Juergen Spitzmueller wrote: > (but if you insist, I'll change only the German varieties and leave the > others to their users).
I don't insist, I was just wondering. If you like to change all languages please go ahead. >> Unfortunately this change does also affect the conversion from pre >> unicode to utf8 in lyx2lyx, since lyx2lyx gets the encoding from the >> languages file. If you now change encodings in the languages file this >> conversion will be incorrect. Therefore this patch should only go in with >> an additional part that freezes the encodings in lyx2lyx to the current >> values. > > What do you have in mind? A file format change? No, see below. > I don't see another way to > decide which encoding was actually used. And how do you handle the case > where a user changed the languages file himself (which several have done > to get the euro afaik)? I don't know. And that is not handled by lyx2lyx btw. I was wrong, lyx2lyx does not read the languages file directly but the generated file lib/lyx2lyx/lyx2lyx_lang.py. That means that changes to the languages file are currently ignored. That does also mean that as long as lib/lyx2lyx/lyx2lyx_lang.py is not regenerated all is OK and the languages file can be changed. > And if he sends me his file then? That is a problem. This is the same as changing a layout file and then sending a .lyx file that uses that layout. I don't have any better idea than to tell users to run lib/lyx2lyx/generate_encoding_info.py on their old languages file (if they modified it) and delete it afterwards. > Will this ever > work correctly? Isn't there a way to get the encoding from the file? Only if it was explicitly set, and in that case that value is used. Georg
