PhpDig.net - View Single Post

baskamer · 12-19-2004, 03:27 AM

hello all!

i think i asked this before and saw at-least one other (unanswered) post on this topic, i never got an satisfying reply. I currently have a working solution, but i don't believe it is very elegant.

If you have pages removed from a site, search results can still point to those page, even after the site is respidered. This is not a situation we want. A brute force solution is to truncate most phpdig tables just before respidering. Allthough this does work it is not ideal (any searches done at that time will result in incomplete search results). This is acceptable for me, because the indexing is done at night and only takes about 2.5 hours. (±1000 pages), but if your site (or your isp slower) is much bigger this could become a problem.

Now my question: is there a build-in phpdig way to approach this problem. Is there any situation in which phpdig removes old spidered pages that should be considered garbage? I image it should work the following way: phpdig to spider a complete site every now and then, only indexing those that have the 'LIMIT_DAYS' configurable constant set. Then it should remove all pages from the indexes that are *not* in the temp table (why is this not a mysql TEMPORARY table anyway?).

This is just my idea of how this could work and it is probably is incorrect with the phpdig way, but i would really like someone to explain what the *best practise* way of doing this is...

thanks

bas

12-19-2004, 03:27 AM	#1
baskamer Green Mole Join Date: Feb 2004 Posts: 9	garbage collection hello all! i think i asked this before and saw at-least one other (unanswered) post on this topic, i never got an satisfying reply. I currently have a working solution, but i don't believe it is very elegant. If you have pages removed from a site, search results can still point to those page, even after the site is respidered. This is not a situation we want. A brute force solution is to truncate most phpdig tables just before respidering. Allthough this does work it is not ideal (any searches done at that time will result in incomplete search results). This is acceptable for me, because the indexing is done at night and only takes about 2.5 hours. (±1000 pages), but if your site (or your isp slower) is much bigger this could become a problem. Now my question: is there a build-in phpdig way to approach this problem. Is there any situation in which phpdig removes old spidered pages that should be considered garbage? I image it should work the following way: phpdig to spider a complete site every now and then, only indexing those that have the 'LIMIT_DAYS' configurable constant set. Then it should remove all pages from the indexes that are not in the temp table (why is this not a mysql TEMPORARY table anyway?). This is just my idea of how this could work and it is probably is incorrect with the phpdig way, but i would really like someone to explain what the best practise way of doing this is... thanks bas