Home » Text » Text_LanguageDetect

Language detection class

This package is unmaintained.

Download latest release 1.0.2 (there might be newer Composer installable version)

License: BSD

Description

Text_LanguageDetect can identify 52 human languages from text samples and return confidence scores for each.

Documentation

Manual

Maintainers

• Nicholas Pisarro (taak) — lead, inactive

Dependencies

Required

• PHP (>= 5.4)
• PEAR Installer (>= 1.9.0)
• pcre (extension)

Optional

• mbstring (extension)

Releases

• 1.0.2 (stable, 2025-07-19)
  Release notes

  License: BSD

  Fix PHP 8.2 warning: Creation of dynamic property Text_LanguageDetect…

  …_Parser::$_trigram is deprecated

  The warning found the bug that the class property had the wrong name.

• 1.0.1 (stable, 2025-07-19)
  Release notes

  License: BSD

  - Support PHP 7.4

• 1.0.0 (stable, 2017-03-02)
  Release notes

  License: BSD

  - Add PHP5 constructors
  - Fix bug #21189: Language detection on PHP7 broken
  - Fix coding style problems
  - BC break: Add real visibility to properties and methods

• 0.3.0 (alpha, 2012-01-16)
  Release notes

  License: BSD

  - BC break: Return lowercase language names
  - BC break: Use exceptions instead of PEAR_Error
  - Implement request #19221: Return ISO 639-1 or ISO 639-2 language codes

• 0.2.3 (alpha, 2008-07-30)
  Release notes

  License: BSD

  updated package requirements

• 0.2.2 (alpha, 2008-07-29)
  Release notes

  License: BSD

  * Fixed Bug #13385

• 0.2.1 (alpha, 2006-12-03)
  Release notes

  License: BSD

  * Fix: Now uses conventionalized include path
  * Fix: Won't cause error if input is purely symbol-range characters
  * Better error reporting if error in unicode db loading

• 0.2.0 (alpha, 2006-01-18)
  Release notes

  License: BSD

  * Added unicode block range identification
  * unicode block ranges used to optimize language detection by pre-selecting which known language trigram profiles in the database to attempt comparison with
  * Added several utf8/unicode utility functions
  * new Parser class for building data profiles from text samples
  * Fix: mb_convert_encoding() now used correctly
  * Fix: basic case-folding for Cyrillic alphabet implemented

• 0.1.1 (alpha, 2006-01-06)
  Release notes

  License: BSD

  * Output of clustering now cached
  * Fix: better error checking in clustering functions
  * Fix: clusteredSearch() now handles null strings gracefully
  * Compare order is preserved in clusteredSearch() results
  * Slight speed improvement to utf8 iterator

• 0.1.0 (alpha, 2006-01-04)
  Release notes

  License: BSD

  Initial PEAR release

Bugs