Enhancement idea: conditional ignores

[ethus3h]

It would be very handy to be able to apply ignore patterns based on the URL of the page the URL matching the pattern was found in.

For example:

/foo/bar.htm links to /foo/baz.htm and /foo/qux.htm

/foo/baz.htm links to /foo/blahblah-1.htm and /foo/blahblah-2.htm

/foo/qux.htm links to /foo/blahblah-3.htm and /foo/blahblah-4.htm

I don't want URLs matching /foo/blahblah-\d+.htm$ if they were linked from /foo/qux.htm, unless they were also linked from /foo/baz.htm

So, it would be very useful to be able to specify the source document URL for applying ignore patterns at link extraction time as opposed to at crawl time.

[ivan]

I agree and this probably isn't that hard to do.

Perhaps an optional second (tab-delimited) column in in the ignores file could specify the required source URL as another regexp.