Can you explain that why you avoid converting to lowercase?
/* The file system is case-insensitive but case-preserving.
So we do not lowercase the path. Otherwise, we might not
be able to display the original casing in a given path. */
I just thought that lowercase is another canonization of path, so you
can have a uniform path when you set breakpoints(file specification)
Two reasons, mostly:
. It is unnecessary;
. It changes the filename casing when displaying the name of
the file where the breakpoint has been inserted.
Several Windows users at AdaCore often complain that GNU tools do not
properly preserve the filename casing, so I did not think it was
proper for us to do so. Even as a Unix user, I do feel that it is
important to preserve the casing as well.
I should say that it does not affect filename matching, which is
performed through another specialized function.
The original code in iberty library is like:
Honestly, I do not understand why they do it. There is no correlation,
IMO, between the fact that the FS is case-insensitive and the fact
that you'd want to change the filename casing (again, doing something
extra which is unnecessary, since it does not make a difference to
the filesystem).