That seems to have resolved the issue!
Tue, Jun 18
I'm going to try replacing DictCursor with SSDictCursor and see if that helps.
May 12 2019
Hi @Kelson , I think you gave extra parameters to your restart command. I tried:
Mar 28 2019
Okay, I had previously added a setenv command for adding the PERL5LIB to the environment, but I didn't add it for list2.fcgi and log.fcgi, which now have their own sections in their configs for the environment variable. Things seem to be working now
This isn't entirely accurate. The tool was in fact migrated to stretch last week. However, certain pages may still be erroring out, and that needs to be addressed. I will look at it later today when I have my laptop.
Jan 7 2019
And yes, the ORM I was speaking of was SQLAlchemy. It was much too slow with the session management strategy I was using, and manually managing multiple sessions and object eviction in order to speed it up would make the code too complicated in my opinion and negate the benefits of using the ORM.
Thanks for the comments. I realized that the test db never needs to be live in production and that helped me figure out the solution.
That looks promising for the Travis CI environment yes. I guess my problem is that I'm not developing locally, I'm developing directly on the Toolforge instance and I guess I just need to provision a database for test there?
Jan 6 2019
Dec 17 2018
The web tool (https://tools.wmflabs.org/enwp10) looks up to me. What specifically is not working?
Dec 3 2018
Restarted it again. Let me know if it works.
Oct 24 2018
I just restarted it. The tool seems to die every 30 days or so, this last happened on Sept 28.
Mar 10 2018
Actually, I believe plaintext would be more useful, but HTML would be a good starting point since it's reasonable to strip out.
Mar 31 2017
In fact the XML is not the problem, it's the wiki markup. From what I understand, the wiki markup is not defined by any formal grammar, so parsing it is non-trivial. Tools like mwparserfromhell help, but can be difficult to configure and not always accurate. There's also of course the problem of recursive template expansion.