Mallard and Ubuntu-Translators and Translation Regressions

[Kyle Nitzsche]

So scripting this to get as far as possible sounds great.

The project needs definition and clarity of rules for converting each db 
tag to mallard. I'm happy to help with the scripting, but defining and 
testing the rules is a big and critical part of the job that would need 
attention/definition.

Cheers,
Kyle

Milo Casagrande wrote:
> 2010/2/25 Shaun McCance <shaunm at gnome.org>:
>   
>> Here's the command I used:
>>
>> grep -o '<[^/][^>/ ]*' es.po | grep -v '@' | sort -u
>>
>> And this is what I got:
>>
>> <anchor
>> <application
>> <citetitle
>> <command
>> <computeroutput
>> <email
>> <EMAIL
>> <emphasis
>> <filename
>> <firstterm
>> <glossterm
>> <guibutton
>> <guilabel
>> <guimenu
>> <guimenuitem
>> <guisubmenu
>> <imagedata
>> <imageobject
>> <indexterm
>> <interface
>> <itemizedlist
>> <keycap
>> <keycombo
>> <keysym
>> <link
>> <listitem
>> <literal
>> <mediaobject
>> <menuchoice
>> <option
>> <para
>> <phrase
>> <placeholder-1
>> <primary
>> <remark
>> <replaceable
>> <screenshot
>> <secondary
>> <see
>> <shortcut
>> <tertiary
>> <textobject
>> <ulink
>> <uri
>> <userinput
>> <xref
>>
>> Note that <placeholder-$i/> elements are an artifact of xml2po.
>> Also note that PO files can split strings anywhere, so the above
>> command could give you partial tag names.  It just happened that
>> it didn't with the PO file I used.
>>     
>
> Almost forgot about this: you can use:
> msgattrib --no-wrap file.po > new-file.po
>
> to get a non-wrapped po file, with msgid and msgstr not split over
> different lines.
>
> Ciao.
>
>