[Yum] Suggestion about yum updater

[Edinelson.Shimokawa at mp.mt.gov.br]

Hello.
I am here suggest a modification to correct a bug.
There is a bug in the yum  _mirror_try method in
/usr/lib/python2.4/site-packages/urlgrabber/mirror.py that works like that:
- the list of mirrors are stored in a array (class URLGrabber?) gr and then
when the socket doesnt conect to first mirror it removes that mirror from
the array and get the next one. But getting always the first one.
- after it passed over all the array, i. e., when a array of zero lenght is
achieved, it populate the array again. but the _next pointer is pointing to
a mirror that doesnt exists (the last position plus one).
This gives a error like this:

 Traceback (most recent call last):
   File "/usr/bin/yum", line 15, in ?
     yummain.main(sys.argv[1:])
   File "/usr/share/yum-cli/yummain.py", line 149, in main
     base.doTransaction()
   File "/usr/share/yum-cli/cli.py", line 592, in doTransaction
     problems = self.downloadPkgs(downloadpkgs)
   File "__init__.py", line 565, in downloadPkgs
   File "repos.py", line 605, in get
   File "/usr/lib/python2.4/site-packages/urlgrabber/mirror.py", line 414,
in
 urlgrab
     return self._mirror_try(func, url, kw)
   File "/usr/lib/python2.4/site-packages/urlgrabber/mirror.py", line 392,
in
 _mirror_try
     mirrorchoice = self._get_mirror(gr)
   File "/usr/lib/python2.4/site-packages/urlgrabber/mirror.py", line 290,
in
 _get_mirror
     return gr.mirrors[gr._next]
 IndexError: list index out of range

I have did this to solve but I dont know if is the best solution:


    def _mirror_try(self, func, url, kw):
        gr = GrabRequest()
        gr.func = func
        gr.url  = url
        gr.kw   = dict(kw)
        self._load_gr(gr)
        gr._next = 0 #Added line

One more thing, I suggest a option of timeout of connection. Because many
linux servers are under a anti-virus appliance that make the real download,
scan and deliver the file. When a file is big size the conection timeout
occurs. There is a timeout argument in kwargs of URLGrabber.

Thank you,

Edinelson Keiji Shimokawa