[asterisk-dev] language-specific sound directories

[Luigi Rizzo]

looking at file.c it seems that language-specific
sound files are intermixed - e.g. we have

	.../sounds/digits/*	english digits
	.../sounds/digits/it/*	italian digits
	.../sounds/letters/*	english letters
	.../sounds/letters/it/*	italian letters

and so on. I would have expected separate trees for
each language i.e.

	.../sounds/en/*		everything in english
	.../sounds/it/*		everything in italian

which would be way easier to manage if you have to install/delete
third-party voice files.

So, apart from legacy issues (which could be easily solved
with a backward compatibility flag) where else is the
mapping done ? I see two places in file.c namely

ast_fileexists() and ast_openstream_full()

(the code is repeated so it needs to be collapsed in one place,
anyways). Are there any other places that (by design, not by coding
mistake) do need to know the layout of files ?
Some management tools maybe ?

	cheers
	luigi