On Thu, 2005-02-17 at 08:17 +0100, Pavel JanÃk wrote: > From: Dwayne Bailey <[EMAIL PROTECTED]> > Date: Thu, 17 Feb 2005 08:40:54 +0200 > > > The following is now part of oo2po - I cannot access CVS at the moment > > so am not sure if this is part of 0.8rc2 or is only in CVS. > > > > You can use: > > > > oo2po --duplicates=msgid_comment > > > > to create separate entries for any duplicates like "Title" in a specific > > PO file. This should now also be the default behaviour. > > > > This will add a KDE style comment that contains the location > > information. Such a comment looks like this: > > > > "_: PLACE_THAT_USES_TITLE\n" > > "Title" > > Yes, but this is the opposite extreme. This is now what we would like to > see ;-) > > Imagine string like this one: > > #: 03091500.xhp#par_id3155764.474.help.text > 03091500.xhp#par_id3147272.477.help.text > 03091500.xhp#par_id3158137.480.help.text > 03091500.xhp#par_id3156750.483.help.text > 03091500.xhp#par_id3144064.486.help.text > 03091500.xhp#par_id3167680.489.help.text > 03091500.xhp#par_id3162156.492.help.text > 03091500.xhp#par_id3180753.495.help.text > 03091500.xhp#par_id3184389.498.help.text > msgid "Meaning" > msgstr "VÃznam" > > There can be several different meanings for the word "Meaning". The amount > and distribution of them could be different for various languages. > > Right now, we have one msgid for this. This is bad. > > With your proposed solution, we will have eight of them where > approx. 6 will be translated as word1 and the rest as word2. This is very > inefficient/inconsistent.
Is it? Once it is translated it will be fine and as consistent as your translations are. You would still need to analyse each to determine what type of "Meaning" it was anyway. So I'm not sure why seeing all 8 separate would be less efficient then looking at each one on a msgid and doing the same thing. With them separate at least they will appear in the correct context. > I'd like to see different system: > > 1. POT files will be generated like they are now, one msgid for every > string where there is only one msgid for every unique word. > > 2. When the translator/localizer have to divide different meanings of the > same word, he will do that manually. E.g. in the above mentioned msgid: > > #: 03091500.xhp#par_id3155764.474.help.text > 03091500.xhp#par_id3147272.477.help.text > 03091500.xhp#par_id3158137.480.help.text > 03091500.xhp#par_id3156750.483.help.text > 03091500.xhp#par_id3144064.486.help.text > 03091500.xhp#par_id3167680.489.help.text > 03091500.xhp#par_id3162156.492.help.text > 03091500.xhp#par_id3180753.495.help.text > msgid "Meaning" > msgstr "VÃznam" > > #: 03091500.xhp#par_id3184389.498.help.text > msgid "_: The word meaning as in "UGLY language depending description" :-)" > "Meaning" > msgstr "VÃznamnÃ" > > > What has to be done so this can be used? We have to keep the split of > one msgid while upgrading PO files from new POT files, etc... David? Manually splitting could be problematic as you could break things. Maybe adding them to a file as suggested by Ain could allow them to be divided into: Meaning1, Meaning2 or Unkown. This same file is then used to split your existing PO file and to parse a POT file before merging it with your PO file. -- Dwayne Bailey 083 443 7114 >From a mesh network in White River, Mpumalanga, South Africa --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
