Overview

Advanced i18n/l10n features useful in a CMS environment.

Copyright / Licenses

The code in this distribution is copyright Plone Foundation and licensed under the GPL version 2 only.

The national flags icon set included in this distribution is copyright by Jakub Steiner () and we have been given permission to include it under the GPL license.

This distribution depends on the Unidecode Python code by Tomaz Solc repackaged by Ben Bengart for setuptools. The code is a port of the Perl module Text-Unidecode-0.04 by Sean M. Burke licensed under the same terms as Perl itself.

Changelog

5.0.2 (2023-10-07)

Internal:

• Update configuration files. [plone devs] (cfffba8c)

5.0.1 (2023-03-21)

Bug fixes:

• Correctly specified dependencies. plone.supermodel and a few others were missing. [plone devs] (#49)

Internal:

• Update configuration files. [plone devs] (25f71b87)

5.0.0 (2022-11-30)

Bug fixes:

• Fixed error in listAvailableLanguageInformation (#40)
• Final release. [gforcada] (#600)

5.0.0b1 (2022-09-01)

New features:

• Add some more native names to language-country combinations [erral] (#45)
• Fix the missing native names for language-country variants [erral] (#46)

Bug fixes:

• Added native names for German and Dutch country-specific language variants. Then all related language variants are shown together in the language control panel. [maurits] (#824)

5.0.0a5 (2022-06-07)

Breaking changes:

• Use SVG Flags in Language Selector Update Name of Flags Prepare of use the Icon Resolver in Plone 6 [1letter, jensens] (#42)

5.0.0a4 (2022-01-28)

Bug fixes:

• Tests: call self.publish with handle_errors=False. [maurits] (#41)

5.0.0a3 (2021-10-20)

New features:

• Let the sr language be Latin: Sprski. If you want to use the Cyrillic character set variant, set environment variable zope_i18n_allowed_languages to sr@Cyrl. [maurits] (#326)

5.0.0a2 (2021-10-16)

New features:

• Support Cyrillic and Latin character sets for Serbian. See locales issue 326. [maurits, fredvd] (#326)

5.0.0a1 (2021-09-01)

Breaking changes:

• Drop support of Python 2 [jensens] (#31)

New features:

• Make it work with newwest Unidecode 1.2.0. Use its method instead of duplicating it in here. Also support newer unicode chars, like with Bei Jing in the example. [gogobd, jensens, alert] (#31)

4.0.6 (2020-04-20)

Bug fixes:

• Minor packaging updates. (#1)

4.0.5 (2019-06-27)

Bug fixes:

• bugfix Shortname starts "b" in Japanese

  [terapyon] (#32)

4.0.4 (2019-03-03)

New features:

• Move the interface ILanguageSchema from Products.CMFPlone.interfaces to this package to avoid install dependency on Products.CMFPlone and the whole Plone stack. This also breaks the circular reference between those packages. [sallner] (#29)

4.0.3 (2019-02-13)

Bug fixes:

• Be more verbose when Japanese normalizer fails in tests. Test that the length of the returned string is lower but not always equal to the given max_length. Changed tested string. (#25)

4.0.2 (2018-12-11)

Bug fixes:

• Code and test cleanuo. [jensens]

4.0.1 (2018-10-31)

Bug fixes:

• Fix py3 tests and functionality [pbauer]

4.0.0 (2018-02-02)

Breaking changes:

• Move PTS Negotiator over form PlacelessTranslationService over to here in order to get rid of PTS itself. [jensens, ksuess]

Bug fixes:

• Add Python 2 / 3 compatibility [pbauer]

3.0.7 (2017-05-16)

Bug fixes:

• Add missing entries to requirements in setup.py. [dhavlik]

3.0.6 (2016-10-03)

Bug fixes:

• Code Cleanup: isort, pep8, manual cleanup [jensens]

3.0.5 (2016-08-15)

Fixes:

• Use zope.interface decorator. [gforcada]

3.0.4 (2016-03-31)

Fixes:

• URLNormalizer now removes leading underscores. They are disallowed for content objects in zope. [do3cc]

3.0.3 (2016-02-19)

Fixes:

• Fixed test (don't expect role from pac-tests). [pbauer]

3.0.2 (2015-05-04)

• Allow calling getPreferedLanguage without parameter again [tomgross]

3.0.1 (2015-03-30)

• Avoid loading of registry on BeforeTraversal breake the request [bloodbare]

3.0.0 (2015-03-26)

• Adapted for Plone5 [bloodbare]
• Added the tests from Products.PloneLanguageTool [bloodbare]
• Added the Products.PloneLanguageTool on this package [bloodbare]
• Added the negotiation traverse hook for language negotiation [bloodbare]

2.0.9 (2013-08-13)

• German URL normalizer: handle German ‚single‘ and „double“ quotation marks, em dash –, paragraph § and € sign. [jnachtigall]

2.0.8 (2013-05-26)

• Add Portuguese normalizer. [hvelarde]

2.0.7 (2013-05-23)

• Fix regression in normalizer: handle POSIX locale identifiers in addition to IETF language codes. [davisagli]

2.0.6 (2013-04-06)

• Fixed does not correctly handle client HTTP_ACCEPT_LANGUAGE is language/country combination. [jianaijun]

2.0.5 (2013-01-13)

• Revert 97645c76f5e0cf14e525f702c66fe0c4de8cb0fc. [esteele]

2.0.4 (2013-01-13)

• Add Latvian flag to language list [maartenkling]

2.0.3 (2012-10-16)

• Removed unknown cctld entry for um. [hannosch]
• Added new cctld entries for: asia, kp, ss and xxx. [hannosch]
• Added new country codes and flags for: bl, bq, cw, mf, ss and sx. [hannosch]
• Clarify and test availability of reserved country codes an and cs. [hannosch]
• Clarify and test availability of deprecated language codes mo and sh. [hannosch]
• Removed invalid me language code added in 2007. me is only a country code. [hannosch]
• Correct language code for Javanese from jw to jv. [hannosch]
• Added missing ISO-639-1 language codes: ae, ak, an, bm, ce, ch, cr, cu, cv, dv, ee, ff, ho, ht, hz, ig, ii, io, kg, ki, kj, kr, kv, lg, lu, mh, nb, ng, nv, ny, oj, os, pi, sc, vk. [hannosch]
• URLNormalizer should remove all ignored characters before making any substitutions. [esteele]
• Added three new countries and its corresponding flags. Also updated the internet top level domains list. Added countries are Kosovo, Montenegro and Serbia. [alecghica]
• Fixed tests not to fail when a new country, language or domain is added. [alecghica]

2.0.2 (2012-08-29)

• Avoid infinite loop if buggy queryUtility() returns normalizer instance of wrong class. Fixes . [patch by Sardtok, applied by kleist]

2.0.1 (2012-07-02)

• Use zope.browserresource. [hannosch]
• Converted most tests from doctests to proper unit tests. [hannosch]
• Fixed various native names of African languages. Added nd, nr and ve language codes. Thanks to Dwayne Bailey and Roche Compaan. [hannosch]
• The max_length argument was ignored by plone.i18n.normalizer.ja.Normalizer. [rossp]

2.0 - 2010-07-18

• Use the standard libraries doctest module. [hannosch]
• Update license to GPL version 2 only. [hannosch]

1.1b1 - 2010-01-24

• In practice the Unidecode data didn't produce good enough results for various languages. We therefore limit the transliteration approach again to latin-like languages and introduce a UNIDECODE_LIMIT. This closes . [hannosch]

1.1a3 - 2009-12-27

• Use the flag of Bangladesh for the Bengali language (code: bn) in general and not only for the bn-bd variant. This closes . [hannosch]
• Fixed the IIDNormalizer to generate valid CSS ID or Python variable names as specified in its docstring. It no longer tries to preserve filename extensions. This closes . [hannosch]
• Added a new explicit base normalizer for Thai, as the Unidecode based transliteration isn't good enough. [hannosch]
• Added new specific normalizer for Japanese, which avoids the Unidecode based transliteration. This refs . [hannosch]

1.1a2 - 2009-12-02

• Depend on and include Unidecode based transliterations. These provide more meaningful results than unicodedata NFKD normalizations or hex codes. [hannosch]
• Added cs-cz combined language code. It solves problem with default language on new Plone site creation (Safari/Mac). [naro]
• Added catalan flag. This closes [ramon]
• Added missing 'native' descriptions to pt-* combined languages. [igbun]

1.1a1 - 2009-04-04

• Removed the negotiator sub-package, as it hasn't been enabled or used yet. The functionality is better placed as a WSGI-middleware. [hannosch]
• Register all dependencies in setup.py. Move test dependencies into a separate extra to keep the dependencies low so things like the normalizer can be used in non-zope contexts. [wichert]

1.0.9 - Unreleased

• Added Romanian language as the language for the .ro TLD. This closes [vincentfretin]

1.0.8 - 2009-10-15

• Added new UNDERSCORE_START_REGEX to the file normalizer. This removes any leading underscores from uploaded file names. Objects in Zope cannot start with an underscore, so it makes little sense to generate suggested file names which cannot work. [hannosch]
• Added bulgarian normalizer. [vlado]

1.0.7 - 2008-11-05

• Allow _ as a valid character in file names and URLs. Do not remove # from file names. It only has a special meaning for URLs. [hannosch, sidnei]

1.0.6 - August 18, 2008

• Added normalization for a French-only character (igature of o and e) which isn't part of ISO 8859-1. This closes . [dbaty, hannosch]
• Fixed the greek character normalization based on a patch by ggozad. This closes . [hannosch]
• Changed the default normalization of characters used in Scandinavian languages to meet the most common rules. This is based on a discussion with translators from all Scandinavian countries. [hannosch]
• Added a subdomain language negotiator (e.g. de.plone.org). [stefan]

1.0.5 - May 22, 2008

• Added a new max_length argument to the normalize method. This allows you to override the default values for the maximum length on a call basis. [hannosch, fschulze]
• Added a new MAX_URL_LENGTH constant used by the URL normalizer. It defaults to 255. [hannosch]
• Added '`' to the list of dangerous chars, which will be removed by the url and be replaced with a dash by the file name normalizer now. [hannosch, mj]

1.0.4 - April 19, 2008

• If a dot was used in a url, the url was not lowercased. This closes . [hannosch]

1.0.3 - February 13, 2008

• Updating the flag/language listings. Updating readme to include the flag mapping logic. This closes . [limi]
• Adding all the Arabic-speaking countries with their respective flags, and adding a generic flag to represent Arabic in general (verified to be OK with two independent, native residents). [limi]
• Added tests for the filename and url request adapters. We have now 100% test coverage. [hannosch]
• Wrote tests for the locale-aware id normalizer and fixed a bug in it. [hannosch]
• Added more tests. [hannosch]
• Removed unused and untested 'native' from country information. [hannosch]
• Don't allow double quotes in normalized urls. This closes . [hannosch]

1.0.2 - November 24, 2007

• Remove those [] brackets from file names as well. [hannosch]
• Increase the maximum filename size to 1023 and make it independently configurable. [hannosch]

1.0.1 - October 7, 2007

• Extend polish normalizer 'dashed L' to L. This closes . [hannosch]
• Added test for filename with non-ascii character and extension. Fixes . [dreamcatcher]

1.0 - August 13, 2007

• Added Polish normalizer which normalizes 'dashed l' to l. This fixes . [hannosch]

1.0rc1 - July 9, 2007

• Merged udg-sprint branch. This adds a Zope3-based configurable language negotiator with similar functionality as PloneLanguageTool. It is not enabled by default. [hannosch]
• Added the reference to the Faroese flag. [deo]
• Remove more punctuation characters as for example using a comma in the title results in a not so friendly ID. This closes . [hannosch]

1.0b1 - March 5, 2007

• Initial implementation. [hannosch]