Jump to content

User talk:Ohconfucius/script/archive1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Sources italicisation changes

[edit]

From what I gather, the area of interest is in User:Ohconfucius/test/Sources subscript2.js where you have a table of work names (presumably subject to italics) and only wish to italicise when one of those names match. Otherwise, the name would be left alone (or I suspect it would be reasonable practice not to automagically de-italicise names in case someone intended that).

First, you may wish to be careful of the order of outer/inner looping. Your outer loop ("i") goes through the table, while the inner loop ("j") goes through each potential reference title expression. In most cases, there are likely fewer matching title expressions in an article than table items, so you could consider switching the inner/outer loops for (typically) more efficiency. If you isolate the candidate expression finely enough, you can consider breaking the inner (italic table) loop once a match is made, then add the italics.

Complicating matters is that parameters among the citation templates can be quite ambiguous for formatting e.g. title in {{cite book}} would almost always be italicised, while title of {{cite news}} would be in quotes (which I believe is applied as part of those templates). If you are handling details for each type of citation template, I may have missed that in my initial look at the script. Dl2000 (talk) 04:22, 17 November 2012 (UTC)[reply]

  • Thanks. I think you understand correctly. As I wrote in the documentation, the source names are included on an opt-in basis, but has a few 'generic' regex catches.

    I think a single loop structure may well be adequate, and this was implemented for subscript1 (although I don't know anything about efficiency). For subscript2, it's true that one loop may eventually suffice. I wasn't aware of what purpose the "i" and "j" served and will now try swapping the two loops around. I'm a real dunce when it comes to programming – would that be a simple swap-over of the lines (i=0; i<table.length; i++) for (j=0; j<expr.length; j++) and vice versa?

    I'm currently treating the issue globally, ignoring the different uses in {{cite book}}. That's why I've also chosen to use |work= as the default parameter to italicise; I was also assuming that the titles of sources are likely to be titles of books in an insignificant number of instances within these walls, though I'm prepared to revisit the scripts treatment of source names within |title= should this prove to be a problem. -- Ohconfucius ping / poke 05:21, 17 November 2012 (UTC)[reply]

  • Just think of this as a nested loop - the the inner loop runs for each instance of the outer loop. Actually, switching the inner and outer loops won't automatically improve efficiency on its own (i times j equals j times i); but it may be a better algorithm to consider that the main loop should run through the expressions. Then within the loop, the necessary substitutions can be done. Haven't done a table as intense as the one you're proposing, but might be inclined to look at associative arrays but you may find that a bit heavy for now. If you do the table search in the inner loop, be sure to exit the loop once you've found a match i.e. if you are trying to match on "Financial Times", once it is found in the table, get out of the loop and don't try to test anything else. That could be done by adding a "found" flag to the loop condition - note this example is not fully-coded - but note how the for loop and found flag might work:
  found = 0; // stays 0 until something is found
  for (j=0; ((j < table.length) && (found == 0)); j++) {
     if ( <<<key matches table[j]>>> ) {
        found = 1; // no point staying in the j loop any more
     }
  }
Hope this provides some further insight. Dl2000 (talk) 03:19, 23 November 2012 (UTC)[reply]

Publisher to agency

[edit]

I notice that the new script automatically converts the "publisher" parameter to "agency" for groups like Reuters, RIA Novosti, etc. Sometimes these groups are legitimately the publisher, though (each has its own website with news stories on it). More importantly, changing these could create a problem when editors are using the common "cite web" template, which has no "agency" parameter; the publisher would then disappear from the citation. I like the new script generally, but I wonder if this change will do more harm than good.

Thanks for another great tool! -- Khazar2 (talk) 13:38, 30 November 2012 (UTC)[reply]

  • I never realised that {{cite web}} doesn't hold that parameter. Thanks for pointing that out. I think it's often necessary to have more than one field that isn't italicised, so I'll need to think about that a little. Maybe the simplest will be to just get someone to add that parameter to the template... -- Ohconfucius ping / poke 16:55, 30 November 2012 (UTC)[reply]
That would be a good fix if it's possible. Anyway, thanks for all your efforts. -- Khazar2 (talk) 17:12, 30 November 2012 (UTC)[reply]

New York Daily News

[edit]

Hi, you might be interested to know that your script caused a wikilink in a citation for the New York Daily News, which redirects to Daily News (New York), to instead go to the Daily News disambiguation page. Maybe the script needs a tweak for this paper? See http://en.wikipedia.org/enwiki/w/index.php?title=Pussy_Riot&diff=prev&oldid=527463305 line 127. Regards, MaxBrowne (talk) 13:48, 12 December 2012 (UTC)[reply]

Mysterious disappearance

[edit]

Removed "work=Evening Herald}}</ref><ref>{{cite news|" in this edit. -- Ohconfucius ping / poke 08:29, 7 January 2013 (UTC)[reply]

Priyanka Chopra edits

[edit]

Your script changed a lot of news sources from publisher="x" to work=x, which is good I guess, but it removed italics from Daily News and Analysis, without using work=x. DNA is a newspaper and is italicized in its own article. You might want to look at that. BollyJeff | talk 13:27, 14 January 2013 (UTC)[reply]

And what the heck is this: "rôle" in Sholay? BollyJeff | talk 13:33, 14 January 2013 (UTC)[reply]
Thanks for that. I was trying some stuff out. Now fixed. -- Ohconfucius ping / poke 14:09, 14 January 2013 (UTC)[reply]
Thank you. Better check all the articles you touched with that. I just fixed some in Rajinikanth. BollyJeff | talk 14:17, 14 January 2013 (UTC)[reply]
Use of the word 'rôle' (with circumflex) isn't actually improper. It is often seen in English, and is apparently de rigeur in certain circles, so its existence within these walls is not a problem. The loss of italics of DNA was because it was incorrectly put into "|publisher=" but was toggled into italic using the quote marks; my script removed the toggling but DNA was not in my publications dictionary. I will check previous edits for any instances of such italicisations undone in this way. -- Ohconfucius ping / poke 14:38, 14 January 2013 (UTC)[reply]

rages separated by dashes caught up in realignment

[edit]

diff: [[PACAF|Far East Air Forces]] 29 June - 15 November 1950 becomes : [[PACAF|Far East Air Forces]] 29 – 15 June November 1950

$x replaced by comma

[edit]

@Peter Garrett. -- Ohconfucius ping / poke 07:39, 30 January 2013 (UTC)[reply]

New line after {Use dmy dates}

[edit]

Hello,

Some of the edits adding {{Use dmy dates}} also add an unnecessary new line after the template, like this, which leaves a blank space at the top of the article. Is that intentional?

Best regards, Intery 11:16, 31 January 2013 (UTC)

Do you plan on removing it from the already edited articles? They look odd beginning with an empty line. --Intery 16:09, 31 January 2013 (UTC) — Preceding unsigned comment added by Intery (talk • contribs)
The blank lines don't do any harm, and a simple edit to remove such is not encouraged, so I will amend the code to do the job as and when the articles are next edited by the script. -- Ohconfucius ping / poke 03:07, 17 February 2013 (UTC)[reply]

BAG?

[edit]

Have you run these scripts past the WP:Bot Approval Group? NE Ent 23:10, 16 February 2013 (UTC)[reply]

  • The scripts are in daily use, and are used by more than a few editors. Updating is a dynamic and very reactive process. Even if submitting the script for approval doesn't put bureaucracy before maintainability, the considerable user-judgement that is necessary makes it a semi-automatic tool rather than a bot. There are a large number of modules and over a dozen custom functions that users may choose from. So, although some tasks within the script are potentially capable of being run within a bot, the script has no chance of passing as a bot as it is currently configured, much as I would love it to be. Having said all that, there is currently an open BRFA that is related. -- Ohconfucius ping / poke 03:01, 17 February 2013 (UTC)[reply]
  • See WP:BOTASSIST, specifically "While such contributions are not usually considered to constitute use of a bot, if there is any doubt, you should make an approval request." When editors start opening WP:AE requests, even if they're summarily closed, it's prudent to cross your i's and dot your t's. I'm not suggesting you ask BAG to sign off on the implementation of the script, but rather the intent or tasking. It would be a one time wikitime investment instead of repeatedly having to address concerns like you had to do at User_talk:Ohconfucius/archive25#MOSNUM. NE Ent 14:27, 17 February 2013 (UTC)[reply]
  • Thanks for your advice. I need to think about how I'm going to frame the request. You probably realise that I know the various style guidelines inside out, and that the editor who filed that last AE request was dead wrong and didn't want to admit it. IMHO, he manifests ownership tendencies. As soon as he found out I had been involved in some Arbcom case, he dropped all assumptions of good faith and backed off from civil discussion. So I'm not so much worried about my scripts per se – the real issue is the use to which they are put. -- Ohconfucius ping / poke 15:35, 17 February 2013 (UTC)[reply]

De-wikilinking

[edit]

I understand the need to delink generic words like railway or republic. But I don't understand the delinking of specific links such as United Kingdom or in Deputy Prime Minister of Zimbabwe this edit. Those seemed useful. Babakathy (talk) 10:00, 20 May 2013 (UTC)[reply]

Likewise here, delinking Prime Minister of Zimbabwe? Or here delinking birthplace but not here and not delinking death place here.
Or for that matter this edit which delinks Romanian language but not French???? Babakathy (talk) 10:17, 20 May 2013 (UTC)[reply]
  • Thanks for your query. I do indeed run a script that I have tried to make contextual. In addition to removing common words that are often linked, I have written it to remove links to major languages such as English, French, Spanish. It doesn't remove languages like Romanian. Also targeted are chained links, so that [[South Africa]]n [[President of South Africa|President]] [[Thabo Mbeki]] will be stripped down to South African President [[Thabo Mbeki]]; similarly, for chained link [[Prime Minister of Zimbabwe|Prime Minister]] [[Morgan Tsvangarai]], the script will remove the position, leaving the name thus: Prime Minister [[Morgan Tsvangarai]].

    I have the script remove country links where it follows another link. That's why you see [[Harare]], [[Zimbabwe]] become [[Harare]], Zimbabwe. OTOH, [[Alamein Farm]], [[Beatrice, Zimbabwe]] remains unchanged because there is no country link. I trust that answers your query. Regards, -- Ohc ¡digame!¿que pasa? 13:01, 20 May 2013 (UTC)[reply]

Thanks for clarification. If there's one or two places where I think a link is still(again) needed, will the script come back and remove it again? Babakathy (talk) 15:10, 20 May 2013 (UTC)[reply]
By all means put the links back. I'll take a look at what you have redone and try to avoid the article. Whilst I might pass by with the script again, there are about two or three million other articles I hope to visit before I come back to that one. -- Ohc ¡digame!¿que pasa? 12:01, 21 May 2013 (UTC)[reply]

EngvarB script doesn't show up

[edit]

Hi! I tried installing your EngvarB script by following the installation instructions, copying the relevant line etc.; but it hasn't appeared. Your excellent date script showed up fine when I installed it. Are there any known incompatibilities? (I also have the Regex editor and Segregate references scripts installed). Or have I just copied it wrong? Or rather, what information would I need to provide for you to begin to be able to answer that question? OS 10.6.8, Safari 5.1.9.

Just in passing, were you aware that Bearcat goes round changing {{EngvarB}} to {{Use British English}} – as here, for example? Justlettersandnumbers (talk) 21:53, 21 May 2013 (UTC)[reply]

The script now shows up correctly and works too. Many thanks indeed, Justlettersandnumbers (talk) 09:35, 22 May 2013 (UTC)[reply]

Vice-Presidents

[edit]

There's a Template:Infobox Vice President, which unfortunately no longer codes when replaced by Template:Infobox Vice-President, see this script edit. I've fixed the Zimbabwe pages, but no idea how many others the script might have changed. Babakathy (talk) 15:58, 23 May 2013 (UTC)[reply]

Bug with Fix sources script

[edit]

Hi again! When I ran your Fix sources script on Christina Hendricks, it tries to change: 

|title=Mad Men's Christina Hendricks voted best-looking woman in US | Entertainment | People |publisher=The First Post

to 

|title=Mad Men's Christina Hendricks voted best-looking woman in US | Entertainment|work= People

It also doesn't change |publisher=Esquire to |work=Esquire. Thanks! GoingBatty (talk) 03:40, 3 June 2013 (UTC)[reply]

  • When I wrote that line of code, I made it note that it catches "sources misplaced by reflinks preceded by spaced hyphens or '&#124;'" (the vertical pipe symbol). I've now disabled it. Let me know if you get an increased number of work titles not coming through.

    I seem to recall 'Esquire' was deliberately left off, but I can't recall why. I'll try to recall, and also to fix it. Thanks, -- Ohc ¡digame!¿que pasa? 04:46, 3 June 2013 (UTC)[reply]

Could you please double check your edit to Christina Hendricks? Your script changed |publisher=[[Guardian Media Group|GMG]] to |GMG]]. I've fixed the article. Thanks for your continued work on this script! GoingBatty (talk) 22:01, 3 June 2013 (UTC)[reply]
Surprise piped link. Will amend code accordingly, thanks. -- Ohc ¡digame!¿que pasa? 02:19, 4 June 2013 (UTC)[reply]

Just passing it on

[edit]
Whack!
You've been whacked with a wet trout.

Don't take this too seriously. Someone just wants to let you know you did something silly.

for needing someone to make this edit for you. GoingBatty (talk) 03:57, 3 June 2013 (UTC)[reply]
Note that this trout is only half the size of the trout on my talk page. GoingBatty (talk) 04:30, 3 June 2013 (UTC)[reply]
Thanks for bearing the brunt of the trouting! -- Ohc ¡digame!¿que pasa? 04:49, 3 June 2013 (UTC)[reply]

Bugs with Fix sources script

[edit]

1. When I run the Fix sources script on Eric Clapton, it wants to change ''[[Billboard (magazine)|Billboard]]'' to ''[[Billboard]]''. GoingBatty (talk) 02:05, 9 June 2013 (UTC)[reply]

Similarly, it wants to change ''[[Q Magazine]]'' to ''[[Q]]'' on Rod Stewart. GoingBatty (talk) 02:26, 9 June 2013 (UTC)[reply]
Confirmed the fix - thanks! GoingBatty (talk) 23:47, 9 June 2013 (UTC)[reply]

2. It removes the italics when changing ''[[The Daily Telegraph|The Telegraph]]'' to [[The Daily Telegraph]]. That would be fine to do if this was in a |work= or |newspaper= parameter, but in this example there's no citation template used. GoingBatty (talk) 02:35, 9 June 2013 (UTC)[reply]

Oops, I should have mentioned that this happens on Phil Collins. GoingBatty (talk) 23:47, 9 June 2013 (UTC)[reply]
 Done Now fixed. -- Ohc ¡digame!¿que pasa? 14:26, 13 June 2013 (UTC)[reply]

Another surprise link?

[edit]

Hi again! When running the Fix sources script on Minecraft, it wants to change |work=[[About.com]]|publisher=[[IAC (company)|InterActiveCorp]] to |publisher=[[About.com]]|InterActiveCorp]]. Another surprise link? Thanks! GoingBatty (talk) 00:36, 11 June 2013 (UTC)[reply]

Could you please double check that this was fixed? I'm getting the same error, and I don't see any changes to User:Ohconfucius/script/Sources.js. Thanks! GoingBatty (talk) 03:41, 11 June 2013 (UTC)[reply]
 DoneSorry, I changed the test script only. Should be done now. -- Ohc ¡digame!¿que pasa? 03:55, 11 June 2013 (UTC)[reply]
Which part of the script changes |work=[[Metacritic]]|publisher=CBS Interactive to |publisher=[[Metacritic]]|publisher=CBS Interactive - I don't see "Metacritic" or "CBS" in User:Ohconfucius/script/Sources.js. Thanks! GoingBatty (talk) 22:53, 11 June 2013 (UTC)[reply]
The library is located at User:Ohconfucius/test/Sources_subscript1.js. -- Ohc ¡digame!¿que pasa? 00:59, 12 June 2013 (UTC)[reply]

"Use dmy dates" script request

[edit]

Hi Ohconfucius! When I used your "Use dmy dates" script on You Broke My Heart So I Busted Your Jaw, it didn't change the accessdate in the {{Allmusic}} template. Is there any way to have the script update those accessdates as well? Thanks! GoingBatty (talk) 03:06, 15 June 2013 (UTC)[reply]

Any thoughts on this request? Thanks! GoingBatty (talk) 01:20, 14 May 2015 (UTC)[reply]

Logic for adding "colwidth=30em"?

[edit]
Hi Ohconfucius! Should there be some logic added to your script so it doesn't add "colwidth=30em" to
when there are a small number of references, such as Ringo Rama? Thanks! GoingBatty (talk) 22:04, 28 June 2013 (UTC)[reply]

Astana

[edit]

I'm not sure what happened in this edit, but on my screen, html codes are visible in the Geography and climate section. Best, 67.163.124.220 (talk) 15:56, 29 June 2013 (UTC)[reply]

(talk page stalker) Looks like the script changed negative numbers beginning with "-" to "−" in {{climate chart}}. I reverted the edits to restore the table. When I ran your Fix sources and All dates to dmy scripts, it didn't cause the same issue. GoingBatty (talk) 23:23, 29 June 2013 (UTC)[reply]

Bug with All dates to dmy script

[edit]

When I run your All dates to dmy script on Sun Myung Moon, it wants to change "$500,000" to ",00,000". GoingBatty (talk) 13:25, 5 July 2013 (UTC)[reply]

Titles to add to Sources.js library

[edit]

Hi Ohconfucius! Is the Sun Sentinel notable enough for you to add to your script? Thanks! GoingBatty (talk) 03:41, 6 June 2013 (UTC)[reply]

Me again! Could you please make a slight change to User:Ohconfucius/test/Sources_subscript1.js so that it will also change thedenverpost.com to The Denver Post? (see 2012 Aurora Shooting reference 62). Thanks! GoingBatty (talk) 01:41, 14 June 2013 (UTC)[reply]

Hi again! Is the Pittsburgh Tribune-Review big enough for you to add to your script that changes |publisher= to |work=, such as in List of furry conventions? Thanks! GoingBatty (talk) 23:56, 29 June 2013 (UTC)[reply]

Hello again! Is Fast Company notable enough for you to add it to your script to change |publisher= to |work=? (e.g. in the references for The Hershey Company) Thanks! GoingBatty (talk) 15:20, 4 July 2013 (UTC)[reply]

Your script now changes |publisher=Fast Company to |work=Fast Company (Magazine), which would be displayed as Fast Company (Magazine). Would it be better to change it to |work=Fast Company, so it's displayed as Fast Company? Thanks!
It needs to have the parenthetic disambiguator there or else it would screw up the wikilinks if they exist – Fast Company is a disambiguation page. I think the problem was because I capitalised the 'M'. Now fixed. -- Ohc ¡digame!¿que pasa? 23:43, 4 July 2013 (UTC)[reply]
Maybe it would be more appropriate to have two rules: One for Fast Company and fastcompany.com, and a second for [[Fast Company]] and [[Fast Company (magazine)|Fast Company]]. I'd think you'd have the same issue with Time (magazine), Life (magazine), etc., but I haven't noticed that. Thanks! GoingBatty (talk) 13:23, 5 July 2013 (UTC)[reply]

As you did with The Denver Post above, could you please make a slight change to User:Ohconfucius/test/Sources_subscript1.js so that it will also change theguardian.co.uk to The Guardian? (see "Back in the U.S.S.R."). Thanks! GoingBatty (talk) 15:25, 5 July 2013 (UTC)[reply]


Hi again! Is The Village Voice notable enough for you to add to your Fix sources script? For example, on the Semi-Pro article, it would be great if your script could change |publisher=Village Voice to |work=The Village Voice. Thanks! GoingBatty (talk) 22:30, 17 July 2013 (UTC)[reply]

Date range formatting on "ALL date to"

[edit]

I'm still having the problem I mentioned at User talk:Ohconfucius#Date ranges. The article I am trying to fix is 41st Canadian Parliament, and for good measure, my vector.js is here. And sorry for not noticing your editnotice about posting bugs here the first time around. -Rrius (talk) 11:23, 18 July 2013 (UTC)[reply]

  • No harm done. In fact, I now have splintered talks that make monitoring a bit more tiresome. But anyway...

    I loaded your vector, and the MOSNUM script runs but is not removing the 'present' like it was doing before. I suspect that perhaps you have not refreshed the cache of your browser. Each browser has a slightly different way of clearing the cache. Follow the relevant instructions at the top of your vector file, and that should do the trick. -- Ohc ¡digame!¿que pasa? 11:49, 18 July 2013 (UTC)[reply]

    • Now it is removing the spaces around the dash in "June 11, 2011 – present". I'm not sure what this stylistic choice has to do with making date formats uniform on the page. -Rrius (talk) 02:09, 21 July 2013 (UTC)[reply]
      • That is correct per WP:ENDASH. "The en dash in a range is always unspaced, except when the endpoints of the range already include at least one space". In this case, one end of the range is "present", which is without spaces. So the lack of spaces is appropriate in this case. -- Ohc ¡digame!¿que pasa? 02:57, 21 July 2013 (UTC)[reply]
        • Sorry, but no. Your script is removing spaces from "YYYY – present" even when the YYYY is preceded by MMMM, D (or D MMMM). When "January 4, 2001" is one of the endpoints, because it includes a space, it demands a spaced dash. What the material quoted is saying is that if there is a space in an endpoint, the dash should be spaced. It is not saying that if both of the endpoints contain spaces, then the dash should be spaced. Note "include at least one space". -Rrius (talk) 11:30, 25 July 2013 (UTC)[reply]
          • No, not by my reading of the guidance. WP:ENDASH says "The en dash in a range is always unspaced, except when the endpoints of the range already include at least one space" (my emphasis). It doesn't say "one of the endpoints" but simply "endpoints" – the plural for me implies both end points need to have spaces for the dash to need spaces on either side. Maybe Tony can enlighten? -- Ohc ¡digame!¿que pasa? 13:39, 25 July 2013 (UTC)[reply]
            • That is a weird reading. It is saying that if the class of endpoints for the given range contains at least one space, then the dash should be spaced. That is standard usage. When you write "12 January 2010 – 2011", you are clearly signaling that "10 January 2010" and "2011" have equal standing as elements in the range. If you write "12 January 2010–2011", you are showing 2010 and 2011 as elements of the range, which forces the reader to correct the range for him- or herself. You are either over thinking this or or getting caught up in a semantic distinction that doesn't exist. Either way, even if you won't accept that your interpretation is wrong, perhaps you will at least recognise that your interpretation isn't necessarily correct, so you shouldn't make it part of the standard action of the main tools in the script. -Rrius (talk) 03:07, 31 July 2013 (UTC)[reply]
  • Well-spotted, Rrius. I've changed MOSDASH, which has been tampered with since the agreed version was posted some time ago. The example given, 1492? – 7 April 1556, is correct, and was not consistent with the lead. The lead now reads "The en dash in a range is always unspaced, except when at least one component of the range includes at least one space." Tony (talk) 08:52, 31 July 2013 (UTC)[reply]

Allmusic is now AllMusic

[edit]

Since the Allmusic article has been moved to AllMusic to reflect the website's capitalization, could you please also change User:Ohconfucius/test/Sources_subscript1.js so your script doesn't change "AllMusic" to "Allmusic" (as it does at Memory Almost Full)? Thanks! GoingBatty (talk) 05:45, 20 July 2013 (UTC)[reply]

More suggestions

[edit]
  • Hello again! After running your Fix sources script on Cory Monteith, I made these edits manually - you might want to incorporate some of these publications and variations into your scripts. Also, if you run the script against the article now, you'll see that it wants to change |title='Glee' Cast, Creators Remember Cory Monteith With 'Emotional Celebration' to |title=Glee' Cast, Creators Remember Cory Monteith With 'Emotional Celebration. Is it possible to only remove the beginning and ending quotes if there are no quotes in the middle? Thanks! GoingBatty (talk) 01:29, 27 July 2013 (UTC)[reply]
  • The quote removal is actually bad practice, and should be either removed from the script, or adjusted as follows: For opening single quotes, it's best to use the {{-'}} template to get spacing between the title field's inherent double quote and the title's opening single quote. For ending single quotes, the template is {{'-}} to get the right spacing before the field's inherent ending double quote. BlueMoonset (talk) 16:15, 27 July 2013 (UTC)[reply]

Date change errors

[edit]

Ohconfucius, in this edit, your script changed three 2013 dates to 2012. Please fix the problem caused by the script there, determine why it happened, and check and fix any other date errors introduced by the bug in your script. Thank you. BlueMoonset (talk) 16:16, 27 July 2013 (UTC)[reply]

I notice you have done extensive editing using your script in the past couple of days, yet you have not yet fixed the damage it did in this case. Presumably, this means you haven't fixed the script to prevent additional errors from being made. How can you justify such behavior? BlueMoonset (talk) 22:23, 29 July 2013 (UTC)[reply]

iOS in the Car

[edit]

Why have you Anglicised the dates in this article using this tool? I created the article (I also happen to be British), however, it's about software from a US company, hence dates used were correctly done using US data formatting. Please explain why &/or undo within 24 hours, or I'll manually do so myself accordingly. Thank you. Jimthing (talk) 22:04, 29 July 2013 (UTC)[reply]

Cricket articles

[edit]

Hi. I can see your bot is doing some useful work on cricket articles, but it's also creating some problems with grammar, the insertion of American English and the replacement of hyphens with en dashes in some places where that hasn't been usual style. I started a small discussion that explains further at WT:CRIC if you'd care to drop by and give us your views. Most of the changes look sensible and make us more orderly, which has got to be good. Thanks. Johnlp (talk) 22:06, 13 August 2013 (UTC)[reply]

Bug report

[edit]

Hi Ohconfucius! When I run your Fix sources script on Jeff Bezos, it changes "Amazon.com" to "Amazon.com.com" in the |title= of one of the references. Is this something you'd be able to fix? Thanks! GoingBatty (talk) 02:58, 15 August 2013 (UTC)[reply]

Lost script

[edit]

Hi! I seem to have lost your MOSNUM script. It was working fine until a few days ago, but now does not show up. I don't think I've changed anything, so was wondering if you had, or if there has been some change in the way the wiki works? Thanks, Justlettersandnumbers (talk) 13:00, 16 August 2013 (UTC)[reply]

Same here - see Wikipedia:Date formattings/script/MOSNUM dates/bugs#Not loading in toolbox. GoingBatty (talk) 21:18, 16 August 2013 (UTC)[reply]
Thanks, good to know it isn't something I did, which was my my first assumption. Justlettersandnumbers (talk) 21:25, 16 August 2013 (UTC)[reply]
Thanks to User:Keith D for fixing this! GoingBatty (talk) 22:06, 18 August 2013 (UTC)[reply]

problem with one script

[edit]

Not sure which of your scripts is the culprit, but it's leaving the moniker of "Script-assisted style fixes: mainly date formats." The complaints are coming in at Tennis and Golf projects for the removal of International nationality flags in the infobox (which we use by consensus), and the removal of the initial term of "tennis" being linked in tennis prose. Could you please tweak it before anymore damage is done? Thanks, as there is a lot to correct already. Fyunck(click) (talk) 21:11, 16 August 2013 (UTC)[reply]

  • the edits were manual and the removals deliberate. according to my understanding, the rather liberal use of flags is a violation of wp:mosicon. Also, my understanding is that the mos is the result of global consensus, and that cannot be negated by local consensus. Regards, -- Ohc ¡digame!¿que pasa? 18:53, 19 August 2013 (UTC)[reply]
    • Actually guidelines and they do not say "don't use". We use it to show international tournament nationality, which is a requirement for the Majors. Also the link to "tennis" on first use is very important and is being removed. Thanks. Fyunck(click) (talk) 19:59, 19 August 2013 (UTC)[reply]

Bug reports

[edit]

Hi Ohconfucius! Two bug reports for your Fix sources script for today:

  • When I run Fix sources on Paris and Kevin Smith, it wants to change |publisher=Time Out to |work=Time Out (magazine) - I don't think that "(magazine)" is needed when the parameter value is not a wikilink.
  • When I run Fix sources on Sam and Diane, it wants to change |title='Are You There, Chelsea?' Star Laura Prepon Compares New NBC Show To 'Cheers' to |title=Are You There, Chelsea F.C.?' Star Laura Prepon Compares New NBC Show To 'Cheers - Besides the incorrect removal of the single quotes, I hope you could restrict the F.C. rule from editing titles.

Thanks! GoingBatty (talk) 15:47, 17 August 2013 (UTC)[reply]

Suggestions

[edit]

Hi Ohconfucius! After running your Fix sources script on the Mariah Carey article, I made the following additional reference fixes. You may want to include some of these in your script. Keep up the good work! GoingBatty (talk) 22:58, 1 September 2013 (UTC)[reply]

Bug report - Ringo Rama

[edit]

Hi Ohconfucius! In this edit to Ringo Rama, your scripts changed one of three reference names from "defense" to "defence". I'm all for changing the text of the article to British English, but would it be possible to change the script so it doesn't change the spelling of the ref names (or at least change them all)? Thanks! GoingBatty (talk) 03:46, 7 September 2013 (UTC)[reply]

  • Thanks. I can't work out why it's doing that. Usually, strings within quote marks are protected, as are strings within ref tags. There are four instances of that ref in the article, and it converted only one. I'll take a closer look when I have time. -- Ohc ¡digame!¿que pasa? 05:12, 7 September 2013 (UTC)[reply]

Bug report - script changes within song title

[edit]

Hi Ohconfucius! I reverted your edit to Tripping the Live Fantastic because it changed "Back in the U.S.S.R." to "Back in the USSR". GoingBatty (talk) 12:52, 8 September 2013 (UTC)[reply]

Please try The Bondman (novel) again

[edit]

Hi again! Could you please try running your style script against The Bondman (novel) again? It appears that there are still a number of curly quotes to be changed to straight quotes. Thanks! GoingBatty (talk) 01:49, 10 September 2013 (UTC)[reply]

  • I'm splitting the rule up so that the leading open quote is treated independently of the end quote. I'm not confident enough to simply swap all of these (‘`´’′) for '. Could you give me your risk assessment if I were to do that? Also, I hate to admit that I actually quite like directional quotes. Not only do they look nice, it would be much easier to code for... ;-) -- Ohc ¡digame!¿que pasa? 02:33, 10 September 2013 (UTC)[reply]
Oh, I thought your goal was to change all of the quotation marks to straight marks. In your most recent edit to The Bondman (novel), I see you changed “ to " but did not change the corresponding ” to ", which looks kind of odd now. Thanks! GoingBatty (talk) 12:27, 11 September 2013 (UTC)[reply]
I still don't know for sure what people use these curlies for, so I'm wary of making simple substitutions for the straight quote marks. Maybe I'm being overcautious, but I've noticed certain non-character marks including quote marks used for example in urls, which is why I asked for a risk assessment from your end. I've further tweaked the code, which should now be more exact in picking them up, but I still dare not make conversion wide open. -- Ohc ¡digame!¿que pasa? 00:14, 12 September 2013 (UTC)[reply]
I agree that changing URLs should be off limits, just like you don't change date formats in URLs. I appreciate your tweaks, but noticed that your changes to The Bondman (novel) changes two ” to " without changing the corresponding “ to ", and still left several “...” pairs unchanged. Thanks! GoingBatty (talk) 16:50, 12 September 2013 (UTC)[reply]

Do not perform edits for -ize -> ise

[edit]

Your edits a few days ago to OCAD University have broken a number of links on the page due to your changing -ize to -se. Please do not perform these edits unattended. Actually please do not perform these edits at all; we normally spell these words with -ize, not -ise.

The same applies for your date format changes. For dates we follow the US convention of month, day, year when the month is spelled out. Please stop changing perfectly unambiguous ISO 8601 dates to British format.

If your script changes the page again I’ll consider your edits vandalism. Thank you very much.—Al12si (talk) 17:18, 12 September 2013 (UTC)[reply]

Suggestion

[edit]

Hi Ohconfucius! Is The Courier notable enough for adding to your Fix Sources script to change |publisher=The Courier.co.uk to |work=The Courier (e.g. the Minecraft article). Thanks! GoingBatty (talk) 02:27, 23 September 2013 (UTC)[reply]

The rule for The Courier works great when I run the Fix SOURCES script on Minecraft, but could you please check the rule for fixing references to Metacritic? It seems your script handles it differently depending on whether it preceeds or follows the publisher parameter. Thanks! GoingBatty (talk) 23:44, 9 October 2013 (UTC)[reply]

Bug report

[edit]

In the changeset at https://en.wikipedia.org/enwiki/w/index.php?title=Athens_Metro&oldid=573445902, the script did not detect images in the infobox. It should leave file names alone.

Sorry for the hurry, I only just spotted a bug while doing something else important. --Marianian(talk) 05:07, 23 September 2013 (UTC)[reply]

Script and diffs

[edit]

Whilst most of your changes are good, a few neutral and only the occasional one debateable, the script you use does have an unfortunate side effect. The script seems to lose the double space between sentences. This is not a problem for the rendered output, but does have two undesireable effects: reduced readability of the source (when viewed in monotype) and excessive diff size. See http://en.wikipedia.org/enwiki/w/index.php?title=History_of_Rochester%2C_Kent&diff=574294190&oldid=566916776 for example. Look at the four modified paragraphs under Toponymy and not a single red mark appears; very difficult to check.

In passing, I've reverted two minor changes. IMHO the titles of 19C articles when quoted directly (as, for instance, in a bibliography) ought to follow the original style. In the present case (Felling mine disaster) two articles from the Newcastle Guardian of 1847 were changed to "Dreadful Colliery Accident at Felling" and "The Late Explosion at Felling Colliery" which no 19C typesetter would have ever countenanced. Regards, Martin of Sheffield (talk) 12:28, 24 September 2013 (UTC)[reply]

  • I've fixed the double-space issue of the script, so that it will no longer remove these where they are preceded by a full stop. OTOH, I don't see the downcasing of "At" is a real issue even though it's an ancient and obsolete practice, but I'm not bothered by its reinstatement although this would appear to violate MOS:CAPS. Regards, -- Ohc ¡digame!¿que pasa? 02:38, 25 September 2013 (UTC)[reply]
Thanks for sorting out the double spacing. As regards the titles: WP:ALLCAPS bullet point 1 mandates sentence or title case - the trouble being that title case is open to a wide range of interpretations. Anyhow, thanks again for the double space issue. Regards, Martin of Sheffield (talk) 09:46, 25 September 2013 (UTC)[reply]
Another little issue that has come to light: when you change author to publisher or work you break the Harvard citation linkage. Specifically CITEREFDurham_Mining_Museam2012a became CITEREF2012a which is not found when the user clicks on the short form citation. I know it is an effort, but perhaps when the script changes the author, then it could insert a line such as | ref = CITEREFDurham_Mining_Museum2012a into the citation template. See Template:Citation#Anchors_for_Harvard_referencing_templates for a full explanation. One wonders if there is there any reason for the author parameter other than for corporate author since real people ought to be referenced by last and first? I don't know about North American library practice, but when I worked in a UK library we certainly indexed corporate authors under their name in the author catalogue, otherwise you would end up with drawer after drawer of "Anon" which does not make life easy for readers. Is there a MOS statement that bans this practice? Regards, Martin of Sheffield (talk) 11:54, 25 September 2013 (UTC)[reply]
I can easily stop converting the |author= field. Actually, you'll be pleased to know that that part of the edit was entirely manual, and I wasn't aware of that issue when I made that change. The concept of corporate authors doesn't get used a lot here, in my experience, and most people, myself included would use |publisher= for consistency/simplicity's sake. I guess I just learned that the {{citation}} template works that way, so I will avoid it in future. -- Ohc ¡digame!¿que pasa? 14:56, 25 September 2013 (UTC)[reply]
It is good when discussion leads to us both learning! I'm probably a bit old fashioned, when I worked in the county central reference library we had two large (may be 8' x 4') units containing wooden card index drawers. One one side of the reception desk was the classified (Dewey) index, the other the author index. Cards were typed on a manual machine, computerisation was only being talked about. Regards, Martin of Sheffield (talk) 19:57, 25 September 2013 (UTC)[reply]

Removing work parameter for Hung Medien

[edit]

Hi Ohconfucius! When I run your Fix Sources script on The Prodigy discography, it removes most |work= parameters in citations with |publisher=Hung Medien, but does not remove |work=charts.org.nz. (Which part of your script controls this?) Thanks! GoingBatty (talk) 23:50, 26 September 2013 (UTC)[reply]

  • That will be in the main part of the script, in the later cleanup section.

    The early (pre-)cleanup rules should convert the italics into straight text, and then subscript 1 should turn |work=charts.org.nz into |publisher=charts.org.nz and then the second |publisher= should be removed by the later cleanup function. I'll look at it later to see what needs to be done. -- Ohc ¡digame!¿que pasa? 02:02, 27 September 2013 (UTC)[reply]

Oh, OK - I was looking for code that contained the word "chart", but your code is more general than that. I just noticed that the domain names your script removes have one dot (e.g. norwegiancharts.com), and the one it doesn't change has two dots (charts.org.nz). Hope this helps! GoingBatty (talk) 04:05, 27 September 2013 (UTC)[reply]

Script problems

[edit]

Hi, in this edit two problems were caused

  1. dashes were changed in a file name
  2. a comma was inserted in an OS reference making a bad map link

Keith D (talk) 12:00, 7 October 2013 (UTC)[reply]

Thanks for that. Just to note that OS references are not always wrapped in the {{gbmapping}} template. Keith D (talk) 12:22, 8 October 2013 (UTC)[reply]
I feared as much. The script should now only convert space separators into commas where preceded by a 'Euro' (€) notation. -- Ohc ¡digame!¿que pasa? 12:40, 8 October 2013 (UTC)[reply]

Fixing newspapers and magazines

[edit]

Hi again! When I run your Fix SOURCES script on United States federal government shutdown of 2013, it removes the italics from |publisher=''[[The Hill (newspaper)|The Hill]]'' but doesn't change it to |work=. Would it be safe to change any instance of |publisher=[[Foobar (newspaper)|Foobar]] and |publisher=[[Foobar (magazine)|Foobar]] to |work=? Thanks! GoingBatty (talk) 23:50, 9 October 2013 (UTC)[reply]

[edit]

Hi I noticed on my watchlist you had removed the wikilinks to Somerset on a range of articles. I know about the policies on not linking common terms, like countries but for many users around the world Somerset, as a county, may not be so well known and they may have other local meanings - see Somerset (disambiguation). I had let it go until now but an article I wrote about a local listed building appeared on the main page today (in DYK) and another user relinked Somerset, England with this edit. Can you point me to any discussion about linking or removing links to counties?— Rod talk 18:36, 15 October 2013 (UTC)[reply]

  • WP:Linking tells us to avoid linking "everyday words understood by most readers in context; the names of major geographic features and locations". England would seem to fall into that category of word. In addition, the section on General points on linking style advises us to: "When possible, avoid placing links next to each other so that they look like a single link... Consider rephrasing the sentence, omitting one of the links, or using a more specific single link instead". My script seeks to apply these rules. Although you did not supply a diff, my script is written to only unlink counties where the preceding term is linked. Chances are, your text probably read something like "[[Taunton]], [[Somerset]], [[England]]", which was unlinked to "[[Taunton]], Somerset, England" through application of those principles. In the example given, I would consider relinking "Somerset, England" neither desirable from an aesthetics ergonomic viewpoint, nor is the link rendered necessary when the Barrow Gurney article is already linked to Somerset down the line. Regards, -- Ohc ¡digame!¿que pasa? 02:13, 16 October 2013 (UTC)[reply]
I think your rationale is wrong here. The example given in the General points on linking style that you selectively quote from is that of Irish chess championship and it quite reasonably suggests that linking to Irish Chess Championship is likely to be better, since it gives a link through to the specific event, rather than to three generic and rather unrelated articles. I don't think a geographical address or location is the same kind of thing at all: there is a natural hierarchic sequence of placenames and to unlink some of them, particularly where, as in the case of Somerset, there could be confusion, is not helpful. There are hundreds of thousands of articles with these kinds of links in them, possibly millions, and my view is that they are helpful. WP:Linking is, in any case, a guideline, not an instruction. You should discuss before diving in. Johnlp (talk) 10:58, 16 October 2013 (UTC)[reply]

Latest edit on Graffiti on the Train

[edit]

Hi Ohconfucius, thanks for using your script on the album's article :) I have a concern however. In the 'Personnel' section you removed the links for Stereophonics member's respective instruments, on Wikipedia:WikiProject Albums/Album article style guide#Personnel it states that they should be "linked on the first occurrence only" in the section. Other than that, again, thanks for the edit :) Ultra Violet Light 17:52, 20 October 2013 (UTC)[reply]

Bug report

[edit]

Hi Ohconfucius! Any idea why your Fix Sources script wants to delete |work=Hartford Courant from Sandy Hook Elementary School shooting? Thanks! GoingBatty (talk) 22:02, 11 November 2013 (UTC)[reply]

 Done bug now fixed. -- Ohc ¡digame!¿que pasa? 03:26, 12 November 2013 (UTC)[reply]
And if you have any scripts that will change the recently added YYYY-DD-MM on this same article to mdy, that would be great too. Thanks! GoingBatty (talk) 22:06, 11 November 2013 (UTC)[reply]
 Done One of the MOSNUM script modules should change those dates. The non-conversion may have been due to the existence of dates with erroneous format (yyyy-dd-mm)! -- Ohc ¡digame!¿que pasa? 03:26, 12 November 2013 (UTC)[reply]
Could you please check that bug again? Your Fix sources script wants to remove |work=Hindustan Times from Shreya Ghoshal. Thanks! GoingBatty (talk) 02:58, 14 November 2013 (UTC)[reply]
Oh, what a clot am I! I adjusted the test script without changing the production script. Duh! -- Ohc ¡digame!¿que pasa? 03:03, 14 November 2013 (UTC)[reply]
Ah, much better - thanks for the quick service! GoingBatty (talk) 03:21, 14 November 2013 (UTC)[reply]

Jus Hus Memorial

[edit]

"The huge monument depicts victorious Hussite warriors and Protestants who were forced into exile 200 years after Hus and a young mother which symbolises national rebirth."

The first part of the sentence does not relate to the last part of the sentance. "The huge monument depicts victorious Hussite warriors and Protestants who were forced into exile (.period??) 200 years after Hus (WHAT?) and a young mother" "WHAT??"

"young mother (,comma??) which symbolises national rebirth." (then what happened)

Sounds like 2 or 3 sentences were combined??.

Opin45 (talk) 11:00, 14 November 2013 (UTC)[reply]

IAEA, U.S., co-operate

[edit]

You recently applied your scripts to the article on the IAEA. Among other things, this imposed British spellings, and changed U.S. to US. The latter is an error. The United States does not use "US" as an abbreviation. We use "U.S." Also, I don't believe "co-operate" is the proper spelling. At a minimum it is not the spelling used by the IAEA itself. I plan to reverse those. I note this as a suggestion that you change your script. NPguy (talk) 18:41, 16 November 2013 (UTC)[reply]

MOS:ACRO states that either "U.S." or "US" is acceptable, and In articles with "UK", "UAE", etc., use "US". Since the International Atomic Energy Agency article uses "UN" and "IAEA", is that justification to use "US" instead of "U.S." in the article? A quick search on iaea.org shows they use both "U.S." and "US" in their web pages. GoingBatty (talk) 20:57, 16 November 2013 (UTC)[reply]
Use of the form with full stops is declining, and if I'm not mistaken, the CMOS now deprecates that cumbersome form. -- Ohc ¡digame!¿que pasa? 01:47, 18 November 2013 (UTC)[reply]

Bug report - not all wikilinks with parentheses contain disambiguators that should be hidden

[edit]

Hi Ohconfucious! Your Fix sources script now updates links with parentheses in them. While sometimes this is great (e.g. Billboard (magazine) to [[Billboard (magazine)|Billboard]]), it causes problems with links to works where the parentheses are part of the title. For example, on Jimi Hendrix, your script tries to change Voodoo Child (Slight Return) to [[Voodoo Child (Slight Return)|Voodoo Child]]. You might want to restrict your fix somehow - maybe only links that end with "magazine)" or "newspaper)" or another disambiguator. Thanks! GoingBatty (talk) 22:49, 17 November 2013 (UTC)[reply]

Bug report - does not remove all instances of "Bot generated title"

[edit]

Hi again! I noticed that your Fix sources script doesn't remove all of the "Bot generated title" comments in Batkid. Thought you'd want to know. Thanks for continuing to improve these valuable scripts! GoingBatty (talk) 00:21, 20 November 2013 (UTC)[reply]

  • Thanks. You seem to imply that some but not all instances are removed. I can't explain why that happens. I've noticed that some of the rules don't seem to work that well even though they work in theory and in the regex tester. Case in point are the rules for removing citations with Wikipedia, Facebook and Twitter links to the extent that I had to create function Ohc_noverify within the script. I suspect the problem may be due to other as yet undetected bugs in the script. -- Ohc ¡digame! 01:20, 20 November 2013 (UTC)[reply]

Script trouble

[edit]

Hi Ohconfucius, somehow, this edit broke the table formatting, resulting in the first column to be omitted. You might want to investigate, and perhaps go through your contributions list to check whether the script caused problems elsewhere, too. Keep up your good and valued work! Schwede66 03:56, 23 November 2013 (UTC)[reply]

Bug report - Time Out

[edit]

Hi Ohconfucius! Could you please tweak your script so it properly fixes (or ignores) Time Out (company) in The Beatles in India? Thanks! GoingBatty (talk) 05:25, 6 December 2013 (UTC)[reply]

I agree that the publisher should be removed, but your script changes |publisher=[[Time Out (company)|Time Out]] → |Time Out]]. Could your script remove the whole link? GoingBatty (talk) 04:36, 7 December 2013 (UTC)[reply]
 Done Finally managed to figure out what was causing the problem. -- Ohc ¡digame! 04:01, 11 December 2013 (UTC)[reply]

Bug report - removing title parameter

[edit]

Hi again! On 2009–11 Toyota vehicle recalls your Fix SOURCES script wants to change |title=CNBC.com |publisher=CNBC.com to |publisher=CNBC.com. Instead of removing the title parameter, could it instead replace it with the "ARTICLE TITLE GOES HERE" message that you do for other references? (Sorry that I don't know the actual words off the top of my head. Thanks! GoingBatty (talk) 04:32, 14 December 2013 (UTC)[reply]

Same request for The Doors and |title=RIAA.com. Thanks! GoingBatty (talk) 18:33, 14 December 2013 (UTC)[reply]
Belated thank you for making this change! GoingBatty (talk) 01:53, 14 May 2015 (UTC)[reply]

|work= to |publisher=

[edit]

At this edit the script inappropriately converts |work= to |publisher= within {{cite news}}. |work= is an alias of |newspaper= and is an appropriate alternative when the source isn't a newspaper: BBC, CNN, Reuters, AP, etc.

I have reverted that part of the edit.

—Trappist the monk (talk) 12:26, 19 December 2013 (UTC)[reply]

although the ones you list are not newspapers, none of them should be italicised per MOS:ITALICS. Reuters, AP get put into |agency= by my script. toggling italicisation with two pairs of single quotemarks gets removed by bot. what to do?? -- Ohc ¡digame! 13:32, 19 December 2013 (UTC)[reply]
While I agree that the sources I listed aren't "newspapers", in today's world they are essentially synonymous with newspapers. MOS:ITALICS seems to agree:
Online magazines, newspapers, and news sites with original content should generally be italicized (Salon.com or The Huffington Post).
Another conversation towards the top of this page also comments on the Reuters / AP / AFP etc. issue. I concur with that editor who points out that these agencies often have their own websites and can be the directly cited source. When an editor is citing one of these agencies directly (not citing a newspaper that ran an agency provided article) then the appropriate parameter is |work=. When an editor is citing a newspaper that ran an agency provided article then |agency= is the appropriate parameter along with |work= (or one of its aliases).
Wikipedia's markup in parameter values should be avoided if at all possible in CS1 citations because such use runs the risk of corrupting the citation's COinS metadata. The next question I have is: Should the value in |agency= render in italicized font? That is a question that might better be raised in a larger venue that this and should be implemented (if it is implemented) in Module:Citation/CS1.
—Trappist the monk (talk) 14:08, 19 December 2013 (UTC)[reply]

Question - Scribd

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Scribd, it wants to remove the |publisher=Scribd parameters. Is this by design, or is this a bug? Thanks! GoingBatty (talk) 00:19, 31 December 2013 (UTC)[reply]

Fix SOURCES no longer working?

[edit]

Happy New Year Ohconfucius! When I've gone into edit mode and tried to run your Fix Sources script recently (Firefox 26.0), nothing seems to happen. I've tried clearing my cache, but that didn't fix the problem. Any suggestions? Thanks! GoingBatty (talk) 01:53, 9 January 2014 (UTC)[reply]

  • I've been trying to understand the problem. I last used my test script without problems two days ago. I've rolled back changes to versions of early December, but I still can't get it to load. I tried the production version after realising the problem with my test script, but noticed it was down, so I'm a bit at a loss. When was the last time you used it? -- Ohc ¡digame! 02:49, 9 January 2014 (UTC)[reply]
Looks like it was this edit from January 5. GoingBatty (talk) 03:29, 9 January 2014 (UTC)[reply]

Bug report - Super Bowl XLVIII

[edit]

Hi Ohconfucius! When I tried running your Fix SOURCES script on Super Bowl XLVIII, it made the following incorrect changes:

  • Changed |publisher=''[[Sports Illustrated]]'' to |publisher=[[Sports Illustrated]] instead of |work=[[Sports Illustrated]]
  • Changed |work=Huffington Post |date=January 21, 2014 to |work=Huffington Postington date=January 21, 2014
  • Changed |publisher=''[[NY Daily News]]'' to |work=[[Daily News (New York)|Daily News]] without adding |location=New York
  • Removed <div style="float:left; width:48%;"></div>, which messes up the formatting of the rest of the article

Your help to resolve these issues would be greatly appreciated, as always. Thanks! GoingBatty (talk) 13:55, 3 February 2014 (UTC)[reply]

Work on other projects

[edit]

At this wonderful page, I noticed your script links are red. Might want to set it to prefix them where necessary. --Gryllida (talk) 01:52, 7 February 2014 (UTC)[reply]

Questions about Fix SOURCES

[edit]

Hi again! Some questions about your Fix SOURCES script:

  • When I run the script on iPhone 5S, it wants to change |publisher=CBS Interactive|work=CNET to |publisher=CBS Interactive. Should it instead remove the publisher and keep the more specific |work=CNET? (There are several variations of this behavior when processing this article.)
  • Also on iPhone 5S, it wants to change |location=[[Moscone Center|Moscone West]], [[San Francisco]], [[California]] to |location=Moscone West, San Francisco, [[California]]. I'm surprised you're not removing the link for California. What logic are you using to remove the wikilinks?

Thanks! GoingBatty (talk) 02:43, 12 February 2014 (UTC)[reply]

  • The script will first convert |work=CNET to |publisher=CNET, and then remove the second placed |publisher=, which in this case would be |publisher=CNET. If the order of your example is reversed, the |publisher=CBS Interactive would be removed instead.
  • I wrote the script to address what goes on in reference sections, and not linking specifically. But the rules will blanket remove links within |location=. I only anticipated there would ever be two links. I guess I neglected the more anal-retentive instincts of some editors. ;-) Removal of the link of [[California]] is targeted in the Common Terms script. -- Ohc ¡digame! 03:52, 12 February 2014 (UTC)[reply]
  • I've now patched my test script and ran it on the iPhone5S article. I want to test it out more thoroughly. If all goes without hitch, I will port the changes to the production script maybe in two weeks. -- Ohc ¡digame! 09:56, 13 February 2014 (UTC)[reply]

Bug report - The Verge

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on The Verge, I get these strange results:

  • It changes www.area17.com to area17.com within a |title= parameter.
  • It changes |publisher=theverge.com|accessdate=September 3, 2012 to |work=The Vergeaccessdate=September 3, 2012

Could you please see if you can fix these bugs? Thanks! GoingBatty (talk) 03:40, 2 March 2014 (UTC)[reply]

Mixed content blocking with Sources.js

[edit]

User:Ohconfucius/script/Sources.js gives me a mixed-content blocking notice in my address bar whenever I load a page following installation; I think it comes the http-explicit protocol in User:Ohconfucius/script/foreigndates.js: 

importScriptURI('http://meta.wikimedia.org/enwiki/w/index.php?title=User:Pathoschild/Scripts/Regex_menu_framework.js&action=raw&ctype=text/javascript');

I'd appreciate it if you could change it to 

importScriptURI('//meta.wikimedia.org/enwiki/w/index.php?title=User:Pathoschild/Scripts/Regex_menu_framework.js&action=raw&ctype=text/javascript');

if you get a moment, and see if that works. Not terribly harmful – still works and all – but it might hinder functionality (haven't checked thoroughly). Thanks, Meteor sandwich yum (talk) 05:25, 5 April 2014 (UTC)[reply]

MOSNUM dates

[edit]

Just FYI. I have installed the script to my commons.js and ran it in the edit window. The edit summary said it ran. But the "changes" view showed no differences. Firefox 28.0, Windows 8.1 When I saved, it was a null edit. However, I got it to work by unclicking WikEd under Preferences/Gadget. — Maile (talk) 11:55, 9 April 2014 (UTC)[reply]

  • @Maile66:Using your common.js, I get the buttons I expect. I was also able to run the script on an article with dates and it works both ways (dates converted to dmy and to mdy) as desired. But I don't have WikiEd installed anywhere in my interface. An edit summary is usually a clear indication that the script has run its full course. You were astute enough to disengage WikiEd to find the underlying conflict. I wasn't aware of it in the past. It's not something I know how to resolve at script level. I will ask around. In the meantime, you might have to continue disabling WikiEd to use my script. If you need to have the formatting functions run concurrently with the MOSNUM fixes, you might be interested to know that I have other complementary scripts. -- Ohc ¡digame! 13:41, 9 April 2014 (UTC)[reply]
Yeah, you have some cool tools. The issue with WikEd is apparently because of a Code Editor from Aug-Sept 2013. It doesn't seem to like JS editing.— Maile (talk) 21:22, 9 April 2014 (UTC)[reply]
Thanks. It's slightly more reassuring to know that it's not just my script that has incompatibility with WikEd. -- Ohc ¡digame! 02:13, 10 April 2014 (UTC)[reply]

Forcing image size

[edit]

It seems that one of your scripts changes image sizes to fixed values. This violates IMGSIZE. RockMagnetist (talk) 14:35, 11 April 2014 (UTC)[reply]

I was going by the summary in this edit. I suppose the editor could have added a few changes. RockMagnetist (talk) 17:47, 11 April 2014 (UTC)[reply]
  • Thanks. I can once again confirm that the script doesn't resize images. The edit summary of "Script-assisted fixes" deliberately allows for some manual intervention, which seems to have been the case. I know that Tony frequently makes manual changes to articles whilst making his script-based edits, and I think he rightly decided that the default size for that lead image was too small. -- Ohc ¡digame! 01:48, 12 April 2014 (UTC)[reply]
Manual, of course. And there are indeed good reasons, as provided for in the policy and guideline, for resizing. Those pics are plain indecipherable at 220. And we do not cap models, hypotheses, theories, theorums, etc. Tony (talk) 02:53, 12 April 2014 (UTC)[reply]
O.k., fair enough. A lot of sources capitalize all the words in PREM, but now I see that many just capitalize Earth. I suppose Preliminary Reference Earth Model needs to be changed as well. I do wonder whether having a fixed image size messes things up on some devices. RockMagnetist (talk) 19:21, 12 April 2014 (UTC)[reply]

Bug report - The University of Chicago Law Review

[edit]

Hi Ohconfucius! On Thomas F. Bayard your Fix SOURCES script wants to change |journal=The University of Chicago Law Review to |publisher=The University of Chicago Law Review, even though University of Chicago Law Review seems to be the name of the journal. GoingBatty (talk) 03:29, 13 April 2014 (UTC)[reply]

@Ohconfucius: Any thoughts on this bug? Thanks! GoingBatty (talk) 01:49, 29 May 2014 (UTC)[reply]

Comment on Sources subscript3.js

[edit]

FYI - it appears you have "Variety" and "Wired" listed twice in the first section of User:Ohconfucius/test/Sources subscript3.js. GoingBatty (talk) 15:37, 13 April 2014 (UTC)[reply]

Bug report - running Fix SOURCES on BP

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on BP, it tries to change | work=Huffington Post Canada | date=12 February 2014 to | work=Huffington Postington date=12 February 2014. Maybe the section of User:Ohconfucius/test/Sources_subscript1.js that contains 'huff(ington ?|)post(?: ?canada| uk)' should be 'huff(?:ington ?|)post(?: ?canada| uk)' instead? Thanks! GoingBatty (talk) 23:14, 13 April 2014 (UTC)[reply]

Thanks for that fix. Now when I run your Fix SOURCES script on BP, it tied to change |title=The Board |publisher=BP to |publisher=The Board. GoingBatty (talk) 19:39, 18 April 2014 (UTC)[reply]
@Ohconfucius: Any thoughts on this second bug? Thanks! GoingBatty (talk) 01:51, 29 May 2014 (UTC)[reply]
Confirmed - thank you! GoingBatty (talk) 16:23, 29 May 2014 (UTC)[reply]

Bug report: Austal

[edit]

Script may be replacing template material with stray span class constructs, at least for [1] Dl2000 (talk) 22:15, 15 April 2014 (UTC)[reply]

Bug report - On the Media

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Lolita, it wants to delete |publisher=[[On the Media]]. Maybe it would be better to change it to |work=[[On the Media]]? Thanks! GoingBatty (talk) 19:33, 18 April 2014 (UTC)[reply]

@Ohconfucius: Similar issue - When I run your Fix SOURCES script on Oil shale in Estonia, it wants to delete two instances of |publisher=Ministry of Economic Affairs and Communications. Thanks! GoingBatty (talk) 01:31, 28 May 2014 (UTC)[reply]
 Done -- Ohc ¡digame! 13:10, 28 May 2014 (UTC)[reply]
@Ohconfucius: Thanks for fixing the "On the Media" bug. Any thoughts on the second bug here? Thanks! GoingBatty (talk) 01:54, 29 May 2014 (UTC)[reply]
I thought I fixed it yesterday, but I must have gotten confused over something else or another browser window. I've tested it just now and the Ministry string is now not removed. -- Ohc ¡digame! 02:04, 29 May 2014 (UTC)[reply]
Thanks! GoingBatty (talk) 02:48, 29 May 2014 (UTC)[reply]

Bug Report - Noah (2014 film)

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Noah (2014 film), it wants to change |work=[[Time (magazine)|Time]]|publisher=[[TIME]] to the incorrect |work=[[Time (magazine)|Time]]|Time]]. Should it just remove the publisher altogether? Thanks! GoingBatty (talk) 01:39, 22 April 2014 (UTC)[reply]

@Ohconfucius: Any thoughts on this bug? Thanks! GoingBatty (talk) 01:56, 29 May 2014 (UTC)[reply]
Thank you! GoingBatty (talk) 16:24, 29 May 2014 (UTC)[reply]

Spaces

[edit]

Re: Space neutral – there should be no impact on the disposition of spaces before or after parameters in edit mode

Citation templates with no spaces can be very difficult to read. In particular, word-wrapping both in the edit window and in diffs does not work in a useful manner. Long URLs are not a problem. Wrapping occurs on at least colon, slash, hyphen and period. I would like to suggest that where there are currently no spaces around any of the pipe symbols that a space is inserted before each |pipe symbol. - 79.67.241.210 (talk) 08:58, 23 April 2014 (UTC)[reply]

  • Thanks for your feedback. Inserting and removals of blank spaces by bot or script is problematic and highly controversial. Even if each edit makes a change to the article in read mode, spaces inserted systematically at each pipe symbol would flood the diffs from which editing can be checked. I would prefer to keep the script space neutral for these reasons. -- Ohc ¡digame! 09:46, 28 May 2014 (UTC)[reply]

Bug report - The Marshall Mathers LP

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on The Marshall Mathers LP, {{start date|2000|5|16}} (which displays as May 16, 2000 (2000-05-16)) is changed to 2000-5-16, which does not meet MOS:DATEFORMAT. Also, could you please set up your script to protect the {{Listen}} template, so your script doesn't remove the quotation marks from its |title= parameter? Thanks! GoingBatty (talk) 13:05, 23 April 2014 (UTC)[reply]

  • I've disabled that template removal from the Sources script. As for the Listen template, I seem to recall another similar case that we decided to leave. Perhaps we could have the template output changed to include the quote marks systematically? -- Ohc ¡digame! 09:41, 28 May 2014 (UTC)[reply]
@Ohconfucius: Thanks for disabling the removal of {{start date}}. Template:Listen#Examples shows several instances where automatically adding quotation marks around the entire title would not be appropriate. Thanks! GoingBatty (talk) 01:59, 29 May 2014 (UTC)[reply]

Bug report - 1500 or Nothin'

[edit]

Hi Ohconfucius! When I ran your Fix SOURCES script on 1500 or Nothin', it wants to change |title=Artist Profile to |work=Artist Profile. Thanks! GoingBatty (talk) 01:25, 1 June 2014 (UTC)[reply]

Confirmed - thanks! GoingBatty (talk) 01:36, 2 June 2014 (UTC)[reply]

Bug report - Posted by:

[edit]

Hi Ohconfucius! When I ran your Fix SOURCES script on Sleeping Dogs (video game), it wanted to change |author=Posted by: to |author=by:. Looks like an issue with the "eliminating time of day" and/or the "removing redundancies" part of User:Ohconfucius/script/Sources.js. Thanks! GoingBatty (talk) 18:49, 9 June 2014 (UTC)[reply]

Possible bugs in script when editing Absolutely Fabulous

[edit]

Hi, The script appears to have changed "Use British English|date=July 2011" to "EngvarB|date=June 2014" rather than just adding the new template. Since the article's talk page does not have a "British English" template, the connection with British English, as opposed to, for example, "British English Oxford spelling" has been lost. Peter Loader (talk) 14:12, 12 June 2014 (UTC)[reply]

The script changed "The 23 July 2012 episode featured the 2012 Summer Olympics which were being held in London that week.", which makes sense, with "23 July 2012 episode featured the 2012 Summer Olympics which were being held in London that week." which does not make sense (at least to me). I think that it may need changing to be better at handling a date used as an adjective. Peter Loader (talk) 14:12, 12 June 2014 (UTC)[reply]

Question about Twitter citations

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on The Interview (2014 film), it replaces a {{cite news}} template for a tweet with an undated {{cn}} template. Could you please help me understand why you would want to remove a citation template? Thanks! GoingBatty (talk) 13:15, 26 June 2014 (UTC)[reply]

Same issue on 24: Live Another Day. GoingBatty (talk) 13:19, 26 June 2014 (UTC)[reply]
  • Like Facebbook, Twitter fails WP:RS – it is a self-published source, and subject to change at a whim and ultimately unverifiable. I believe that referring to reliable third party sources that cite certain Twitter feeds is acceptable, but not the Tw feed themselves. -- Ohc ¡digame! 15:39, 26 June 2014 (UTC)[reply]
Thanks for letting me know. Could you please add this to User:Ohconfucius/script/Sources? Also, could you please use
{{citation needed|date={{subst:CURRENTMONTHNAME}} {{subst:CURRENTYEAR}}}} instead of {{cn}}? Thanks! GoingBatty (talk) 22:39, 26 June 2014 (UTC)[reply]
Thanks for the suggestion. It has now been incorporated. Regards, -- Ohc ¡digame! 01:26, 27 June 2014 (UTC)[reply]

Misspelling in script

[edit]

Just a heads up, but one of your scripts just added the misspelled word "politican" to several articles. I've edited the articles so the word is spelled correctly, but the script should probably be corrected. Spirit of Eagle (talk) 03:41, 14 July 2014 (UTC)[reply]

Dates in document titles

[edit]

What is your date script intended to do about dates in document titles?

You have just gone through the Armagh railway disaster article making dates consistent with the MOS. Unfortunately, the Railway Inspector who investigated the accident worked to a different MOS (Victorian values?) and therefore wrote his report as "Report of the Board of Trade (Railway Department) into the circumstances of the collision near Armagh on 12th June 1889" not " ... on 12 June 1889". Clearly (I would have thought) the title of a document is a quote and (in order) to maintain accuracy should be left alone when making text compliant with MOS. Regards Rjccumbria (talk) 18:23, 11 August 2014 (UTC)[reply]

  • There is nothing to indicate that that was a quote. If the relevant sentence was within quotes, it would not have been changed. However, even looking at it now, I see it could easily be paraphrased; the format change wouldn't matter in such case. -- Ohc ¡digame! 01:47, 12 August 2014 (UTC)[reply]
Your reply rather talks past the point. The Railway Inspector wrote a document. He gave that document a title. Titles of references are intrinsically quotes. When the document is cited, its title is quoted verbatim, and - as for a quote in quotation marks - should not be paraphrased or reformatted for any reason (including/especially because someone thinks it looks neater/cooler reworded). You (or your script) has changed the title of the document. You (or it) shouldn't. Rjccumbria (talk) 14:13, 12 August 2014 (UTC)[reply]
I have no gripes about not changing quotes. BUT if it's a quote, please ensure it's properly formatted (such as using quotes or italicisation), because it isn't at all obvious on the surface that it's a title. My script will change dates within citations (because the majority of dates reside within citations) unless there are specific clues to indicate that it shouldn't be touched. And if it isn't properly formatted (no specific clues), then don't be surprised that somebody comes along, mistakes it for a piece of ordinary text that should be subject to normal formatting, and makes it compliant with policies and guidelines. -- Ohc ¡digame! 15:16, 12 August 2014 (UTC)[reply]
Well, I take your reproof, and may well act upon it when I remember. But (and I had better stress this is in a friendly tone of voice) it does rather sound as though you have a script that looks for a particular type of non-compliance and then assumes no other coincident non-compliances (and that there is no other non-compliance which if rectified would remove the non-compliance being patrolled for). It might perhaps be safer to assume that a non-compliant editor is more likely than average to be multiply non-compliant. Rjccumbria (talk) 21:16, 12 August 2014 (UTC)[reply]

Note to self: dashes script

[edit]

change within dts template causes display error. -- Ohc ¡digame! 08:05, 15 August 2014 (UTC)[reply]

Problem with Sources script

[edit]

Hi Ohconfucius! Your sources script seems to only be displaying a "CITE name" link in my Tools sidebar, and not the "Fix SOURCES" link. Could it be something in this edit? Thanks! GoingBatty (talk) 22:28, 1 September 2014 (UTC)[reply]

@Ohconfucius: Are you having this issue? Thanks! GoingBatty (talk) 03:02, 12 September 2014 (UTC)[reply]
Sorry, I was asleep ;-) It seems to work now. I just ran it on Sarah Palin, and it ran OK. -- Ohc ¡digame! 03:23, 12 September 2014 (UTC)[reply]
Confirmed - thanks! GoingBatty (talk) 21:40, 14 September 2014 (UTC)[reply]

Suggestion - remove deadurl=no when removing archiveurl/archivedate

[edit]

Hi Ohconfucius! When I ran your Fix SOURCES script on FAC Sleeping Dogs (video game), I noticed that your script removed empty |archiveurl= and |archivedate=. When it does so, should it also remove |deadurl=no? Thanks! GoingBatty (talk) 02:02, 23 September 2014 (UTC)[reply]

|deadurl= is ignored when |archiveurl= is missing or empty and so when that occurs may safely be removed.
—Trappist the monk (talk) 23:05, 27 September 2014 (UTC)[reply]
I would like to assume that |deadurl=no would be removed in the archiving process, but that is not always the case. As a quick suggestion, assuming you only wanted to remove |deadurl= while removing empty |archiveurl=, you could add two tests before the blank parameter removal: (tested only lightly)
regex(/\|[ ]*archiveurl[ ]*=\s*((\|[^|}]*)*)\|[ ]*deadurl[ ]*=[^|}]*([|}])/,'$1$3');
regex(/\|[ ]*deadurl[ ]*=[^|}]*((\|[^|}]*)*)\|[ ]*archiveurl[ ]*=\s*([|}])/,'$1$3');
The blank parameter removal regex would stay as is. I don't think more than that is needed. In fact, to remove all unused |deadurl= occurrences, you would need to find the bounds of the template, which could be hard, since templates could be nested in between. If this nesting was not possible, you could use a function as the replace parameter to filter [in place of unsupported negative lookbehind,] like this:
regex(/(\{\{[^}]*)\|[ ]*deadurl[ ]*=[^|}]*([^}]*\}\})/,function($0,$1,$2){return $0.search(/\|[ ]*archiveurl[ ]*=/)<0?$1+$2:$0;});
This would be placed after removing empty parameters, but if another template is transcluded inside the citation template between |deadurl= and |archiveurl=, it will problematically remove |deadurl=. Citation templates do not usually have this problem, but it's a possibility. I wrote a script to handle nesting, but it's complicated, and best avoided. BTW, as for removing explicit |deadurl=yes, it could simply be removed en masse in a separate replacement, if there is no reason to use this marking. Its purpose would simply be for human use, since the code seems to ignore it. —PC-XT+ 04:39, 28 September 2014 (UTC) 04:50, 28 September 2014 (UTC)[reply]
I found a potential problem with the code I suggested. If there were duplicate |archiveurl= parameters, and one of them was blank, but not both, the blank one would be correctly removed, but so would |deadurl=, even though it may be relevant to the other parameter. There may be no simple solution. —PC-XT+ 07:11, 28 September 2014 (UTC)[reply]
Here we go. This should remove only |deadurl= (set to yes, no, empty or something else), when |archiveurl= is not present, and handles 3 levels of nested templates in multiple groups. If there are more than 3 levels of nested templates anywhere in the citation, it fails by deleting nothing. The number of levels could be expanded if necessary, but the regex is getting quite long, and I rarely find that many nested templates in citations. This code goes after removing empty |archiveurl=.
regex(
/(\{\{[^{}]*(\{\{[^{}]*(\{\{[^{}]*(\{\{[^{}]*\}\}[^{}]*)*\}\}[^{}]*)*\}\}[^{}]*)*)\|\s*deadurl\s*=[^|}]*((\|[^{}]*(\{\{[^{}]*(\{\{[^{}]*(\{\{[^{}]*\}\}[^{}]*)*\}\}[^{}]*)*\}\}[^{}]*)*)?\}\})/g
,function($0,$1,$2,$3,$4,$5,$6,$7,$8,$9){return $0.search(/\|\s*archiveurl\s*=/)<0?$1+$5:$0;});
I'm still testing it, but it seems to work. To simplify the regex, the number of levels could even be reduced, to make it fail at only 2 or 3 levels of nested templates, without affecting many citations. —PC-XT+ 18:42, 30 September 2014 (UTC)[reply]

Bug report - Fix SOURCES script changes infobox parameters

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on The Blacklist (TV series), it wants to change the |website= parameter of {{Infobox television}}. Similarly, your script wants to remove |language=English from {{Infobox film}} on JLA Adventures: Trapped in Time. Are you able to adjust your script so it only changes citation templates? Thanks! GoingBatty (talk) 03:26, 6 October 2014 (UTC)[reply]

Thanks for fixing the first issue. Now when I run the script on The Blacklist (TV series), it wants to remove the empty |editor= parameter from {{Infobox television}} and messes up the Metacritic reference. It still tries to remove |language=English from {{Infobox film}} on JLA Adventures: Trapped in Time. GoingBatty (talk) 02:38, 7 October 2014 (UTC)[reply]
@Ohconfucius: Any progress on these remaining issues? Thanks! GoingBatty (talk) 17:53, 9 January 2015 (UTC)[reply]

Bug report - adding "The"

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on James Frey, it wants to change the ''New York Times'' to the ''The New York Times''. Could you please tweak your code so that it would change it to ''The New York Times'' instead? Thanks! GoingBatty (talk) 17:53, 24 October 2014 (UTC)[reply]

@Ohconfucius: Any update on this? Thanks! GoingBatty (talk) 03:21, 18 November 2014 (UTC)[reply]
I've tweaked the code so that it should now remove the leading "the". -- Ohc ¡digame! 04:08, 18 November 2014 (UTC)[reply]
Confirmed - thanks! GoingBatty (talk) 04:20, 18 November 2014 (UTC)[reply]

"Edition of" (Date script), "deadurl" (Sources script)

[edit]

Hello Ohconfucius, and thank-you once again for your helpful scripts! Two issues: (1) The Date script changes "The front page of the 1 May 1930 edition of Weekly News" to "The front page of the edition of 1 May 1930 of Weekly News". I have seen worse examples. It just never works when it moves "edition of" like this; I always have to move it back. Please consider simply removing this feature. (2) The Sources script routinely deletes "deadurl=yes". The deadurl parm is very useful when the url is dead and we wish the reader to use the provided archive link instead. Removing it causes the reader to click on the dead url and become disappointed. It just never works when it deletes "deadurl=yes" like this; I always have to put it back. Please consider simply removing this feature. (3) The Sources script routinely changes Template:Infobox graphic novel (and perhaps other infoboxes) parameter from "date" to "year"; however the correct parameter is "date" according to its doc (compare the similar "origdate" parameter). Cheers. Prhartcom (talk) 16:25, 17 November 2014 (UTC)[reply]

@Prhartcom: I don't think |deadurl=yes does anything. For example, I don't see a difference between these two:
{{cite news | url=http://www.udreview.com/2.1980/behind-the-stone-balloon-1.142251 | title=Behind the Stone Balloon: Part 1 | author=Markovetz, Jessie | newspaper=[[University of Delaware#Student media|The Review]] | location=University of Delaware | date=November 21, 2006 | accessdate=July 28, 2010 | archiveurl=http://web.archive.org/web/20100213214833/http://www.udreview.com/2.1980/behind-the-stone-balloon-1.142251 |archivedate=February 13, 2010 |deadurl=yes}} generates:
Markovetz, Jessie (21 November 2006). "Behind the Stone Balloon: Part 1". The Review. University of Delaware. Archived from the original on 13 February 2010. Retrieved 28 July 2010. {{cite news}}: Unknown parameter |deadurl= ignored (|url-status= suggested) (help)
{{cite news | url=http://www.udreview.com/2.1980/behind-the-stone-balloon-1.142251 | title=Behind the Stone Balloon: Part 1 | author=Markovetz, Jessie | newspaper=[[University of Delaware#Student media|The Review]] | location=University of Delaware | date=November 21, 2006 | accessdate=July 28, 2010 | archiveurl=http://web.archive.org/web/20100213214833/http://www.udreview.com/2.1980/behind-the-stone-balloon-1.142251 |archivedate=February 13, 2010}} generates:
Markovetz, Jessie (21 November 2006). "Behind the Stone Balloon: Part 1". The Review. University of Delaware. Archived from the original on 13 February 2010. Retrieved 28 July 2010.
GoingBatty (talk) 03:19, 18 November 2014 (UTC)[reply]
|deadurl=yes is the default case. It serves no useful purpose and only adds to the clutter in edit mode. -- Ohc ¡digame! 04:12, 18 November 2014 (UTC)[reply]
Thank-you, Ohconfucius and GoingBatty, you are both right. I see it is the deadurl=no that is the critical parameter, not deadurl=yes, I was wrong. Of course deadurl=yes does something, it just does it by default. (I was thinking it was the other one that was by default; sorry.) Could you please comment on issue number (1) above and I have added issue number (3). Again, thank-you for creating these useful scripts! Cheers. Prhartcom (talk) 05:03, 18 November 2014 (UTC)[reply]

Sources script caused error to be reported on my Talk page

[edit]

Documentation at Template:Cite web indicates parameter website= has aliases work= and therefore one or the other should be used but not both. Fix SOURCES script used both causing redundant parameter error with this edit, causing bot to report this message on my Talk page. Prhartcom (talk) 03:54, 21 November 2014 (UTC)[reply]

Bug report - removes publisher

[edit]

Hi Ohconfucius! Could you please help me understand why your Fix SOURCES script removes |publisher=AGS Publishing in the article for The Who? Thanks! GoingBatty (talk) 23:33, 26 November 2014 (UTC)[reply]

Bug report - Religion News Service

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Who's Bigger?, it wants to change |work=Religion News Service|publisher=Washington Post to |work=Religion News Service. Could you please update your code so it changes it to |agency=Religion News Service|work=Washington Post instead? Thanks! GoingBatty (talk) 17:33, 29 November 2014 (UTC)[reply]

Bug report - only changes first set of quotation marks in title parameter

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Homeland (TV series), it changes |title=Showtime Picks Up "House of Lies" and "Homeland" to Series to |title=Showtime Picks Up 'House of Lies' and "Homeland" to Series. If I run it a second time, it doesn't change the second set of quotation marks. Could you tweak the script so it changes both sets of quotation marks? Thanks! GoingBatty (talk) 21:54, 30 December 2014 (UTC)[reply]

Any thoughts on this? Thanks! GoingBatty (talk) 02:28, 14 May 2015 (UTC)[reply]

The edition of DATE of

[edit]

The Date script changes "The front page of the DATE edition of News" to "The front page of the edition of DATE of News". Please remove this code, it isn't working; thanks. Prhartcom (talk) 13:04, 9 January 2015 (UTC)[reply]

"date=" to "year="

[edit]

The Sources script changes Infobox parameter "date=" to "year=" if its data contains only a year; however many Infobox templates do not contain a "year=" parameter according to their docs (i.e. Infobox Language, Infobox Graphic novel, etc. specify a year should be the data in its "date="). Please remove this code; thanks. Prhartcom (talk) 13:10, 9 January 2015 (UTC)[reply]

Another option would be to tweak the code so it modifies citation parameters but not infobox parameters, which would also solve other issues above. Thanks! GoingBatty (talk) 17:54, 9 January 2015 (UTC)[reply]
Ohconfucius, for an example, please go to E language and try running the Sources script without saving your work, and see what it tries to do to {{Infobox Language}}. Prhartcom (talk) 14:00, 12 January 2015 (UTC)[reply]
@Ohconfucius: Please consider addressing this issue, which continues to occur, and responding to my notice. The suggestion of GoingBatty would solve the problem. Thank-you. Prhartcom (talk) 14:11, 2 February 2015 (UTC)[reply]
 Done I have now amended that line of the regex accordingly. For some reason, it doesn't work at all with vertical citation templates, but what the heck... -- Ohc ¡digame! 03:27, 3 February 2015 (UTC)[reply]

|year= vs. |date= citation usage

[edit]

It seems users of your script are incorrectly replacing |year= with |date=. Please review Template:Citation Style documentation#date to readjust your script, and review the other Template:Citation Style documentation parameters to make sure they are all up-to-date. Thanks.
This is a copy of the message left at Wikipedia:Date formattings/script/MOSNUM dates/bugs#year vs. date citation usage since it's not clear where the master bug summary page is (the edit summary links there, while another bug report button links here).   ~ Tom.Reding (talk ⋅contribs ⋅dgaf)  01:41, 12 January 2015 (UTC)[reply]

(talk page stalker) It was removed from AWB's general fixes in April 2014 - see Wikipedia talk:AutoWikiBrowser/Archive 27#Cite book date - Year. GoingBatty (talk) 03:57, 13 January 2015 (UTC)[reply]
So that mean's it's no longer necessary to change it, and that BattyBot no longer does it either? -- Ohc ¡digame! 04:54, 13 January 2015 (UTC)[reply]
I'm not aware of any benefit in changing it, and bots such as BattyBot that AWB's use general fixes don't change it either. GoingBatty (talk) 20:57, 13 January 2015 (UTC)[reply]
I just seem to remember it throwing cs1 error messages, and that some bot or another had been changing |date=YYYY to |year=YYYY for 4-digit parameters. I've just disabled that line of the script. -- Ohc ¡digame! 03:36, 14 January 2015 (UTC)[reply]
It doesn't generate CS1 errors now: {{cite book|title=Title|date=2015}} generates Title. 2015.. GoingBatty (talk) 03:46, 14 January 2015 (UTC)[reply]

Sources script holds up well to source-heavy article with multiple contributors

[edit]

Just a note to say good work: the Sources script holds up well when running it on the Charlie Hebdo shooting article, which has heavy editor contributions and has ~300 source references (so many, that the article has a "too many sources" tag). I ran it twice there over a period of a few days: Many important corrections were made and no errors or poor script behavior took place. That is, with the exception of these two occurrences that you may be interested in:

  • For some reason, it tried to completely remove a {{Wikinews}} template. I prevented it from happening. Not sure why code would be in place to do this.
  • It tried to rename a [[File]], which would have broken the link to the image (the file name contained the word "janvier" and the script tried to change it to "January"). I prevented it from happening (the second time; I made a mistake and didn't catch it the first time). This is admittedly a tough one. Code would have to first check to see if it was modifying within the "xxx" of [[File:xxx]]

You can see these by running the Sources script at this article yourself (without saving your work). With the exception of these, the Source script seems bulletproof. I run it occasionally on various articles, am very happy with it, and would like to see it in widespread, even automated use. In the recent Wiki survey I recommended this script (and the Date script). Prhartcom (talk) 19:48, 18 January 2015 (UTC)[reply]

Thanks for your feedback. I disabled the function that converts month names on a non-specific basis, so the second issue you mentioned should not recur. As to removal of the WN link, I have never found any WP article where the link to the embedded WN link actually brings anything. At best, WN coverage is patchy, and pales into insignificance in terms of both coverage and any sources that the corresponding WP article has, so it's a feature that I find helps remove some of the screen clutter when in edit mode with little detriment. Regards, -- Ohc ¡digame! 10:10, 20 January 2015 (UTC)[reply]
You're welcome, thanks again for the script, and good to hear the first issue is fixed. Since I have wholeheartedly stated support for a script that upholds established editing policy I hope I can say that is not the same as upholding a single editor's opinion about an entire entity within Wikimedia. I don't much follow Wikinews either but in this case the news is still quite current and the template helpful. Maybe if the script first checked to see if the date of the news was old, but that would still be implementing opinion, not policy. Suggest disabling. Note: I ran the script on a different article with 300+ references and I'm happy to say it behaved flawlessly. Prhartcom (talk) 13:54, 20 January 2015 (UTC)[reply]
I think you are correct. I've now disabled that line of code. Oh, I don't know if you've come across it... I mentioned in the documentation that the script may stall occasionally, and all you have to do is to click on "continue". Regards, -- Ohc ¡digame! 01:54, 21 January 2015 (UTC)[reply]

author=

[edit]

Unfortunately, I found another bug, one that I hope is easy to fix. The Sources script deleted "author=BBC News staff" from the cite template invoked in the article Tintin in the Congo, but that author's name is the name of the anchor used in the inline reference, i.e. {{sfn|BBC News staff|2004}}. Deleting the "author=BBC News staff" from the cite template broke the shortened footnote hyperlink three times, in this case. I have left the error in place at that article so you can see it, but I will be reverting my change after you do so. I see from the cite template documentation that the "author=" parameter is a valid alias of "last=", so shouldn't the script leave it alone? Thanks. Prhartcom (talk) 23:14, 21 January 2015 (UTC)[reply]

  • Author names like "staff reporter" "BBC reporter" are removed by design because they are utterly meaningless and add nothing to the quality of referencing. I took the cue to remove these from template documentation. The problem with your Tintin article lies within the rather obsessive box-filling of an editor who populated authors parameter within the {{cite news}} template with meaningless bbc author credits without giving those references a name using {{sfnref}}. I have now fixed these manually. -- Ohc ¡digame! 01:58, 22 January 2015 (UTC)[reply]
Thanks for your edit to that article; I agree with that fix. The {{SfnRef}} template is how I handle that kind of situation in articles I edit. Prhartcom (talk) 02:04, 22 January 2015 (UTC)[reply]

Script still holding up extremely well, more feedback

[edit]

This is just another note to express support for this Sources script as well as Date script and General Formatting script and to offer feedback. I wish I could list the successes that these scripts catch and solve; too many to list here but it does give me the opinion that there must be a limit to what a human editor can catch on their own; these scripts take over at that point. I believe the feedback below is nit picky but I considered that perhaps you would urge me to provide it regardless, therefore:

  • Just as the edit summary for the Date script begins "Date formats per ..." and for the General script begins "General ...". perhaps for the Sources script could begin "Sources per ...". After the script's message there should be a full stop and space, therefore if we run each script sequentially the edit summary will be three sentences. Each sentence should start with a capital letter. You could even consider putting a verb before the subject ("Fix date formats per...", "Fix sources per...", etc.) but this may not be necessary. The General script could end with "per script" as the others.
  • Unlike the Sources script and the Date script, the General script does not automatically push the Show Changes button; perhaps this could be made consistent.
  • The Date script changes the "date=" parameter of the "Use dmy dates" (or mdy) template to the current date. To me, this is unnecessary and even a bit rude; actually I always change it back to the previously reported date. Perhaps the script could leave the Use dates parameter alone and only do this if the template needs to be added.
  • I have brought this up before and was shot down, but in my opinion the "deadurl=yes" parameter/value should be kept and not removed. I realize that "yes" is the default value and therefore supposedly unnecessary, however there is such a thing as self-documentation. It calls attention to editors that the link is dead and that the archive link is active. Therefore I always laboriously restore the "deadurl=yes" parameter for them before saving changes the script made. Perhaps the script could leave the "deadurl=yes" parameter alone or somehow provide a user option to keep or not.
  • I had never used the Date scripts various user option "Body dates" etc. before today and normally use only the "ALL dates" but assumed the provided options allow the user to format dates in the cite template separately from the dates in the visible prose body; however I see that is not the case? If this is not already provided, perhaps the script should allow the user to separately format dates in citation references and dates in prose.
  • I occasionally try running the Linkspam script but it has not yet found any link spam and therefore I often forget to try running it. Perhaps this entire script can simply be made part of the General script instead of a separate script.
  • The General script does a good job of lowering the case of "Of", "The", etc. in a cite's title to proper title case but FYI it occasionally misses some; I will start collecting examples of missed opportunities here and give them to you later. Note that this feature does not do the opposite: Raise the case of "This is the title" to "This is the Title" (occasionally while running the script I will spot this and fix it manually); perhaps this feature could be added.
  • Occasionally the script simply moves the "work=" or the "publisher=" to a different location within the cite template. I have seen it move work before publisher and I have seen it move publisher before work (try North American Piedmontese cattle). I have not saved my changes when it does this as I would not be able to defend the reason for the seemingly unnecessary move. Perhaps the Sources script should not move these parameters.
  • The Date script removes the non-breaking space from dates, despite the MOS giving examples of inserting the non-breaking space into dates (MOS:NBSP and WP:DATERANGE, try running the script on Stuckism). I have not saved my changes when it does this as I would not be able to defend the reason for this removal, even though I personally do not insert the non-breaking space into my dates. Perhaps the Date script sound not remove this formatting.
  • Ooo, here's a good one that the General script missed: diff.
  • Just to end on a high note I will try listing just some my favorite successes these scripts fix: stripping tracking information from a URL, catching repeated words, removing manually applied italics in cite template parameters, replacing curly quote marks with standard quote marks, replacing mdash with ndash, removing Twitter and Facebook references, and of course fixing confusion between "work=" and "publisher=".

Cheers. Prhartcom (talk) 15:01, 8 February 2015 (UTC)[reply]

Thanks for your comments. Replies to your points are as follows:
  • perhaps for the Sources script could begin "Sources per ...".  Done
  • The punctuation is prepended only when there is something in the edit summary box, and I chose the comma because it looks weird otherwise.
  • The General script could end with "per script" as the others.  Done
  • the General script does not automatically push the Show Changes button;  Done
  • Changing the "date=" parameter of the "Use dmy dates" (or mdy) template. These are maintenance templates, which nobody else and no other bots use. The objective of dating them is to ensure most efficient updating. Please refer to Template:Use dmy dates/doc
  • the "deadurl=yes" parameter/value should be kept and not removed.  Done no longer removed
  • I’ve been trying to cut the number of buttons and so eliminated that function. Ref dates sometimes need fixing and body dates ALWAYS need to be aligned, so there’s no benefit to having a separate button to call up reference formatting only
  •  Done linkspam script now part of the General script.
  • Raising the case of "This is the title" to "This is the Title" – I have been downcasing only specific strings but cannot reliably make generic upcasing for titles.
  • As written, the script ought to move the "|work=" parameter before "|publisher=" within the cite template. It seems to do that consistently for the two instances in North American Piedmontese cattle.
  • The Date script removes the non-breaking space from dates to ensure that all dates are aligned. NB spaces written in html cause unnecessary screen clutter, and are only optional and occasional. Thus they are removed for easy parsing and consistent correction. Nobody has ever complained to me of their removal. Retaining them would entail a significant rewrite, which I am not prepared to do at present due to the “cost” significantly outweighing the benefit
  • diff. That was an original feature that I disabled due to false positives. I have found in the past that the double hyphens are sometimes used for other situations. I might rethink the possible solutions a bit more...
-- Ohc ¡digame! 09:07, 13 February 2015 (UTC)[reply]
Ohconfucius, thanks very much for seeing to these suggestions and for the feedback. Prhartcom (talk) 23:38, 13 February 2015 (UTC)[reply]
Could you remind me why we occasionally delete the "publisher=[data]"? I understand why in the cases where the publisher is obvious because it is nearly the same data as "work", but what about when it is not obvious, i.e. "publisher=NewBay Media" was deleted when "work=Broadcasting & Cable", "publisher=Cox Media Group" was deleted when "website=Austin360.com", "publisher=Hearst Magazines UK" was deleted when "website=Digital Spy", "publisher=Fairfax Media" was deleted when "website=Standard-Examiner" or "website=The Age" or "website=The Sydney Morning Herald", "publisher=Prometheus Global Media" was deleted when "work=Billboard (magazine)", "publisher=Viacom" was deleted when "work=MTV News", "publisher=Mortimer Zuckerman" was deleted when "work=Daily News (New York)" and "location=New York". Thanks. Prhartcom (talk) 13:07, 16 March 2015 (UTC)[reply]
(talk page watcher) @Prhartcom: Template:Cite web#Publisher states that |publisher= is not normally used for periodicals. Similarly, Help:Citation Style 1#Work and publisher states that |publisher= should not be included for mainstream newspapers. GoingBatty (talk) 00:52, 17 March 2015 (UTC)[reply]
Thanks GoingBatty, good to know. Prhartcom (talk) 03:30, 17 March 2015 (UTC)[reply]

Script removes citation book publisher

[edit]

Hello, I have been wondering why on occasion the Sources script removes the citation's book publisher, and today this was challenged here. Checking all the usual docs (Template:Citation, Help:Citation Style 1, WP:CITEHOW) does not reveal the answer. I have told the editor I am checking into it. Thanks. Prhartcom (talk) 15:54, 12 February 2015 (UTC)[reply]

Thanks very much for resolving this. I have reported back to the editor. Prhartcom (talk) 23:38, 13 February 2015 (UTC)[reply]

Bare URLs removed

[edit]

I honestly didn't think I would be back so fast. Running the General script on this article: Sleaford removes several dozen bare URL references. WT-? Prhartcom (talk) 00:07, 14 February 2015 (UTC)[reply]

Oh no, that was my fault then for suggesting it. Tried running Reflinks first and it didn't help, I backed out. I am so sorry. Prhartcom (talk) 04:08, 14 February 2015 (UTC)[reply]
I've disabled it for now. I think I have a solution, and am testing it now. Give me a couple of weeks! -- Ohc ¡digame! 11:12, 14 February 2015 (UTC)[reply]
@Prhartcom: The formatting script now incorporates linkspam removal as default. -- Ohc ¡digame! 12:04, 5 March 2015 (UTC)[reply]
Well, trying it on the test case I mentioned above still tries to remove several bare URLs. Prhartcom (talk) 15:40, 5 March 2015 (UTC)[reply]
I see. I have fixed it, so try it again now. -- Ohc ¡digame! 03:32, 6 March 2015 (UTC)[reply]
It's causing some unnecessary white space, right? And it's still deleting a couple of bare URLs there. This article must be a good test case. Also, I thought I remembered that it was finding some legitimate general fixes in that article and now it isn't finding a single thing, unless I am mis-remembering. Prhartcom (talk) 15:51, 6 March 2015 (UTC)[reply]
It was link within an unexpected {{refn}} template that caused the false positive. I also fixed the whitespace issue while I was at it. As to other fixes, the linkspam removal is a separate module, so it operates independently of the other formatting fixes. But it's possible that there were other minor formatting changes that were neutralised because I realigned the entire code to my test script two days ago. -- Ohc ¡digame! 02:30, 7 March 2015 (UTC)[reply]
Try it on David Hume; it tries to remove one bare URL, and Night of the Living Dead, it tries to remove two, so perhaps it's not fixed yet. I wish editors wouldn't create links like those. Prhartcom (talk) 06:35, 7 March 2015 (UTC)[reply]
One was caused by the link being inside a "Works" section, now protected; the other instance was repeated use of the "/" inside a ref name. Never seen it before, but that has now been protected as well. Thanks for your help. -- Ohc ¡digame! 11:13, 8 March 2015 (UTC)[reply]
Still not working; try it at Tintin (character) (Someone came along to this article I had rewritten and added the bare URLs. I'll be removing them in the future also, just not like this.) Prhartcom (talk) 23:04, 10 March 2015 (UTC)[reply]
Nothing wrong with those removals that I can see. That's pretty classic linkspam, and what the script was designed to remove. Disabling that type of removal would defeat the whole point of the script. -- Ohc ¡digame! 02:48, 12 March 2015 (UTC)[reply]
Good to know, thanks. Prhartcom (talk) 05:09, 12 March 2015 (UTC)[reply]

Bug report - removes publisher when work doesn't exist

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Senate Intelligence Committee report on CIA torture, it removes |publisher=The McClatchy Company from a reference that doesn't contain |work= or any other information on what company created the source. Is there a way you can adjust your script to make a better change in this case? Thanks! GoingBatty (talk) 01:28, 5 March 2015 (UTC)[reply]

Any thoughts on this? Thanks! GoingBatty (talk) 02:30, 14 May 2015 (UTC)[reply]

Also: this diff (search for "a clever twist on Rapunzel") changed the parameters "website" and "publisher" to "work" and "website", causing a redundant parameter error. Prhartcom (talk) 03:30, 17 March 2015 (UTC)[reply]

Correcting capitalisation within hidden text

[edit]

Looks like one of your scripts corrected "improper capitalisation" with italics within hidden text, which obviously would accomplish nothing. See here; https://en.wikipedia.org/enwiki/w/index.php?title=Mariah_Carey&diff=652040391&oldid=652019247 Thanks. --Escape Orbit (Talk) 17:46, 19 March 2015 (UTC)[reply]

formatgeneral script – don't remove spaces between refs in {{reflist|refs=…}} sections

[edit]

Hi,

When using the "refs" argument of {{reflist}}, it's common to put newlines between each separate reference – the whitespace isn't displayed so separating out the references makes it easier to read. As such, I think the script should spot when it's working on references within a {{reflist}} section, and not remove the whitespace between the refs.

For an example of what I mean, take a look at the source of CN Lester.

Thanks!

—me_and 10:26, 27 April 2015 (UTC)[reply]

Bug report - incorrect removal of part of URL

[edit]

Hi Ohconfucius! I like that your Fix SOURCES script gets rid of some of the garbage in citation URLs. Unfortunately, the script is being too greedy in List of killings by law enforcement officers in the United States prior to 2009 - please see the newamericamedia.org link at line 485. Could you please let me know when this is fixed? Thanks! GoingBatty (talk) 00:50, 14 May 2015 (UTC)[reply]

It's also being too greedy in List of killings by law enforcement officers in the United States, 2009 - please see the New York Times link at line 300. Thanks! GoingBatty (talk) 00:53, 14 May 2015 (UTC)[reply]

  • Can you give me another example, please, or a more detailed description of what the string removal is being produced? There is no transformation of the link to newamericamedia.org that I can see at the former example, and likewise the link to nytimes in the latter is unchanged when I run the script. -- Ohc ¡digame! 02:05, 14 May 2015 (UTC)[reply]
In List of killings by law enforcement officers in the United States prior to 2009, the Fix sources script changes |url=http://news.newamericamedia.org/news/view_article.html?article_id=a8f8bfd273dd9321528c98ceb070d0c1 to |url=http://news.newamericamedia.org/news/view_article.html. Looks like the script removes information after the question mark, even if it's important to the URL. Thanks! GoingBatty (talk) 02:34, 14 May 2015 (UTC)[reply]
Twas due to a regex that was not properly mirrored to my test script. Now fixed. -- Ohc ¡digame! 05:39, 14 May 2015 (UTC)[reply]
Confirmed - thanks so much! GoingBatty (talk) 12:41, 25 May 2015 (UTC)[reply]

Bug report: Changes Stylus Magazine to Stylus

[edit]

Hi Ohconfucius! When running your Fix SOURCES script against Vocal Studies + Uprock Narratives, I saw it changes |work=[[Stylus Magazine]] to |work=[[Stylus (magazine)|Stylus]], thus changing Stylus Magazine to the redlink Stylus. Could you please add an exception to your script so it doesn't change this link? Thanks! GoingBatty (talk) 12:38, 25 May 2015 (UTC)[reply]

now fixed. -- Ohc ¡digame! 05:50, 26 May 2015 (UTC)[reply]
Now it's changing |work=[[Stylus Magazine]] to |work=[[Stylus Magazine (magazine)|Stylus Magazine]]. Could your script just ignore [[Stylus Magazine]]? Thanks! GoingBatty (talk) 01:44, 27 May 2015 (UTC)[reply]
@Ohconfucius: Any thoughts on this bug and the others I've posted here? Thanks! GoingBatty (talk) 23:30, 18 September 2015 (UTC)[reply]
I've disabled this line for now. -- Ohc ¡digame! 08:32, 23 September 2015 (UTC)[reply]

Bug report: online.wsj.com

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Christine O'Donnell, it wants to change | work = online.wsj.com | publisher = Wall Street Journal to the redundant | work = The Wall Street Journal | work = The Wall Street Journal. Could you please tweak the code so it will only include one |work= parameter? Thanks! GoingBatty (talk) 14:20, 31 May 2015 (UTC)[reply]

Bug report: Changes citation title

[edit]

Hi Ohconfucius! When I try running your Fix SOURCES script on Charleston church shooting, it tries to change |title=Walmart, Amazon, Sears, eBay to stop selling... to |title=Walmart, Amazon.com, Sears, eBay to stop selling... in the Retailer bans section. Could you please change your script so that it doesn't try to change the |title= parameter? Thanks! GoingBatty (talk) 22:20, 9 July 2015 (UTC)[reply]

Bug Report: People

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on Kelly Osbourne, it wants to change [[People magazine]] to [[People]]. Could you please change your script so it changes it to [[People (magazine)|People]] instead? Thanks! GoingBatty (talk) 12:54, 7 September 2015 (UTC)[reply]


Bug report: Removes URLs

[edit]

Hi Ohconfucius! When I try running your wonderful Fix SOURCES script on D.I. (band), it removes the MySpace URLs. Could you please let me know when this is fixed? Thanks! GoingBatty (talk) 23:27, 18 September 2015 (UTC)[reply]

Bug report: script creates a duplicate argument error

[edit]

here, by changing |work=sky.com to |publisher=BSkyB when publisher= is already present. Regards, NSH002 (talk) 08:02, 8 October 2015 (UTC)[reply]

P.S. Not sure that "BSkyB" iis the right name when according to its wiki article it's now called "Sky". Point may be moot in this case, as the link appears to be dead, but I'm sure you'd like to avoid creating new errors for people to investigate!

Also here, where you added |language=zh right next to its twin. (error has been fixed by Sporkbot) --NSH002 (talk) 08:39, 8 October 2015 (UTC)[reply]

Script stripping URL from Creative Commons attributions

[edit]

In my watchlist, I have found your script removing URLs from a number of articles I created yesterday from the subsection of References used to attribute Creative Commons sources. Please see this diff as an example. These articles are created from a CC-BY licensed source material (an arrangement made with the content provider so the text could be used as the basis for Wikipedia articles). As part of the obligations of using CC-BY content, we have to attribute them and this requires URLs to point to the original content. When this project started (there are around 1700 articles of which about half now exist and the rest are to come), Project Australia had a very long discussion about the structure of these articles and the correct way to handle the attribution, which included the inclusion of the archived version of the original sources material (for which the URLs have also been stripped). Our understanding is that external links are permitted in the References section yet the bot is stripping these URLs. Could you please update your bot to stop stripping these URLs as it is creating copyright violations (failure to properly attribute sources). If a short-term temporary fix is required, then whitelist anything in Category:Articles incorporating text from the Queensland Heritage Register (please review) or its parent category Category:Articles incorporating text from the Queensland Heritage Register. However, I presume anyone attributing CC content in a similar way will be affected by these changes. I note your bot has processed hundreds of other articles with these attributions without stripping these URLs, so I assume there has been some recent change. Thanks Kerry (talk) 23:08, 2 November 2015 (UTC)[reply]

(talk page watcher) @Kerry Raymond: I think you mean this diff as an example. GoingBatty (talk) 02:48, 3 November 2015 (UTC)[reply]
@GoingBatty: Yes, I probably did! Kerry (talk) 09:21, 3 November 2015 (UTC)[reply]
Now fixed. thanks for picking up the error! -- Ohc ¡digame! 20:18, 3 November 2015 (UTC)[reply]
Thanks Kerry (talk) 22:58, 3 November 2015 (UTC)[reply]

Feature request: Please change "Connecting to the iTunes Store"

[edit]

Hi Ohconfucius! Per this conversation, Could you please update your great Fix SOURCES script so that when it fixes iTunes references it will also change |title=Connecting to the iTunes Store to |title=ACTUAL ARTICLE TITLE GOES HERE! (like it does for several other publishers)? Harold and the Purple Crayon is an example of an article with such a reference. Thanks! GoingBatty (talk) 00:49, 15 November 2015 (UTC)[reply]

This is worrying - your script is still generating errors

[edit]

I reported this bug above, but I see your script is still generating new errors in citation templates. This is worrying, because the purpose of scripts and bots is to correct errors, not to generate new ones: latest example, fixed here.

Basically, your script needs to be modified so that when it changes "wrongparm = something" to "correctparm = something", it needs to check that "correctparm = already_there" is not already there. Obviously, if "already_there" is identical to "something", or is identical ignoring white space and/or hidden comments, then your script doesn't need to make any change to "correctparm", but it should probably still get rid of "wrongparm". If "something" differs from "already there", then the problem is much harder to fix. Easiest thing would just be to make no change at all, except possibly adding a hidden comment that it's found an error. Otherwise, careful thought needed.

User:Frietjes is aware of this problem, and is very fast, efficient and helpful (apart from her lack of edit summaries!). She is expert in Javascript, and can probably give good advice on how to modify your script, if needed.

Regards, NSH002 (talk) 09:38, 20 November 2015 (UTC)[reply]

NSH002, where is the script exactly? is it User:Ohconfucius/test/Sources.js? do we know which line? basically, a simple fix would be to first "mangle" cases which will create duplicates, then "unmangle" them after you are done processing. I added work to my list of duplicates by parameter. I only create sublists for ones that are requested. Frietjes (talk) 14:48, 20 November 2015 (UTC)[reply]
Frietjes - Well it's certainly Sources.js or some version thereof. The diff I cited above was by Indopug, so looking at his common.js should tell you the version.
I know nothing of Javascript, apart from what I can surmise by looking at the .js code lying around here, so I haven't tried to find the precise location of the problemm. That's User:Ohconfucius's responsibility, I think - I doubt I'd be very productive trying to track down the error. (I first started programming nearly 50 years ago, at university, but haven't done any serious programming for about 25 years. Now that I'm retired, I'm trying to learn some programming again, but not in Javascript - just to try and keep the brain cells from dying off) NSH002 (talk) 15:29, 20 November 2015 (UTC)[reply]
NSH002, okay, found the problem, it's in User:Ohconfucius/test/Sources subscript2.js. it changes known italicised source names to use |work= rather than |author= or |publisher= or .... the problem is the replacement is "global" which means it will transform both author and publisher at the same time, and there is no test for collisions. a solution would be to have it (1) "mark" the parameters for transformation, e.g., by prefixing the parameters to be transformed with some unique string (e.g., OHC_WORK_ or something), then (2) check for collisions, and then (3) transform if there are no collisions. if there are collisions, it would just unprefix the colliding ones and issue a warning or something. Frietjes (talk) 16:30, 20 November 2015 (UTC)[reply]
@Frietjes and NSH002: Sorry about all that detective work you had to do. The global replacement of author/publisher is indeed a core part of the script, and there is indeed a test for collisions near the end of the run sequence, so I couldn't understand why the script wasn't correcting it. This test and elimination is meant to be located within the cleanup part of the script proper – search for the string "removing identical/similar entries in 'work' and 'publisher', and in 'work' and 'website'". Maybe I have a blind spot, but I can't fix a problem that I can't see. I'd be grateful if you could cast your expert eye on the script in light of my explanation as to the script structure. Of course, any other suggestions for corrections and improvement would be most welcome. Regards, -- Ohc ¡digame! 17:36, 20 November 2015 (UTC)[reply]
@Ohconfucius and NSH002:, as indicated above, the problem is when it transforms things like <ref>{{cite web | publisher = NME | date = 2008-02-05 | title = Camp Bestival add more acts to lineup | url=http://www.nme.com/news/nme/34104 | accessdate = 2008-02-05 |author= NME}}</ref>. the User:Ohconfucius/test/Sources subscript2.js replaces BOTH publisher = NME and author= NME with work = NME, which creates a duplicate parameter error. the solution I am proposing would be to (1) transform publisher = NME and author= NME to OHC_work_OHC_publisher = NME and OHC_work_OHC_author= NME, (2) check for duplicate 'OHC_work_OHC' within the same citation, and (3) if there are no duplicates, change 'OHC_work_OHC_para' to 'para', but if there are duplicates, simply undo the 'OHC_work_OHC_' prefix. Frietjes (talk) 17:42, 20 November 2015 (UTC)[reply]

@Frietjes and Ohconfucius: Thanks for all your work. I won't even try to go into the nitty-gritty detail here, but will make some general observations:

  1. The check for collisions needs to work for any parms, whatever they are (for starters, I reported| language = zh above).
  2. If you get a collision, the possibilities are (i) new and old are the same (ignoring whitespace, etc) - easy, don't do anything except remove "wrongparm" (ii) they differ, but the existing one is empty or blank - easy, just replace it (this won't arise if the script starts off by removing empty or blank parms - which is what I would do anyway, to reduce clutter in the edit window)
  3. If (i) and (ii) don't apply, then you get the difficult case. Easy thing is to give up and do nothing, possibly leaving an error message somewhere. Or it might even be better just to do what you're doing at the moment, which is to create an error - admittedly, this is annoying for people trying to clean out the error-tracking category, but at least it ensures that someone will come along eventually and look at the error.

Regards, NSH002 (talk) 19:01, 20 November 2015 (UTC)[reply]

NSH002, if no changes are made to the script, then the transclusions are put in the duplicate parameter category. if they values are entirely the same, then a bot will fix them. if they are different, someone will probably find them eventually as you suggested. Frietjes (talk) 19:09, 20 November 2015 (UTC)[reply]
Frietjes, yes I'm aware of that, but it's still better if it doesn't create errors for the simple case in the first place (fewer errors for people to have to look at or the bot to fix). Plus I'm trying to prompt some thought about what to do in the difficult case. --NSH002 (talk) 19:16, 20 November 2015 (UTC)[reply]

Bug report: Spectator

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on the Minecraft article, it wants to change {{cite web |url=http://minecraft.gamepedia.com/Spectator |title=Spectator |work=Minecraft Wiki |publisher=Gamepedia}} to {{cite web |url=http://minecraft.gamepedia.com/Spectator |title=ACTUAL ARTICLE TITLE BELONGS HERE! |work=The Spectator |publisher=Gamepedia}}. Could you pleas fix your script so this doesn't happen? Thanks! GoingBatty (talk) 22:55, 22 November 2015 (UTC)[reply]

  • Here is where I would appreciate your input: I have seen many many examples where editors have wrongly input journal names into the title field, but there are a very small number of instances where the title resembles the name of an established journal. So should I be aiming for zero errors and avoid converting titles at all costs, or ought I to aim for editors to notice that the script is not 100 percent correct? Or what other ways are there to deal with this issue? Regards, -- Ohc ¡digame! 12:01, 23 November 2015 (UTC)[reply]
Ohconfucius, your experience with other examples is broader than this one example, therefore I would stay with your first instinct to assume the journal name had been inserted into the title field. Prhartcom (talk) 12:20, 23 November 2015 (UTC)[reply]
When editors wrongly input journal names into the title field, are they also populating the work field? If not, maybe the rule could skip those references where work is populated? Thanks! GoingBatty (talk) 13:42, 29 November 2015 (UTC)[reply]
Currently reviewing the script to see if I can get it to do what you suggested. Regards, -- Ohc ¡digame! 07:49, 1 December 2015 (UTC)[reply]

Click "General formatting" - nothing happens

[edit]

I installed this script a few days ago and cannot use it because nothing happens when I click on the "General formatting" link. I've tried this on Chrome and Firefox on my Mac and it does not work in either browser. Both are the latest versions. And the OS is 10.11.2. Any idea what it could be? (Please ping me if you reply) Dismas|(talk) 03:18, 25 December 2015 (UTC)[reply]

Someone complaining about your general script

[edit]

Cheers. Walter Görlitz (talk) 15:46, 13 January 2016 (UTC)[reply]

Bug report: Adding second exclamation point for Yahoo!

[edit]

Hi Ohconfucius! In MC Hammer, your Fix SOURCES script is changing "Yahoo! Music" to "Yahoo!! Music" in the title parameter. Would it be better to change |title=anytext - Yahoo! Music |publisher=New.music.yahoo.com to |title=anytext |publisher=Yahoo! Music instead? Thanks! GoingBatty (talk) 17:44, 5 February 2016 (UTC)[reply]

Confirmed - thanks! GoingBatty (talk) 02:06, 9 February 2016 (UTC)[reply]

Suggestion to improve edit summary

[edit]

Hi Ofconfucius! When I use your wonderful Fix SOURCES script, the edit summary kindly provides a link to User:Ohconfucius/script. Could you please change it to User:Ohconfucius/script/Sources to make it easier for interested people to find the objectives and principles you documented there? Thanks! GoingBatty (talk) 20:08, 7 February 2016 (UTC)[reply]

Inappropriate actions

[edit]

Based on a few recent edits I've seen, this script as of now seems to be adding more, and worse, problems than it might improve. Based on the recent edit on Camden Town:

  • Dubious capitalisation which in some cases breaks links. E.g., change "Tinker, Tailor, Soldier, Spy" to "Tinker, Tailor, soldier, Spy", "The Camden Town Murder" to "The Camden Town murder", and others which were dubious even though the link was salvaged by a redirect. In another article a section about the company named "Charisma Records" was recapitalised as "Charisma records".
  • Change of original date on templates to current date, e.g., {{EngvarB|date=December 2013}} to {{EngvarB|date=February 2016}}, clearly inappropriate.
  • Loss of information; changing "Camden Town with Primrose Hill (ward)" to display only "Camden Town with Primrose Hill" loses the information that the level of this constituency is a ward, as distinct from a borough or parliamentary constituency. "[[Camden Town with Primrose Hill (ward)|Camden Town with Primrose Hill ward]]" would have been acceptable, though not in my opinion strictly necessary.
  • Changing {{Use British English}} to {{EngvarB}} in Slade and many other England-related articles is wrong because British English, specifically, is appropriate (the date was also changed).
  • I think that changing "U.S." to "US", as was done in one article, is ill-advised. I personally prefer "US", and use this in my own additions, but some people are very insistent that "U.S." (and "U.S.A.") should be used. So this is a contentious change.

The above I think are clearly undesirable changes. Some others are a matter of opinion:

  • I think the unlinking of common terms has been too aggressive. From various articles, "rock", "pop", and "classical" (but not "jazz"?!) have been unlinked in relation to music; "label" has been unlinked from "record label"; I think English counties such as Lincolnshire have been unlinked. I think these links are potentially useful depending on people's backgrounds (e.g., links might help someone from rural USA, or China). I agree with unlinking most country names, but personally think "London" and such should be linked (there are several), but that is arguable.
  • Some words are acceptable unhyphenated. Underrated is in dictionaries, and I think more common than under-rated. Lineup and breakuo are commonly used, though perhaps arguable. (These are all words that have had hyphens inserted by this script.) As a rough guide, I'd suggest that any word that is in spelling-checker dictionaries should be left alone.

I hope these comments are useful. Best wishes, Pol098 (talk) 13:54, 24 February 2016 (UTC)[reply]

(talk page watcher) @Pol098: Please note that Template:EngvarB#Usage indicates that it is acceptable to change the date on the template. I believe the goal is for wikignomes to review the articles with the oldest dates, make any updates to the article to ensure the language is correct, and then redate the template to show when the last review was done. This may be explained a little better on the similar Template:Use mdy dates#Usage. Hope this helps! GoingBatty (talk) 21:34, 24 February 2016 (UTC)[reply]
(talk page watcher) @GoingBatty: Thanks. Changing dates on templates is usually a bad idea, but the date on a "use language" template is not particularly important, I actually see no reason to preserve the original date, despite my comment above. I hadn't come across EngvarB before; I'm not sure that, as a maintenance template adding articles to hidden Category:EngvarB it should replace templates like {{Use British English}}, an instruction to editors—it would seem that both templates have their place. Maybe this has been discussed somewhere? Best wishes, Pol098 (talk) 22:07, 24 February 2016 (UTC)[reply]
Have to agree that the script should not switch {{Use British English}} to the generic {{EngvarB}} once a language variant has been selected for an article then the {{EngvarB}} should be changed and should remain at the selected variant. In theory once all articles have had a selected language variant then no article should use {{EngvarB}}. Keith D (talk) 23:28, 24 February 2016 (UTC)[reply]
Thanks. My understanding of language templates was incomplete and incorrect and I withdraw my previous comments about that particular issue. Keith D says essentially that EngvarB is a temporary template, to be replaced by "Use <xxx> English" by a human editor; this makes sense. If this is a consensus opinion it would be useful to add it to the {{EngvarB}} documentation, for the guidance of those such as me who come across the template and find no documentation on why and when it should be used. As far as this script is concerned: I don't think any script should change a language template. Best wishes Pol098 (talk) 12:23, 25 February 2016 (UTC)[reply]

Bug report: breaks URL

[edit]

Hi Ohconfucius! When I run your Fix SOURCES script on John Stamos, it breaks a zap2it.com URL. Could you please ping me when this is fixed? Thanks! GoingBatty (talk) 01:37, 28 February 2016 (UTC)[reply]

Ohc script (Jesus)

[edit]
Greetings, Ohconfucius! I already tried to contact you via your User Page about two weeks ago,[2] but never got any reply. It concerns your script - which I find by the way of a great help - and therefore I thought maybe you'd be more comfortable to reply through this page. So, my original question goes as follows:

Whereas I completely agree that Jesus should be among the WP:OVERLINKed ones as the main character of the most wide-spread religion of the world, Christianity, there is now discussion at Talk:Mandaeism#A Wikilink for Jesus? about the issue. Could you please have a look and share your opinion?

I am not asking you to get involved with the article discussion in question, but to share your opinion here at his very forum. I hope this helped to clarify! Cheers! Jayaguru-Shishya (talk) 20:38, 25 May 2016 (UTC)[reply]

Removal of templates

[edit]

I have observed twice that editors removed {{wikinews}} template from See also or External Links section using this script (Fix SOURCES is affecting it). That is against WP:SISTERS. Please fix it.
Agastya Chandrakant (talk) 08:01, 18 July 2016 (UTC)[reply]

  • Most WP articles are more all-encompassing and better sourced than Wikinews articles on the same o related subjects. I'd be interested to know if you managed to find any links that are useful to Wikipedia readers that the script removed? I don't think blindly claiming fraternal relations matter when it is of no service to Wikipedia users.-- Ohc ¡digame! 10:24, 19 July 2016 (UTC)[reply]
I don't know why you think so, but if you are saying so, you are not following WP:SISTERS. Did I manage, yes! Wikinews articles are indeed helpful and I request you to fix the script.
Agastya Chandrakant (talk) 08:03, 23 July 2016 (UTC)[reply]

Script leaving subst:lc

[edit]

In Twenty20 International matches results (2010–14) and Twenty20 International matches results (2005–09), the script left {{subst:lc:Place}}. CheckWiki looks for this, so I see it every couple of days. I hadn't looked at who made the errors before. I am cleaning it up. Bgwhite (talk) 06:05, 22 July 2016 (UTC)[reply]

Script changing comment markers

[edit]

Hi, there appears to be a problem with one of the scripts changing the opening comment marker by making the second dash into a space followed by en-dash. See example which causes a cite error. Keith D (talk) 23:59, 22 July 2016 (UTC)[reply]

EngvarO, EngvarC and EngvarA

[edit]

Hello. I have constructed additional generic spelling templates that are intended to be used exactly as EngvarB is being used:

  • {{EngvarB}} - generic British/Australian/Commonwealth spelling (en-GB) (the existing template that you created)
  • {{EngvarO}} - generic British/Australian/Commonwealth spelling, but with -ize instead of -ise (en-GB-oxendict, "Oxford spelling")
  • {{EngvarC}} - generic Canadian spelling (en-CA)
  • {{EngvarA}} - generic American spelling (en-US)

These are the four main spelling standards in the English language. I would like to use EngvarO and EngvarC in a similar fashion as you are using EngvarB, for maintenance and to prevent "spelling drift" in consistently edited articles. I'm aware that there are templates such as "Use Canadian English", but I would want to indicate only spelling, not necessarily terminology or grammar, hence the need for a generic spelling template. What do you think? Would you be willing to extend your spelling standardisation efforts to these templates? Acopyeditor (talk) 07:57, 9 September 2016 (UTC)[reply]

  • @Acopyeditor: Oh, I let a genie out of the bottle... I only intended to create a spelling script (with script markers) that merely distinguished between "British vs. American spelling". It was never meant to go venture into what tiny differences with other codes and certainly other vernacular differences. WP is a collection of individual initiative, and the ENGVAR tagging was born out of this philosophy. It is now much bigger than me, and you are welcome to discuss this more widely, perhaps at the Engvar template talk page. I would, however, just say that the proliferation of such new templates will eventually complicate the task of maintaining articles and leave a messy trail of [possibly] redundant templates, and may not be worth the marginal possible advantages of the precision you expect to create. -- Ohc ¡digame! 08:10, 19 September 2016 (UTC)[reply]
Hello and thanks for the reply. Yes, you may be right, perhaps it's too much detail in proportion to how much users actually care about such issues. But with your current system of using EngvarB for everything that is not en-US, how do you indicate whether -ize or -ise is used, or whether Canadian spelling is used in an article? Also, I think that the current (talk page) templates with the national flags are fine for articles with strong ties to a certain nation, but for "neutral" topics, such templates can easily evoke the impression of national ownership of articles, which is why I came up with more neutral versions ({{EngvarB spelling}}...). Acopyeditor (talk) 19:52, 23 September 2016 (UTC)[reply]
PS: Re: "proliferation of such new templates": The four templates above {{EngvarB}}, {{EngvarO}}, {{EngvarC}} and {{EngvarA}} only describe the spelling system (and NOT grammar, word choice, other style elements). I'm quite certain that there are only these four, so they will be sufficient. Acopyeditor (talk) 19:59, 23 September 2016 (UTC)[reply]

Script breaks file names

[edit]

Your recent edit to Umbrella Square has broken several file names by changing to day-month-year order. Acopyeditor (talk) 20:20, 23 September 2016 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
Retrieved from "https://en.wikipedia.org/enwiki/w/index.php?title=User_talk:Ohconfucius/script/archive1&oldid=1144428723"