TYPO3 Core

attached is a patch which adds the ISO code and use it in core

At the moment when static_info_tables is installed, the uid of the static_language record is saved in the field. With the patch the 2-char ISO code directly is used.

So there is an incompatibility which could be solved with update-routine from static_info_tables.

Hi,

By reading:

Patch, of t3lib_flexformtools.php, around line 306. Are you sure $row[$row['static... is correct? I guess it should be only $row['static_lang_isocode'], just as for t3lib_tceforms.php Furthermore, this can't compile as a semicolon is missing.

t3lib_transl8tools, third chunk, missing space after elseif. elseif($isLoaded && $row['static_lang_isocode']

t3lib/stddb/tables.sql, you define static_lang_isocode as varchar(3) but I read that you use a 2-char ISO code, as such I would like to read char(2) instead.

I corrected the 2 chunks, thx.

I used varchar(3) as static_info_tables use integers which might need 3 chars for saving.

hm transl8tools checks two times for "$isLoaded && $row['static_lang_isocode']" ... imho elseif block will never be reached

Would it be possible to use a locale for the language?

If you have a website for branches in USA, Germany, Australia using the alpha2 code might still lead to problems with TV because "en" is used twice.

IMO using locales like en_US and en_AU would be better.

What do you think?

Wrong example given by me.

Switzerland or Belgium will suit as a better example. You could have a TYPO3 installation for one of these countries and use more than one language.

How to solve that with alpha2 codes?

That would be the next step. Take german speaking languages, you may use
de_DE
de_AT
de_CH
so in general langKey_ISOcode

btw - this examples couldn't be used with static_info_tables, AT and CH are missing there!

Maybe we should move the langKey from BE_USER to sys_language as well.

@Tolleiv Nietsch - thanks, flux destroys concentration.

Andreas and Steffen, there's already some stuff in the work for "Dialect support" in TYPO3.

Here we need FR_CA because you know, "courriel" isn't "email" in french... Except in France. ;)

The sad thing is that there's no ISO for Countries AND dialect. It's a mix like the one made by Mozilla or OOo:
http://l10n.openoffice.org/L10N_Framework/iso_code_build2.html
https://wiki.mozilla.org/L10n:Simple_locale_names

Contact me if you need more information. firstname.name@infoglobe.ca.

i don't get it - why should we use ISO 3166 in sys_languages? languages (2 letter codes) are definied in ISO 639-1

else the patch (13539.diff) is a bit unlucky since iso 639 is lowerchase. If done, there should also be a field for ISO 639-2.

see http://www.loc.gov/standards/iso639-2/php/code_list.php