A problem and a suggestion

I’ve had some problems with trying to create a feed where items contain only one image and nothing else. I’ve tried to do it with the following line but it seems to always return an empty description.

body: //img[contains(@src, ‘whatever’)]

However, when this should return more than one image it seems to work.

I’ve tried setting the prune and tidy options to no but that doesn’t help.

I’ve tried using something like

body: //img[contains(@src, ‘whatever’)]/…

or

body: //img[contains(@src, ‘whatever’)]/…/…

but this may include some other stuff besides the image that have to be stripped. Is it really this hard to get only one image?

Then another thing…

Would it be possible to have a GUID element added to feeds that are created from a web page (not from an RSS feed) that is not only the URL of the page. Perhaps it could be a combination of the page URL and some random string of characters.

When the webpage gets updated, the RSS reader (Tiny Tiny RSS) seems to only change the content of the previous item from the feed. I think it would be better if it would be treated as new item when the webpage is updated and if I understand correctly, having a different GUID would solve this.

Hi Matti, thanks for the report. We’re going to look into this and see what we can do for the next release.

As for detecting updates to pages. That’s not something that Full-Text RSS does yet. We have no record of the previous contents of the page we’re processing. So no way to know if there have been updates to it or not. Of course it’s possible to create a string based on the hash value of the content we’re processing, but I’d have to look into that some more.

Just an update to say the problem related to extracting a single item has been fixed in version 3. We hope to have a release soon.

Thanks again for reporting this.

After quick testing this seems to work fine now. Thanks for the fix.

Matti

Good to hear. Thanks for the update!