Re: not ignoring content (leave those files alone!)
"Linda W. (that's swishey, not squishey!)" <[email protected]>
| Newsgroups | gmane.comp.web.swish-e |
|---|---|
| Message-ID | <[email protected]> |
> Yep, NoContents should index just the file path. Sounds like that is > working since you're seeing "5 words" for those files. ---- NoContents indexes part of the _content_ if the file has the default filetype (HTML*) -- it looks for some random HTML fields. > As for why one file appeared to take 5 minutes to index perhaps the file > immediately following or preceding it took 5 minutes to index? Might be > worth trying to index that one file by itself to verify. --- I used Process Explorer to look at the process -- it still had the file open to the "rogue .mbf" file. The files before and after were fairly short. The problem was/is that for every file extension one includes in "NoContents" one must also change the filetype from the Default HTML*. Changing it to TXT should be sufficient to stop the scanning, since HTML* is currently the only type that is scanned for Indexing keywords I'd feel safer being able to assign some "Opaque" filetype to files, and not calling a file "TXT" when it really isn't (future wishlist item?)... How about a "ReallyNoContent" switch that doesn't look for HTML tags for headers within a binary....:-).. -l