r00t2/repomirror
1
0
Fork 0
Code Issues Pull Requests Projects Releases Activity

fucking fedora.

Browse Source
This commit is contained in:
brent s. 2020-06-18 04:14:18 -04:00
parent 3ece313a6f
commit 1a5068d77d
Signed by: bts
GPG Key ID: 8C004C2F93481F6B
1 changed files with 47 additions and 40 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

87
repomirror/utils/find_fastest_upstream/centos_epel.py
View File

@ -25,47 +25,54 @@ class Ranker(classes.Ranker):
if preferred_proto not in ('rsync', 'ftp'): if preferred_proto not in ('rsync', 'ftp'):
raise ValueError('Invalid preferred_proto; must be one of rsync or ftp') raise ValueError('Invalid preferred_proto; must be one of rsync or ftp')
non_preferred = ('rsync' if preferred_proto == 'ftp' else 'ftp') non_preferred = ('rsync' if preferred_proto == 'ftp' else 'ftp')
mirror_section = self.bs.find('h2', string = 'Public active mirrors') print(('Fedora (who maintains EPEL) do their mirroring in an extremely weird way.\n'
mirror_table = mirror_section.find_next('table') 'See https://fedoraproject.org/wiki/Infrastructure/Mirroring and '
if mirror_table is None: 'https://fedoraproject.org/wiki/Infrastructure/Mirroring/Tiering#Tier_1_Mirrors for which mirrors and '
return(None) 'how to sync.'))
# https://stackoverflow.com/a/56835562/733214
headers = [h.text for h in mirror_table.find_all('th')]
rows = [m for m in mirror_table.find_all('tr')][1:]
for row in rows:
mirror = {}
do_skip = False
for idx, cell in enumerate(row.find_all('td')):
k = headers[idx]
v = cell.text.strip()
if k == 'Country' and v != self.my_info['country']:
do_skip = True
continue
if k == 'Categories' and not do_skip:
if 'EPEL' not in v:
do_skip = True
continue
pref_proto = cell.find('a', attrs = {
'href': re.compile(r'^{0}://'.format(preferred_proto), re.IGNORECASE)})
non_pref = cell.find('a', attrs = {
'href': re.compile(r'^{0}://'.format(non_preferred), re.IGNORECASE)})
if pref_proto is not None:
v = pref_proto['href']
elif non_pref is not None:
v = non_pref['href']
else:
v = None
mirror['url'] = v
# Fedora team can't spell.
elif k in ('Bandwidth', 'Bandwith'):
mirror['bw'] = int(v)
if do_skip:
continue
if not mirror['url']:
continue
self.raw_mirrors.append(mirror)
self.mirror_candidates.append(mirror['url'])
return(None) return(None)
# mirror_section = self.bs.find('h2', string = 'Public active mirrors')
# mirror_table = mirror_section.find_next('table')
# if mirror_table is None:
# return(None)
# # https://stackoverflow.com/a/56835562/733214
# headers = [h.text for h in mirror_table.find_all('th')]
# rows = [m for m in mirror_table.find_all('tr')][1:]
# for row in rows:
# mirror = {}
# do_skip = False
# for idx, cell in enumerate(row.find_all('td')):
# k = headers[idx]
# v = cell.text.strip()
# if k == 'Country' and v != self.my_info['country']:
# do_skip = True
# continue
# if k == 'Categories' and not do_skip:
# # TODO: DO THIS BETTER! Their mirrorlist sucks and is not easily parsed at all.
# # I need to check and try to grab the specific URL that contains "epel".
# if 'EPEL' not in v:
# do_skip = True
# continue
# pref_proto = cell.find('a', attrs = {
# 'href': re.compile(r'^{0}://'.format(preferred_proto), re.IGNORECASE)})
# non_pref = cell.find('a', attrs = {
# 'href': re.compile(r'^{0}://'.format(non_preferred), re.IGNORECASE)})
# if pref_proto is not None:
# v = pref_proto['href']
# elif non_pref is not None:
# v = non_pref['href']
# else:
# v = None
# mirror['url'] = v
# # Fedora team can't spell.
# elif k in ('Bandwidth', 'Bandwith'):
# mirror['bw'] = int(v)
# if do_skip:
# continue
# if not mirror['url']:
# continue
# self.raw_mirrors.append(mirror)
# self.mirror_candidates.append(mirror['url'])
# return(None)
def parseArgs(): def parseArgs():