Description

After upgrading to 1.5.5a and cleaning up the cache with the cachecleaner (or more accuratly with the ScriptMarket/AllWikisHack), some pages don't show up in linkto: search results. After throughout investigation, it appears that some pages have empty (or almost empty: there's just a \n) cache/pagelinks. Those are generated when I did the first search, which generated thousands of such messages:

Traceback (most recent call last):
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/Page.py", line 1529, in parsePageLinks
    page.send_page(request, content_only=1)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/Page.py", line 1235, in send_page
    start_line=pi_lines)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/Page.py", line 1314, in send_page_content
    self.format(parser)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/Page.py", line 1335, in format
    parser.format(self.formatter)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/parser/wiki.py", line 1017, in format
    self.processor_is_parser)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/formatter/base.py", line 317, in processor
    p.format(self)
  File "/usr/local/www/wiki/wikis/koumbitwiki/data/plugin/parser/sctable.py", line 567, in format
    wikiizer.format(formatter)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/parser/wiki.py", line 1102, in format
    formatted_line = self.scan(scan_re, line)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/parser/wiki.py", line 878, in scan
    result.append(self.replace(match))
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/parser/wiki.py", line 907, in replace
    result.append(replace(hit))
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/parser/wiki.py", line 354, in _word_repl
    return (self.formatter.pagelink(1, word, anchor=anchor) +
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/formatter/pagelinks.py", line 15, in pagelink
    FormatterBase.pagelink(self, on, pagename, page, **kw)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/formatter/base.py", line 82, in pagelink
    pagename = self.request.normalizePagename(pagename)
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/request.py", line 866, in normalizePagename
    if wikiutil.isGroupPage(self, page):
  File "/usr/local/lib/python2.4/site-packages/MoinMoin/wikiutil.py", line 616, in isGroupPage
    filter = re.compile(request.cfg.page_group_regex, re.UNICODE)
  File "/usr/local/lib/python2.4/sre.py", line 180, in compile
    return _compile(pattern, flags)
  File "/usr/local/lib/python2.4/sre.py", line 222, in _compile
    if not sre_compile.isstring(pattern):
RuntimeError: maximum recursion depth exceeded

Update:

Here is the 300 last lines of my apache error log: backtrace300.log

-- TheAnarcat 2006-09-30 21:36:46

• Those logs are not enough to see the recursion, please change your Page.py parsePageLinks method to use traceback.print_exc(200) and send us this (longer) traceback. A (minimal) sample of a page causing this error would be nice, too.

Steps to reproduce

1. clean the cache
2. make a search

Example

http://wiki.koumbit.net/

Details

Workaround

Deleting the cache from the pages that have corrupted caches makes search include those pages properly.

Discussion

Plan

• Priority:
• Assigned to:
• Status: fixed in 1.5 and 1.6 branch

CategoryMoinMoinBugFixed