Add new supported languages by updating lunr-languages to 1.12.0

Bump from 1.4.0 to 1.12.0

This adds support for:

- hi
- hy
- kn
- ko
- sa
- ta
- te
- zh
This commit is contained in:
Ultrabug
2023-08-18 15:51:45 +02:00
committed by GitHub
parent c8bd1aff53
commit 9cc8f24981
13 changed files with 972 additions and 25 deletions

View File

@@ -143,7 +143,7 @@
var wordCharacters = "\u0621-\u065b\u0671\u0640";
var testRegex = new RegExp("[^" + wordCharacters + "]");
self.word = self.word
.replace('\u0640', '');
.replace(new RegExp('\u0640', 'g'), '');
if (testRegex.test(word)) {
return true;
}

View File

@@ -0,0 +1,123 @@
/*!
* Lunr languages, `Hindi` language
* https://github.com/MiKr13/lunr-languages
*
* Copyright 2020, Mihir Kumar
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.hi = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.hi.trimmer,
lunr.hi.stopWordFilter,
lunr.hi.stemmer
);
// change the tokenizer for japanese one
// if (isLunr2) { // for lunr version 2.0.0
// this.tokenizer = lunr.hi.tokenizer;
// } else {
// if (lunr.tokenizer) { // for lunr version 0.6.0
// lunr.tokenizer = lunr.hi.tokenizer;
// }
// if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
// this.tokenizerFn = lunr.hi.tokenizer;
// }
// }
if (this.searchPipeline) {
this.searchPipeline.reset();
this.searchPipeline.add(lunr.hi.stemmer)
}
};
/* lunr trimmer function */
lunr.hi.wordCharacters = "\u0900-\u0903\u0904-\u090f\u0910-\u091f\u0920-\u092f\u0930-\u093f\u0940-\u094f\u0950-\u095f\u0960-\u096f\u0970-\u097fa-zA-Z--0-9-";
// lunr.hi.wordCharacters = "ऀँंऄअआइईउऊऋऌऍऎएऐऑऒओऔकखगघङचछजझञटठडढणतथदधनऩपफबभमयरऱलळऴवशषसहऺऻ़ऽािीुूृॄॅॆेैॉॊोौ्ॎॏॐ॒॑॓॔ॕॖॗक़ख़ग़ज़ड़ढ़फ़य़ॠॡॢॣ।॥१२३४५६७८९॰ॱॲॳॴॵॶॷॸॹॺॻॼॾॿa-zA-Z--0-9-";
lunr.hi.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.hi.wordCharacters);
lunr.Pipeline.registerFunction(lunr.hi.trimmer, 'trimmer-hi');
/* lunr stop word filter */
lunr.hi.stopWordFilter = lunr.generateStopWordFilter(
'अत अपना अपनी अपने अभी अंदर आदि आप इत्यादि इन इनका इन्हीं इन्हें इन्हों इस इसका इसकी इसके इसमें इसी इसे उन उनका उनकी उनके उनको उन्हीं उन्हें उन्हों उस उसके उसी उसे एक एवं एस ऐसे और कई कर करता करते करना करने करें कहते कहा का काफ़ी कि कितना किन्हें किन्हों किया किर किस किसी किसे की कुछ कुल के को कोई कौन कौनसा गया घर जब जहाँ जा जितना जिन जिन्हें जिन्हों जिस जिसे जीधर जैसा जैसे जो तक तब तरह तिन तिन्हें तिन्हों तिस तिसे तो था थी थे दबारा दिया दुसरा दूसरे दो द्वारा न नके नहीं ना निहायत नीचे ने पर पहले पूरा पे फिर बनी बही बहुत बाद बाला बिलकुल भी भीतर मगर मानो मे में यदि यह यहाँ यही या यिह ये रखें रहा रहे ऱ्वासा लिए लिये लेकिन व वग़ैरह वर्ग वह वहाँ वहीं वाले वुह वे वो सकता सकते सबसे सभी साथ साबुत साभ सारा से सो संग ही हुआ हुई हुए है हैं हो होता होती होते होना होने'.split(' '));
/* lunr stemmer function */
lunr.hi.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
var segmenter = lunr.wordcut;
segmenter.init();
lunr.hi.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
});
var str = obj.toString().toLowerCase().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
lunr.Pipeline.registerFunction(lunr.hi.stemmer, 'stemmer-hi');
lunr.Pipeline.registerFunction(lunr.hi.stopWordFilter, 'stopWordFilter-hi');
};
}))

View File

@@ -0,0 +1,98 @@
/*!
* Lunr languages, `Armenian` language
* https://github.com/turbobit/lunr-languages
*
* Copyright 2021, Manikandan Venkatasubban
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.hy = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.hy.trimmer,
lunr.hy.stopWordFilter
);
};
/* lunr trimmer function */
// http://www.unicode.org/charts/
lunr.hy.wordCharacters = "[" +
"A-Za-z" +
"\u0530-\u058F" + // armenian alphabet
"\uFB00-\uFB4F" + // armenian ligatures
"]";
lunr.hy.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.hy.wordCharacters);
lunr.Pipeline.registerFunction(lunr.hy.trimmer, 'trimmer-hy');
/* lunr stop word filter */
// https://www.ranks.nl/stopwords/armenian
lunr.hy.stopWordFilter = lunr.generateStopWordFilter('դու և եք էիր էիք հետո նաև նրանք որը վրա է որ պիտի են այս մեջ ն իր ու ի այդ որոնք այն կամ էր մի ես համար այլ իսկ էին ենք հետ ին թ էինք մենք նրա նա դուք եմ էի ըստ որպես ում'.split(' '));
lunr.Pipeline.registerFunction(lunr.hy.stopWordFilter, 'stopWordFilter-hy');
/* lunr stemmer function */
lunr.hy.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
lunr.Pipeline.registerFunction(lunr.hy.stemmer, 'stemmer-hy');
};
}))

View File

@@ -0,0 +1,110 @@
/*!
* Lunr languages, `Kannada` language
* https://github.com/MiKr13/lunr-languages
*
* Copyright 2023, India
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.kn = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.kn.trimmer,
lunr.kn.stopWordFilter,
lunr.kn.stemmer
);
if (this.searchPipeline) {
this.searchPipeline.reset();
this.searchPipeline.add(lunr.kn.stemmer)
}
};
/* lunr trimmer function */
lunr.kn.wordCharacters = "\u0C80-\u0C84\u0C85-\u0C94\u0C95-\u0CB9\u0CBE-\u0CCC\u0CBC-\u0CBD\u0CD5-\u0CD6\u0CDD-\u0CDE\u0CE0-\u0CE1\u0CE2-\u0CE3\u0CE4\u0CE5\u0CE6-\u0CEF\u0CF1-\u0CF3";
lunr.kn.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.kn.wordCharacters);
lunr.Pipeline.registerFunction(lunr.kn.trimmer, 'trimmer-kn');
/* lunr stop word filter */
lunr.kn.stopWordFilter = lunr.generateStopWordFilter(
'ಮತ್ತು ಈ ಒಂದು ರಲ್ಲಿ ಹಾಗೂ ಎಂದು ಅಥವಾ ಇದು ರ ಅವರು ಎಂಬ ಮೇಲೆ ಅವರ ತನ್ನ ಆದರೆ ತಮ್ಮ ನಂತರ ಮೂಲಕ ಹೆಚ್ಚು ನ ಆ ಕೆಲವು ಅನೇಕ ಎರಡು ಹಾಗು ಪ್ರಮುಖ ಇದನ್ನು ಇದರ ಸುಮಾರು ಅದರ ಅದು ಮೊದಲ ಬಗ್ಗೆ ನಲ್ಲಿ ರಂದು ಇತರ ಅತ್ಯಂತ ಹೆಚ್ಚಿನ ಸಹ ಸಾಮಾನ್ಯವಾಗಿ ನೇ ಹಲವಾರು ಹೊಸ ದಿ ಕಡಿಮೆ ಯಾವುದೇ ಹೊಂದಿದೆ ದೊಡ್ಡ ಅನ್ನು ಇವರು ಪ್ರಕಾರ ಇದೆ ಮಾತ್ರ ಕೂಡ ಇಲ್ಲಿ ಎಲ್ಲಾ ವಿವಿಧ ಅದನ್ನು ಹಲವು ರಿಂದ ಕೇವಲ ದ ದಕ್ಷಿಣ ಗೆ ಅವನ ಅತಿ ನೆಯ ಬಹಳ ಕೆಲಸ ಎಲ್ಲ ಪ್ರತಿ ಇತ್ಯಾದಿ ಇವು ಬೇರೆ ಹೀಗೆ ನಡುವೆ ಇದಕ್ಕೆ ಎಸ್ ಇವರ ಮೊದಲು ಶ್ರೀ ಮಾಡುವ ಇದರಲ್ಲಿ ರೀತಿಯ ಮಾಡಿದ ಕಾಲ ಅಲ್ಲಿ ಮಾಡಲು ಅದೇ ಈಗ ಅವು ಗಳು ಎ ಎಂಬುದು ಅವನು ಅಂದರೆ ಅವರಿಗೆ ಇರುವ ವಿಶೇಷ ಮುಂದೆ ಅವುಗಳ ಮುಂತಾದ ಮೂಲ ಬಿ ಮೀ ಒಂದೇ ಇನ್ನೂ ಹೆಚ್ಚಾಗಿ ಮಾಡಿ ಅವರನ್ನು ಇದೇ ಯ ರೀತಿಯಲ್ಲಿ ಜೊತೆ ಅದರಲ್ಲಿ ಮಾಡಿದರು ನಡೆದ ಆಗ ಮತ್ತೆ ಪೂರ್ವ ಆತ ಬಂದ ಯಾವ ಒಟ್ಟು ಇತರೆ ಹಿಂದೆ ಪ್ರಮಾಣದ ಗಳನ್ನು ಕುರಿತು ಯು ಆದ್ದರಿಂದ ಅಲ್ಲದೆ ನಗರದ ಮೇಲಿನ ಏಕೆಂದರೆ ರಷ್ಟು ಎಂಬುದನ್ನು ಬಾರಿ ಎಂದರೆ ಹಿಂದಿನ ಆದರೂ ಆದ ಸಂಬಂಧಿಸಿದ ಮತ್ತೊಂದು ಸಿ ಆತನ '.split(' '));
/* lunr stemmer function */
lunr.kn.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
var segmenter = lunr.wordcut;
segmenter.init();
lunr.kn.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
});
var str = obj.toString().toLowerCase().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
lunr.Pipeline.registerFunction(lunr.kn.stemmer, 'stemmer-kn');
lunr.Pipeline.registerFunction(lunr.kn.stopWordFilter, 'stopWordFilter-kn');
};
}))

File diff suppressed because one or more lines are too long

View File

@@ -490,9 +490,9 @@
if (!sbp.eq_s_b(1, "u"))
break;
}
case 2:
sbp.slice_del();
break;
case 2:
sbp.slice_del();
break;
}
}
sbp.limit_backward = v_1;

View File

@@ -249,9 +249,9 @@
if (!sbp.eq_s_b(1, "\u044F"))
return false;
}
case 2:
sbp.slice_del();
break;
case 2:
sbp.slice_del();
break;
}
return true;
}

View File

@@ -0,0 +1,110 @@
/*!
* Lunr languages, `Sanskrit` language
* https://github.com/MiKr13/lunr-languages
*
* Copyright 2023, India
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.sa = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.sa.trimmer,
lunr.sa.stopWordFilter,
lunr.sa.stemmer
);
if (this.searchPipeline) {
this.searchPipeline.reset();
this.searchPipeline.add(lunr.sa.stemmer)
}
};
/* lunr trimmer function */
lunr.sa.wordCharacters = "\u0900-\u0903\u0904-\u090f\u0910-\u091f\u0920-\u092f\u0930-\u093f\u0940-\u094f\u0950-\u095f\u0960-\u096f\u0970-\u097f\uA8E0-\uA8F1\uA8F2-\uA8F7\uA8F8-\uA8FB\uA8FC-\uA8FD\uA8FE-\uA8FF\u11B00-\u11B09";
lunr.sa.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.sa.wordCharacters);
lunr.Pipeline.registerFunction(lunr.sa.trimmer, 'trimmer-sa');
/* lunr stop word filter */
lunr.sa.stopWordFilter = lunr.generateStopWordFilter(
'तथा अयम्‌ एकम्‌ इत्यस्मिन्‌ तथा तत्‌ वा अयम्‌ इत्यस्य ते आहूत उपरि तेषाम्‌ किन्तु तेषाम्‌ तदा इत्यनेन अधिकः इत्यस्य तत्‌ केचन बहवः द्वि तथा महत्वपूर्णः अयम्‌ अस्य विषये अयं अस्ति तत्‌ प्रथमः विषये इत्युपरि इत्युपरि इतर अधिकतमः अधिकः अपि सामान्यतया ठ इतरेतर नूतनम्‌ द न्यूनम्‌ कश्चित्‌ वा विशालः द सः अस्ति तदनुसारम् तत्र अस्ति केवलम्‌ अपि अत्र सर्वे विविधाः तत्‌ बहवः यतः इदानीम्‌ द दक्षिण इत्यस्मै तस्य उपरि नथ अतीव कार्यम्‌ सर्वे एकैकम्‌ इत्यादि। एते सन्ति उत इत्थम्‌ मध्ये एतदर्थं . स कस्य प्रथमः श्री. करोति अस्मिन् प्रकारः निर्मिता कालः तत्र कर्तुं समान अधुना ते सन्ति स एकः अस्ति सः अर्थात् तेषां कृते . स्थितम् विशेषः अग्रिम तेषाम्‌ समान स्रोतः ख म समान इदानीमपि अधिकतया करोतु ते समान इत्यस्य वीथी सह यस्मिन् कृतवान्‌ धृतः तदा पुनः पूर्वं सः आगतः किम्‌ कुल इतर पुरा मात्रा स विषये उ अतएव अपि नगरस्य उपरि यतः प्रतिशतं कतरः कालः साधनानि भूत तथापि जात सम्बन्धि अन्यत्‌ ग अतः अस्माकं स्वकीयाः अस्माकं इदानीं अन्तः इत्यादयः भवन्तः इत्यादयः एते एताः तस्य अस्य इदम् एते तेषां तेषां तेषां तान् तेषां तेषां तेषां समानः सः एकः च तादृशाः बहवः अन्ये च वदन्ति यत् कियत् कस्मै कस्मै यस्मै यस्मै यस्मै यस्मै न अतिनीचः किन्तु प्रथमं सम्पूर्णतया ततः चिरकालानन्तरं पुस्तकं सम्पूर्णतया अन्तः किन्तु अत्र वा इह इव श्रद्धाय अवशिष्यते परन्तु अन्ये वर्गाः सन्ति ते सन्ति शक्नुवन्ति सर्वे मिलित्वा सर्वे एकत्र"'.split(' '));
/* lunr stemmer function */
lunr.sa.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
var segmenter = lunr.wordcut;
segmenter.init();
lunr.sa.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
});
var str = obj.toString().toLowerCase().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
lunr.Pipeline.registerFunction(lunr.sa.stemmer, 'stemmer-sa');
lunr.Pipeline.registerFunction(lunr.sa.stopWordFilter, 'stopWordFilter-sa');
};
}))

View File

@@ -0,0 +1,123 @@
/*!
* Lunr languages, `Tamil` language
* https://github.com/tvmani/lunr-languages
*
* Copyright 2021, Manikandan Venkatasubban
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.ta = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.ta.trimmer,
lunr.ta.stopWordFilter,
lunr.ta.stemmer
);
// change the tokenizer for japanese one
// if (isLunr2) { // for lunr version 2.0.0
// this.tokenizer = lunr.ta.tokenizer;
// } else {
// if (lunr.tokenizer) { // for lunr version 0.6.0
// lunr.tokenizer = lunr.ta.tokenizer;
// }
// if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
// this.tokenizerFn = lunr.ta.tokenizer;
// }
// }
if (this.searchPipeline) {
this.searchPipeline.reset();
this.searchPipeline.add(lunr.ta.stemmer)
}
};
/* lunr trimmer function */
lunr.ta.wordCharacters = "\u0b80-\u0b89\u0b8a-\u0b8f\u0b90-\u0b99\u0b9a-\u0b9f\u0ba0-\u0ba9\u0baa-\u0baf\u0bb0-\u0bb9\u0bba-\u0bbf\u0bc0-\u0bc9\u0bca-\u0bcf\u0bd0-\u0bd9\u0bda-\u0bdf\u0be0-\u0be9\u0bea-\u0bef\u0bf0-\u0bf9\u0bfa-\u0bffa-zA-Z--0-9-";
lunr.ta.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.ta.wordCharacters);
lunr.Pipeline.registerFunction(lunr.ta.trimmer, 'trimmer-ta');
/* lunr stop word filter */
lunr.ta.stopWordFilter = lunr.generateStopWordFilter(
'அங்கு அங்கே அது அதை அந்த அவர் அவர்கள் அவள் அவன் அவை ஆக ஆகவே ஆகையால் ஆதலால் ஆதலினால் ஆனாலும் ஆனால் இங்கு இங்கே இது இதை இந்த இப்படி இவர் இவர்கள் இவள் இவன் இவை இவ்வளவு உனக்கு உனது உன் உன்னால் எங்கு எங்கே எது எதை எந்த எப்படி எவர் எவர்கள் எவள் எவன் எவை எவ்வளவு எனக்கு எனது எனவே என் என்ன என்னால் ஏது ஏன் தனது தன்னால் தானே தான் நாங்கள் நாம் நான் நீ நீங்கள்'.split(' '));
/* lunr stemmer function */
lunr.ta.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
var segmenter = lunr.wordcut;
segmenter.init();
lunr.ta.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
});
var str = obj.toString().toLowerCase().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
lunr.Pipeline.registerFunction(lunr.ta.stemmer, 'stemmer-ta');
lunr.Pipeline.registerFunction(lunr.ta.stopWordFilter, 'stopWordFilter-ta');
};
}))

View File

@@ -0,0 +1,122 @@
/*!
* Lunr languages, `Hindi` language
* https://github.com/MiKr13/lunr-languages
*
* Copyright 2023, India
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball JavaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory()
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function() {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/* register specific locale function */
lunr.te = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.te.trimmer,
lunr.te.stopWordFilter,
lunr.te.stemmer
);
// change the tokenizer for japanese one
// if (isLunr2) { // for lunr version 2.0.0
// this.tokenizer = lunr.hi.tokenizer;
// } else {
// if (lunr.tokenizer) { // for lunr version 0.6.0
// lunr.tokenizer = lunr.hi.tokenizer;
// }
// if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
// this.tokenizerFn = lunr.hi.tokenizer;
// }
// }
if (this.searchPipeline) {
this.searchPipeline.reset();
this.searchPipeline.add(lunr.te.stemmer)
}
};
/* lunr trimmer function */
lunr.te.wordCharacters = "\u0C00-\u0C04\u0C05-\u0C14\u0C15-\u0C39\u0C3E-\u0C4C\u0C55-\u0C56\u0C58-\u0C5A\u0C60-\u0C61\u0C62-\u0C63\u0C66-\u0C6F\u0C78-\u0C7F\u0C3C\u0C3D\u0C4D\u0C5D\u0C77\u0C64\u0C65";
lunr.te.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.te.wordCharacters);
lunr.Pipeline.registerFunction(lunr.te.trimmer, 'trimmer-te');
/* lunr stop word filter */
lunr.te.stopWordFilter = lunr.generateStopWordFilter(
'అందరూ అందుబాటులో అడగండి అడగడం అడ్డంగా అనుగుణంగా అనుమతించు అనుమతిస్తుంది అయితే ఇప్పటికే ఉన్నారు ఎక్కడైనా ఎప్పుడు ఎవరైనా ఎవరో ఏ ఏదైనా ఏమైనప్పటికి ఒక ఒకరు కనిపిస్తాయి కాదు కూడా గా గురించి చుట్టూ చేయగలిగింది తగిన తర్వాత దాదాపు దూరంగా నిజంగా పై ప్రకారం ప్రక్కన మధ్య మరియు మరొక మళ్ళీ మాత్రమే మెచ్చుకో వద్ద వెంట వేరుగా వ్యతిరేకంగా సంబంధం'.split(' '));
/* lunr stemmer function */
lunr.te.stemmer = (function() {
return function(word) {
// for lunr version 2
if (typeof word.update === "function") {
return word.update(function(word) {
return word;
})
} else { // for lunr version <= 1
return word;
}
}
})();
var segmenter = lunr.wordcut;
segmenter.init();
lunr.te.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
});
var str = obj.toString().toLowerCase().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
lunr.Pipeline.registerFunction(lunr.te.stemmer, 'stemmer-te');
lunr.Pipeline.registerFunction(lunr.te.stopWordFilter, 'stopWordFilter-te');
};
}))

View File

@@ -60,22 +60,22 @@
/* register specific locale function */
lunr.th = function() {
this.pipeline.reset();
this.pipeline.add(
/*lunr.th.stopWordFilter,*/
lunr.th.trimmer
);
this.pipeline.reset();
this.pipeline.add(
/*lunr.th.stopWordFilter,*/
lunr.th.trimmer
);
if (isLunr2) { // for lunr version 2.0.0
this.tokenizer = lunr.th.tokenizer;
} else {
if (lunr.tokenizer) { // for lunr version 0.6.0
lunr.tokenizer = lunr.th.tokenizer;
}
if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
this.tokenizerFn = lunr.th.tokenizer;
}
if (isLunr2) { // for lunr version 2.0.0
this.tokenizer = lunr.th.tokenizer;
} else {
if (lunr.tokenizer) { // for lunr version 0.6.0
lunr.tokenizer = lunr.th.tokenizer;
}
if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
this.tokenizerFn = lunr.th.tokenizer;
}
}
};
/* lunr trimmer function */
@@ -85,13 +85,15 @@
var segmenter = lunr.wordcut;
segmenter.init();
lunr.th.tokenizer = function (obj) {
lunr.th.tokenizer = function(obj) {
//console.log(obj);
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function (t) { return isLunr2 ? new lunr.Token(t) : t })
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t) : t
})
var str = obj.toString().replace(/^\s+/, '');
return segmenter.cut(str).split('|');
}
};
}))
}))

View File

@@ -0,0 +1,145 @@
/*!
* Lunr languages, `Chinese` language
* https://github.com/MihaiValentin/lunr-languages
*
* Copyright 2019, Felix Lian (repairearth)
* http://www.mozilla.org/MPL/
*/
/*!
* based on
* Snowball zhvaScript Library v0.3
* http://code.google.com/p/urim/
* http://snowball.tartarus.org/
*
* Copyright 2010, Oleg Mazko
* http://www.mozilla.org/MPL/
*/
/**
* export the module via AMD, CommonJS or as a browser global
* Export code from https://github.com/umdjs/umd/blob/master/returnExports.js
*/
;
(function(root, factory) {
if (typeof define === 'function' && define.amd) {
// AMD. Register as an anonymous module.
define(factory)
} else if (typeof exports === 'object') {
/**
* Node. Does not work with strict CommonJS, but
* only CommonJS-like environments that support module.exports,
* like Node.
*/
module.exports = factory(require('@node-rs/jieba'))
} else {
// Browser globals (root is window)
factory()(root.lunr);
}
}(this, function(nodejieba) {
/**
* Just return a value to define the module export.
* This example returns an object, but the module
* can return a function as the exported value.
*/
return function(lunr, nodejiebaDictJson) {
/* throw error if lunr is not yet included */
if ('undefined' === typeof lunr) {
throw new Error('Lunr is not present. Please include / require Lunr before this script.');
}
/* throw error if lunr stemmer support is not yet included */
if ('undefined' === typeof lunr.stemmerSupport) {
throw new Error('Lunr stemmer support is not present. Please include / require Lunr stemmer support before this script.');
}
/*
Chinese tokenization is trickier, since it does not
take into account spaces.
Since the tokenization function is represented different
internally for each of the Lunr versions, this had to be done
in order to try to try to pick the best way of doing this based
on the Lunr version
*/
var isLunr2 = lunr.version[0] == "2";
/* register specific locale function */
lunr.zh = function() {
this.pipeline.reset();
this.pipeline.add(
lunr.zh.trimmer,
lunr.zh.stopWordFilter,
lunr.zh.stemmer
);
// change the tokenizer for Chinese one
if (isLunr2) { // for lunr version 2.0.0
this.tokenizer = lunr.zh.tokenizer;
} else {
if (lunr.tokenizer) { // for lunr version 0.6.0
lunr.tokenizer = lunr.zh.tokenizer;
}
if (this.tokenizerFn) { // for lunr version 0.7.0 -> 1.0.0
this.tokenizerFn = lunr.zh.tokenizer;
}
}
};
lunr.zh.tokenizer = function(obj) {
if (!arguments.length || obj == null || obj == undefined) return []
if (Array.isArray(obj)) return obj.map(function(t) {
return isLunr2 ? new lunr.Token(t.toLowerCase()) : t.toLowerCase()
})
nodejiebaDictJson && nodejieba.load(nodejiebaDictJson)
var str = obj.toString().trim().toLowerCase();
var tokens = [];
nodejieba.cut(str, true).forEach(function(seg) {
tokens = tokens.concat(seg.split(' '))
})
tokens = tokens.filter(function(token) {
return !!token;
});
var fromIndex = 0
return tokens.map(function(token, index) {
if (isLunr2) {
var start = str.indexOf(token, fromIndex)
var tokenMetadata = {}
tokenMetadata["position"] = [start, token.length]
tokenMetadata["index"] = index
fromIndex = start
return new lunr.Token(token, tokenMetadata);
} else {
return token
}
});
}
/* lunr trimmer function */
lunr.zh.wordCharacters = "\\w\u4e00-\u9fa5";
lunr.zh.trimmer = lunr.trimmerSupport.generateTrimmer(lunr.zh.wordCharacters);
lunr.Pipeline.registerFunction(lunr.zh.trimmer, 'trimmer-zh');
/* lunr stemmer function */
lunr.zh.stemmer = (function() {
/* TODO Chinese stemmer */
return function(word) {
return word;
}
})();
lunr.Pipeline.registerFunction(lunr.zh.stemmer, 'stemmer-zh');
/* lunr stop word filter. see https://www.ranks.nl/stopwords/chinese-stopwords */
lunr.zh.stopWordFilter = lunr.generateStopWordFilter(
'的 一 不 在 人 有 是 为 為 以 于 於 上 他 而 后 後 之 来 來 及 了 因 下 可 到 由 这 這 与 與 也 此 但 并 並 个 個 其 已 无 無 小 我 们 們 起 最 再 今 去 好 只 又 或 很 亦 某 把 那 你 乃 它 吧 被 比 别 趁 当 當 从 從 得 打 凡 儿 兒 尔 爾 该 該 各 给 給 跟 和 何 还 還 即 几 幾 既 看 据 據 距 靠 啦 另 么 麽 每 嘛 拿 哪 您 凭 憑 且 却 卻 让 讓 仍 啥 如 若 使 谁 誰 虽 雖 随 隨 同 所 她 哇 嗡 往 些 向 沿 哟 喲 用 咱 则 則 怎 曾 至 致 着 著 诸 諸 自'.split(' '));
lunr.Pipeline.registerFunction(lunr.zh.stopWordFilter, 'stopWordFilter-zh');
};
}))

View File

@@ -70,7 +70,7 @@ class SearchConfigTests(unittest.TestCase):
def test_lang_missing_and_with_territory(self):
option = search.LangOption()
value = option.validate(['zh_CN', 'pt_BR', 'fr'])
value = option.validate(['cs_CZ', 'pt_BR', 'fr'])
self.assertEqual(['fr', 'en', 'pt'], value)