Both hgwebdir.cgi and webserve-dir.cgi are mbcs broken

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Goffredo Baroncelli wrote:
> On Thursday 01 February 2007, Dongsheng Song wrote:
>> For mercurial:
>> When I use mbcs file name and logs, the hgwebdir.cgi DO NOT display
>> use same char encoding, see:
>>     http://www.foresee.com.cn:9999/hg/scratch
>>
>> For bazaar:
>> webserve-dir.cgi DO NOT display mbcs logs correct, see:
>>     http://www.foresee.com.cn:9999/webserve/scratch
> 
> Hi Dongsheng,
> 
> is it possible to get your repository ? For example webserve crash if you 
> browse the directory
> 
> /layout/service/  
> 
> (http://www.foresee.com.cn:9999/webserve/scratch?cmd=inventory;rev=dongsheng.song%40gmail.com-20070202110557-gfgeg6bz14l9fln0;path=layout/service/)
> 
> I am interested in analize the problem
> 
> Goffredo

- From what I could tell it seems that some entries are hex encoded and
other ones are decimal encoded. For example, I'm pretty sure I saw (eg):
 &#1a3b;〶

If I remember my XML, they should all be decimal encoded.

I don't know how webserve is doing it, but if you are using the apis,
you should be getting Unicode strings, and then you need to encode and
escape them. (I doubt you are reading anything directly off of the disk)

If you are getting back something weird (rather than a Unicode string),
then we need to figure out why that is happening.

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFw53AJdeBCYSNAAMRAh6kAKCUE2/Jm8Deq8Y6lJN8Pw668WHMGACfTG+5
v+fKb3k8SdUVIfwYvchrb8k=
=sTW6
-----END PGP SIGNATURE-----