IntenseWebs/sphinx
3
0
Fork 0
mirror of https://github.com/sphinx-doc/sphinx.git synced 2025-02-25 18:55:22 -06:00
Code Issues Packages Projects Releases Wiki Activity

Linkcheck now works with non-ASCII URIs.

Browse Source 
Fixes: #1739
This commit is contained in:
Georg Brandl 2015-02-25 12:49:27 +01:00
parent 7189a2cf2c
commit ea97e6658c
1 changed files with 18 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
sphinx/builders/linkcheck.py
View File

@ -11,12 +11,13 @@
import re
import socket
import codecs
import threading
from os import path
from six.moves import queue
from six.moves.urllib.request import build_opener, Request, HTTPRedirectHandler
from six.moves.urllib.parse import unquote
from six.moves.urllib.parse import unquote, urlsplit, quote
from six.moves.urllib.error import HTTPError
from six.moves.html_parser import HTMLParser, HTMLParseError
from docutils import nodes
@ -131,12 +132,27 @@ class CheckExternalLinksBuilder(Builder):
if rex.match(uri):
return 'ignored', '', 0
# split off anchor
if '#' in uri:
req_url, hash = uri.split('#', 1)
else:
req_url = uri
hash = None
# handle non-ASCII URIs
try:
req_url.encode('ascii')
except UnicodeError:
split = urlsplit(req_url)
req_url = (split[0].encode() + '://' + # scheme
split[1].encode('idna') + # netloc
quote(split[2].encode('utf-8'))) # path
if split[3]: # query
req_url += '?' + quote(split[3].encode('utf-8'))
# go back to Unicode strings which is required by Python 3
# (but now all parts are pure ascii)
req_url = req_url.decode('ascii')
# need to actually check the URI
try:
if hash and self.app.config.linkcheck_anchors:
@ -261,7 +277,7 @@ class CheckExternalLinksBuilder(Builder):
self.app.statuscode = 1
def write_entry(self, what, docname, line, uri):
output = open(path.join(self.outdir, 'output.txt'), 'a')
output = codecs.open(path.join(self.outdir, 'output.txt'), 'a', 'utf-8')
output.write("%s:%s: [%s] %s\n" % (self.env.doc2path(docname, None),
line, what, uri))
output.close()