This is an archive of Module:Citation/CS1/Feature requests for those requests that have been completed.

Asian titles

Main work titles are formatted in italics. This is not appropriate for Asian scripts such as kanji, hangul and the like. Asian titles may also be underlined or placed in brackets 『』 or 《》.

See ((Asiantitle)) for current support. --— Gadget850 (Ed) talk 14:36, 29 March 2013 (UTC)

And such styling should be done with CSS per #Presentation and content. --  Gadget850 (Ed) talk 14:42, 1 April 2013 (UTC)

See also: Module talk:Citation/CS1/Archive 11#non-italic titles

—Trappist the monk (talk) 13:13, 12 September 2014 (UTC)

Error message help pages

 Done

Each error message should link to a help page. I will take on the task of creating the help. --  Gadget850 (Ed) talk 12:05, 3 April 2013 (UTC)

Page url

Editors often link to specific pages in an online book by wrapping a |page= or |pages= parameter in an external link. This practice can corrupt COinS data for the citation. This feature request suggests the creation of a new parameter that allows editors to continue linking to individual pages without corrupting COinS data. Linking to individual pages is supported at the WP:PAGELINKS guideline, which see.

|pageurl=: URL of an online book's page or pages where the cited text can be found. While not required, if provided, |url= must link to the same source as |pageurl=.

Examples:
Single-page link –
|page=18
|pageurl=http://books.google.com/books?id=kvpby7HtAe0C&pg=PA18
renders as: p. 18.
Page-range link (to the first page in the range) –
|pages=18–24
|pageurl=http://books.google.com/books?id=kvpby7HtAe0C&pg=PA18
renders as: pp. 18–24.
Multiple page links (only with |pages=; urls listed in |pageurl= must follow the same order as the pages listed in |pages=) –
|pages=18–24, 56
|pageurl=http://books.google.com/books?id=kvpby7HtAe0C&pg=PA18 http://books.google.com/books?id=kvpby7HtAe0C&pg=PA56
renders as: pp. 18–24, 56.
–or–
Multiple page links (forces pp. prefix unless |nopp=y; |pageurln= matches the last of |pagen= or |pagesn= or |atn=) –
|pages=18–24
|pages2=34 —this parameter ignored because it is followed by |page2=
|page2=56
|pageurl=http://books.google.com/books?id=kvpby7HtAe0C&pg=PA18
|pageurl2=http://books.google.com/books?id=kvpby7HtAe0C&pg=PA56
renders as: pp. 18–24, 56.

While two possible multiple page link handling methods are described, only one should be implemented.

Categories

—Trappist the monk (talk) 18:55, 3 April 2013 (UTC)

Me gusta This is needed. I think I like the pageurl1, pageurl2, etc version. Jason Quinn (talk) 00:35, 16 September 2013 (UTC)
I've been recently wondering if |pageurl= is the wrong solution to the problem. The problem is that urls in |page=, |pages=, and |at= corrupt COinS metadata. Because specifying external links using wiki-markup is standardized, it seems to me that its possible that with the application of a little bit of code, the page numbers can be extracted from the external link wiki-markup for use in COinS without the need for new parameters.
The benefit is that nothing except the template code needs to be reworked.
—Trappist the monk (talk) 12:12, 16 September 2013 (UTC)

website for cite web

 Done

Add 'website' as an alias for 'work'. Many editors seem confused by the use of 'work' to indicate the website. --  Gadget850 (Ed) talk 16:28, 6 April 2013 (UTC)

Done in sandbox. Dragons flight (talk) 19:56, 8 April 2013 (UTC)
((cite web)) documentation updated.
—Trappist the monk (talk) 14:34, 27 April 2013 (UTC)

Move toward uniform parameter name style

Add these parameter names to the whitelist. These replace their counterparts which are hyphenated, spaced, underscored, or camelcased. Template documentation should be updated to use these parameter names. The purpose is to move toward uniform parameter-name style. Other variants of these names should then, over time, be deprecated (I can always hope).

['authorfirst#'] = true,
['authorlast#'] = true,
['authornamesep'] = true,
['authorseparator'] = true,
['doibrokendate'] = true,
['doiinactivedate'] = true,
['editorfirst#'] = true,
['editorgiven#'] = true,
['editorlast#'] = true,
['editornamesep'] = true,
['editorseparator'] = true,
['namesep'] = true,
['seriessep'] = true,
['templatedocdemo'] = true,
['transchapter'] = true,
['transtitle'] = true,

—Trappist the monk (talk) 18:27, 6 April 2013 (UTC)

I agree we are very inconsistent about word separators, either nothing, hyphen, underscore, or space, and moving toward consistency is an excellent goal. That said, I'm not sure that removing the separator actually makes the most sense. For example, would it be easier for users if we consistently supported the use of a hyphen as the preferred word separator in parameter names? I have to think that having some separator is easier to read than having no separator. Dragons flight (talk) 19:19, 8 April 2013 (UTC)
Infoboxes (and navboxes, IIRC) seem to be standardising, de facto, on an underscore. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:15, 17 May 2013 (UTC)

 Done. This was resolved in favor of hyphenated multi-word parameters in a July 2014 RFC. – Jonesey95 (talk) 06:38, 12 December 2014 (UTC)

Check for uri scheme in url parameters

 Done

One of the more common url errors is the omission of the uri scheme in |url=-type parameters.

((cite web |url=www.example.com |title=Missing scheme |accessdate=2013-04-10)) → [www.example.com "Missing scheme"]. Retrieved 2013-04-10. ((cite web)): Check |url= value (help)

CS1 can check for this and report an error |url= requires http:// and categorize into Category:CS1 url missing uri scheme errors or some such.

—Trappist the monk (talk) 11:29, 10 April 2013 (UTC)

See Help:External link icons for the list of supported URIs. --  Gadget850 (Ed) talk 11:32, 10 April 2013 (UTC)
"requires http://" would not be appropriate phrasing; as another scheme may apply. Were that not the case, we could programmatically apply http://. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:38, 10 April 2013 (UTC)
Ok, how about this: |url= missing URI scheme?
—Trappist the monk (talk) 11:53, 10 April 2013 (UTC)
OK, but it is URI scheme. Note to implementer: news: and mailto: do not require the //. --  Gadget850 (Ed) talk 12:02, 10 April 2013 (UTC)
Ok, changed.
—Trappist the monk (talk) 13:10, 10 April 2013 (UTC)
Given the reality of URI schemes, I think the best we are likely to do is check that a URL either starts with "//" (protocol relative URLs) or has a ":" occurring in it somewhere. That covers most uses though it wouldn't catch the case of a bad "scheme", like htp://www.foobar.com/. Dragons flight (talk) 20:20, 10 April 2013 (UTC)
Check added in sandbox. Dragons flight (talk) 23:32, 10 April 2013 (UTC)

Date: error checking

Currently, and invalid date shows as entered:

Markup Renders as 
((cite book |last=Drucker |title=Book |date=Flubtember 2013 |ref=harv))

Drucker (Flubtember 2013). Book. ((cite book)): Check date values in: |date= (help); Invalid |ref=harv (help)

But the year is not extracted and inserted into the anchor as expected. There is no error message on this, so the editor is not aware of the issue.

I don't know how much checking can be done on 'date', but at a minimum if |ref=harv and the extracted year is blank, then an error message should be generated. --  Gadget850 talk 15:46, 6 May 2013 (UTC)

Is it also an error if |ref=harv and no date / year is given at all? Dragons flight (talk) 22:02, 20 May 2013 (UTC)
Yes. --  Gadget850 talk 22:08, 20 May 2013 (UTC)
What about websites and other resources where the date of publication may be not obvious or missing? Even if there is no date, I believe that ((sfn)), etc. will still work as long as the year is also omitted from the footnote template. At the moment, I'd lean towards reporting an error when date is present and it can't figure out the year, but not report an error if date is blank. Dragons flight (talk) 18:33, 22 May 2013 (UTC)
One problem is if the author has more than one publication, then the id is going to be duplicated. --  Gadget850 talk 18:43, 22 May 2013 (UTC)
You mean if there are multiple works by the same author(s) and several different works lack dates? That seems like a rare edge case. Dragons flight (talk) 19:47, 22 May 2013 (UTC)
It is quite common for dates to be in the format that is not recognised by the date parameter. eg date=January-March 2013, the work around for that is to include a year=value as well as a date parameter. The date displays and the year does not, but the year parameter is used in the CITEREF generated by ref=harv. For author publications in the same year the usual solution is to use year=2013 in the first one year=2013b in the second etc. Another method (which works for no date) is to set ref=((SfnRef)) to whatever is wanted as a CITEREF. Also ff no year or date is set, but either author or last is set then an error should not be generated as harv will work on just author. -- PBS (talk) 10:29, 15 June 2013 (UTC)

Check for missing authors

I posted this on Help talk:Citation Style 1 in a section on a similar error (coauthors without authors). This is a separate error from "coauthors without authors". I believe that it should be detectable.

The general case is that in a list of authors or editors, if one is missing, the remaining authors are not displayed, even though the intent of the editor creating the citation is usually for all of the authors or editors to appear.

Example 1: If there are multiple authors but no "author1" or "last1", no authors are displayed. Like this:

Example 2: The same thing happens with editors as well:

Example 3: If you include author1 and leave out author2, the remaining authors are omitted:

I expect that there are additional variations on this parsing as well.

As for what to do about it, it is clear that the editor adding the citation intended for authors to appear, but they are not appearing. I suggest the following:

 Done. This request has been implemented in the CS1 module, as you can see from the error messages above. The documentation has been mostly updated. – Jonesey95 (talk) 06:34, 12 December 2014 (UTC)

Check for author first names without last names

When a "firstn" parameter is present in a citation without a corresponding "lastn" citation, no author/editor name is displayed for that author/editor. Either the citation's creator wants something to appear there, or there is some other kind of error that should be resolved. An error message will help editors resolve this problem.

Example 1: "first1" but no "last1". No author is displayed.

Example 2: "last1" but no "first1". Author's last name is displayed. This is fine; no error should be displayed.

Example 3: Same as Example 1, but with "editor-last1" missing while "editor1-first" exists.

Recommended behavior:

In my travels through Category:CS1 errors: coauthors without author, I am finding a number of articles with |first= followed by |coauthors=. It appears that editors thought that |first= was for the "first author", but it ends up displaying nothing. It would be useful to show an error message for these cases.
Examples from actual articles:
1. "| first = Christopher D. K. Herd | coauthors = Alexandra Blinova, Danielle N. Simkus, Yongsong Huang, et al. " from Tagish Lake (meteorite). – Jonesey95 (talk) 07:02, 23 January 2014 (UTC)
2. "|first=Thiruvalluvar |coauthors=Rev.G.U.Pope (translation) " from Tamil literature.
These are easy to fix, but without the "coauthors without author" error message, which appears only because |coauthors= is used, citations like these will languish invisibly. – Jonesey95 (talk) 07:02, 23 January 2014 (UTC)
I believe that there are hundreds, if not thousands, of these errors out in the wild. Here's one I just fixed. – Jonesey95 (talk) 14:59, 27 March 2014 (UTC)

Ontology and or schema for Citation template parameters to and from COinS, BIBO?

I'm interested in extending Diberri's citation filler to use Metadata or COinS information in it's cite web generation, to provide appropriate cite web, news, journal, book, ... output. Wikipedia:Scripts/Perl scripts/hdump-head.pl is a sketch that does ok on some journal articles. Is there an ontology or mapping of citation parameters to COinS data that generates the code or is it interleaved in the Lua? EutilsRDF Web Service: an RDF interface to NCBI Entrez Utilities could be interesting for NLM resources, building on the Bibliographic Ontology (BIBO). BIBO mentions projects using it at Worldcat and the UK National Library. Obvious mappings are 

bibo:asin	Citation/core:ASIN ?
bibo:doi == prism:doi	Citation/core:DOI
bibo:edition == prism:edition Citation/core:Edition
bibo:isbn == prism:isbn	Citation/core:ISBN
bibo:issn == prism:issn	Citation/core:ISSN
bibo:issue	Citation/core:Issue
bibo:lccn	Citation/core:LCCN
bibo:oclcnum	Citation/core:OCLC
bibo:pmid	Citation/core:PMID
bibo:uri	Citation/core:URL
bibo:volume == prism:volume	Citation/core:Volume

(not quite to leaf nodes since (bibo:eissn == prism:eissn) is a bibo:issn, similarly with isbn10, isbn13 (Note that my understanding of this is at the little knowledge is a dangerous thing stage.) More table driven citation generation tools could be useful. RDBrown (talk) 14:28, 11 November 2013 (UTC)

See Template:Cite_book#Syntax for a list of parameters that are included in the COinS metadata. Or, search Module:Citation/CS1 for the term "OCinSoutput" (without quotes).
—Trappist the monk (talk) 18:27, 12 November 2013 (UTC)
See Module talk:Citation/CS1/COinS. --  Gadget850 talk 21:45, 14 March 2014 (UTC)

Editions in cite map template

This entry is to track a request made on the Help Talk page: Help_talk:Citation_Style_1#Editions_in_cite_map. – Jonesey95 (talk) 14:16, 19 November 2013 (UTC)

Support for day and year ranges

Support for day and year ranges? Examples of valid day and year ranges were provided in this discussion and this discussion.

—Trappist the monk (talk) 17:30, 8 December 2013 (UTC)

 Done in the sandbox. – Jonesey95 (talk) 21:01, 14 March 2014 (UTC)

Add the new "Draft" and "Draft Talk" namespaces to the module

Minor feature request for the sandbox after this week's updates (I would put it in directly, but I don't want to interfere with this week's migration of code to the live module):

There will soon (December 17, 2013, at this writing) be a "Draft" namespace, with an accompanying "Draft Talk" namespace, in Wikipedia. We need to include it in the list of namespaces that are not included in the CS1 error categories. Errors should be shown in those namespaces but not categorized.

We'll need to update the module itself, along with the module's documentation. I think it makes sense to wait until after this week's module updates before making any changes. – Jonesey95 (talk) 00:28, 13 December 2013 (UTC)

Green tickY Done in the sandbox and in a comment on the Help:CS1 errors page. – Jonesey95 (talk) 00:45, 18 December 2013 (UTC)
Done through ((broken ref)). --  Gadget850 talk 21:43, 14 March 2014 (UTC)

Detect and report duplicated parameters

Duplicated parameters in CS1 citations could be reported so that editors could avoid creating them in error and gnomes could fix them where they occur. I'm not talking about the existing redundant parameters errors. This would simply be reporting where there are multiple instances of e.g. |first= or |url= in the same citation. Duplicated parameters exist for numerous reasons, but only the last instance of the parameter is displayed, which is usually not the intent of the editor who creates or edits the citation. – Jonesey95 (talk) 21:00, 14 March 2014 (UTC)

If I understand correctly, parameters are parsed by the MediaWiki software before they are passed to the CS1 module. By the time it gets to the module, previous parameters are already ignored. There is no way for a template or mudule to do this detection. --  Gadget850 talk 21:42, 14 March 2014 (UTC)
Oof, that's disappointing. I don't feel like starting a giant discussion about this, but would there be a valid reason for having two identical parameters in the same template? It seems to my naive self that this would always be an error that should be flagged by the MW software so that editors could fix this inadvertent mistake. Maybe there are some situations in which this arrangement is not an error, and I am just not clever enough to think of any. – Jonesey95 (talk) 23:11, 14 March 2014 (UTC)

Translator

Add 'translator' parameters. Should show preceded by "Translated by". --  Gadget850 (Ed) talk 19:51, 1 April 2013 (UTC)

I would appreciate this feature. Could you code this in the same way as authors and editors (eg: "last1=|first1=" etc.), as some works have multiple translators. Mindmatrix 03:20, 2 April 2013 (UTC)
Just like 'last' and 'first', there would be an unlimited number. --  Gadget850 (Ed) talk 09:38, 2 April 2013 (UTC)

 Done. This change was implemented in the live CS1 module in September 2015. – Jonesey95 (talk) 18:38, 21 October 2015 (UTC)

Add "authorn=et al." to a maintenance category

Resolved

An idea posted by Trappist the monk elsewhere, recorded here for posterity: "Add code to Module:Citation/CS1 that will put citations with |authorn=et al. into a separate specific category."

I imagine this as a maintenance category rather than a "CS1 error" / "incorrect syntax" category. Let's discuss at some point to see if there are editors interested in filling in the remaining authors, editors who object to people doing so, or other opinions/ideas. – Jonesey95 (talk) 15:53, 14 May 2014 (UTC)

Also when et al. is included in |firstn=.
I think its an error condition; et al. is not an authors name, it is simply an indicator that there are unlisted authors. I'm wondering if there shouldn't be some mechanism by which editors can inform the template that not all authors are listed so that Module:Citation/CS1 will add a properly formatted et al. to the citation but that same et al. won't be included in the COinS metadata (as happens now with |authorn=et al.). This is much like |display-authors= except that all of the authors included in the citation template are displayed followed by et al. – perhaps, |etal=yes or something similar.

arXiv class would be nice

Resolved

Currently, supplying a parameter of the form "|arxiv=1409.7951 [physics.atom-ph]" produces a messy error:

Botermann, Benjamin; Bing, Dennis; Geppert, Christopher; Gwinner, Gerald; Hänsch, Theodor W.; Huber, Gerhard; Karpuk, Sergei; Krieger, Andreas; Kühl, Thomas; Nörtershäuser, Wilfried; Novotny, Christian; Reinhardt, Sascha; Sánchez, Rodolfo; Schwalm, Dirk; Stöhlker, Thomas; Wolf, Andreas; Saathoff, Guido (September 2014). "Test of Time Dilation Using Stored Li+ Ions as Clocks at Relativistic Speed". Physical Review Letters. 113 (120405): 1–5. arXiv:[physics.atom-ph 1409.7951 [physics.atom-ph]]. doi:10.1103/PhysRevLett.113.120405. ((cite journal)): Check |arxiv= value (help)

And trying [ doesn't improve it all that much:

Botermann, Benjamin; Bing, Dennis; Geppert, Christopher; Gwinner, Gerald; Hänsch, Theodor W.; Huber, Gerhard; Karpuk, Sergei; Krieger, Andreas; Kühl, Thomas; Nörtershäuser, Wilfried; Novotny, Christian; Reinhardt, Sascha; Sánchez, Rodolfo; Schwalm, Dirk; Stöhlker, Thomas; Wolf, Andreas; Saathoff, Guido (September 2014). "Test of Time Dilation Using Stored Li+ Ions as Clocks at Relativistic Speed". Physical Review Letters. 113 (120405): 1–5. arXiv:[physics.atom-ph] 1409.7951 [physics.atom-ph]. doi:10.1103/PhysRevLett.113.120405. ((cite journal)): Check |arxiv= value (help)CS1 maint: extra punctuation (link)

It would be nice if there were some way to include the arXiv class in new-style identifiers. It's not critical, but it's normally included in arXiv citations in print, and if it's WP:MOS to omit it, a comment in the template docs would be nice.

In Lua, I imagine simply expanding the syntax of legal arxiv identifiers would be simplest, but an additional template parameter in the style of Template:Cite arxiv is also okay. The latter allows a class, but has the problem that it's not possible to list journal, volume, issue, page numbers, etc. for papers that are published peer-reviewed:

Botermann, Benjamin; Bing, Dennis; Geppert, Christopher; Gwinner, Gerald; Hänsch, Theodor W.; Huber, Gerhard; Karpuk, Sergei; Krieger, Andreas; Kühl, Thomas; Nörtershäuser, Wilfried; Novotny, Christian; Reinhardt, Sascha; Sánchez, Rodolfo; Schwalm, Dirk; Stöhlker, Thomas; Wolf, Andreas; Saathoff, Guido (September 2014). "Test of Time Dilation Using Stored Li+ Ions as Clocks at Relativistic Speed". pp. 1–5. arXiv:1409.7951 [physics.atom-ph]. ((cite arXiv)): Unknown parameter |doi= ignored (help); Unknown parameter |journal= ignored (help); Unknown parameter |number= ignored (help); Unknown parameter |volume= ignored (help)

Thank you! 71.41.210.146 (talk) 12:30, 1 October 2014 (UTC)

What is an arXiv class? --  Gadget850 talk 14:23, 1 October 2014 (UTC)
The arXiv classification is explained here and here. The folks at arXiv.org recommend including it in citations, but we do not render it well. It appears to me that it would best be included in CS1 templates as a separate parameter to make validation, presentation, and linking easier. – Jonesey95 (talk) 15:44, 1 October 2014 (UTC)
The arXiv used to assign submission identifiers that started with a subject classification. This was annoyig if a paper was originally misclassified; reclassifying it required assigning a new identifier. In 2007, they changed to a system where permanent identifiers were purely numerical, with the classification appended as extra information. It's not essential, but customarily appended as additional information about a paper. 71.41.210.146 (talk) 21:33, 1 October 2014 (UTC)
—Trappist the monk (talk) 18:07, 14 May 2014 (UTC)

Improve accessdate checking

Resolved

May be we could improve the accessdate checking to rule out improbable dates of access. Probably the date when wiki was started should be an earliest possible accessdate. Have just seen article List of department stores of the United Kingdom with some accessdates of 1914! Keith D (talk) 19:20, 7 December 2014 (UTC)

Since |access-date= requires URL (and maybe also requires that Wikipedia existed), we could set a minimum year. The World Wide Web article says that the oldest known web page dates from 1991.
We could also test the access-date for future dates. I believe that we already test the year for dates greater than next year (since publications like magazines can have next year in their date). – Jonesey95 (talk) 19:26, 7 December 2014 (UTC)
What should be the limit for future access dates? It would seem that the minimum must be current server date+1 because 12:00:00 UTC today is 00:00:00 tomorrow in New Zealand.
  • 2024-03-14T10:12:28 – UTC time when this page last refreshed (Purge)
  • 2024-03-14T22:12:28 – NZST (UTC+12)
  • 2024-03-14T23:12:28 – NZDT (UTC+13)
Accepting current UTC date+1 allows editors to enter their local 'today's' date without error. If they use their local 'tomorrow's' date, there will be an error message for at most 24 hours after which there is no error.
—Trappist the monk (talk) 20:38, 7 December 2014 (UTC)
Looks good to me. --  Gadget850 talk 22:16, 7 December 2014 (UTC)
Same here. Keith D (talk) 22:36, 7 December 2014 (UTC)
Today+1 is fine with me. I sometimes enter a date that is tomorrow (to me) when it is already tomorrow in the UTC time zone (I know, UTC isn't really a time zone, but you get the idea). – Jonesey95 (talk) 00:04, 8 December 2014 (UTC)
The current method of checking errors only considers information stored in the template (and could presumably consider the system time). If someone enters a date 2 days in the future, and no one does anything about it, after a day passes the error message will go away. I think that really diminishes the usefulness of the message. Jc3s5h (talk) 18:24, 14 January 2015 (UTC)
Do you have a proposal that is better than the one above? – Jonesey95 (talk) 23:28, 14 January 2015 (UTC)
I don't have a better proposal. In a much more restricted editing environment, where the editing software knew that a citation was being entered (as in Microsoft Word) I would block the edit from happening at all, but the Wikipedia editing environment is too flexible for that. Jc3s5h (talk) 21:32, 15 January 2015 (UTC)
So what's the current status of this request? If it makes a difference, I'm fine leaving future alone if it will get it moved in faster...Naraht (talk) 22:44, 15 January 2015 (UTC)

Making notes to myself for this topic. mw:Extension:Scribunto/Lua_reference_manual#mw.language:formatDate refers to mw:Help:Extension:ParserFunctions#.23time.

Wikipedia started 15 January 2001. Using ((#time:)) we can get the number of seconds since 1970-01-01T00:00:00 UTC for the dates 2001-01-15, keyword 'today', and keyword 'tomorrow':

So then the value in |accessdate= must not convert to less than 979516800 seconds nor more than 1710460800 seconds (tomorrow).

—Trappist the monk (talk) 17:56, 2 April 2015 (UTC)

Sounds good to me!Naraht (talk) 17:34, 8 May 2015 (UTC)

et al.

Resolved
see also: Module talk:Citation/CS1/Feature requests#Add "authorn=et al." to a maintenance category

Detect "et al." in an author field. --  Gadget850 talk 02:03, 14 December 2014 (UTC)

And having detected it, what then?
—Trappist the monk (talk) 03:13, 14 December 2014 (UTC)
Trigger an error. I have recently found it in author fields for some reason. --  Gadget850 talk 23:31, 14 December 2014 (UTC)
Yeah, so have I. In fact, Monkbot, when it finds et al. in |coauthors=, is guilty of converting it to |authorn=et al. But I guess I was looking for some sort of idea about what should be done instead of simply adding an error message. What do we suggest that editors should do to get CS1/2 to display et al. outside of misusing |author=, |last=, and |first=? Do we tweak the definition of |display-authors= so that when |display-authors=et al the rendered author list has et al. but the COinS data does not. This might also be applied to |display-editors=. We could invent |et-al-author= and |et-al-editor= or some such similar to do the same sort of thing. We'd need to worry about interaction with |display-authors=n when the author list has more than n authors so that |display-authors= and |et-al-author= don't both add et al. to a rendered citation.
—Trappist the monk (talk) 11:36, 15 December 2014 (UTC)
I usually see |author= or |authors= with multiple authors and et al. --  Gadget850 talk 14:36, 15 December 2014 (UTC)

Extra checking of URL & title

Hi, I think that there should be some additional checking on the |url= and |title= fields and setting up tracking categories so that they can be looked at and fixed appropriately.

For the |url= check if there is text in the field other than the URL, easiest way to do this would be to check for mid-string white space. This would pick up things such as this.

A converse would be to check for a URL in the |title= field as you should not get URLs in the title. This would pick up the inclusion of unnecessary details such as this.

Keith D (talk) 22:44, 17 June 2015 (UTC)

Part of this request was discussed in a previous conversation, although it does not appear that any action was taken based on that discussion. Help talk:Citation Style 1 is usually a better forum for these conversations, in any event; it has more watchers. I recommend that you start two separate threads on that page, since this is really two different feature requests. – Jonesey95 (talk) 04:34, 18 June 2015 (UTC)
 Done. This change was implemented in the live CS1 module in September 2015. – Jonesey95 (talk) 18:38, 21 October 2015 (UTC)

Suppress original URL

Based on this IP edit, I made this change to Da Hip Hop Witch. The cite templates force the display of the original URL once an archived link is provided, and unfortunately in this case, provides a link to a Paypal request. I've seen other original links leading to linkfarms or other undesirable links.

Can we add a feature to suppress the display of the original URL in such cases? Mindmatrix 15:23, 10 August 2015 (UTC)

I presume that we could extend the functionality of |dead-url= to include some sort of meaningful code word that would serve the purpose of rendering the citation with just the archived message without the link. What should that code word be?
—Trappist the monk (talk) 15:51, 10 August 2015 (UTC)
Maybe |dead-url=hide? – Jonesey95 (talk) 18:36, 10 August 2015 (UTC)
Perhaps |dead-url=nolink?
—Trappist the monk (talk) 12:55, 11 August 2015 (UTC)
I think I'd prefer something which is better represents the semantics of the problem, in this case that the original link is spam or at the least "undesirable". Not quite sure of a good keyword for that. "origspam" or "originalspam". --Izno (talk) 13:26, 11 August 2015 (UTC)
I had exactly that thought but my poor little brain isn't finding the right term. I did think of spam but that to me connotes spamming or spam (email), not a usurped or expired domain name. I also thought about advert, but this case isn't that; phishing, fraud also came to mind.
—Trappist the monk (talk) 13:42, 11 August 2015 (UTC)
I was thinking that the parameter value should describe the function, not the purpose. There could be multiple reasons for hiding the original URL. Hence "hide" or "nolink". – Jonesey95 (talk) 14:02, 11 August 2015 (UTC)
The problem I have with function over purpose is that function enables behavior that may not be desirable. For example, I can't think of any reason other than a link being a "bad" link to be correct to hide. --Izno (talk) 14:05, 11 August 2015 (UTC)
unfit meaning not approriate?
—Trappist the monk (talk) 14:12, 11 August 2015 (UTC)
What about usurped? Mindmatrix 15:00, 11 August 2015 (UTC)
Moved to Help talk:Citation Style 1/Archive 9 § Suppress original URL

 Done. This change was implemented in the live CS1 module in September 2015. – Jonesey95 (talk) 18:38, 21 October 2015 (UTC)

Presentation and content

 Not done per MediaWiki_talk:Common.css/Archive_17#The_cite_element_needs_to_not_auto-italicize_any_longer --Izno (talk) 16:38, 16 March 2016 (UTC)

The CSS styling for <cite> has been defaulted, so it now formats the contents as italics, while adding the semantic meaning of a title. Thus the current use of italics to format the main work title can be replaced by <cite>...</cite>.

Markup Renders as 
<cite>Title</cite>

Title

--— Gadget850 (Ed) talk 01:04, 28 March 2013 (UTC)

<cite> would be used internally in the template, so I am confused as to how anyone would be confused. It has an HTML semantic value indicating the title of a work.[1] If we want to add the semantics for an included work title which is marked in quotes, then we can easily style the tag. And with almost half a million uses of cite book alone, I would not call the italic title rare. --— Gadget850 (Ed) talk 09:30, 28 March 2013 (UTC)
Isn't this the problem the IP was discussing earlier, where we're really using the "title" paremeter for multiple semantic functions? Choess (talk) 18:11, 29 March 2013 (UTC)
I refactored the original proposal to indicate that by title I meant the main work title. Currently, we have no separation of presentation and content. That is, the format of the main work title is always italics, and the included work is always in quotes. This presentation should be moved to CSS. Wrapping the main work title in <cite>...</cite> will present the title in italics by default. We can add a class to present the included title in quotes.
Since this would be a new feature, I will be moving this to Module talk:Citation/CS1/Feature requests. --— Gadget850 (Ed) talk 18:53, 29 March 2013 (UTC)
To expand on separation of presentation and content: currently the templates include both content and presentation, that is the markup used to style the content as italics, in quotes or bold. Hard coding the presentation means that readers cannot style citations as they desire and it limits template portability. Presentation should be done in CSS. Currently the <cite> tag has a default style of italics and has the semantic meaning of a title.
For an included work title that is presented in quotes, we can create a class with CSS styling. For example the class includedtitle:
.includedtitle:before {font-style: normal; content: '\22';}
.includedtitle {font-style: normal;}
.includedtitle:after {font-style: normal; content: '\22';}
Then you simply wrap the content in <cite class="includedtitle">...</cite> causing the font to show as normal and the content wrapped in quotes.
The HTML classes discussion does not include a class for the included work title, so I made up an illustrative class. --— Gadget850 (Ed) talk 14:58, 30 March 2013 (UTC)
This allows the style to be customized per Wikipedia language version. In the CSS above, \22 is the hex code for the standard quote mark. This can be replaced with other marks: see Non-English usage of quotation marks. --  Gadget850 (Ed) talk 00:17, 12 April 2013 (UTC)

Fix bug causing spurious "Check |url= value" warnings on single-letter subdomains of 4-or-more-letter gTLDs (e.g. "z.cash")

The is_domain_name function incorrectly considers <single letter>.<4-or-more-letter gTLD> not to be a valid domain.

See https://en.wikipedia.org/wiki/Zcash#References for an example; "z.cash" is incorrectly considered not to be a valid domain. (See https://www.iana.org/domains/root/db/cash.html for the delegation record of ".cash".) The fix is straightforward.

Daira Hopwood ⚥ (talk) 02:36, 8 August 2017 (UTC)

Fixed in the sandbox.
((Cite web/new|url=https://z.cash/about.html|title=Zcash - About))
"Zcash - About".
—Trappist the monk (talk) 11:00, 8 August 2017 (UTC)
Thanks :-) Daira Hopwood ⚥ (talk) 12:44, 8 August 2017 (UTC)

Suppress spurious warning about zero-width joiner

((cite web|title=തിരുവിതാംകൂര്‍|language=Malayalam)) on the Travancore article displays with a warning:

The zero-width joiner is at the very end and is needed for the title to render correctly.

I see that the citation templates support |ignore-isbn-error=true to suppress warnings about the ISBN. Can support for |ignore-title-error=true be added to suppress warnings about the title field's content?

See Zero-width joiner#Examples for how the zero-width joiner and also Zero-width non-joiner#Use of ZWNJ to display alternate forms are used to control the rendering of Indic scripts. http://unicode.org/review/pr-37.pdf goes into technical details.

Another option, would be for you to check the language field and if it's set to one of the Brahmic scripts, Indo-Aryan languages, or Arabic scripts to suppress the warnings about ZWJ and ZWNJ. --Marc Kupper|talk 00:47, 6 June 2016 (UTC)

I thought of a work-around that does not involve changing the template. Instead, the title can be entered using HTML entities. Here's the ((cite web)) used in the previous example but using the entities &#x0D24;...&#x200D; rather than UTF-8 as before.
  • "തിരുവിതാംകൂര്‍" (in Malayalam). ((cite web)): Missing or empty |url= (help)
Or, I can enter the title in UTF-8 except for the zero width joiner which I entered as an HTML entity.
  • "തിരുവിതാംകൂര്‍" (in Malayalam). ((cite web)): Missing or empty |url= (help)
There's no warning about the zero width joiner character though I did use one. The downsides are 1) people are used to copy/pasting UTF-8 strings and would need to convert them to HTML entities. 2) If the entire title is converted then it's no longer human readable when in edit-mode and is confusing to humans should just the zero-width joiners be entered as entities.
Another way to code the template check which is to see if any characters in the string being examined are part language blocks that uses characters such as the zero-width joiner and to then suppress warnings about the zero-width joiners. All 13 characters in the example title prior to the zero-width joiner were in the Malayalam (Unicode block) which runs from U+0D00 to U+0D7F. --Marc Kupper|talk 01:22, 8 June 2016 (UTC)
Please let us not go back to 2004 with HTML entities.
These are valid Unicode characters, which are required in these languages, and which can be easily typed using keyboards for these languages.
The right thing would be to check the language automatically and not show this error if it is one of the languages in which this character is valid, as you suggest initially. --Amir E. Aharoni (talk) 15:28, 20 July 2016 (UTC)

Reporting apparent error in error indication

I was looking at Help:CS1#How the templates work, and saw what looked like unintended error indications there as follows:

CS1 templates present a citation generally as:

Author (Date). "Title". Work. Publisher. Identifiers. ((cite journal)): |last= has generic name (help); Check date values in: |date= (help); Unknown parameter |template doc demo= ignored (|no-tracking= suggested) (help)
"Title". Work. Publisher. Date. Identifiers. ((cite journal)): Check date values in: |date= (help); Unknown parameter |template doc demo= ignored (|no-tracking= suggested) (help)
(note: wikitext for above copied from Help:CS1#How the templates work)

I'm guessing that this comes from the handling of the |template doc demo=true parameter which is passed there to ((Cite journal)). I would be outside my comfort zone to pursue this further myself, so I am reporting it here. Wtmitchell (talk) (earlier Boracay Bill) 22:24, 25 April 2017 (UTC)

Not a feature request, how did you end up here?
|template doc demo=true does not prevent error messages; it only prevents this page from being included in Category:CS1 errors: dates‎. To get rid of the error message, those citations will require a valid date: |date=Date is not a valid date; or, the citations should be handwritten without using the templates.
—Trappist the monk (talk) 22:37, 25 April 2017 (UTC)

prearchive for deadurl

Add |prearchive=yes as an alias to |deadurl=no. Negative logic like setting a parameter to false to enable a feature is foreign to the average editor. 'prearchive' better describes the feature for preemptively archiving the link. --  Gadget850 (Ed) talk 16:30, 6 April 2013 (UTC)

I am seeing a number of uses of |deadurl=yes which is meaningless. --  Gadget850 talk 12:37, 31 October 2014 (UTC)

Check for wikilink to current page

If a citation includes a wikilink to the current page, it becomes bolded:

Markup Renders as 
((cite book |title=[[Module talk:Citation/CS1/Feature requests]]))

Propose: If the wikilink is to the current page, then remove the linking. --  Gadget850 (Ed) talk 16:18, 12 April 2013 (UTC)

If this is done, it should be done via CSS. Removing the irrelevant italics, the output is styled
<span class="citation book"><strong class="selflink">Module talk:Citation/CS1/Feature requests</strong></span>
Setting ".citation .selflink { font-weight:normal; }" will remove the bold. Dragons flight (talk) 18:05, 12 April 2013 (UTC)
If this is done, then it should apply to all links. I've seen lists of books in author pages where the list is constructed with CS1 templates that include |authorlink=. See Andrew Hunt §Bibliography.
—Trappist the monk (talk) 14:43, 27 April 2013 (UTC)
Yes, the suggested CSS would eliminate the bolding associated with such links regardless of where they occur in the citation. Adding that (or not) doesn't actually affect the Module, it would need to go in Mediawiki:Common.css (or similar). Dragons flight (talk) 16:36, 27 April 2013 (UTC)
Yep, it does. Tested in my common.css and citation wikilinks to the current page aren't <strong /> anymore. I gather that this change needs to be made to MediaWiki:Common.css. Shall I make an edit request there?
—Trappist the monk (talk) 11:43, 28 April 2013 (UTC)
It would make more sense to me to throw an error in these cases for correction, as these items of various types (author/title) probably just shouldn't have the links rather than being removed by the template. Where an error might get messy would be in things like cite doi, which are by their nature not necessarily on the page which is linked. Hmm. --Izno (talk) 03:44, 28 April 2013 (UTC)
I'm inclined to disagree. There is nothing wrong with wikilinks that point to the current page as long as the link is disabled in the citation. This allows editors to reuse whole citations; it allows <section> to transclude text from one article to another with self-linking citations intact. I don't see a need for error messages here.
—Trappist the monk (talk) 11:43, 28 April 2013 (UTC)
And citation templates. I agree with Trappist. --  Gadget850 talk 13:06, 28 April 2013 (UTC)
Just exploring another option. I'm not particularly attached to it. --Izno (talk) 13:20, 28 April 2013 (UTC)
The CSS can deal easily with redirects and wikilinks that occur in places other than link parameters. Trying to accomplish the same breadth of error checking in Module logic would be performance prohibitive. One could check a few of the obvious parameters, e.g. authorlink, but I don't really see how that improves over the CSS suggestion. Dragons flight (talk) 22:13, 20 May 2013 (UTC)
I'd suggest that people make a request for this at MediaWiki talk:Common.css. Dragons flight (talk) 22:13, 20 May 2013 (UTC)

Template:Cite wikisource

Could anybody move it to Lua? --DixonD (talk) 09:26, 10 May 2013 (UTC)

it has some novel parameters, including the unique facility of editor-specified page section links (|anchor=). in the doc, this parameter perhaps belongs under "In-source locations" rather than under "Title"? while useful, the parameter name anchor is confusing and should be more user-friendly.
i also like that any or all wikisource-related icons can be hidden.
the doc also misidentifies type: this parameter does not provide "additional information about the media type of the source". the medium is fixed for all wikisource sources: it is a digital, online host, formatted as a wiki. as documented at the template page, type provides information about the work classification or work type of the source.
70.19.122.39 (talk) 17:45, 11 May 2013 (UTC)
I've done a few more conversions this last weekend. I've been handling these roughly in order by number of transclusions. The 1100 uses of cite wikisource is roughly in the middle of the list of what is left. We'll probably get to it eventually, but it will likely be months before all of the 20+ minor templates are converted. Dragons flight (talk) 22:38, 20 May 2013 (UTC)

Chapter and section

Why is there a restriction on using "chapter" and "section" at the same time? It seems to me there are legitimate need for the two to be used at the same time.Here is a book which has volume, sections and chapters:

-- PBS (talk) 13:33, 14 June 2013 (UTC)

'section' originates from ((cite manual)). When I merged it into ((cite book)), I made 'section' and 'chapter' aliases. See below for more. --  Gadget850 talk 17:42, 14 June 2013 (UTC)
I think that is not the best way to go. I have shown you above an example of where a book uses sections and chapters. -- PBS (talk) 09:27, 15 June 2013 (UTC)

strip wikilinks from page, pages, at

This citation:

Cornelius Tacitus, Publius (98), De origine et situ Germanorum (On the Origin & Situation of the Germans), Cap. XL  ((citation)): Check date values in: |year= (help)

uses

|at=[[s:la:De origine et situ Germanorum (Germania)#XL|Cap. XL]]

the module does not know how to strip the markup before adding it to the COinS.

—Trappist the monk (talk) 13:19, 4 June 2015 (UTC)

Sfnref

Integrate ((sfnref)) into the CS1 templates. This would take two parameters: |sfnref= and |sfnyear=. --  Gadget850 talk 12:45, 31 March 2014 (UTC)

postscript check

Check for 'postscript' with more than one character. I'm seeing some odd stuff inserted. --  Gadget850 (Ed) talk 19:30, 6 April 2013 (UTC)

My present reaction is that it may be too late for this. There appear to be too many people doing things like |postcript=<!-- None -->, plus User:Citation bot has been putting a message about ((inconsistent citations)) in the postscript. Insisting that it ought to be only one character seems a bit futile at this point. Dragons flight (talk) 22:25, 20 May 2013 (UTC)
Perhaps just a hidden tracking category to see what is out there. --  Gadget850 talk 02:04, 21 May 2013 (UTC)

Part

A part= parameter would be useful, see for example:

-- PBS (talk) 13:41, 14 June 2013 (UTC)

Also useful would be column (col.), paragraph (para.); track; hours, minutes and seconds; act, scene, canto, book, part, folio, stanza, back cover, liner notes, indicia, colophon, dust jacket, verse and probably a lot more. But this is why we have the catchall 'at' field where you can add free text where 'chapter' and 'page' don't exactly fit. --  Gadget850 talk 17:45, 14 June 2013 (UTC)
By that argument volume may as well be removed! A difference is made between chapter and page. A similar difference needs to be made between volume and part. In parts page numbering often restarts. Most of the items you have listed in your first sentence are attributes of the same physical book (and are other ways of pinpointing information within a physical book). Part is different, it frequently describes a physical book with its own page numbering. -- PBS (talk) 09:26, 15 June 2013 (UTC)

Partial links in title

Moved to Help talk:Citation Style 1 § Partial links in title

—Trappist the monk (talk) 18:57, 30 July 2019 (UTC)

More