Wikipedia:Bots/Requests for approval/WugBot

The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was

Approved.

WugBot

Operator: Wugapodes (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 04:55, Monday, December 5, 2016 (UTC)

Automatic, Supervised, or Manual: Automatic

Programming language(s): Python

Source code available: Github

Function overview: Reads WP:GAN and posts the oldest nominations on Wikipedia:Good article nominations/Report

Links to relevant discussions (where appropriate): Wikipedia_talk:Good_article_nominations#Reports_not_being_updated

Edit period(s): daily

Estimated number of pages affected: 1

Exclusion compliant (Yes/No): No?

Already has a bot flag (Yes/No): No

Function details: The bot that usually compiles these reports, User:StatisticianBot seems to be down and the operator seems to be on break. This will serve as a growing replacement and open source back up should the other bot go down.

The bot reads into memory WP:GAN and, using regular expressions it picks out each entry and what kind of review stage, if any it is on. It then goes through each category of waiting for review, on review, on hold, or looking for a second opinion and determines which are the oldest in each category. It creates a report in the same style as Wikipedia:Good article nominations/Report with notes saying what parts this bot changed and when. It then reads Wikipedia:Good article nominations/Report, copies everything before the "Exceptions report" section, inserts the updated report, and then copies what was not updated below it.

It updates the backlog report and backlog report archive. It reads the 10 current entries, removes the oldest entry to the subpage Wikipedia:Good article nominations/Report/Backlog archive, and then adds the current day's entry at the top of the list.

An example output can be seen at User:Wugapodes/GANReportBotTest and User:Wugapodes/GANReportBotTest/Backlog archive.

Discussion

Approved for trial (21 edits or 7 days). Please provide a link to the relevant contributions and/or diffs when the trial is complete. — xaosflux ^Talk 03:00, 6 December 2016 (UTC)[reply]

After a few hours trying to get things working on Tools, the bot made its first edit. I expect it to run again at midnight, and it will edit once a day for the next 7 days. It should come in at 8 edits. Thanks for taking the time to approve the trial, and I'll be back with the results in a week. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 04:24, 7 December 2016 (UTC)[reply]

It would be more useful if the links were to the GAN page, as StatisticianBot did, rather than directly to the articles. This is a report on the GAN page, after all, and at least when I use it, I click on the Report link and am taken to the GAN section with the nomination. Is there a reason you've changed the link style? BlueMoonset (talk) 08:35, 8 December 2016 (UTC)[reply]

Getting links to the section of the GAN page where the entry is a non-trivial change. I decided that getting the list updated, even if the links are a little more difficult, would be the best first step. This will be my next goal. One thought I had was, instead of linking to the section, link to the article's talk or GAN page (if it exists). Do you think that would be an improvement? Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 16:59, 8 December 2016 (UTC)[reply]

I'd frankly find it less convenient; there's more information available from the GAN page. But I'm just one user; why not ask at WT:GAN to see what other people think. BlueMoonset (talk) 19:42, 8 December 2016 (UTC)[reply]

I also hope that in the list of all old nominations (30 days and more) that you can restore a way to distinguish between unreviewed, reviewed, on hold, and second opinion noms in this list. Thanks. BlueMoonset (talk) 08:40, 8 December 2016 (UTC)[reply]

Honestly I never even noticed those. They should be easy enough to add. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 16:59, 8 December 2016 (UTC)[reply]

Done. Should be included in the next run. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 17:49, 8 December 2016 (UTC)[reply]

Thanks! Good to have them back. I think, though, that your bot is not picking up all of the data it needs. For example, in the Oldest nominations section, you're missing Frankie Ruiz (which should be second), and London mayoral election, 2016, which should be fourth between the Duterte and Wolf nominations. These two are also missing from the Old nominations section, and many more with them (e.g., Simplifly Deccan, John Dacey, Samuel Kipi, and Jack of Fables). BlueMoonset (talk) 19:42, 8 December 2016 (UTC)[reply]

Wugapodes, another feature that would be worth restoring is the daily updates to Wikipedia:Good article nominations/Report/Backlog archive, which is useful for historical analysis. Thanks for taking this on, by the way; StatisticianBot was useful and it's great to have a backup/replacement in the works. Mike Christie (talk - contribs - library) 18:06, 8 December 2016 (UTC)[reply]

One step ahead of you, just put the finishing touches on that. @Xaosflux: This is my first BRFA so I'm not sure, but is it allowed that I add this feature during the trial? I assume yes since it's part of the task StatisticianBot undertook, but if you'd rather I complete a process for it, let me know. It simply takes the oldest of the ten entries on the page and moves it to the end of a subpage archive. It's in git around ln151 if you wanted to look it over. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 18:17, 8 December 2016 (UTC)[reply]

Wugapodes, I think StatisticianBot uses the "GAN counter" template to get the numbers for total nominations and nominations awaiting review (and maybe the other three values as well); at least, it's used on the Wikipedia:Good article nominations/guidelines page to get the most up-to-date numbers. These numbers are far higher than the numbers you came up with in your latest bot test (443 total, 373 waiting, vs. your bot's 256 total, 204 waiting); your bot is missing about 40% of what's out there. Does it make sense for you to use that template for the Backlog report numbers, which would, I think, give consistent results with what StatisticianBot produced? BlueMoonset (talk) 19:42, 8 December 2016 (UTC)[reply]

I thought those numbers were low. I don't use the GAN counter template, I use regular expressions which, honestly, is probably not the best decision but one that works--sometimes.I'll look at the template and see if I can get the results more accurately. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 21:30, 8 December 2016 (UTC)[reply]

@BlueMoonset: I have fixed the issue. It was indeed a problem with the regex (for those who care, I forgot to use a look forward) that is now fixed. The test runs give me 440 which is only 1 fewer than the number on the guidelines page. Seems like a much more accurate number. Should be ready for tonight's run, I'll reset the page and archive since the last number was incredibly inaccurate. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 22:40, 8 December 2016 (UTC)[reply]

@Wugapodes: thanks for asking - you may include this; please update your task description at the top of this page to include the additional function. — xaosflux ^Talk 19:43, 8 December 2016 (UTC)[reply]

Trial complete. I can't count, so it wound up doing an 8th day run. Hopefully that's not a deal breaker. Wugapodes [t^hɔk] [ˈkan.ˌʧɻɪbz] 20:02, 14 December 2016 (UTC)[reply]

Approved. As your bot is project space and low edit rate, you can begin prior to flagging. — xaosflux ^Talk 21:25, 14 December 2016 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.