On 13 July 2015 at 16:19, Chris Cormack <chris@bigballofwax.co.nz> wrote:
On 13 July 2015 at 11:23, Mark Tompsett <mtompset@hotmail.com> wrote:
Note: It seems there is a bug on sending patch using git bz attach -e (works without -e), could someone confirm?
Actually, it works if you don’t Obsolete AND change status at the same time. If you run the same command again after failing the status is changed, but the patch attach wasn't done, and it succeeds.
I concur
It works if you are only doing one or the other, if you do both, it fails. Now I can consistently replicate it, I might be able to fix it.
Chris
And yes it is all to do with the fact they are screen scraping, the code says if not check_for_success(response, response_data, r"<title>\s*Bug[\S\s]*processed\s*</title>"): die ("Failed to update bug %d, status=%d" % (self.id, response.status)) But when you change the status of a bug, it says <title> 6473 processed</title> So this if is failing and the script is dying, before it gets to the code to attach the file. Once I figure out how regex in python work, I'll patch git-bz Sometime, we should rewrite it to use the API instead of screen scraping, so this stops happening Chris