Hi Tal! You can download the file, make the required replacements and upload back.
If unsure about the replacements you can always use the search confirmation method in VIM. If you need any further help I'd love to assist. Kind regards, Yaron Shahrabani <Hebrew translator> On Mon, May 26, 2014 at 12:54 PM, Tal Daniel <[email protected]> wrote: > I've started to look into the Help in Hebrew, and I need help to make > massive Find & Replace in the Hebrew Help translation. Is it possible to > download it, make a batch/bash script and upload it again? Are there any > consequences (e.g. strings will be converted to suggestions, after > uploaded?) > > The last post about it in L10N was from 2013, and wasn't answered. > The Pootle mailing list says it IS POSSIBLE, by downloading the PO files, > changing, and merging them back into SVN (see > > https://www.mail-archive.com/[email protected]/msg02282.html > ) > > As for the Hebrew Help problem, in more detail: > It seems many of the strings are left as-is, in English, but an even worse > problem is that many strings are messed up. Ariel reported that EMPH and > BOOKMARK tags show with error, In Pootle too. > > I've found there are 999+ strings with broken BOOKMARK_VALUE tags: > > https://translate.apache.org/he/aoo40help/translate/helpcontent2/source/#search=bookmark_value&sfields=source,target > and over 999+ strings with escaped EMPH tags: > > https://translate.apache.org/he/aoo40help/translate/helpcontent2/source/#search=%5C%5C%3Cemph%5C%5C%3E&sfields=target > > Thanks for any help, > Tal > > ---------- Forwarded message ---------- > From: Ariel Constenla-Haile <[email protected]> > Date: Sat, May 24, 2014 at 12:29 AM > Subject: Re: OO, Hebrew Help > To: Tal Daniel <[email protected]> > > [...] unzip the *.jar files in the help that comes with OpenOffice in > Hebrew [...] > > \<bookmark - 4007 matches > \<emph - 5355 matches > > As you see, it does not make sense to try to fix them on Pootle; on the > other hand, the help seems to be almost 50% translated, so it would be > nice not to lose all that work. > > May be it's possible to generate a translation memory from the current > translation, then import a clean version without the errors; [...] > > Regards > -- > Ariel Constenla-Haile > La Plata, Argentina >
