Closed jddurand closed 1 year ago
Seems that I have an issue with libthai, where the relocatable thingy patch that I used (c.f. https://github.com/tlwg/libthai/commit/391b612d6916fa34254bdd8f1984cbbe0ac3c14a) fails. Sorry about this issue, I'll come back on it and will keep it updated, regardless that fact that I closed it.
I confirm this was an installation on my side, I installed the thai dictionnary in the wrong path. So please consider that Prima compiles ok and tests ok on Windows using cl, modulo the problem with the CALLBACK definition mentionned in this comment
Yes, I also have an issue with libthai's dll relocation, and also made a issue https://github.com/tlwg/libthai/issues/17 , but that didn't go anywhere so I basically rolled my own libthai.dll, more or less the same way as you did (but never sent a patch, didn't want to step on any toes before a discussion). I references that binary distro in Prima's README if anyone needs it. But I agree the situation is quite shameful for this library, it is a standard implementation of thai breaks on linux, and the author(s) are not responding.
Hello,
After having compiled all the dependencies of Prima on Windows and the cl compiler, in particular libthai version 0.1.29 for which the test suite is successful, Prima test suite fails only thai breaks IMHO:
My question is if you know if libthai is stable concerning the notion of breaks. If yes, I suppose my libthai is wrong (and this is strange since its test suite is ok...), if no then the thai breaks part of Prima test suite should not be compulsory.