franc-min - npm Package Compare versions

Comparing version 3.1.1 to 4.0.0

expressions.js

@@ -23,2 +23,2 @@ // This file is generated by `build.js`.
   ell: /[\u0370-\u0373\u0375-\u0377\u037A-\u037D\u037F\u0384\u0386\u0388-\u038A\u038C\u038E-\u03A1\u03A3-\u03E1\u03F0-\u03FF\u1D26-\u1D2A\u1D5D-\u1D61\u1D66-\u1D6A\u1DBF\u1F00-\u1F15\u1F18-\u1F1D\u1F20-\u1F45\u1F48-\u1F4D\u1F50-\u1F57\u1F59\u1F5B\u1F5D\u1F5F-\u1F7D\u1F80-\u1FB4\u1FB6-\u1FC4\u1FC6-\u1FD3\u1FD6-\u1FDB\u1FDD-\u1FEF\u1FF2-\u1FF4\u1FF6-\u1FFE\u2126\uAB65]|\uD800[\uDD40-\uDD8C\uDDA0]|\uD834[\uDE00-\uDE45]/g
-};
+}

index.js

 // This file is generated by `build.js`
-'use strict';
+'use strict'
 
 /* Load `trigram-utils`. */
-var utilities = require('trigram-utils');
+var utilities = require('trigram-utils')
 
 /* Load `expressions` (regular expressions matching
  * scripts). */
-var expressions = require('./expressions.js');
+var expressions = require('./expressions.js')
 
 /* Load `data` (trigram information per language,
  * per script). */
-var data = require('./data.json');
+var data = require('./data.json')
 
 /* Expose `detectAll` on `detect`. */
-detect.all = detectAll;
+detect.all = detectAll
 
 /* Expose `detect`. */
-module.exports = detect;
+module.exports = detect
 
 /* Maximum sample length. */
-var MAX_LENGTH = 2048;
+var MAX_LENGTH = 2048
 
 /* Minimum sample length. */
-var MIN_LENGTH = 10;
+var MIN_LENGTH = 10
 
 /* The maximum distance to add when a given trigram does
  * not exist in a trigram dictionary. */
-var MAX_DIFFERENCE = 300;
+var MAX_DIFFERENCE = 300
 
 /* Construct trigram dictionaries. */
-(function () {
-  var languages;
-  var name;
-  var trigrams;
-  var model;
-  var script;
-  var weight;
+;(function() {
+  var languages
+  var name
+  var trigrams
+  var model
+  var script
+  var weight
 
   for (script in data) {
-    languages = data[script];
+    languages = data[script]
 
     for (name in languages) {
-      model = languages[name].split('|');
+      model = languages[name].split('|')
 
-      weight = model.length;
+      weight = model.length
 
-      trigrams = {};
+      trigrams = {}
 
       while (weight--) {
-        trigrams[model[weight]] = weight;
+        trigrams[model[weight]] = weight
       }
 
-      languages[name] = trigrams;
+      languages[name] = trigrams
     }
   }
-})();
+})()
 
@@ -67,3 +67,3 @@ /**
 function detect(value, options) {
-  return detectAll(value, options)[0][0];
+  return detectAll(value, options)[0][0]
 }
@@ -81,19 +81,19 @@
 function detectAll(value, options) {
-  var settings = options || {};
-  var minLength = MIN_LENGTH;
-  var script;
+  var settings = options || {}
+  var minLength = MIN_LENGTH
+  var script
 
   if (settings.minLength !== null && settings.minLength !== undefined) {
-    minLength = settings.minLength;
+    minLength = settings.minLength
   }
 
   if (!value || value.length < minLength) {
-    return und();
+    return und()
   }
 
-  value = value.substr(0, MAX_LENGTH);
+  value = value.substr(0, MAX_LENGTH)
 
   /* Get the script which characters occur the most
    * in `value`. */
-  script = getTopScript(value, expressions);
+  script = getTopScript(value, expressions)
 
@@ -105,3 +105,3 @@ /* One languages exists for the most-used script.
   if (!(script[0] in data)) {
-    return script[1] === 0 ? und() : singleLanguageTuples(script[0]);
+    return script[1] === 0 ? und() : singleLanguageTuples(script[0])
   }
@@ -111,5 +111,6 @@
    * normalize the distance values. */
-  return normalize(value, getDistances(
-    utilities.asTuples(value), data[script[0]], settings
-  ));
+  return normalize(
+    value,
+    getDistances(utilities.asTuples(value), data[script[0]], settings)
+  )
 }
@@ -128,12 +129,12 @@
 function normalize(value, distances) {
-  var min = distances[0][1];
-  var max = (value.length * MAX_DIFFERENCE) - min;
-  var index = -1;
-  var length = distances.length;
+  var min = distances[0][1]
+  var max = value.length * MAX_DIFFERENCE - min
+  var index = -1
+  var length = distances.length
 
   while (++index < length) {
-    distances[index][1] = 1 - ((distances[index][1] - min) / max) || 0;
+    distances[index][1] = 1 - (distances[index][1] - min) / max || 0
   }
 
-  return distances;
+  return distances
 }
@@ -151,17 +152,17 @@
 function getTopScript(value, scripts) {
-  var topCount = -1;
-  var topScript;
-  var script;
-  var count;
+  var topCount = -1
+  var topScript
+  var script
+  var count
 
   for (script in scripts) {
-    count = getOccurrence(value, scripts[script]);
+    count = getOccurrence(value, scripts[script])
 
     if (count > topCount) {
-      topCount = count;
-      topScript = script;
+      topCount = count
+      topScript = script
     }
   }
 
-  return [topScript, topCount];
+  return [topScript, topCount]
 }
@@ -177,5 +178,5 @@
 function getOccurrence(value, expression) {
-  var count = value.match(expression);
+  var count = value.match(expression)
 
-  return (count ? count.length : 0) / value.length || 0;
+  return (count ? count.length : 0) / value.length || 0
 }
@@ -196,17 +197,14 @@
 function getDistances(trigrams, languages, options) {
-  var distances = [];
-  var whitelist = options.whitelist || [];
-  var blacklist = options.blacklist || [];
-  var language;
+  var distances = []
+  var whitelist = options.whitelist || []
+  var blacklist = options.blacklist || []
+  var language
 
-  languages = filterLanguages(languages, whitelist, blacklist);
+  languages = filterLanguages(languages, whitelist, blacklist)
 
   for (language in languages) {
-    distances.push([
-      language,
-      getDistance(trigrams, languages[language])
-    ]);
+    distances.push([language, getDistance(trigrams, languages[language])])
   }
 
-  return distances.length ? distances.sort(sort) : und();
+  return distances.length ? distances.sort(sort) : und()
 }
@@ -225,25 +223,25 @@
 function getDistance(trigrams, model) {
-  var distance = 0;
-  var index = -1;
-  var length = trigrams.length;
-  var trigram;
-  var difference;
+  var distance = 0
+  var index = -1
+  var length = trigrams.length
+  var trigram
+  var difference
 
   while (++index < length) {
-    trigram = trigrams[index];
+    trigram = trigrams[index]
 
     if (trigram[0] in model) {
-      difference = trigram[1] - model[trigram[0]] - 1;
+      difference = trigram[1] - model[trigram[0]] - 1
 
       if (difference < 0) {
-        difference = -difference;
+        difference = -difference
       }
     } else {
-      difference = MAX_DIFFERENCE;
+      difference = MAX_DIFFERENCE
     }
 
-    distance += difference;
+    distance += difference
   }
 
-  return distance;
+  return distance
 }
@@ -266,24 +264,21 @@
 function filterLanguages(languages, whitelist, blacklist) {
-  var filteredLanguages;
-  var language;
+  var filteredLanguages
+  var language
 
   if (whitelist.length === 0 && blacklist.length === 0) {
-    return languages;
+    return languages
   }
 
-  filteredLanguages = {};
+  filteredLanguages = {}
 
   for (language in languages) {
     if (
-      (
-        whitelist.length === 0 ||
-        whitelist.indexOf(language) !== -1
-      ) &&
+      (whitelist.length === 0 || whitelist.indexOf(language) !== -1) &&
       blacklist.indexOf(language) === -1
     ) {
-      filteredLanguages[language] = languages[language];
+      filteredLanguages[language] = languages[language]
     }
   }
 
-  return filteredLanguages;
+  return filteredLanguages
 }
@@ -293,3 +288,3 @@
 function und() {
-  return singleLanguageTuples('und');
+  return singleLanguageTuples('und')
 }
@@ -300,3 +295,3 @@
 function singleLanguageTuples(language) {
-  return [[language, 1]];
+  return [[language, 1]]
 }
@@ -306,3 +301,3 @@
 function sort(a, b) {
-  return a[1] - b[1];
+  return a[1] - b[1]
 }

package.json

 {
   "name": "franc-min",
   "threshold": 8000000,
-  "version": "3.1.1",
+  "version": "4.0.0",
   "description": "Detect the language of text",
@@ -6,0 +6,0 @@ "license": "MIT",

readme.md

@@ -7,3 +7,3 @@ <!--This file is generated by `build.js`-->
 
-Built with support for 81 languages (8M or more speakers).
+Built with support for 82 languages (8M or more speakers).
 
@@ -66,15 +66,15 @@ View the [monorepo](https://github.com/wooorm/franc) for more packages and
 | [`fuv`](http://www-01.sil.org/iso639-3/documentation.asp?id=fuv) | Nigerian Fulfulde | 22M |
+| [`bos`](http://www-01.sil.org/iso639-3/documentation.asp?id=bos) | Bosnian (Cyrillic) | 21M |
 | [`bos`](http://www-01.sil.org/iso639-3/documentation.asp?id=bos) | Bosnian (Latin) | 21M |
-| [`bos`](http://www-01.sil.org/iso639-3/documentation.asp?id=bos) | Bosnian (Cyrillic) | 21M |
 | [`hrv`](http://www-01.sil.org/iso639-3/documentation.asp?id=hrv) | Croatian | 21M |
 | [`nld`](http://www-01.sil.org/iso639-3/documentation.asp?id=nld) | Dutch | 21M |
+| [`srp`](http://www-01.sil.org/iso639-3/documentation.asp?id=srp) | Serbian (Cyrillic) | 21M |
 | [`srp`](http://www-01.sil.org/iso639-3/documentation.asp?id=srp) | Serbian (Latin) | 21M |
-| [`srp`](http://www-01.sil.org/iso639-3/documentation.asp?id=srp) | Serbian (Cyrillic) | 21M |
 | [`tha`](http://www-01.sil.org/iso639-3/documentation.asp?id=tha) | Thai | 21M |
 | [`ckb`](http://www-01.sil.org/iso639-3/documentation.asp?id=ckb) | Central Kurdish | 20M |
 | [`yor`](http://www-01.sil.org/iso639-3/documentation.asp?id=yor) | Yoruba | 20M |
+| [`uzn`](http://www-01.sil.org/iso639-3/documentation.asp?id=uzn) | Northern Uzbek (Cyrillic) | 18M |
 | [`uzn`](http://www-01.sil.org/iso639-3/documentation.asp?id=uzn) | Northern Uzbek (Latin) | 18M |
-| [`uzn`](http://www-01.sil.org/iso639-3/documentation.asp?id=uzn) | Northern Uzbek (Cyrillic) | 18M |
+| [`zlm`](http://www-01.sil.org/iso639-3/documentation.asp?id=zlm) | Malay (individual language) (Arabic) | 18M |
 | [`zlm`](http://www-01.sil.org/iso639-3/documentation.asp?id=zlm) | Malay (individual language) (Latin) | 18M |
-| [`zlm`](http://www-01.sil.org/iso639-3/documentation.asp?id=zlm) | Malay (individual language) (Arabic) | 18M |
 | [`ibo`](http://www-01.sil.org/iso639-3/documentation.asp?id=ibo) | Igbo | 17M |
@@ -99,2 +99,3 @@ | [`nep`](http://www-01.sil.org/iso639-3/documentation.asp?id=nep) | Nepali (macrolanguage) | 16M |
 | [`zyb`](http://www-01.sil.org/iso639-3/documentation.asp?id=zyb) | Yongbei Zhuang | 10M |
+| [`pbu`](http://www-01.sil.org/iso639-3/documentation.asp?id=pbu) | Northern Pashto | 10M |
 | [`kin`](http://www-01.sil.org/iso639-3/documentation.asp?id=kin) | Kinyarwanda | 9M |
@@ -101,0 +102,0 @@ | [`zul`](http://www-01.sil.org/iso639-3/documentation.asp?id=zul) | Zulu | 9M |

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

121203

increased by1.58%

Lines of code

679

increased by0.3%

Number of lines in readme file

112

increased by0.9%

No dependency changes