linkchecker bugs

[Ian Smith-Heisters]

The linkchecker package says bugs should be reported here, so here goes:

1) Uncommenting any of the ignoreN= lines in the default linkcheckerrc 
results in the following cryptic error:

Traceback (most recent call last):
   File "/usr/bin/linkchecker", line 511, in ?
     config.read()
   File "/usr/lib/python2.4/site-packages/linkcheck/configuration.py", 
line 243, in read
     self.read_config(cfiles)
   File "/usr/lib/python2.4/site-packages/linkcheck/configuration.py", 
line 262, in read_config
     self.read_filtering_config(cfgparser)
   File "/usr/lib/python2.4/site-packages/linkcheck/configuration.py", 
line 437, in read_filtering_config
     linkcheck.log.error(
TypeError: error() takes at least 2 arguments (1 given)

By changing line 436 in configuration.py from
                 if len(ctuple)!=2:
to
                 if len(ctuple)!=1:
makes the error go away, and the software appear to behave as expected. 
Sorry, no patch since it's so simple, and I have no idea if this "fix" 
has other ramifications.

2) The IGNORE WARNINGS list from "$ linkchecker -h" incomplete. 
According to "grep -R tag= /usr/lib/python2.4/site-packages/linkcheck/*" 
, the complete list of ignore tags is

file-missing-slash
file-system-path
ftp-missing-slash
http-robots-denied
http-no-anchor-support
http-robots-denied
http-moved-permanent
http-wrong-redirect
http-empty-content
http-cookie-store-error
http-decompress-error
http-unsupported-encoding
http-unsupported-encoding
ignore-url
mail-no-addresses
mail-no-mx-host
mail-unverified-address
mail-no-connection
nntp-no-server
nttp-no-newsgroup
nttp-busy
url-effective-url
url-unicode-domain
url-unnormed
url-error-getting-content
url-anchor-not-found
url-warnregex-found
url-content-too-large

While I haven't gone through them all, http-robots-denied is definitely 
missing from the help output.

3) The default config file, linkcheckerrc, uses something like 
url-not-normed as an ignore flag. The correct flag would seem to be 
url-unnormed.

That's it. This package seems to suit me, but it seems a little old and 
crufty. Can anyone recommend a better package for doing automated link 
checking (and perhaps even accessibility checks)? I saw ht://check, but 
I didn't like that it stored everything in MySQL; I'd rather just get a 
nightly email of errors. And it hadn't been updated in even longer than 
linkchecker.

Thanks!
-- 
Ian Smith-Heisters
Webmaster
Marlboro College
(802) 451.7559 (x.559)
ismithhe at marlboro.edu
http://marlboro.edu
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3186 bytes
Desc: S/MIME Cryptographic Signature
URL: <https://lists.ubuntu.com/archives/ubuntu-users/attachments/20060427/771f13f2/attachment.bin>