub#740] [Patch by Naoya Murakami]
[load] Supported Apache Arrow 0.5.0 or later.
[How to analyze error messages] Added howto article to analyze error message in Groonga.
[Debian GNU/Linux] Updated required package list to build from source.
[Ubuntu] Dropped Ubuntu 16.10 (Yakkety Yak) support. It has reached EOL at July 20, 2017.
Fixes#
Fixed to construct correct fulltext indexes against vector column which type belongs to text family (
`ShortTextand so on). This fix resolves that fulltext search doesn’t work well against text vector column after updating indexes. [GitHub#494][thread_limit] Fixed a bug that deadlock occurs when thread_limit?max=1 is requested at once.
[groonga-httpd] Fixed a mismatch path of pid file between default one and restart command assumed. This mismatch blocked restarting groonga-httpd. [GitHub#743] [Reported by sozaki]
Thanks#
Naoya Murakami
Release 7.0.4 - 2017-06-29#
Improvements#
Added physical create/delete operation logs to identify problem for troubleshooting. [GitHub#700,#701]
[in_records] Improved performance for fixed sized column. It may reduce 50% execution time.
[grndb] Added
--log-pathoption. [GitHub#702,#703][grndb] Added
--log-leveloption. [GitHub#706,#708]Added the following API
grn_operator_to_exec_func()grn_obj_is_corrupt()
Improved performance for “FIXED_SIZE_COLUMN OP CONSTANT”. Supported operators are:
==,!=,<,>,<=and>=.Improved performance for “COLUMN OP VALUE && COLUMN OP VALUE && …”.
[grndb] Supported corrupted object detection with
grndb check.[io_flush] Supported
--only_openedoption which enables to flush only opened database objects.[grndb] Supported to detect/delete orphan “inspect” object. The orphaned “inspect” object is created by renamed command name from
inspecttoobject_inspect.
Fixes#
[rpm][centos] Fixed unexpected macro expansion problem with customized build. This bug only affects when rebuilding Groonga SRPM with customized
additional_configure_optionsparameter in spec file.Fixed missing null check for
grn_table_setoperation(). There is a possibility of crash bug when indexes are broken. [GitHub#699]
Thanks#
Release 7.0.3 - 2017-05-29#
Improvements#
[select] Add document about Full text search with specific index name.
[index] Supported to log warning message which record causes posting list overflows.
[cmake] Supported linking lz4 in embedded static library build. [Original patch by Sergei Golubchik]
[delete] Supported to cancel.
[httpd] Updated bundled nginx to 1.13.0
Exported the following API
grn_plugin_proc_get_caller()
Added index column related function and selector.
Added new selector: index_column_df_ratio_between()
Added new function: index_column_df_ratio()
Fixes#
[delete] Fixed a bug that error isn’t cleared correctly. It affects to following deletions so that it causes unexpected behavior.
[windows] Fixed a bug that IO version is not detected correctly when the file is opened with
O_CREATflag.[vector_slice] Fixed a bug that non 4 bytes vector columns can’t slice. [GitHub#695] [Patch by Naoya Murakami]
Fixed a bug that non 4 bytes fixed vector column can’t sequential match by specifying index of vector. [GitHub#696] [Patch by Naoya Murakami]
[logical_select] Fixed a bug that “argument out of range” occurs when setting last day of month to the min. [GitHub#698]
Thanks#
Sergei Golubchik
Naoya Murakami
Release 7.0.2 - 2017-04-29#
Improvements#
[logical_select] Supported multiple drilldowns[${LABEL}].columns[${NAME}].window.sort_keys and drilldowns[${LABEL}].columns[${NAME}].window.group_keys.
[windows] Updated bundled LZ4 to 1.7.5.
[cache] Supported persistent cache feature.
[log_level] Update English documentation.
Added the following APIs:
grn_set_default_cache_base_path()grn_get_default_cache_base_path()grn_persistent_cache_open()grn_cache_default_open()
[
groonga --cache-base-path] Added a new option to use persistent cache.[groonga-httpd] [groonga_cache_base_path] Added new configuration to use persistent cache.
[windows] Updated bundled msgpack to 2.1.1.
[object_inspect] Supported not only column inspection, but also index column statistics.
Supported index search for “
.*” regexp pattern. This feature is enabled by default. SetGRN_SCAN_INFO_REGEXP_DOT_ASTERISK_ENABLE=noenvironment variable to disable this feature.[in_records] Added function to use an existing table as condition patterns.
[Ubuntu] Dropped Ubuntu 12.04 (Precise Pangolin) support because of EOL.
Fixes#
[logical_select] Fixed a bug that wrong cache is used. This bug was occurred when dynamic column parameter is used.
[logical_select] Fixed a bug that dynamic columns aren’t created. It’s occurred when no match case.
[reindex] Fixed a bug that data is lost by reindex. [GitHub#646]
[httpd] Fixed a bug that response of quit and shutdown is broken JSON when worker is running as another user. [GitHub ranguba/groonga-client#12]
Release 7.0.1 - 2017-03-29#
Improvements#
Exported the following API
grn_ii_cursor_next_pos()
grn_table_apply_expr()
grn_obj_is_data_column()
grn_obj_is_expr()
grn_obj_is_scalar_column()
[dump] Supported to dump weight reference vector.
[load] Supported to load
array<object>style weight vector column. The example ofarray<object>style is:[{"key1": weight1}, {"key2": weight2}].Supported to search
!(XXX OPERATOR VALUE)by index. Supported operator is not only>but also>=,<,<=,==and!=.Supported index search for “!(column == CONSTANT)”. The example in this case is:
!(column == 29)and so on.Supported more “!” optimization in the following patterns.
!(column @ "X") && (column @ "Y")(column @ "Y") && !(column @ "X")(column @ "Y") &! !(column @ "X")
Supported to search
XXX || !(column @ "xxx")by index.[dump] Changed to use
'{"x": 1, "y": 2}'style for not referenced weight vector. This change doesn’t affect to old Groonga because it already supports one.[experimental] Supported
GRN_ORDER_BY_ESTIMATED_SIZE_ENABLEenvironment variable. This variable controls whether query optimization which is based on estimated size is applied or not. This feature is disabled by default. SetGRN_ORDER_BY_ESTIMATED_SIZE_ENABLE=yesif you want to try it.[select] Added query log for
columns,drilldownevaluation.[select] Changed query log format for
drilldown. This is backward incompatible change, but it only affects users who convert query log by own programs.[table_remove] Reduced temporary memory usage. It’s enabled when the number of max threads is 0.
[select]
columns[LABEL](N)is used for query log format instead ofcolumns(N)[LABEL]..[Query expansion] Updated example to use vector column because it is recommended way. [Reported by Gurunavi, Inc]
Supported to detect canceled request while locking. It fixes the problem that
request_cancelis ignored unexpectedly while locking.[logical_select] Supported
initialandfilteredstage dynamic columns. The examples are:--columns[LABEL].stage initialor--columns[LABEL].stage filtered.[logical_select] Supported
match_columns,queryanddrilldown_filteroption.[highlight_html] Supported similar search.
[logical_select] Supported
initialand stage dynamic columns in labeled drilldown. The example is:--drilldowns[LABEL].stage initial.[logical_select] Supported window function in dynamic column.
[select] Added documentation about dynamic columns.
[Window function] Added section about window functions.
[CentOS] Dropped CentOS 5 support because of EOL.
[httpd] Updated bundled nginx to 1.11.12
Supported to disable AND match optimization by environment variable. You can disable this feature by
GRN_TABLE_SELECT_AND_MIN_SKIP_ENABLE=no. This feature is enable by default.[vector_new] Added a new function to create a new vector.
[select] Added documentation about
drilldown_filter.
Fixes#
[lock_clear] Fixed a crash bug against temporary database.
Fixed a problem that dynamically updated index size was increased for natural language since Grooonga 6.1.4.
[select] Fixed a bug that “A && B.C @ X” may not return records that should be matched.
Fixed a conflict with
grn_io_flush()andgrn_io_expire(). Without this change, ifio_flushandloadcommand are executed simultaneously in specific timing, it causes a crash bug by access violation.[logical_table_remove] Fixed a crash bug when the max number of threads is 1.
Thanks#
Gurunavi, Inc.
Release 7.0.0 - 2017-02-09#
Improvements#
[in_values] Supported sequential search for reference vector column. [Patch by Naoya Murakami] [GitHub#629]
[select] Changed to report error instead of ignoring on invalid
drilldown[LABEL].sort_keys.[select] Removed needless metadata updates on DB. It reduces the case that database lock remains even though
selectcommand is executed. [Reported by aomi-n][lock_clear] Changed to clear metadata lock by lock_clear against DB.
[CentOS] Enabled EPEL by default to install Groonga on Amazon Linux.
[query] Supported “@X” style in script syntax for prefix(“@^”), suffix(“@$”), regexp(“@^”) search.
[