Hi,
Thank you again for submitting your fix. Now the page is good.
At Sat, 7 Jul 2001 09:29:02 +0200 (CEST),
peter karlsson <[EMAIL PROTECTED]> wrote:
> and those were not matched properly. However, I seem to have missed a
> quotation mark missing in the regexp, it should read:
>
> $title =~ s/^#use .* title="(.+?)(" .*$|"$|"\e.*$)/$1/;
> ^
Imagine an ISO-2022-JP string has a JIS X 0208 part and following
ASCII part. When the JIS X 0208 part ends with 0x22, it matches "\e
and thus the regexp will fail.
Rather, I think the Japanese pages like
> title="<switch to 0208>DBCS<switch to 0201>"<switch to ASCII><space>
is not appropriate, even though it is not illegal. I imagine such
pages were generated by some buggy editors or some unappropriate
usage of editors. Thus, I think the Right Thing is to remove "\e
from the regexp. However, even though I can fix these pages now,
if there are some possibility for such pages to be generated accidentally
in future, the sitemap page will have some possibility to be broken.
Thus, I think your way is appropriate, though not strictly correct.
---
Tomohiro KUBOTA <[EMAIL PROTECTED]>
http://www.debian.or.jp/~kubota/
"Introduction to I18N" http://www.debian.org/doc/manuals/intro-i18n/
