User talk:Citation bot/Archive 3

Bot 579 added doi-broken-date when doi-inactive-date was already present

 * Probably best if bot replaced all instances of doi-inactive-date with doi-broken-date, since that is the standard form that 99% of people and bots use. AManWithNoPlan (talk) 13:58, 29 June 2015 (UTC)
 * The new bot is still doing this (see 10.2903/j.efsa.2004.36 in that edit). Pinging . – Jonesey95 (talk) 03:53, 22 September 2015 (UTC)
 * fixed now AManWithNoPlan (talk) 05:26, 26 November 2015 (UTC)

button no work
You can get around it by editing page and then clicking the citation button next to the preview/save/etc AManWithNoPlan (talk) 03:21, 24 November 2015 (UTC)


 * Your link is the development bot, which I'm assuming still runs against the test Wikipedia server. If I run the standard bot against Lithopsian (talk) 11:53, 24 November 2015 (UTC)


 * Not sure why, but the expand citations item in the left hand column points to the development version. That needs fixed.  AManWithNoPlan (talk) 21:46, 28 November 2015 (UTC)


 * I see the same thing. I think this means that the Gadget, accessible via Preferences, is pointing to the dev version. – Jonesey95 (talk) 01:46, 29 November 2015 (UTC)

Thanks for the report! Our bad--we were making the interface the gadget uses to process the pages more secure and less fragile, and forgot to switch the gadget back from the dev version when we updated the gadget here. It should be fixed in the next few hours. --Fhocutt (WMF) (talk) 03:28, 1 December 2015 (UTC)

Blanking pages on some PC platforms

 * When starting the bot from the "edit article" mode, it consistently blanks the page, which is annoying as I lose my other corrections. This happens on Win XP + Firefox 19, but not on newer platforms with a more recent Firefox. I anticipate an advice to upgrade, aiming for a more recent FF Java, but there is a reason for using FF19 on that laptop: after some Wikimedia updates newer FF versions were spontaneously jumping into a 70-99% CPU consumption mode while editing en.wiki (don't know if this has improved lately, but don't want to test either - FF19 is more robust in this sense, and laptops don't handle CPU surges so well). Materialscientist (talk) 05:37, 5 December 2015 (UTC)
 * Neither Fx 19 nor Win XP are supported by their organizations any longer nor have been for quite a while. As you expected, the solution is to upgrade, use a different browser, use a different OS, or all of the above. Not a bug IMO. --Izno (talk) 16:17, 5 December 2015 (UTC)

Citation bot progress
I've been getting the parts of Citation bot that involve manual activation working again. The development version is up and running on testwiki right now. The development version is set up to only run on testwiki and will neither read nor edit enwiki pages. If you want to help test, create a page on testwiki that has broken citations. I have been mostly using old regression tests in my sandbox.

To test doibot, the tool that is hosted on Labs:
 * Create or find your test page on testwiki.
 * Go to https://tools.wmflabs.org/citations-dev/doibot.html.
 * Follow the directions: enter the page title (for instance "User:Fhocutt (WMF)/Sandbox"), click "Submit query", and wait for the bot to work. It may take up to a minute or two, depending on how many citations are on the page and whether it is in "thorough mode".
 * Please leave comments, bugs, or feedback on the associated task in our issue tracking system.

To test the citation expander gadget:
 * Add the contents of https://test.wikipedia.org/wiki/User:Fhocutt_%28WMF%29/common.js to your own common.js user subpage on testwiki
 * Create or find your test page with broken or incomplete citations on testwiki. If the gadget is loaded properly, there should be a "Citations" button beside "Show changes". Click the button.
 * It should briefly take you to an interstitial page and then, after expanding the citations, back to the "show changes" edit page on testwiki. If you use NoScript or some other extension that blocks clickjacking, this will not work.
 * Fix the citations as needed, and save your changes.
 * Please leave comments, bugs, or feedback on the associated task in our issue tracking system.

My goal here is to get this bot working roughly as it previously was when it was blocked so that it can be unblocked, with a focus on manually activated functions which should be easier to check. If someone familiar with both PHP and current enwiki citation policies is interested in bringing the bot up to date with current policy, I am happy to give them an overview of how it functions and review the changes they make.

--Fhocutt (WMF) (talk) 01:10, 4 September 2015 (UTC)


 * Thanks for your efforts to get this bot up and running again. There are a number of unresolved bugs listed above that need to be resolved before this bot is put in operation. In particuluar:


 * Deprecated cite pmid, cite doi, etc. templates (see discussion). The bot should no longer create templates containing citation data that are transcluded, rather these templates should be substituted.  Please note that  User:Dexbot has been systematically substituting the existing transcluded templates.
 * New vauthors parameter (see discussion). Instead of adding redundant lastn to citations that already contain Vancouver style formatted authors stored in a single author parameter, the bot should replace author parameter name with vauthors, support for which has been recently added to all templates. Boghog (talk) 03:40, 4 September 2015 (UTC)


 * Thanks for your feedback. The bot has four main parts: the manually activated doibot tool, the citation-expander gadget, the automated citation checking bot, and (4) the bot that creates the templates. Due to the deprecation I don't intend to do anything with the template-creating portions. The citations-expander tool/gadget simply fills in the edit box with changes, so it doesn't actually use the bot account and still makes it possible for users to check the citations before submission--I think that a version of that can be up and running very soon, with the caveat that the expander portion hasn't been substantially updated since the bot was blocked. It sounds like you want at least a minimum number of bugs fixed/updates made to the automatic expander before the bots (the doibot tool and the automated citation checker) are unblocked. A list of what is essential and what is nice to update for each of these would be very helpful. Because the doibot tool is manually activated, I wonder if it would be possible to open it for use with appropriate "You are responsible for the edits that you make and this tool is outdated; please check the edits made with this tool to ensure they comply with the relevant policies" warnings on the tool page. Community Tech is currently prioritizing quick and high-impact fixes as we get up and running. --Fhocutt (WMF) (talk) 19:55, 4 September 2015 (UTC)

Likewise, thanks for the explanation. Just to restate the above as a list:
 * 1) citation-expander gadget
 * 2) editing mode, makes changes only to the active edit window and requires the user to save the page
 * 3) viewing mode, makes and commits edits without further user intervention
 * 4) manually activated citation bot tool
 * 5) the automated citation checking/expander bot
 * 6) bot that creates the transcluded cite pmid, etc. templates

Just to be sure, you are proposing to reactivate initially only #1-1 and not #1-2? This in theory is safer, however some editors will undoubtedly use the gadget without carefully checking the edit before saving. Also, it isn't exactly clear which parts are currently blocked. As far as I can determine, none of the parts of the bot are currently in operation. Finally, I would like to see at a minimum, the redundant lastn issue fixed before any part of the bot is put back into operation as this bug made a mess in a significant number of articles that had to be cleanup manually. Boghog (talk) 21:41, 4 September 2015 (UTC)


 * #1-1 is not being blocked, but the gadget and the file it talks to (text.php) both need updating. I believe I've fixed what needs fixing there, leaving aside the issues with the centrally-used expander. #1-2 can't function on enwiki because Citation bot is blocked, but I believe I've fixed the underlying main problems in my fixes to doibot.php. #2 also uses Citation bot and doibot.php so should have the same status as #1-2. #3 uses the Citation bot 1 account, and I doubt I have fixed the underlying problems. #4 will not be worked on unless any of that code can be used for purposes besides updating the deprecated templates.


 * I'd like to activate #1 and #2 whenever possible, but #1-1 is a good first target because it can be activated without acting, strictly speaking, as a bot and so without unblocking any of the bot accounts. I don't currently have access to the central repository, just the dev one, which is why none are actually activated--take a look at the versions that use testwiki if you haven't already, and it'll give you an idea of what it does right now. Regarding people clicking save without checking, I can easily make the redirect for #1-1 not load automatically and can put some scary "Warning, outdated, use at your own risk" text on the interstitial page to address some of that. I'll take a look at the lastn issue and hope to get that fixed reasonably quickly. --Fhocutt (WMF) (talk) 00:03, 5 September 2015 (UTC)
 * : Let me see if I understand the lastn issue. Vancouver style displays the authors as a comma-separated list, but each author's name may vary (may use periods or not, and corporate authors are credited with ). Citation Style 1 templates allow for either the use of the vauthors parameter or for a standard firstn/lastn list when vanc. Looking above, I see there is some disagreement on whether it's preferred to list all the names and only display some of them, or only provide the shortened list. Regardless which approach is taken (I am inclined to use the vanc parameter, but if it has been discussed please point me to the discussion), the bot should pick one and stick with it: properly replace the Vancouver list of names with a list of first and last, and add the   tag, or correctly identify a Vancouver-formatted list (I am not sure how easy this will be, with the allowed variation) and replace authors with vauthors. Am I missing anything? --Fhocutt (WMF) (talk) 01:09, 5 September 2015 (UTC)
 * Thanks for the clarification on what functions of the tool can be reactivated without unblocking. Concerning the format of Vancouver style style authors, the format is very regular and does not vary. In particular, periods are not allowed. If the value of vauthors contains a period, cite journal generates a Vancouver style error.
 * The format is documented here:
 * (Please note that the period at the end of the author list is not included in the vauthor parameter value but is included in the rendered citation.)
 * The  are not very common and in fact the Wikipedia template filling that generated the majority of these Vancouver style citation completely leaves these out (compare for example  with the tool output). Also note that previously, the tool generated a citation using a single author, this has recently changed to vauthors.
 * Finally the various author parameters are compared here and a rationale for vauthors may be found here. Boghog (talk) 07:40, 5 September 2015 (UTC)
 * Patrias requires Romanization to remove diacritics etc. cs1|2 does not follow this rule and allows all letters from the Unicode Latin character sets. See Help:CS1 errors for a list of these characters.


 * There are cases where Vancouver system allows lowercase letters in initials. Module:Citation/CS1 cannot know if a lower case letter is appropriate (multi-letter Romanization of Greek or Cyrillic) or not so these are flagged as errors.  PMID sometimes uses lowercase letters in initials of hyphenated Chinese given names.  If the bot is fetching names from PMID, it should not attempt to correct or mask the error but rather leave it so that human editors can determine if the lower case initial is appropriate.  PMID also has gotten some (particularly Dutch) names with nobiliary particles wrong where the first letter (usually lowercase) of the particle becomes the second initial (see the CS1 errors help page).  These too should not be auto-corrected or masked.  Where Unicode U+2019, right single quotation mark, appears in a name, this should be fixed by replacement with the standard typewriter apostrophe.


 * —Trappist the monk (talk) 11:00, 5 September 2015 (UTC)


 * Basically working OK for me, but then it always worked for me before. Mostly just expanding references from a bibcode or doi.  One worrying "regression" is that "cite conference" references are blanked, which didn't used to happen.  Lithopsian (talk) 13:01, 9 September 2015 (UTC)
 * Something that happens with the gadget, and always did happen, is that it croaks when there are more than a handful of references. Then it blanks the whole article (still in edit mode, not yet saved). The wmflabs tool never did this for me, even on the same article, just the toolbar gadget.  Lithopsian (talk) 13:30, 9 September 2015 (UTC)
 * Thanks for testing! Do you have a diff for the blanked "cite conference" references? Blanking large articles probably has something to do with how the browser is handling the requests, but I'm not sure on the specifics there. If I were rewriting this I would find some other way to handle that. --Fhocutt (WMF) (talk) 04:03, 10 September 2015 (UTC)
 * Good news! I don't think the cite conference problem is being caused by citation bot.  testwiki doesn't seem to have that template so they always display blank.  Just pasting a cite conference template into my sandbox and using the citations gadget on it works fine.  Saving it gives a blank redlink. Lithopsian (talk) 10:16, 10 September 2015 (UTC)


 * I don't think it is going to be feasible to have the bot replace the author parameter name with vauthors when the existing author list is a proper Vancouver-style list. That is a task better suited to humans, IMO. Even if we build something that works today, as soon as the Vancouver error checking is changed the bot will probably be broken again. We should avoid building code that is complicated and fragile, unless we absolutely need to. My suggestion would be to simplify where possible. For example, if the existing citation template includes any author parameters (author, authors, vauthors, last, last1, etc.), Citation bot should just leave them alone and not try to add any author parameters. That way it is less likely to cause unexpected problems. This is just a suggestion though. Ryan Kaldari (WMF) (talk) 23:44, 9 September 2015 (UTC)
 * General comment - leaving author coding untouched is a very good idea, which was suggested here long ago, and was implemented at some time. Then the bot started reformatting authorlists, which caused various errors, partly due the continuous changes in citation templates. Materialscientist (talk) 00:19, 10 September 2015 (UTC)
 * I agree that writing fragile code to constraints that are likely to change is generally a poor idea. In this case I think that it does make more sense to stop the bot from changing existing authorlists. I don't think that should be too much more difficult than figuring out how to mechanically detect a Vancouver style list and implementing that (both sound like they will take some work with the current codebase). --Fhocutt (WMF) (talk) 04:13, 10 September 2015 (UTC)
 * Agreed that it would be far better to leave the author lists alone than to do what the bot had been doing. IMHO, the error checking in vauthors is way more complicated than it needs to be. Worrying about rare capitalization errors in first name initials in PubMed citations is not a good use of anyones time. The much simpler error checking in vcite2 journal that catches 99.9% errors in vauthors lists in my opinion is good enough. But OK, at least if the bot does not introduce new author errors, I will not object if the bot restarts. Boghog (talk) 05:03, 10 September 2015 (UTC)

Marking as resolved to archive this discussion. AManWithNoPlan (talk) 03:13, 18 December 2015 (UTC)

Substitute instead of transcluding new cite pmid templates
Per this consensus, a request should be made that the function of Citation Bot be changed so that cite pmid, cite doi, and cite isbn are substituted instead of transcluded. There is also this consensus that existing cite pmid, etc. templates should be substituted in all WP:MED articles. I am planning to submit a WP:BRFA for a new bot to carry out this task. However before doing that, it would be important that Citation Bot stop creating new transcluded templates and instead substitute them. Boghog (talk) 19:19, 7 February 2015 (UTC)
 * Don't forget to fix cite hdl and cite jstor and any cite pmc that might exist. AManWithNoPlan (talk) 01:22, 8 February 2015 (UTC)
 * Thanks for the reminder. cite hdl – 41 transclusions, cite jstor – 819 transclusions. The cite pmc template doesn't exist. Boghog (talk) 06:59, 8 February 2015 (UTC)
 * Cite pmc became cite pmid per Templates_for_discussion/Log/2012_May_21. -- Ricky81682 (talk) 18:40, 23 February 2015 (UTC)


 * What ever happened to this? It's clear that consensus was achieved in the linked discussion - From the closing: "Existing and future DOI details should be included in articles, however, the bot function should remain, with a BRFA raised to change its function to use cite journal within articles without separate subpages." So it's clear that Citation bot should not be creating new cite doi/pmid/etc subpages, yet here he is, doing it. We're not waiting for consensus on that, we're waiting for the operator, so I changed the "waiting for" parameter on the template. cymru.lass (talk • contribs) 18:01, 15 June 2015 (UTC)
 * I am still waiting for the bot operator to make this change. Without this change, there will be a lot of unnecessary follow-up edits. There is also this related RFC.  If that RFC closes the way the first one did, then there will be an even stronger consensus to substitute instead of transcluding these cite templates. Boghog (talk) 19:36, 15 June 2015 (UTC)

I'm waiting for the discussions on deprecating cite:doi and cite:pmid to close and for there to be clear consensus on what the way forward here is. Once that is clear, the change should be possible, if probably not quite as straightforward as it sounds. --Fhocutt (WMF) (talk) 23:27, 9 October 2015 (UTC)
 * Depreciation won the day again. If there is no matching sub-page then it should be converted into a cite journal and expanded as best as possible with the available metadata.  AManWithNoPlan (talk) 17:56, 25 October 2015 (UTC)
 * Per the consensus to deprecate, I've created T119932, which includes removing all the functionality related to those templates. Regarding substituting existing templates, I think that would be a task best handled by a dedicated new bot (via a WP:BOTR request), rather than having Citation bot do it. Ryan Kaldari (WMF) (talk) 04:40, 1 December 2015 (UTC)
 * can do it. AManWithNoPlan (talk) 05:16, 1 December 2015 (UTC)
 * I've done most of it already. I'll finish it soon. I also have approval to do it. Waiting to remove my cast Ladsgroupoverleg 03:35, 2 December 2015 (UTC)
 * I removed about 1200 more cases. Everything left need human attention Ladsgroupoverleg 09:56, 2 December 2015 (UTC)
 * I assume that most of the human attention in cite pmid and cite doi is acutally just white space like .  Awesome work with your bot by the way. AManWithNoPlan (talk) 17:04, 2 December 2015 (UTC)
 * Most of problems came from using "cite doi" (two spaces) which is strange thing to use. I'm re-running  Ladsgroupoverleg 17:30, 2 December 2015 (UTC)
 * Please also check for "cite PMID", which seem to have been used in several subpages of template:PBB.LeadSongDog come howl!  21:26, 2 December 2015 (UTC)
 * since I don't think he watches this page AManWithNoPlan (talk) 21:41, 2 December 2015 (UTC)
 * One more thing. Some one (bot or human) needs to go back and delete all cite pmid, cite isbn, cite jstor, cite hdl, cite pmc, and cite doi that are not linked to anymore.  AManWithNoPlan (talk) 22:41, 2 December 2015 (UTC)
 * I'll do "cite pmid" soon. We need to delete cite doi subpages that are orphan. I don't have admin access to do it Ladsgroupoverleg 12:22, 3 December 2015 (UTC)

Thanks for all the work. Looks like doi with underscores in them lead to pages with spaces which dexbot didn't like AManWithNoPlan (talk) 05:28, 6 December 2015 (UTC)
 * This code is all now removed from the bot. And cite pmid is getting wiped. AManWithNoPlan (talk) 03:12, 18 December 2015 (UTC)

Google https
Here is the RFC https://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28proposals%29/Archive_127#RfC:_Should_we_convert_existing_Google_and_Internet_Archive_links_to_HTTPS.3F AManWithNoPlan (talk) 02:41, 15 November 2015 (UTC)
 * This should be fixed shortly. Kaldari (talk) 21:42, 15 December 2015 (UTC)
 * Fixed now: Ryan Kaldari (WMF) (talk) 02:54, 18 December 2015 (UTC)

Unhandled error
resolved

Establishing connection to Wikipedia servers with username Citation_bot... Using account Citation bot. Fetching parameter list ... done.Activated by Josve05a

Expanding 'Donald_Trump'; will commit edits. [01:30:04] Processing page 'Donald Trump' — edit—history * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting - https://books.google.com/books * Expanded from Google Books API * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting - https://books.google.com/books [..> process> lebooks> details> ddifnew> add] + Adding isbn * Expanded from Google Books API * Initial authors exist, skipping authorlink in tidy * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting - https://books.google.com/books [..> process> lebooks> details> ddifnew> add] + Adding isbn [..> process> lebooks> details> ddifnew> add] + Adding date * Expanded from Google Books API * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting
 * Donald Trump: Master Apprentice
 * Think Big and Kick Ass in Business and in Life
 * Time to Get Tough: Making America #1 Again
 * Donald Trump: Profile of a Real Estate Tycoon: Easyread Super Large 20pt Edition

- Checking AdsAbs database Similar title not found in database [..> process> rossref] - Checking CrossRef database for doi. [..> process> indpmid] - Searching PubMed... nothing found. * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting - https://books.google.com/books * Expanded from Google Books API * Initial authors exist, skipping authorlink in tidy * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting - https://books.google.com/books * Expanded from Google Books API * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting ! Unhandled error. Please copy this output and report a bug. ! Unhandled error. Please copy this output and report a bug. ! Unhandled error. Please copy this output and report a bug. history / last edit
 * Expand citation: Donald Trump emphasizes plans to build 'real' wall at Mexico border
 * Money
 * The Trumps: Three Generations That Built an Empire

(t) Josve05a  (c) 01:33, 12 December 2015 (UTC)

Turns Wikilink for author into broken brackets

 * For
 * the expected output from the bot would be
 * Stacy Mintzer Herlihy Herlihy Stacy Mintzer E. Allison Hagood Hagood E. Allison Paul A. Offit Offit Paul A.
 * The Graphene example is not the same problem: the ref had an unbalanced  and the bot simply removed that without altering the rest of the ref. -- Red rose64 (talk) 08:13, 25 March 2014 (UTC)
 * The Graphene example is not the same problem: the ref had an unbalanced  and the bot simply removed that without altering the rest of the ref. -- Red rose64 (talk) 08:13, 25 March 2014 (UTC)


 * , Graphene is the same problem as there was no unbalanced bracket in the reference.
 * Original ref
 * After Citation bot:
 * Bgwhite (talk) 17:29, 25 March 2014 (UTC)
 * Ah yes: I had assumed that since your first link (Vaccine controversies) was to the problem edit, the second link (Graphene) would also be to the problem edit. Instead, it seems that it's a link to your fix for the previous edit to that page. -- Red rose64 (talk) 17:36, 25 March 2014 (UTC)
 * Before the bot edit, most of the cite journal templates in the Graphene article used a single author parameter to store the authors. Furthermore the wikilinks were fully functional before the bot edit.  The bot is inserting a ridiculous number of new parameters in these templates in an attempt to produce clean metadata that no one will use.  It would be better to leave the author parameters in these templates untouched. Boghog (talk) 07:01, 27 March 2014 (UTC)
 * Ah yes: I had assumed that since your first link (Vaccine controversies) was to the problem edit, the second link (Graphene) would also be to the problem edit. Instead, it seems that it's a link to your fix for the previous edit to that page. -- Red rose64 (talk) 17:36, 25 March 2014 (UTC)
 * Before the bot edit, most of the cite journal templates in the Graphene article used a single author parameter to store the authors. Furthermore the wikilinks were fully functional before the bot edit.  The bot is inserting a ridiculous number of new parameters in these templates in an attempt to produce clean metadata that no one will use.  It would be better to leave the author parameters in these templates untouched. Boghog (talk) 07:01, 27 March 2014 (UTC)

Using author (singular) to store multiple names doesn't seem like a good idea. Before Citation Bot got to work, Revision 600716072 contained stuff like this:
 * - with a name and "et al" in "author" and then a load of first, but not last, names later in the list.
 * - with multiple names in "author" separated by commas.
 * - with multiple names in "author" separated by semi-colons.
 * - lists of names in "author" with some of the names wikilinked.

I prefer using "last"/"first" for persons and "author" for committees, departments and organisations. Using "authorlink" is more robust and this works with both the "last"/"first" and "author" parameters.

Citation bot made a in the Graphene article.

Why did it do this to the patent?

Why does the "last2" parameter get added at the end of the list of names instead of at the beginning? Why is "et al" not cleared from "author"? Why is "author" not changed to "last"/"first" to match the rest?

Why did the bot duplicate the name found in "last2"/"first2" into "last3"/"first3"?

Why are only "last2" to "last6" created and not "first2" to "first6"? Why is the "author" parameter with six names in it left untouched? This causes duplication in display.

Why was the working Wiley URL changed to a DOI attribute and then immediately marked as "dead"?

I have those and very many other errors. Although the parameter names are now the same for all references, there is very little consistency in the format of some of the data in the parameters. I have fixed all the dates, but "first names" are a mixture of either first name or initials, the latter found both with or without periods. -- 79.67.241.76 (talk) 00:43, 28 March 2014 (UTC)


 * Another article today, Delimiter. It messed up three refs.  Bgwhite (talk) 06:35, 28 March 2014 (UTC)


 * "Using author (singular) to store multiple names doesn't seem like a good idea." – Why not? Using a single parameter to store multiple authors produces more compact templates that don't overwhelm the surrounding wikitext. The only down side is that is doesn't produce clean author metadata.  However how many consumers of Wikipedia citation metadata are there?  I suspect not very many.  I agree that it is perhaps more logical to store multiple authors in authors (plural).  Nevertheless, per consensus and long established usage and consistent with the current cite journal documentation, full author lists can be stored in a single field called either "authors" or "author" without need for additional numbered author parameters. Boghog (talk) 07:07, 28 March 2014 (UTC)

With a free form input using "authors", there will be no consistency of display. Before Citation Bot got to work, the Graphene article contained the following names in references:


 * Comma between names (last-first):


 * - no period after initial
 * - period after initial


 * Comma in name (last-first)


 * - period after each initial, but no space between initials
 * - period after each inital and space between initials
 * - plus "et al." in italics
 * - one initial, with "and" between names
 * - two initials, with "and" between names


 * comma in and between names (last-first)




 * Semi-colon between names (last-first):


 * - plus "et al.", but not in italics
 * - plus "et al.", but not in italics
 * - plus "et al.", but not in italics


 * Comma between names (first-last):


 * - space, or not, between initials, with "and" before final entry
 * - initials before last name
 * - full first name before last name
 * - list of names, comma separated
 * - either initials or full first name, with "and" before final entry
 * - full first name, with "&" before final entry
 * - full first name, with "and" before final entry
 * - initial and last name
 * - one name plus "et al." in italics
 * - list of names plus "et al." not in italics


 * errors:


 * - with "et al." followed by extra "first" parameters, but no matching "last" parameters
 * - first entry spread over "author1" and "author2"
 * - two full names in "last" then "first1", "last2" and "first2" repeating parts of those names


 * separate last and first parameters with coauthors:


 * - deprecated parameter containing "et al."
 * - deprecated parameter containing list of authors


 * individual parameters, almost the "right" way:


 * - first names (mostly) as initials
 * - first names (mostly) in full

There were also references that would not display because the reference name had been duplicated, four or five different date formats including dates like "SEP，03，2013" with Chinese characters in them and many other issues. Half the "et al." were in italics and half were not.

By changing to separate parameters for names, all names display in "last, first" order with the same separators throughout. The only variation is whether the first name is stated in full or is initials, and whether there are periods after initials or not. A bot can fix those entries to be consistent. If "et al." is specified it is currently in authorn+1 where n is the highest numbered "lastn"/"firstn" parameter. The number of authors to display can also be set using the display-authors parameter. -- 79.67.241.76 (talk) 11:19, 28 March 2014 (UTC)


 * The format of the author names could just as easily been standardized using a single author parameter which would have avoided all the parameter bloat. There is no "house style" for citations, hence there is no single "right way" to format citations. A single author parameter was the predominate style before your edits. Per WP:CITEVAR, if you want to change this style, you should have obtained consensus for this change on the article talk page before your edits. Boghog (talk) 12:35, 28 March 2014 (UTC)
 * If you want to list more than one author in a parameter, then use authors. If you want to use a single author, use author.  Martin  (Smith609 – Talk)  11:28, 12 July 2014 (UTC)

author= converted to authors= and author=
See the diff. The bot converted  to.

There is a similar problem here, where author was left in place while last1 etc. were added. The bot should choose one or the other. – Jonesey95 (talk) 20:56, 13 October 2014 (UTC)
 * This edit is similar, but it also added "author2=et al." instead of the author's actual name. Setting displayauthors to 1 generates "et al." automatically. If someone manually removes displayauthors, the author names will be wrong. – Jonesey95 (talk) 05:59, 20 October 2014 (UTC)
 * This bug is still present. – Jonesey95 (talk) 23:39, 13 January 2015 (UTC)
 * As shown here the bot hasn't been edited since 1 September. Still needs a relief operator. LeadSongDog come howl!  04:45, 14 January 2015 (UTC)

Butchered author names
By the way, how can you justify leaving an unmaintained bot in service?

9 no longer necessary for exactly nine authors

 * The bot is still inserting this parameter (see the Hongying citation), but it is not necessary., can you disable this portion of the code? – Jonesey95 (talk) 03:29, 22 September 2015 (UTC)


 * The reasoning for keeping this appears to be that it is better to show the 9th name rather than having et al. hide only one name. I can take it out easily, but has that reasoning changed? --Fhocutt (WMF) (talk) 02:00, 23 October 2015 (UTC)
 * Looks like the testwiki cite journal template may be out of date? https://test.wikipedia.org/w/index.php?title=User:Fhocutt_%28WMF%29/Citation_bot_test&oldid=244444 shows the et al., which the same citation on enwiki does not. --Fhocutt (WMF) (talk) 02:23, 23 October 2015 (UTC)
 * The CS1 citation module (and its predecessor code) that renders cite journal and similar templates used to display "et al." when exactly nine authors were listed, for reasons too boring to go into. The module no longer does that; it displays all authors unless x is used (where x is a number less than the number of authors in the cite template). So 9 is redundant when there are exactly nine authors listed. – Jonesey95 (talk) 02:46, 23 October 2015 (UTC)
 * Great, thanks for the explanation. I'll just take that special case out, then. I've updated the CS1 module on testwiki as well. --Fhocutt (WMF) (talk) 23:14, 23 October 2015 (UTC)
 * BTW, I wasn't trying to be coy or evasive by saying it was too boring to get into. Here's the boring reason relating to backwards compatibility and eventually updating the module after we cleaned up all of the old citations with a ton of help from Citation Bot. – Jonesey95 (talk) 05:07, 24 October 2015 (UTC)

Blanking pages on some PC platforms

 * When starting the bot from the "edit article" mode, it consistently blanks the page, which is annoying as I lose my other corrections. This happens on Win XP + Firefox 19, but not on newer platforms with a more recent Firefox. I anticipate an advice to upgrade, aiming for a more recent FF Java, but there is a reason for using FF19 on that laptop: after some Wikimedia updates newer FF versions were spontaneously jumping into a 70-99% CPU consumption mode while editing en.wiki (don't know if this has improved lately, but don't want to test either - FF19 is more robust in this sense, and laptops don't handle CPU surges so well). Materialscientist (talk) 05:37, 5 December 2015 (UTC)
 * Neither Fx 19 nor Win XP are supported by their organizations any longer nor have been for quite a while. As you expected, the solution is to upgrade, use a different browser, use a different OS, or all of the above. Not a bug IMO. --Izno (talk) 16:17, 5 December 2015 (UTC)
 * The same happens now with me on the latest Windows + latest Firefox, so it might be a bug. Note that WMF programmers took over Citation Bot code maintenance after the bot was blocked for same behavior - page blanking. Materialscientist (talk) 00:26, 18 December 2015 (UTC)
 * Looking into this currently. Does it happen on every page or just certain pages? Ryan Kaldari (WMF) (talk) 00:36, 18 December 2015 (UTC)
 * Actually right now it fails with a message "Citation request failed". Materialscientist (talk) 00:46, 18 December 2015 (UTC)
 * Yeah, I just added some code to the gadget to keep it from blanking pages if the replacement text is empty. I'm not sure exactly what's going on here. I'm sometimes seeing the same bug you describe and sometimes it works fine (on the same article), and the data being returned by the API seems to be valid in either case. May be some kind of race condition or bug in the AJAX request. Still troubleshooting it. Ryan Kaldari (WMF) (talk) 01:06, 18 December 2015 (UTC)
 * This is (un?)fortunately not the same bug that had the bot consistently blanking pages. That one was caused by the HTTPS-only change, leading to nonexistent responses that the bot translated into blank pages. This one is apparently more subtle. Can you provide links to the revisions where this happens and as much browser information as possible? --Fhocutt (WMF) (talk) 01:07, 18 December 2015 (UTC)
 * We are talking about launching the bot from the "edit article" mode, and I obviously don't save those edits. The operation returned to normal while I was typing this reply, so I suspect (i) the problem is related to Wikimedia servers and doesn't occur when they run normally, and (ii) the latest instability (but not my original complaint on top of this thread) is hardly platform dependent. Materialscientist (talk) 01:17, 18 December 2015 (UTC)
 * I think we have it fixed now (it seems to have been an issue with the content headers in the API response). Also, the gadget code is now more defensive, and if it fails to receive valid data back, it should now give an error message instead of blanking the page. Thanks again for the bug report. Ryan Kaldari (WMF) (talk) 01:38, 18 December 2015 (UTC)
 * Yes, from "edit article"--if page blanking shows up again, a link to the revision you were trying to edit (since you couldn't save the blanked page) is probably the most useful debugging information available. But as Kaldari says, the gadget now fails more robustly so you at least won't lose your changes now. Fhocutt (WMF) (talk) 23:54, 18 December 2015 (UTC)

Update: Security fixes and code cleanup
I just finished and merged some substantial updates to the Citation bot codebase.
 * I found and fixed many cross-site scripting vulnerabilities
 * Citation bot will now entirely ignore any remaining Template:Cite doi/pmid/jstor templates, and related code has been removed
 * As there is no one currently interested in taking on maintainership, I removed all code related to the previous automated citation-fixing capabilities. It should be easy for a new maintainer to write an automated bot that does not rely on the Template:Cite doi-type system.

If anything is newly broken, you can leave a comment here or file a ticket on Phabricator. --Fhocutt (WMF) (talk) 21:24, 17 December 2015 (UTC)

Flagging for archiving, now that discussion is done notabug

Cite PMID disabled

 * Expansion of cite pmid has been disabled for unclear reasons. Can you clarify?  Why was it disabled? Unclear to me.  Fhocutt (WMF) indicates that security problems were resolved at the same time it was disabled, but it's unclear if there were any security problems in the cite pmid-expanding code that was disabled.--Elvey(t•c) 17:45, 20 December 2015 (UTC)


 * As Template:Cite pmid has been deprecated (along with Template:Cite doi and Template:Cite jstor), I disabled the functions that create new Template:Cite-type subpages to avoid creating new pages from a deprecated templating system. The security issues were separate and mostly involved the Labs-hosted tool. --Fhocutt (WMF) (talk) 20:15, 20 December 2015 (UTC)


 * you can still insert  into a page, or into your sandbox, and then use Citation Bot to expand that into a full journal citation. – Jonesey95 (talk) 20:41, 21 December 2015 (UTC)


 * That's not my experience. https://tools.wmflabs.org/citations/doibot.php?edit=toolbar&slow=1&user=Elvey&page=User%3AElvey%2Fsandbox does nothing. / remains untouched.--Elvey(t•c) 00:35, 25 December 2015 (UTC)
 * My reading of the RFC comments is that what most people felt should stop was the expansion onto separate pages of the template. Not that expansion of template should stop or it should be deleted.--Elvey(t•c) 00:45, 25 December 2015 (UTC)

It was decided that the existing ones would be substituted in and then the sub-templates once unused would be deleted, since they we not be watched by anyone. Continuing to support this template gives the false appearance that it is supported. Other than your sandbox, there are no pages on all of Wikipedia that used cite pmid. AManWithNoPlan (talk) 03:46, 25 December 2015 (UTC)


 * OK. Not convinced that's a good enough reason to disable the expansion, but there's no consensus, it seems, so let's drop it.  True.  --Elvey(t•c) 20:58, 26 December 2015 (UTC)


 *  Still remains untouched.--Elvey(t•c) 20:58, 26 December 2015 (UTC)
 * of course it is untouched, because it is invalid. The second pipe should be an equals. AManWithNoPlan (talk) 02:35, 27 December 2015 (UTC)
 * It works fine if you provide a valid PMID in the format I showed above. The example provided by, "1942838746454", is not a valid PMID, and you need an equals sign (=) between "pmid" and the number. – Jonesey95 (talk) 15:10, 2 January 2016 (UTC)

wontfix

error
Works on current version with current bot AManWithNoPlan (talk) 23:51, 5 February 2016 (UTC)

Questions: Interest in and prerequisites for starting up automated citation fixing
I've gotten Citation Bot back up and running for the manually activated tools (MediaWiki:Gadget-citations and https://tools.wmflabs.org/citations/doibot.html). Previously, Citation bot 1 also used maintenance categories to guide its task of continuous automatic citation fixing. I'm investigating interest in and requirements for bringing back this function of Citation bot. My questions: --Fhocutt (WMF) (talk) 22:31, 30 November 2015 (UTC)
 * Is there community interest in bringing back a constantly running, fully automated Citation bot?
 * What minimum standard would the bot need to meet to run automatically?
 * Are there any current problems with the bot that mean that it should be limited to running when manually activated?
 * Although an already-approved bot may not need an active maintainer, citation style appears to be an active and contentious enough area that an automatically running Citation bot should have one. I will soon be busy with work identified in the Community Wishlist Survey and will not be able to maintain this bot. However, I am willing and able to help a new maintainer become familiar with the code. Is anyone interested in maintaining Citation bot?


 * On your last point: yes, citations are sufficiently critical that any automated bot should have an active maintainer in the house. On your second point, we might consider what period of time – a weak proxy for number of edits – is sufficient to have a good probability of hitting any problems. ~ J. Johnson (JJ) (talk) 22:52, 30 November 2015 (UTC)
 * On the first question, I do belive so. As long as it is not making too many bad fixes, there is community interest for there to be a bot that fixes stuff like this. (There are intrest for all things which makes the encyclopedia better, problem is only if it going to cause problems while doing so. Hency why people oppose(d) new extentions).
 * On the second question I belive that would be something for the BAG (Bot Approval Group) to detemine, but I'd say a trial run of ~50-100 pages and evaluate.
 * (t) Josve05a  (c) 03:18, 6 December 2015 (UTC)
 * It's running. Flag for archive.  notabug

Certain things stop non-interactive bot from running
https://en.wikipedia.org/w/index.php?title=Battle_of_France&oldid=696441783 AManWithNoPlan (talk) 14:25, 27 December 2015 (UTC)
 * there was a book with two titles--the bot when run in an edit window finds this for me. I fixed that and now the bot runs.  There is nothing for it to do though. AManWithNoPlan (talk) 14:31, 27 December 2015 (UTC)
 * Not sure if this is a bug or a feature of running the bot in the 'I do not want to preview my changes' mode AManWithNoPlan (talk) 16:37, 2 January 2016 (UTC)

No discussion so flag as notabug AManWithNoPlan (talk) 19:39, 14 July 2016 (UTC)

Bot added "comment placeholder"
Looks like someone is working on the comment code AManWithNoPlan (talk) 00:58, 6 February 2016 (UTC) fixed AManWithNoPlan (talk) 19:26, 14 July 2016 (UTC)

Category:Cite doi templates and Category:Cite pmid templates
I don't know if citation bot has admin rights (I doubt it) but there's about 28k orphaned template:cite doi subpages at Category:Cite doi templates and about 7k similar template:cite pmid subpages at Category:Cite pmid templates, all of which I believe were created by citation bot. See Bot_requests. There's a discussion at WT:CSD about the orphaned ones for the moment. I'm wondering if we can get an agreement that the 'creator' here on the orphaned ones being deleted and if possible, to substitute the remainder based on the deprecation decisions at both templates and to support deletion for those. -- Ricky81682 (talk) 22:57, 11 February 2016 (UTC)
 * I read your final sentence three times and cannot make sense of it. What do you want?


 * If you're looking for the operator of Citation Bot to agree that something should happen to templates created by the bot, I don't think that makes sense. The bot created templates only at the request of individual editors. Those individual editors created cite doi or similar templates to be filled in by the bot, so the bot is not the creator of any of the templates in any meaningful way. Yes, the template's history may say that the bot created the template, but the bot did not do so of its own volition. – Jonesey95 (talk) 23:53, 11 February 2016 (UTC)
 * Let me try again. (1) there are orphaned template subpages created by citation bot. Orphaned either by substitution or more often by deletion of the page/change in citation/other reasons. I asked if we could delete those by G6. There's opposition and if deletion isn't going to be considered by CSD under G6, then I'm trying to get it by G7 deletion. I don't want to be hunting down the actual editors who created each request. Else there could be a TFD request for above 35k pages to be deleted and that will require some adminbot to enforce anyways. (2) the cite doi template usage by subpage has been considered deprecated so I was wondering if the bot operator would consider using the bot to substitute the usage. That's secondary but if combined with (1) when those pages become orphaned, that would speed up the deprecation. -- Ricky81682 (talk) 09:45, 12 February 2016 (UTC)
 * The creator (or at least the long-time maintainer) of this bot,, is no longer maintaining this bot. It has been resurrected by some WMF staff, I believe, but they are looking for an en.WP maintainer to take over. You are unlikely to get Citation Bot to perform any new functions, but another bot operator may be willing to finish the substitutions. It looks like all of the Cite PMID templates have been substituted already. – Jonesey95 (talk) 16:27, 12 February 2016 (UTC)

notabug AManWithNoPlan (talk) 19:33, 14 July 2016 (UTC)

Bot has been logged out
Output from running "Expand citations" from my left toolbar:

Welcome to Citation Bot Please wait while the Citation bot processes the page you requested.

Establishing connection to Wikipedia servers with username Citation_bot... Using account Citation bot. Fetching parameter list ... done.Activated by Jonesey95

Expanding 'IceCube_Neutrino_Observatory'; will commit edits. [17:53:26] Processing page 'IceCube Neutrino Observatory' — edit—history

. Checking that DOI 10.1126/science.1242856 is operational... DOI ok. [..> ndbydoi] - Expanding from crossRef record (ok) - Checking AdsAbs database [..> yadsabs] Match for pagination but database journal "Applied Surface Science" didn't match "journal = Science". [..> indpmid] - Searching PubMed... no results. nothing found. * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting
 * Expand citation: Evidence for High-Energy Extraterrestrial Neutrinos at the IceCube Detector [..> rifydoi]

~ Recognized DOI in URL; dropping URL * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting [..> rifydoi] . Checking that DOI 10.1103/PhysRevLett.113.101101 is operational... DOI ok. [..> ndbydoi] - Expanding from crossRef record (ok) * Initial authors exist, skipping authorlink in tidy - Checking AdsAbs database [..> yadsabs] [..> indpmid] - Searching PubMed... no results. nothing found. * Initial authors exist, skipping authorlink in tidy * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * initial authors exist, not correcting * Getting data from arXiv 1410.7227 [..> add] + Adding last1 [..> add] + Adding first1 [..> add] + Adding last2 [..> add] + Adding first2 [..> add] + Adding last3 [..> add] + Adding first3 [..> add] + Adding last4 [..> add] + Adding first4 [..> add] + Adding last5 [..> add] + Adding first5 [..> add] + Adding last6 [..> add] + Adding first6 [..> add] + Adding last7 [..> add] + Adding first7 [..> add] + Adding last8 [..> add] + Adding first8 [..> add] + Adding last9 [..> add] + Adding first9 [..> add] + Adding last10 [..> add] + Adding first10 [..> add] + Adding last11 [..> add] + Adding first11 [..> add] + Adding last12 [..> add] + Adding first12 [..> add] + Adding last13 [..> add] + Adding first13 [..> add] + Adding last14 [..> add] + Adding first14 [..> add] + Adding last15 [..> add] + Adding first15 [..> add] + Adding last16 [..> add] + Adding first16 [..> add] + Adding last17 [..> add] + Adding first17 [..> add] + Adding last18 [..> add] + Adding first18 [..> add] + Adding last19 [..> add] + Adding first19 [..> add] + Adding last20 [..> add] + Adding first20 [..> add] + Adding last21 [..> add] + Adding first21 [..> add] + Adding last22 [..> add] + Adding first22 [..> add] + Adding last23 [..> add] + Adding first23 [..> add] + Adding last24 [..> add] + Adding first24 [..> add] + Adding last25 [..> add] + Adding first25 [..> add] + Adding last26 [..> add] + Adding first26 [..> add] + Adding last27 [..> add] + Adding first27 [..> add] + Adding last28 [..> add] + Adding first28 [..> add] + Adding last29 [..> add] + Adding first29 [..> add] + Adding last30 [..> add] + Adding first30 [..> add] + Adding display-authors [..> add] + Adding class [..> add] + Adding doi [..> rifydoi] . Checking that DOI 10.1103/PhysRevD.91.072004 is operational... DOI ok. [..> ndbydoi] - Expanding from crossRef record [..> add] + Adding journal [..> add] + Adding volume [..> add] + Adding issue (ok) [..> ndbydoi] - Expanding from crossRef record (ok) [..> forget] - Dropping parameter class ! LOGGED OUT: The bot has been logged out from Wikipedia servers ! LOGGED OUT: The bot has been logged out from Wikipedia servers ! LOGGED OUT: The bot has been logged out from Wikipedia servers history / last edit
 * Expand citation: Observation of High-Energy Astrophysical Neutrinos in Three Years of IceCube Data

Views

User page Discussion Report error Contributions

See the final message: " ! LOGGED OUT: The bot has been logged out from Wikipedia servers" – Jonesey95 (talk) 17:58, 12 February 2016 (UTC)


 * Is nobody else seeing this? I've been getting it for about a week and assumed it was just some administrative thing.  Maybe not? Lithopsian (talk) 17:45, 15 February 2016 (UTC)


 * It's been doing that for me as well. Been a couple of days now. Headbomb {talk / contribs / physics / books} 18:15, 15 February 2016 (UTC)
 * Haven't been able to expand the latest citations added to Dodo and Rodrigues solitaire. FunkMonk (talk) 18:20, 15 February 2016 (UTC)
 * This is probably due to the forced logout of all users. --Izno (talk) 18:36, 15 February 2016 (UTC)
 * Is it possible to modify bot to check if logged out before checking citations? Spirit Ethanol (talk) 05:33, 17 February 2016 (UTC)
 * Still seeing it - does it work for anyone? Is there some manual workaround? Richiez (talk) 20:57, 19 February 2016 (UTC)
 * can you get someone to log the bot back in or poke at why it's logged out? I realize you're busy but the bot is basically broken. --Izno (talk) 21:59, 19 February 2016 (UTC)
 * And since you both seem to be related to the mass-logout. --Izno (talk) 22:00, 19 February 2016 (UTC)
 * The mass logout was a one-time thing; maybe this is related to the memcached server upgrades which I think is still ongoing. OAuth/Owner-only consumers is an alternative to traditional login that should be more resilient. --Tgr (WMF) (talk) 22:08, 19 February 2016 (UTC)
 * Or it could be that the bot isn't properly handling cookies: it copies the Set-Cookie headers from the initial call to action=login (without correctly handling deleted cookies), then it manually constructs cookies from the deprecated response values which are likely to be failing to be valid anymore with the increased validation of the session data. Anomie⚔ 01:14, 20 February 2016 (UTC)
 * Still not logged in :-( GoingBatty (talk) 01:16, 22 February 2016 (UTC)
 * The bot can still be activated from the (optional) edit toolbar button. Which is nice because that never used to work for me. Lithopsian (talk) 19:36, 23 February 2016 (UTC)
 * how exactly do you use this method? —Chris Capoccia  T&#8260;C 14:29, 25 February 2016 (UTC)
 * In the Editing section on the the gadgets tab in preferences, there is a "Citation expander" option. Choose that and a button will appear on the toolbar when you are editing an article.  If you press that button, cite templates will be expanded by Citation Bot within your existing editing session.  You can then examine and save the changes.  For me, this only seems to work once per editing session, but it works even when the bot is logged out. Lithopsian (talk) 14:18, 1 March 2016 (UTC)
 * Still showing the same error, which is preventing the bot writing any edits. Modest Genius talk 13:21, 1 March 2016 (UTC)

Frances Hocutt is on leave. So I guess should be pinged about this? --Closedmouth (talk) 05:03, 4 March 2016 (UTC)
 * Thanks for the ping. I created and will see if someone can look into this further. Ryan Kaldari (WMF) (talk) 23:49, 4 March 2016 (UTC)
 * In the meantime, I've temporarily removed the "Expand citations" link in the toolbar (since it doesn't work). You can still expand citations from the editing interface, however. Kaldari (talk) 01:20, 25 March 2016 (UTC)
 * You mean click on expand citations, without it working? FunkMonk (talk) 03:32, 7 April 2016 (UTC)
 * No, I mean click the "Citations" button from the bottom of the editing interface. This does all the citation expanding within the editing interface and let's you review it before it is saved. Ryan Kaldari (WMF) (talk) 16:24, 2 May 2016 (UTC)
 * Cool, someone else told me about that method in the meantime, I had never heard of it... FunkMonk (talk) 16:27, 2 May 2016 (UTC)
 * I've fixed the bug and submitted a pull request (https://github.com/ms609/citation-bot/pull/106). Now we just have to wait for Martin to merge it. Ryan Kaldari (WMF) (talk) 16:25, 2 May 2016 (UTC)

fixed AManWithNoPlan (talk) 19:37, 14 July 2016 (UTC)

Cleanup Bot Request
A bot request (Bots/Requests for approval/AnomieBOT III 2) to remove thousands of pages you created is in progress, please place any comments about this at the bot request page. — xaosflux  Talk 14:32, 27 March 2016 (UTC)

notabug This is now done and over AManWithNoPlan (talk) 19:36, 14 July 2016 (UTC)

Correct DOI from invalid Bibcode
, can you provide an example diff to illustrate what you mean? – Jonesey95 (talk) 18:21, 27 April 2016 (UTC)
 * I've set up a minimal example in my sandbox. It shows two citations each containing only a bibcode tag whose value is actually a DOI instead of a bibcode, being converted magically to a full citation, but with no hint that the bibcodes are still invalid.  Note that clicking the first one actually loads an ADS page for the paper, while clicking the second does not.  Lithopsian (talk) 18:37, 27 April 2016 (UTC)
 * It looks like Citation Bot is doing a fabulous job with this citation, and the invalid Bibcode still takes you to the right article on the adsabs site. I think Citation Bot's job is done here.


 * It sounds like you would like an enhancement to cite journal that checks the Bibcode for validity. You should request this at Help Talk:Citation Style 1, where we discuss error checks like this all the time. We have implemented tests for doi, isbn, arxiv, issn, language, and other parameters. Please provide a link to a source that describes the Bibcode specification so that the template code can identify invalid Bibcodes. Thanks. – Jonesey95 (talk) 18:56, 27 April 2016 (UTC)


 * Except that the bibcode doesn't always link to the journal. That's why I gave two citations.  For whatever reason, clicking one gets to the correct journal, while the second just gives a not-found page. Maybe it should be fixed in the template, but it makes me uncomfortable that the bot is skating over something invalid and giving the impression that all is well.  Lithopsian (talk) 13:47, 28 April 2016 (UTC)


 * Right, which is why I recommended posting a request at Help Talk:Citation Style 1, which I have done for you. – Jonesey95 (talk) 16:57, 28 April 2016 (UTC)

notabug templates now flag invalid bibcodes and put them is a maintenance category. AManWithNoPlan (talk) 19:36, 14 July 2016 (UTC)

end-page bug
anyone here? need more details?Ihaveacatonmydesk (talk) 19:42, 28 May 2016 (UTC) notabug the cross-ref data is not always complete. AManWithNoPlan (talk) 19:28, 14 July 2016 (UTC)

comment inside citation template breaks the bot and causes unexpected exit
Flag as notabug, since duplicate bug. AManWithNoPlan (talk) 19:33, 14 July 2016 (UTC)

End of run, link back to start page
Flag as notabug, since it is duplicate of other bug AManWithNoPlan (talk) 19:26, 14 July 2016 (UTC)

Bot adds a redundant encyclopedia tag to cite encyclopedia template with journal tag

 * IMO this is user error. Use cite encyclopedia with encyclopedias and cite journal with journals. --Izno (talk) 21:22, 6 June 2016 (UTC)
 * Okay then notabug AManWithNoPlan (talk) 19:26, 14 July 2016 (UTC)

Adding author initials as lower case
I highly doubt this is a bug. Every other time this bug is reported it always turns out to be the DOI database has bad data. AManWithNoPlan (talk) 19:18, 14 July 2016 (UTC) notabug

Bot doesn't run
This is a duplicate of another bug (The comments kill the bot bug). Flagging as notabug to close AManWithNoPlan (talk) 19:56, 16 July 2016 (UTC)

Other languages
Does this bot operate in other wikis? I am importing the FAC toolbox to the Spanish wikipedia and I would like to know if it works. Triplecaña (talk) 10:16, 6 March 2016 (UTC) notabug

Bot adds issue where number is present
This bug has been reported as fixed once before, but it appears that something happened to that fix or that the fix does not work in all cases. – Jonesey95 (talk) 12:52, 9 June 2016 (UTC)
 * The better fix would be to standardize on the use of issue, changing instances of number to issue where found (provided this does not result in a no-displayed-change edit). I've manually done so at that article. LeadSongDog come howl!  15:56, 9 June 2016 (UTC)
 * It should be noted that it does the same thing for pages when pp is already present Ihaveacatonmydesk (talk) 16:33, 9 June 2016 (UTC)
 * duplicate bug. Flag as notabug for archiving.  Will copy link to other bug AManWithNoPlan (talk) 20:04, 16 July 2016 (UTC)

Not helpful
It's simply a waste of server time and resources to, especially if it's the only edit it makes to the article. --RexxS (talk) 10:50, 20 June 2016 (UTC)
 * Broken DOIs can be reported to the publisher so that they can be fixed. Marking the DOI as broken puts the article in a hidden maintenance category, . – Jonesey95 (talk) 13:54, 20 June 2016 (UTC)
 * I see you used the passive voice in the first sentence. It does rather beg the question who is going to report a broken doi to the American Review of Respiratory Disease for an article it published in 1974? I'm also pretty sure that you can put an article in hidden maintenance category without polluting the data in the reference. --RexxS (talk) 15:20, 20 June 2016 (UTC)
 * I have reported broken DOIs. I just reported the one in the citation listed above. – Jonesey95 (talk) 16:01, 20 June 2016 (UTC)
 * Odd, the url seems ok, though it's ATS membership-required for access. Either they already fixed it, or the bot only checked crossref.org, which does not correctly report it as available. LeadSongDog come howl!  16:17, 20 June 2016 (UTC)
 * Well if they have just fixed it, that was quick. Well done Jonesey95! I'd better put the doi back in the reference. Cheers --RexxS (talk) 17:50, 20 June 2016 (UTC)
 * (e/c) This DOI also does not work at http://dx.doi.org, which usually means that it is not registered correctly in the right databases. Hence the usefulness of the broken doi parameter and category: getting these broken DOIs fixed these helps readers verify statements in articles more easily. – Jonesey95 (talk) 17:52, 20 June 2016 (UTC)
 * In general, I certainly have, and will continue to, report DOIs given on publisher's websites that don't appear at crossref. Unfortunately mixed results getting any repaired, never hear back from some publishers, though have had fixes and polite responses from the likes of Nature and Science in the past though. Rjwilmsi  18:34, 20 June 2016 (UTC)

Discussion ended, so flag to archive notabug AManWithNoPlan (talk) 19:57, 16 July 2016 (UTC)

Only partially fixes malformed book-title-in-publisher citation
That might be hard, let me think about it. AManWithNoPlan (talk) 20:05, 8 November 2022 (UTC)