diff options
| author | ThibG <thib@sitedethib.com> | 2019-03-26 19:51:28 +0100 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2019-03-26 19:51:28 +0100 |
| commit | bdf4f56f9a9052668e7cce1ce7863268e03d9c86 (patch) | |
| tree | 9952f2e08780455dc1af0febe4b374f5b8a60e2e /app/lib/language_detector.rb | |
| parent | 9f56511860493b770c2e535be0f8ce4777c8d373 (diff) | |
| parent | 9e80b34a95c39d0a1a944713f0f4d6cc5e8b8c43 (diff) | |
Merge pull request #966 from ThibG/glitch-soc/merge-upstream
Merge upstream changes
Diffstat (limited to 'app/lib/language_detector.rb')
| -rw-r--r-- | app/lib/language_detector.rb | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/app/lib/language_detector.rb b/app/lib/language_detector.rb index 70a9084d1..1e90af42d 100644 --- a/app/lib/language_detector.rb +++ b/app/lib/language_detector.rb @@ -3,7 +3,7 @@ class LanguageDetector include Singleton - CHARACTER_THRESHOLD = 140 + WORDS_THRESHOLD = 4 RELIABLE_CHARACTERS_RE = /[\p{Hebrew}\p{Arabic}\p{Syriac}\p{Thaana}\p{Nko}\p{Han}\p{Katakana}\p{Hiragana}\p{Hangul}]+/m def initialize @@ -37,7 +37,7 @@ class LanguageDetector end def sufficient_text_length?(text) - text.size >= CHARACTER_THRESHOLD + text.split(/\s+/).size >= WORDS_THRESHOLD end def language_specific_character_set?(text) |