Richard KLINDA writes:
Eval the following lisp code: (require 'un-define)
Mule-UCS is deprecated. I don't think there has anybody who actually
knows that code working on it for 15 years. If you want to use XEmacs
with Unicode, the use of XEmacs 21.5 is highly recommended.
Here is an example file I created that contains corrupted sequences,
1.4
MiB:
https://www.dropbox.com/s/2kprme5e4m3b2cc/report-bigfile.txt?dl=0
I don't have an XEmacs 21.4 to try to reproduce right away, but
looking at your example file, the first instance is exactly on a 65536
(2^16) byte boundary (the 11th) and the later instances occur at
intervals of 65537 bytes after that. There are three possibilities I
can think of:
1. there's an old bug due to "const propagation" in C that caused
problems with stream output at buffer-size intervals (usually a
power of 2), but that was fixed a long time ago;
2. the CCL byte-code interpreter used to implement Mule-UCS uses
16-bit arithmetic and it might be buggy; and
3. finally the Mule-UCS coding system written in CCL might be failing
to clean up after itself when called repeatedly or some similar bug.
I'll try to build an XEmacs 21.4 and reproduce but I doubt I'll have
time to take a close look at the code for a while (I'm not familiar
with it). Maybe Aidan has some idea, though.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://lists.xemacs.org/mailman/listinfo/xemacs-beta