From: Don A. <dal...@us...> - 2003-07-27 19:16:23
|
This discussion has got me wondering if we should make Date parsing locale dependent, like we have with the relationship calculator. Trying to handle all possible the possible combinations of parsing with all the combinations of languages has made the code overly complex and slow. Most translation errors that we have encountered have been related to the regular expression translation. Thoughts? Don On Thu, 2003-07-24 at 08:19, Lorenzo Cappelletti wrote: > Alex Roitman <sh...@al...>, Mon 21 Jul 2003 19:44 -0500: > > > > In Italian we say "1854 ca." which stands for "abt. 1854". Note that > > > also "ca. 1854" is understandable. > > > > I think this can be dealt with. Right now we have "abt." as > > translatable. We can change it to "abt. %s" which you will translate > > into Italian as "%s ca." What other modifiers are like this? > > I found this topic interesting. Hence I started a new thread... > > I was noticing that some terms are part of more than variable in > Calendar.py. For example, the word "about" is part of _modifiers, M2V > and MODE. > > I think it's wise to distinguish the use of words as regexp from that as > mode. For example, you could add a trailing "|" for each regexp atom: > > _("abt\.?|") + \ > _("about|") + \ > _("est\.?|") + \ > _("circa|") + \ > ..... > > This way a translator can customize their regexps. > > > BTW, what variable M2V and MODE used for? Do they need further > differentiation? -- Don Allingham <dal...@us...> GRAMPS OpenSource Genealogy |