'bzr get' fails over sftp in arch-conversion branch.

[John Arbash Meinel]

Aaron Bentley wrote:
> John Arbash Meinel wrote:
> 
>>>The problem being that he reads the XML directly, expecting all
>>>characters to be interesting and valid. And he doesn't try to un-escape
>>>them.
> 
> 
> Is it possible they also need utf8-decoding?
> 
> Aaron

Well, so far we don't allow non-ascii characters in file ids.
And we've never allowed them in Arch, so I don't think it is a problem.

Right now in bzrlib.workingtree.gen_file_id we use the sequence:
name = re.sub(r'[^\w.]', '', name)

without using re.UNICODE, the above pattern will only match ascii
characters.

Eventually that will be part of my 'encoding' branch, so that we can
have file-ids (and maybe revision-ids) which have unicode characters in
them.

But so far, just ascii, no need to decode.

John
=:->
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 256 bytes
Desc: OpenPGP digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060119/d0cb1d8e/attachment.pgp