Spider mode: Follow links in files linked via "file://"

[enc24]

Hello,

i have a problem indexing HTML-files linked via "file://". The spider doesn't follow the links in this files. Can I configure ZoomSearch to follow links in this files?

Thanks for an answer,

enc24

[Ray]

On the "Scan Options" tab of the Configuration window, there's a checkbox for "Scan files linked via 'file://' URLs in spider mode".

[enc24]

Hi Ray,

Thanks for the answer. I saw this option befor, and the file itself is scanned and added to the index. But unfortunately the spider do not follow links in this file!! Could you please check this? The file is a HTML-File.

[Ray]

Yes, this is the current designed behaviour. The "Scan files linked via file:// URLs" option will only index the content of files which have been linked with a file:// style URL. It will not crawl the page for subsequent links to follow.

We would consider adding this for a future version, but there is some complexity involved. First of all, the concept of a base URL needs to be enforced, and usually, the file:// link that has taken us off the original base URL already - in which case, at that point, how do we work out whats a URL we should follow, and what we should not? We can't just follow every link from it, otherwise one single link to an external website may end up attempting to index the whole Internet.

A new base URL would need to be determined, possibly from the file:// link, but that's not always obvious. And asking the user to specify all the possible file:// base URLs that he/she may be linking to is probably too complex and demanding for the end-user.