Help talk:Citation Style 1/Archive 12

Archive 5

←

Archive 10

Archive 11

Archive 12

Archive 13

Archive 14

Archive 15

Cite web needs a section parameter

We need a |section= parameter in {{cite web}}. We used to have better support for this, it seems to have been "rationalised".

I'm just trying to reference a PDF hundreds of pages long. It has chapters and legalistic paragraph or section numbering, yet it's not published except on the web, it has no ISBN, it's not a "book". |section= or |chapter= annotation would be useful and appropriate here. Merging the section numbers into the title (as suggested) breaks downstream metadata handling. Andy Dingley (talk) 13:38, 21 February 2016 (UTC)

This cite? So what if it doesn't have an ISBN? That source is book length, has chapters, and pages, is in electronic form which would seem to pass the duck test for an e-book. Therefore, use {{cite book}}.

—Trappist the monk (talk) 14:03, 21 February 2016 (UTC)

Similar issues arise with books and other documents. Rather than sections or chapters, a more general solution might be to allow |at= to co-exist with |page=/|pages=. In this case, one would write |at=Rule 184.3|page=208. In others, it would be |at=Table 3.1 or |at=Map 9 in conjunction with a page number. Kanguole 14:05, 21 February 2016 (UTC)

|at= would work for me (better than calling something that isn't a book a book). I tried that, but it's either/or with |pages=.

Incidentally (AIUI) this isn't a book and can't be readily bought as a book. That itself is a source of some annoyance to those who'd like to be able to buy it as a bound paper book, but instead keep ending up with ragged printouts. Andy Dingley (talk) 14:15, 21 February 2016 (UTC)

If you just dislike calling it a book, use {{Citation}}, with |mode=CS1 if you must scatter stops everywhere. :-) Peter coxhead (talk) 22:23, 21 February 2016 (UTC)

A |section= parameter would be useful only in those cases (analoguous with the use of |pages=) where the source as a whole is a sub-unit of a larger work (such as an article in a journal, or contribution that is published separately). Where one is citing a specific section (or page) then it is best not to merge a specific detail into the full citation. In the typical case where a source is cited only once it is quite acceptable to append the specific location after the template. Which is a tad cumbersome with {cite} as you have defeat the automatic terminal punctuation (with |postscript=none). And another reason for using {citation}. Or (as Peter suggests) |mode=CS1. ~ J. Johnson (JJ) (talk) 21:05, 22 February 2016 (UTC)

Proper use of Website not explained.

From the CS1 messages it appears that it is not proper to use a URL in the website= entry for cite web. I think that needs to be *much* better explained if that is a problem. A would expect a new user to be much more likely to do {{cite web|url=http://www.example.com/widgets|title=Widgets through the ages|website=http://www.example.com}} than {{cite web|url=http://www.example.com/widgets|title=Widgets through the ages|website=Example Company}}. IMO, *either* website= needs to be specifically added to both Help:CS1 errors#External link in |<param>= (which doesn't show website as one that would be a problem and the docs on Cite web, or this CS1 error needs to be disabled.Naraht (talk) 21:13, 22 February 2016 (UTC)</nowiki>— Preceding unsigned comment added by ‎Naraht (talk • contribs) 19:45, 22 February 2016

• In the doc for the templates, the description of |website= says "title of the website". I don't see how that could be interpreted as URL.
• Of the multiple occurrences of |website= in the Examples section, none show a URL.
• The error shows |website= as the problem parameter.
• The error description lists several problem parameters and says "or any of their aliases". The doc shows that |website= is an alias of |work=, which is one of the parameters listed in the error description.

That's adequate in my opinion. Any more would be unwarranted instruction creep, and people do need to learn to use the doc that's available to them. ―Mandruss ☎ 21:50, 22 February 2016 (UTC)

I've calmed down now. I agree that the docs are as good as they could be, but it is still a fairly likely mistake. Any ideas that may help?Naraht (talk) 15:25, 23 February 2016 (UTC)

I've added the aliases to Help:CS1 errors#External link in |<param>=. This is about users who don't read the template documentation, but adapt existing examples and are then confused by error messages. The first documentation they encounter is a subsection of Help:CS1 errors that is linked from the error. Kanguole 15:54, 23 February 2016 (UTC)

Status of filtering square brackets in URLs

Are we planning to handle brackets in URLs? I found an article (at redirect "EAPPI") where the "url=" contains sets of single brackets ("[...]") and should be encoded by a Lua filter (as '%7b' and '%7d' values?). Apparently those are very rare inside a URL, but they should be filtered by the Lua module, some day. This problem goes back 15 years due to the poor design of the MediaWiki markup language which should have used 2-character tokens to denote an external URL link, such as with both angle+square brackets, "<[http:...]>" rather than just single brackets "[http...]" as now unable to include each ']' inside a URL address. Anyway, 15 years later, now the cite templates should handle "[...]" inside each URL parameter. -Wikid77 (talk) 11:37, 27 January 2016 (UTC)

Technically, unencoded single brackets are actually forbidden in the path part of a URL according to the URL specification. Of course, in practice anything is allowed provided that browsers and servers support it. Wikipedia also have issues with angle brackets ("<", ">") breaking urls. There might be other examples too. Dragons flight (talk) 12:39, 27 January 2016 (UTC)

The embedded brackets "[...]" are so rare, it can wait to be handled, along with other URL characters. -Wikid77 (talk) 14:00, 27 January 2016 (UTC)

Forbidden in the path part yes, but I think that they are legal in the query string. When used there they will still break the MediaWiki parsing: I have come across examples of such use in the past, occasionally when somebody posts to VPT with "why doesn't my URL work?". --Redrose64 (talk) 21:53, 27 January 2016 (UTC)

Here's an example of a URL containing square brackets (from Pass Me By (R5 song)) that does not render correctly:

Cite web comparison

Wikitext	{{cite web|accessdate=2012-09-13|date=2012-09-08|publisher=Muzikkitabi|title=R5 Performs Pass Me By, Loud [Morning Show Toronto] 8/26/13|url=http://muzikkitabi.com/Video/VIDEOIDEDGno_UD0U4/R5-performs-Pass-Me-By-Loud-[Morning-Show-Toronto]-8-26-13}}
Live	"R5 Performs Pass Me By, Loud [Morning Show Toronto] 8/26/13". Muzikkitabi. 2012-09-08. Retrieved 2012-09-13.
Sandbox	"R5 Performs Pass Me By, Loud [Morning Show Toronto] 8/26/13". Muzikkitabi. 2012-09-08. Retrieved 2012-09-13.

We should probably do something to either render this reasonably or flag it as an error. – Jonesey95 (talk) 03:41, 10 February 2016 (UTC)

Following up on a comment at Help_talk:Citation_Style_1#Update to the live cs1|2 modules weekend of 20–21 February 2016:

For a very long time, the template docs have advised editors to percent-encode certain characters when they occur in a url; see Template:Cite_book#URL for the list. :Characters allowed in the various parts of a url are defined in Uniform Resource Identifier (URI): Generic Syntax. Essentially, the characters defined in the table at Template:Cite_book#URL must be percent-encoded when used in the path portion of a url (not allowed in the scheme and domain parts).

That leaves us three options:

1. ignore the content and make up of the url path – this is what we do now
2. detect the presence of these characters in the url path and then emit an error message
3. detect the presence of these characters in the url path and then percent encode them

What do?

—Trappist the monk (talk) 18:31, 14 February 2016 (UTC)

I think option 1 is not good, since the citation renders quite poorly.

We have been encouraged many times on this page to pursue option 3, i.e. be permissive in what we accept, even if it is technically invalid. Our historical practice in most cases has been option 2, detect input that does not comply with guidelines and emit an error messages, letting human editors fix the errors.

If option 3 (reliably detect and replace) is technically feasible, I recommend it. If not, option 2 (emit an error message). – Jonesey95 (talk) 23:54, 14 February 2016 (UTC)

I'll explore the latter two options after the next update.

—Trappist the monk (talk) 16:29, 15 February 2016 (UTC)

I too would prefer option 3. The FAQ database on various Konica Minolta web sites is another example of urls containing square brackets (see f.e. http://www.konicaminoltasupport.com/index.php?id=4569&L=2 ).

--Matthiaspaul (talk) 21:03, 15 February 2016 (UTC)

I've added a bit of code to external_link() so that square brackets occurring in the path portion of a url are percent encoded.

Cite web comparison

Wikitext	{{cite web|language=de|title=Was ist die Zonenwahl-Funktion?|trans-title=What is the Zone Matching function?|url=http://www.konicaminoltasupport.com/index.php?id=4569&tx_faqmanager_pi1[question]=3640&tx_faqmanager_pi1[product]=136&tx_faqmanager_pi1[producttype]=10&tx_faqmanager_pi1[category]=&L=2}}
Live	"Was ist die Zonenwahl-Funktion?" [What is the Zone Matching function?] (in German).
Sandbox	"Was ist die Zonenwahl-Funktion?" [What is the Zone Matching function?] (in German).

Because the module creates a label from |url= when |title= is missing or empty, the label it creates is not encoded:

Cite web comparison

Wikitext	{{cite web|language=de|url=http://www.konicaminoltasupport.com/index.php?id=4569&tx_faqmanager_pi1[question]=3640&tx_faqmanager_pi1[product]=136&tx_faqmanager_pi1[producttype]=10&tx_faqmanager_pi1[category]=&L=2}}
Live	(in German) http://www.konicaminoltasupport.com/index.php?id=4569&tx_faqmanager_pi1[question]=3640&tx_faqmanager_pi1[product]=136&tx_faqmanager_pi1[producttype]=10&tx_faqmanager_pi1[category]=&L=2. {{cite web}}: Missing or empty |title= (help)
Sandbox	(in German) http://www.konicaminoltasupport.com/index.php?id=4569&tx_faqmanager_pi1[question]=3640&tx_faqmanager_pi1[product]=136&tx_faqmanager_pi1[producttype]=10&tx_faqmanager_pi1[category]=&L=2. {{cite web}}: Missing or empty |title= (help)

—Trappist the monk (talk) 13:13, 22 February 2016 (UTC)

Looks good to me. – Jonesey95 (talk) 14:56, 22 February 2016 (UTC)

Thanks. --Matthiaspaul (talk) 00:51, 24 February 2016 (UTC)

Template {{Cite episode}} undocumented parameter / error

In researching the answer to the query above, I discovered that the |city= parameter in this template doesn't seem to function, nor is it documented. Would someone investigate, please?
—D'Ranged 1 VT^alk 23:56, 23 February 2016 (UTC)

This comparison is the current Lua version against the last {{citation/core}} version:

Cite episode comparison

Wikitext	{{cite episode|city=Chicago|date=April 14, 1958|first=Alfred|last=Hitchcock|network=NBC|number=311|quote=Mary had a secret.|series-link=Alfred Hitchcock Presents|series=Alfred Hitchcock Presents|station=WMAQ|title=What Mary Knew}}
Live	Hitchcock, Alfred (April 14, 1958). "What Mary Knew". Alfred Hitchcock Presents. Episode 311. NBC. WMAQ. Mary had a secret. {{cite episode}}: Unknown parameter |city= ignored (|location= suggested) (help)
Sandbox	Hitchcock, Alfred (April 14, 1958). "What Mary Knew". Alfred Hitchcock Presents. Episode 311. NBC. WMAQ. Mary had a secret. {{cite episode}}: Unknown parameter |city= ignored (|location= suggested) (help)

|city= was removed from {{cite episode}} with this edit; three years before the Lua conversion.

—Trappist the monk (talk) 01:20, 24 February 2016 (UTC)

I've edited the documentation to reflect the deletion. The parameter no longer appears in the documentation. Is this template part of any tools that you know of that need to be checked?

—D'Ranged 1 VT^alk 01:34, 24 February 2016 (UTC)

multiple author names in |author= and aliases

One of the things that I noticed recently while cleaning up Category:CS1 maint: display-editors is that there are a lot of cs1|2 templates that use |author= to hold multiple author names. This is semantically incorrect and I can't see in our documentation that we explicitly permit multiple author names in a |last= alias – yeah, that practice isn't specifically proscribed either but should be. The simple fix is to change |author= to |authors= but that's not really a fix. It's not a fix because |authors= is not made part of the metadata:

{{cite book |title=Title |authors=one, two, three}}

'"`UNIQ--templatestyles-00000028-QINU`"'<cite class="citation book cs1">''Title''.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+12" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

I've added code to extract_names() that calls a new function name_has_mult_names() which counts the number of separator characters (comma and semicolon). If there is more than one of these characters in a |last= alias then the page is added to a new (not yet existing) maintenance category:

{{cite book/new |title=Title |author=one, two, three}}

one, two, three. Title.{{cite book}}: CS1 maint: multiple names: authors list (link)

the same applies to |editor-last= aliases:

{{cite book/new |title=Title |editor=one, two, three}}

one, two, three (ed.). Title.{{cite book}}: CS1 maint: multiple names: editors list (link)

I don't know how many of these kinds of improper uses there are but I'm guessing that there are a lot of them. Fixing these will be a task similar to the (still incomplete) |coauthor(s)= task. I suspect that a bot can cleanup the low-hanging fruit by converting these kinds of parameters to |vauthors= or to individual |authorn= parameters. When that is done, we should then convert this code so that it emits an error message. —Trappist the monk (talk) 16:32, 22 February 2016 (UTC)

What is your plan for corporate authors that have two or more commas as part of their name? Jc3s5h (talk) 17:27, 22 February 2016 (UTC)

Examples?

—Trappist the monk (talk) 17:48, 22 February 2016 (UTC)

It's obvious that corporations and partnerships can choose names that include several commas. You're the one that wants to make the templates harder to use; you should prove there are no legitimate institutional names that contain more than one comma. Jc3s5h (talk) 20:09, 22 February 2016 (UTC)

It might be something like Abel, Baker, Charley and partners --Redrose64 (talk) 21:36, 22 February 2016 (UTC)

So no examples then? [Proving] there are no legitimate institutional names that contain more than one comma is a bit like proving that Russell's teapot doesn't exist, isn't it? You have claimed that these names exist so the burden to prove that is yours.

But, to forestall a pointless discussion about orbiting crockery and the like, I will stipulate that such corporate names exist. I can think of no better way to know if |author= has more than one author name assigned to it. Can you? The method won't find the cases where two names are separated a single separator character (|author=A. Smith, B. Jones) so it is just as likely that the method will find names that legitimately have multiple separators.

—Trappist the monk (talk) 21:42, 22 February 2016 (UTC)

My suggestion is don't implement unreliable checks and make editors do unnecessary troubleshooting when there is really nothing wrong with an author's name. Jc3s5h (talk) 14:39, 23 February 2016 (UTC)

So no idea for a better method then?

—Trappist the monk (talk) 11:24, 24 February 2016 (UTC)

Trappist, are you seriously trying to claim that there are no corporate or partnership names with more than one comma in them? We have articles here on Adelson, Testan, Brundo & Jimenez – Baker, Donelson, Bearman, Caldwell & Berkowitz – Donovan, Leisure, Newton & Irvine – Finnegan, Henderson, Farabow, Garrett & Dunner – Robinson, Silverman, Pearce, Aronsohn, and Berman – Sullivan, Papain, Block, McGrath, & Cannavo – to pick a few examples found from briefly poking around some likely categories. —David Eppstein (talk) 01:02, 24 February 2016 (UTC)

No.

—Trappist the monk (talk) 11:24, 24 February 2016 (UTC)

I agree that no parameter (except in vcite) should have multiple authors. However, is it proper or useful for the cs code the enforce this? It seems to me this is not such a great problem that the cs code needs to be further complicated in a quest to Right All (citation) Wrongs. An occassional run with a special purpose bot would suffice. And then Trappist would have little more time to expand the |title=none feature. ~ J. Johnson (JJ) (talk) 21:09, 22 February 2016 (UTC)

I wouldn't mind a maintenance category, with the understanding that there is not yet a consensus for "fixing" the condition identified. The description of the category, and the linked help text, could make it clear that this message is merely a tracking category to identify the scale of this type of parameter usage and to look for potential false positives that could be used to refine the error check. – Jonesey95 (talk) 04:04, 23 February 2016 (UTC)

Yes, sounds like a reasonable approach to me. --Matthiaspaul (talk) 00:50, 24 February 2016 (UTC)

For those cases where any single entity's name requires multiple separators, we can employ a similar trick that is used in |vauthors= to instruct the module to skip the multiple names test:

{{cite book/new |author=(([[Baker, Donelson, Bearman, Caldwell & Berkowitz]])) |title=Title}}

Baker, Donelson, Bearman, Caldwell & Berkowitz. Title.

The doubled parentheses must wrap the entire parameter value. If they do not, then they are not stripped from the rendering:

((Baker, Donelson, Bearman, Caldwell & Berkowitz)) more author stuff. Title. {{cite book}}: |author= has generic name (help)CS1 maint: multiple names: authors list (link)

—Trappist the monk (talk) 14:11, 24 February 2016 (UTC)

I think we should avoid introduction of such type of text (vauthors excepted, of course) since there's no demonstrable consensus above for editing the articles which exhibit this concern. As with Jonesey and Matthias, I think a maintenance category is desirable. --Izno (talk) 14:18, 24 February 2016 (UTC)

Unrecognized language

Is there some reason that Filipino is triggering the unrecognized language error message? It's a valid language with the code [fil], apparently. Maybe some list needs to be updated?
—D'Ranged 1 VT^alk 14:25, 23 February 2016 (UTC)

fil is not an ISO 639-1 code which is a requirement of |language=. Neither of fil nor Filipino are recognized by MediaWiki. See names.php for the current list of supported language names and codes. Consider Tagalog (tl).

—Trappist the monk (talk) 14:39, 23 February 2016 (UTC)

More confusing information, then. If you look at the entry for Tagalog at List of ISO 639-1 codes, you find the following in the Notes column:

Note: Filipino (Pilipino) has the code [fil]

Apparently ISO 639-1 is limited to two-letter codes, although they have three-letter codes under ISO 639-2. In skimming the Wikipedia articles on Tagalog and Filipino, apparently Filipino is the standardized version of Tagalog. That's all well and good, but shouldn't we be accommodating to what people want to call their own language? I can't see us editing every citation that lists Filipino as a language to read Tagalog instead. I realize this isn't your circus, nor your monkeys, but once again I'm up against not knowing how to effect a change. How does one request that Mediawiki adopt the ISO 639-2 standard to be more inclusive? Should one even bother?

—D'Ranged 1 VT^alk 23:27, 23 February 2016 (UTC)

What the reader sees when you use |language=Filipino is (in Filipino). When Module:Citation/CS1 doesn't recognize the content of |language=, it does categorize the page and, for those who have turned on messaging, shows the maintenance category message. Maintenance messages are not necessarily error messages.

When I introduced ISO 639-1 as an option, I created a table of all of the ISO 639-1 codes and their languages. I abandoned that when I discovered the Scribunto library function mw.language.fetchLanguageName( code, inLanguage ). It works well enough for ISO 639-1 and to some extent with ISO-639-2. The library fails for ISO 639-2 codes what also have an ISO 639-1 code. The library appears to use the same data set as the magic word {{#language}}. As an example,

{{#language:is|en}} – ISO 639-1

Icelandic

{{#language:ice|en}} – ISO 639-2

ice

mw.language.fetchLanguageName('is', 'en') returns 'Icelandic' but when the language code is ice, it returns nil.

That 'anomaly' was sufficient for me to not attempt to support ISO 639-2.

The other issue that comes to mind is categorizing. Module:Citation/CS1 categorizes all templates that use |language= (except English). For the ISO-639-1 languages there are 185 of these categories. There are some 500-ish ISO 639-2 language codes 185 of which duplicate ISO 639-1 codes. Supporting all of them might be more work than it is worth.

I have thought, on and off, of supporting certain ISO 639-2 codes where those codes or their associated language names appear commonly in Category:CS1 maint: Unrecognized language; fil and Filipino would be one such example.

—Trappist the monk (talk) 00:57, 24 February 2016 (UTC)

I would definitely support that solution. I agree that supporting an additional 300+ language codes is burdensome and unnecessary, but since ISO 639-1 isn't ever likely to include Filipino, it would be great to add it to our list somehow. Thanks for your very educational response!

—D'Ranged 1 VT^alk 19:04, 24 February 2016 (UTC)

In perusing the pages in the maintenance category, I've found two anomalies already.

• If the language in |language= is Wikilinked, the maintenance tag is added, as in |language=[[Japanese language|Japanese]].
• If the language is the English equivalent, it may trigger the maintenance tag, as in |language=German

While the first example is easily fixed by unlinking the language (although that's a pain), the second is more problematical. Purging the page cache removes the green maintenance tag notifications, but doesn't remove the page from the category. The relevant article is 1. FSV Mainz 05. Other articles using |language=German don't fall into the maintenance category, though, see Attila.

Something goofy is obviously going on, but don't ask me what. It would seem that there are numerous pages in the category that don't belong there.

19:28, 24 February 2016 (UTC)

At the last update of Module:Citation/CS1 a bug in the code caused the language_parameter() to improperly categorize proper and correct language names (not codes) as unrecognized. That has been fixed. MediaWiki sometimes takes its time when removing pages from a category. The usual solution is to simply wait for MediaWiki to get around to doing whatever it is that needs doing to remove the page from the category. After all, there are thousands of us all making little modifications which all need MediaWiki's attention, so some things can be put off until later.

—Trappist the monk (talk) 19:58, 24 February 2016 (UTC)

A null edit will typically remove the page from the category. – Jonesey95 (talk) 21:27, 24 February 2016 (UTC)

The problem in the 1. FSV Mainz 05 article was apparently the use of {{de icon}} in some raw html references. Once I converted them to Cite templates (which were widely used in the rest of the article), the page was de-listed from the Category. I'll try to remember to check back in a few days and see if the Category has been purged.

I want to expressly thank you, Trappist, for your patience and all the time you spend composing detailed, educational responses to my queries. I know I'm probably on your "HME list" (High Maintenance Editors), but I hope my wanting to improve my editing skills and Wikipedia as a whole is some sort of balance for all the energy you expend. You're very much appreciated! Thank you. —D'Ranged 1 VT^alk 21:51, 24 February 2016 (UTC)

If that were the problem, you would see the category and the error message when looking at the previous version of the article in the article's History. I do not see the category or the error message. Your edits forced the page to reload all of the templates in the page, which is what fixed the category membership.

Standardizing the citations was the right thing to do, but it was not related to this discussion. – Jonesey95 (talk) 22:54, 24 February 2016 (UTC)

Thanks! Since I'm mostly clueless about this stuff, all I can do is relate my actions and wait for someone to explain what impact they did or didn't have. I appreciate the education!

—D'Ranged 1 VT^alk 23:03, 24 February 2016 (UTC)