Wikipedia talk:AutoWikiBrowser/Typos

Home
Introduction and rules
User manual
How to use AWB
Discussion
Discuss AWB, report errors, and request features
User tasks
Request or help with AWB-able tasks
Technical
Technical documentation

WT:AWB/T

Archives

This page has archives. Sections older than 40 days may be automatically archived by .

Tie scores being treated as false positives

1-1, 2-2, etc. scores are being overlooked for ndash replacement and with all the sports articles I work on, this is getting a little too frustrating. Tom.Reding - I found out you had made a change to make this a false positive in March 2020. Is there any way this can be made to bypass fewer of these (i.e. look for additional text to match)? A *lot* of tie scores aren't getting a correction (unless I catch it and perform the correction manually). Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 23:02, 3 February 2024 (UTC)[reply]

@StefenTower: the relevant part of rule "0–0", (?<!\b\1[-—]\1\b), was moved the next day into rule "2–1", precisely so that "0–0" can find draws and ties. If "0–0" isn't finding ties now, it's not because of that lookbehind; it's because "0–0" needs to be expanded with more relevant keywords. ~ Tom.Reding (talk ⋅dgaf) 13:33, 4 February 2024 (UTC)[reply]

Tom.Reding Thanks for the reply. (?<!\b\1[-—]\1\b) is still in "2–1", and stopping corrections of draws/ties that don't meet "0–0". "0–0" currently doesn't cover a lot of scenarios I'm seeing in my typo correction work, but I don't know why the draws/ties need to be avoided in "2–1" 's general case in the first place. Is there any harm from removing that code? That's what I'm driving at. Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 17:37, 4 February 2024 (UTC)[reply]

@StefenTower: the reason that lookbehind exists is because the rule was incorrectly catching journal volume numbers, e.g. "Some Journal 5-5", preceded by "5-4" and succeeded by "5-6" etc., which should not be changed/en-dashed. ~ Tom.Reding (talk ⋅dgaf) 17:46, 4 February 2024 (UTC)[reply]

I can see why we don't want a journal volume number en-dashed, but I don't quite understand why volume numbers that look like draws/ties are more problematic than those where the numbers aren't equal (e.g. "5-4"). If the "5-4" is in the middle of the series instead of "5-5", wouldn't it be falsely corrected? Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 17:58, 4 February 2024 (UTC)[reply]

Tom.Reding Is it that a draw/tie is guaranteed to not be a number range if referring to a volume? If that's the case, I understand it better now after thinking more about it. At any rate, I am working on more cases to convert the ties to use ndash outside of this false positive check. But I wish this check wasn't needed - we really miss a lot of genuine draws/ties as it stands. Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 00:01, 5 February 2024 (UTC)[reply]

@StefenTower: do you have a list of pages where ties are known to have been missed? ~ Tom.Reding (talk ⋅dgaf) 11:59, 5 February 2024 (UTC)[reply]

@Tom.Reding I don't keep a list of these. I just keep seeing these missed, and I have to correct them manually. It's any draw/tie that the "2–1" rule would ordinarily catch but doesn't due to the false positive catch. If you run AWB typo checks in sports articles especially, it can get rather frustrating. Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 16:26, 5 February 2024 (UTC)[reply]

Here is an example. While it shows "17-17" and "7-7" being corrected, that happened only because I saw in my AWB viewer they weren't corrected, and I manually placed the en dash there. Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 16:56, 5 February 2024 (UTC)[reply]

@StefenTower: rule amended to find "17-17" and "7-7" in that example, but with enough specificity, I think, to avoid most/all freeform journal citations, which are unlikely to end the line at a journal volume (if so, they can/should have the page # appended). ~ Tom.Reding (talk ⋅dgaf) 21:32, 19 June 2024 (UTC)[reply]

Tom.Reding Thanks for addressing this. While I usually have no issue reading RegEx, it has been many months since I looked at this. So, I guess what you did is add code to catch more cases for the "0–0" rule, so any tie ranges within those cases won't be skipped in the "2–1" rule. Do I have it right? Stefen Tower_{s among the rest!} ^{Gab • Gruntwerk} 05:25, 20 June 2024 (UTC)[reply]

@StefenTower: correct for the "0–0" rule, which I did isolate & test. I did not look at the "2–1" rule, since it wasn't relevant to, nor triggered for, the "17-17"/"7-7" example. Also, rules aren't necessarily run in the order they appear on the typo page, and so should be coded in a way that is independent of the firing sequence of other rules. ~ Tom.Reding (talk ⋅dgaf) 18:39, 20 June 2024 (UTC)[reply]

Misspelled CNN reporter name

I don't know if this is the right place for this, but there are 47 articles that misspell a CNN reporter's name in references as Arinne de Vogue instead of Ariane de Vogue. Annoyedhumanoid (talk) 02:35, 10 June 2024 (UTC)[reply]

Eurodance

Shouldn't Eurodance be capitalized always? In the article itself, it always remains capitalized even when it's not at the start of the sentence. Wiiformii (talk) 03:27, 27 July 2024 (UTC)[reply]

Washington, D.C

To find: insource:/Washington, D\.C[^\.]/
Search: Washington, D\.C[^\.]
Replace: Washington, D.C.$1

Headbomb {t · c · p · b} 23:49, 19 August 2024 (UTC)[reply]

U.S.A

To find: insource:/U\.S\.A[^\.[^\.]/ (~4500 hits)
Regex Search: U\.S\.A[^\.]
Regex Replace: U.S.A.$1

Headbomb {t · c · p · b} 06:39, 20 August 2024 (UTC)[reply]

Were they won

I've just changed 18 instances of "were they won" to "where they won". No false positives, worth a rule? Ϣere SpielChequers 08:12, 22 August 2024 (UTC)[reply]

Pick up "where the won" as well. Neils51 (talk) 16:47, 22 August 2024 (UTC)[reply]