AndrewShkrob/organicmaps
1
Fork 0
forked from organicmaps/organicmaps
Code Pull requests Activity

[crawler] Handle correctly paths with symbols '(' and ')'.

Browse source
This commit is contained in:
Yury Melnichek 2012-09-17 12:12:12 +02:00 • committed by Alex Zolotarev
parent c97750449b
commit f8d90e92ce
1 changed files with 6 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
crawler/wikitravel-optimize-articles.py
View file

@ -7,6 +7,9 @@ import sys
myPath = os.path.dirname(os.path.realpath(__file__))
def formatPath(s):
return s.replace('(', '\\(').replace(')', '\\)')
for i, line in enumerate(sys.stdin):
(url, title, fileBase) = json.loads(line)
fileName = fileBase + '.article'
@ -19,4 +22,6 @@ for i, line in enumerate(sys.stdin):
'--remove-intertag-spaces --simple-bool-attr --remove-quotes '
'--remove-js-protocol --type html '
'-o {outFileName} {fileName}'
.format(myPath = myPath, fileName = fileName, outFileName = outFileName))
.format(myPath = myPath,
fileName = formatPath(fileName),
outFileName = formatPath(outFileName)))