Page MenuHomePhabricator

Pagelinks tables sometimes not being populated for newly-imported pages
Open, LowPublic


After T107969, I found lots of articles in azbwiki that after import has links.

SELECT count(distinct page_id) FROM page LEFT OUTER JOIN pagelinks ON (page_id = pl_from) WHERE pl_from IS NULL AND page_namespace in (0) AND page_is_redirect = 0 LIMIT 10;

Event Timeline

Mjbmr raised the priority of this task from to Unbreak Now!.
Mjbmr updated the task description. (Show Details)
Mjbmr added subscribers: Mjbmr, Krenair.
Mjbmr added a subscriber: Legoktm.

Saving null edits would work which I'm doing it right now.

I just tried importing User:This, that and the other/importtest from enwiki to testwiki. As soon as I imported the page, correctly listed the link from the newly import page.

In other words: can't reproduce...

I found lots of articles in azbwiki that after import has links.

Are you saying because they didn't have their links in the pagelinks table, they weren't counted? If newly imported pages aren't getting counted in the statistics, that's a separate issue, one that we have traditionally had a lot of trouble with, but that should have been resolved since the start of this year.

@TTO Not at all, only null edits do it.

Could you please be a little more specific? :)

@TTO When you save a page without changing the content, is called null edit.

Yes, I'm aware of what a null edit is. I'll try again:

  • What are you referring to when you say "only null edits do it"? What other things, besides null editing, are you expecting to "do it"?
  • Could you please give a clear statement of the actual behaviour you are seeing? And what you are expecting to see? In a clear, detailed "steps to reproduce" type format if possible.

@TTO I was very clear, probably pagelinks missed links or links were not added while importing from incubator.

So what you're saying is: "When I import a page from incubator to azbwiki, the links tables are not updated."

Please try to be a bit clearer. Speaking one sentence at a time is not helpful.

@TTO I'm not a talking person, you should be an expert to understand what I'm saying and I think you understood very well. But as I said, I'm not aware if that's with import or not, there is a debug needed here to see why pagelinks table is missing rows.

Legoktm lowered the priority of this task from Unbreak Now! to High.Aug 18 2015, 8:03 AM
TTO renamed this task from pagelinks table don't get updated on import to Pagelinks tables sometimes not being populated for newly-imported pages.Aug 18 2015, 8:05 AM

I tried importing an XML dump of some The Big Bang Theory season listings to testwiki. And lo and behold, WhatLinksHere did not update; see, for instance,

Actually, scratch that: [[Chuck Lorre]]'s link is always within a transclusion of a non-existent template, so it apparently doesn't get counted. Leonard Hofstadter, on the other hand, is working fine:

TTO lowered the priority of this task from High to Low.Oct 10 2015, 2:21 AM

No real evidence of what is going on here, or whether it is still happening.