Explanation and exampleEdit

Hi, sorry to bother you, but you are obviously the best person to write an "official" description of the neat solution you proposed for labeled section transclusion. It would be great if you could summarize the idea in one or two paragraphs, and give some links to current online examples. (You had said earlier you were working on some examples, but know for some reason I can't seem to find them... sorry...)

If you can do it, the description should go here. Thanks, Dovi 18:28, 10 September 2006 (UTC)


Hi, saw there's been a lot of recent activity at the bug report. Thanks for all your effort.

Would it be possible for you to provide brief updates at Wikisource talk:Labeled section transclusion for all us non-techies who don't know how to read code? Dovi 17:27, 7 October 2006 (UTC)

Bug report for availability at WikisourceEdit

Hi Steve, please see my note to you here. Dovi 03:58, 20 November 2006 (UTC)

Bot flag?Edit

Hi, Sanbeg,

I was wondering if, in accordance with our bot policy, you could apply for a bot flag for User:Sanbeg (bot)? That way, the community will have some information about the bot, and also we can hide its edits from the Recent Changes list. :) Thanks!—Zhaladshar (Talk) 15:51, 30 June 2007 (UTC)


A suggestion was given (in French) that we could use WebCodex to take the Bible verses from the data bases at Sefarim.fr and judeopedia.org I am not sure what he means by that and I thought you might have a better idea.--BirgitteSB 21:28, 9 July 2007 (UTC)

  • Sefarim.fr
    • Hebrew: Le texte hébraïque a été mis à disposition par l’association Mechon Mamre. Il s'inspire de la méthode suivie par Rav Mordechai Breuer (Mosad ha Rav Kook and Chorev) et son disciple Yosef (Ofer (Keter Yerushalayim) . C'est à dire qu'il ne reprend pas un codex en particulier et ses inévitables erreurs, mais effectue la synthèse de plusieurs manuscrits de l'école de Tibériade, remontant au 8 ème siècle. Plus d'une centaine d'erreurs ont ainsi été éliminées, par rapport aux codex les plus fréquemment utilisés (BHS) poor translation of key parts the synthesis of several manuscripts of the school of Tibériade going up to the 8(?) 2nd century. More than one hundred of errors were thus eliminated This hebrew version is probably a copyright problem.
    • English: Le texte anglais provient de l’édition JPS 1917. Il a été mis à disposition par l’association Mechon Mamre This english verison is the text we have at Bible (Jewish Publication Society 1917) But it is both incomplete and a questionable transcription.
  • judeopedia.org
    • Hebrew: I am unsure of what version it is.
    • English: King James Version.
    • Latin: Vulgate
The JPS 1917 is something we really need (Wikisource:Possible copyright violations#JPS 1917) even outside of the verse by verse project. If this webcodex will allow automatic importation by verse that would obviously be helpful for the Vulgate and the Hebrew (assuming is a good copyright version). And if that can be used to insert verse markers easier on the JPS 1917 that would also be great. I will ask Dovi to comment about the Herbew version.--BirgitteSB 13:20, 11 July 2007 (UTC)

With regard to the Hebrew versions, we have had extensive and exhaustive discussion of this at Hebrew Wikisource, and come to some practical conclusions. An explanation of the issue is in order.

The Aleppo Codex (A) is considered by scholars, both traditional and modern, to be a more accurate work than the Leningrad Codex (L). The former is also considered authoritative in traditional Jewish law (for writing scrolls). Only the latter, however, is currently complete, while about 1/3 of the former was lost or destroyed in 1947 when the Arabs of Aleppo, Syria, rioted against the city's Jewish community in protest against the creation of the State of Israel. When the manuscript was finally smuggled to Israel in the 1950s, about 1/3 of it was missing.

This leaves scholars with two options:

  • To simply use the Leningrad Codex as the basis for editions of the Hebrew Bible, as it is an excellent and generally accurate manuscript reflecting the Tiberian tradition, and the oldest complete manuscript currently extant of the masoretic Hebrew Bible. This has been the practical decision of most Christian scholars, adopted in w:BHS.
  • To use the nearly perfect, error-free Aleppo Codex as the text where it is extant, and to reconstruct its missing parts based on parallel manuscripts (among them the Leningrad Codex). This reconstruction has been objectively proven to be nearly 100% accurate for the over two million orthographic elements in the masoretic text. The proof is simply that when the same method is applied to the extant portions of the Aleppo Codex, the results match perfectly. In addition, witnesses who took notes on the Aleppo Codex (before it was damaged in 1947) agree with the reconstuction as well. The method of A + reconstruction has been adopted in no less than four different Israeli published versions of the full Hebrew Bible, and a two more incomplete ones (by Bar-Ilan University & Hebrew University). This text also reflects nearly the exact letter-text actually used by Jews in Torah scrolls for the past several centuries (the scrolls were aligned with reports of the mesorah and A).

The question, of course, is whether the reconstruction of the missing parts of A is copyrighted? I personally am of the opinion that it is not, because not only has the reconstruction method been used with the same objective results in six different printed editions and a seventh online edition, but mostly because it is based on documented data that anyone can use with identical results.

However Mechon Mamre, which created the online version, objects. For the sake of peace we decided not to use their version for vowels and cantillation, but to instead substitute for the meantime the L-based version at tanach.us, which has a liberal usage license.

Nonetheless, for the letter-text itself we decided to use the same A version that you see reproduced at the French site. The reasoning is that when it comes to the letter-text, the vast majority of what is missing in A comes from the Torah, which has been copiously documented and compared with Torah scrolls (on which there is no copyright). The very few questionable places (where a single letter here or there is in doubt) can then be edited by hand, with the editorial decisions documented on the talk pages.

On the French page, when they talk about numbers of errors, that is what they are talking about: The number of single-letter-errors found in L but not in A. They have also reproduced the MM letter-text version, but also with MM's somewhat strange punctuation, which is surely copyrighted. Therefore, if you choose to reproduce that text here at en.wikisource (and I personally recommend that you do so because it is the best), you must make sure to automatically remove the puctuation marks and leave only the Hebrew letters (replacing the punctuation with a space when needed). This is easy to do automatically. Dovi 06:36, 12 July 2007 (UTC)

A Trip to VenusEdit


Why did you create pages just to put "speedy delete"? Regards, Yann 18:04, 5 September 2007 (UTC)

Why the speedy delete request?Edit

This appears to come from Baird's Manual of American College Fraternities (1879). Is there a problem that I'm missing? FloNight 20:43, 23 October 2007 (UTC)

I thought maybe you noticed something specific about the entry that I missed like it was a duplicate or something. Or maybe did not realize it was part of a larger work since it was ill formatted and not linked. Since this page is part of a book that has been here for a long while with different editors as contributors, I think a deletion discussion would be better than a speedy delete request if you still think deletion is best. FloNight 23:10, 23 October 2007 (UTC)


Hi, I see you created LST ... thank you! {{no header}} is using it to display the header template usage in a funky way. John Vandenberg 01:22, 12 December 2007 (UTC)

More LSTEdit

Hi, Sanbeg. First of all, as JV says, thanks for software implementations like LST.

I have found today some bugs at diferent projects (:w:ca, :wikt:ca, :it:s, :fr:s), and I think they are related to meta:Migration_to_the_new_preprocessor. And finally something rare at :ca:s:

At [1] there is a transclusion that uses the lst tags. I think it worked days ago, and now it does not transclude. But there is another page (ca:Romans_fet_per_Johan_Fogassot) which transcludes a section from the same page as the one that does not work (but from a second section label).

Could you please have a look at test page. Does the LST works with only one section label but not with more??

Thanks in advance! -Aleator (talk) 23:32, 24 February 2008 (UTC)

please acceptEdit

Adminship! ; please accept ;-) John Vandenberg (chat) 20:51, 28 February 2008 (UTC)

  • Congrats you are now a sysop!--BirgitteSB 17:40, 9 March 2008 (UTC)

Collaboration projectEdit

This weeks collaboration project is G. W. Bush. Please take the time this week to identify and/or transcribe one important work by, or involving, this very prominent person who is relevant to us all. John Vandenberg (chat) 03:12, 7 March 2008 (UTC)

compact bugEdit

I have set up a demo of the bug at User:Sanbeg/poem. At present, the use of "compact" disables results in <poem> formatting only taking effect for the first line. The rest are unformatted. John Vandenberg (chat) 01:51, 5 April 2008 (UTC)

It looks like I stripped out a few too many newlines, which messed up indentation, but shouldn't affect other formatting. I've commited a fix for that. -Steve Sanbeg 15:45, 8 April 2008 (UTC)
I just purged that page, it looks fine now. -Steve Sanbeg 19:49, 8 April 2008 (UTC)

Footnotes question at ScriptoriumEdit

If you have a moment, could you take a look at Wikisource:Scriptorium#How to preserve source text’s erroneous footnote numbering? Your expertise in the footnote system would be pretty enlightening one way or the other, I suspect. Many thanks! Tarmstro99 06:13, 10 April 2008 (UTC)

Thanks for the input, it’s very helpful. Tarmstro99 16:27, 10 April 2008 (UTC)

Scripts, againEdit

Hi Steve. Thank you very much for answering both of my queries at the Scriptorium, for scripts and for text quality. I just have one more question, where can I find the available edit buttons images on Wikisource for my extra edit buttons? As you can see from my monobook, I've added quite a few based on your suggestion. I found this list on commons, but the normal links don't work. I believe the commons.wikimedia.org links don't work, but upload.wikimedia.org seem to work fine (an example is commons:Image:Button link to Wikipedia.png) Any ideas? Many thanks, Mtmelendez 19:19, 15 April 2008 (UTC)

The commons.wikimedia.org links are the image description pages; the upload.wikimedia.org links are the image files, which are what you need. So doing a right click and "view image" or similar to the the actual file URL on upload should give you the URL you need. -Steve Sanbeg 21:23, 15 April 2008 (UTC)
Thanks! They work great now! - Mtmelendez 14:37, 16 April 2008 (UTC)


hi Sanberg,

I just noticed that you modified Template:Page There are a bunch of deprecated things in it, that could be removed. The image used to be a quality indicator (it worked with javascript) The "from ... to" parameters should be replaced by LST

ThomasV 18:48, 16 May 2008 (UTC)

It looks like there are a bunch of pages still using the from..to, but it should be manageable; I've fixed a few already.. -Steve Sanbeg 22:46, 16 May 2008 (UTC)

Thank you for changing thatEdit

Just wanted to say thanks for changing the proofreading status markers. It probably isn't a huge thing, but they now give users a better idea what they're indicating when they attempt to pick between them. --Mkoyle 22:08, 2 July 2008 (UTC)

The Mahabharata: thank youEdit

Thank you for these. They have been on my todo list for a while! :/ Im glad to see they are getting done. John Vandenberg (chat) 23:51, 21 August 2008 (UTC)

Cooking raw html in wiki-textEdit

Hi, when you have some time, I'd like to continue this discussion. A few links would be great so I can go read-up on it. Cheers, Jack Merridew 07:04, 3 September 2008 (UTC)

Sure. mw:Manual:$wgUseTidy would be a good place to start, since that seems to link the relevant docs. -Steve Sanbeg 18:51, 3 September 2008 (UTC)
Thank you; I'm finding interesting things off there. Cheers, Jack Merridew 08:06, 4 September 2008 (UTC)

Wikisource:Scriptorium archivationEdit

Hi, is User:Sanbeg (bot) currently automatically archiving the Scriptorium? It would be considerably easier to stick a

Attention: The “section resolved” template needs a valid date. The section will not be archived otherwise. Please use {{Section resolved|1=~~~~}} instead.

tag on sections rather than do them manually. :-) Jude (talk) 22:45, 7 September 2008 (UTC)

Lately, I've been running a script around the end of each month that parses the time stamps and archives sections that weren't touched recently. I'm thinking about setting that up to run automatically once I've done some more testing on it, and made sure it does the right thing in any odd cases that could arise. In the past I've occasionally done manual archiving with LST; it should also be possible to have the script use a tag instead of a timestamp to detect the status of a thread. -Steve Sanbeg (talk) 17:19, 8 September 2008 (UTC)
Hm. I've still got the source code for the archivation program I wrote for User:Xenophon (bot) about two years ago, which I could probably bring back into action pretty easily... So long as I can find it on the toolserver. :-) Jude (talk) 00:05, 9 September 2008 (UTC)
No rush. Actually, the time stamp method seems to work pretty well on the scriptorium, so it's nice to have archiving with no human intervention, although I could see a leave thread here until x/y/z kind of tag could in theory be useful. Other pages where the threads reach a more definite conclusion would probably benefit from an archive tag, though. -Steve Sanbeg (talk) 16:39, 12 September 2008 (UTC)
I've gotten it to cooperate better with partial, manual archives, so it should be able to run more regularly now. -Steve Sanbeg (talk) 01:14, 28 September 2008 (UTC)

Sections and templates: a questionEdit

Here the copy of a talk between me and John Vandenberg.

Hi John, my boldness is severely worrying my new Italian friends. ;-)

My last "discovery" is a novel use of selective transclusion; t.i., any page can be used as a "record" containing any number of "fields", identified by a named section tag. This could perhaps be a key to manage in a polite way some "user-defined variables", or better some "user-defined constants". Take a look, if you like, to it:Utente:Alex brollo/Gaio Giulio Cesare and to its code; I hope, it will be self-explanatory for you. The problem is that section tags are to be written into the page; it would be great if they could be put into a template, but wiki software doesn't seem to see sections coming from the code of a template; therefore I can't find a way to help an unexperienced user to use this trick. Nevertheless, perhaps some use of this trick could be found - some to think about in my opinion. --Alex brollo (talk) 06:35, 23 September 2008 (UTC)

The "section" tags are evaluated before templates are evaluated. This does limit its usefulness.
You will probably enjoy Page:College Songs (Waite, 1887).djvu/22 used on Oh My Darling, Clementine#College Songs (Waite, 1887), and Template:Bible/Joel which is used on Bible/Joel/1/1
John Vandenberg (chat) 06:57, 23 September 2008 (UTC)
Pretty examples - I have to admit, difficult to understand at a first look. About the order of evaluation for sections and templates: OK, I understand the reason why my tries failed. But - there's a deep-rooted reason I suppose to follow that order - or could it be changed? --Alex brollo (talk) 12:32, 24 September 2008 (UTC)
User:Sanbeg is the man to talk to about the ordering. John Vandenberg (chat) 13:06, 24 September 2008 (UTC)

So here I am to ask you about the order of evaluation of sections and templates. I imagine that there is a sound reason, but... I'd like to be sure about. Thanks! --Alex brollo (talk) 14:23, 24 September 2008 (UTC)

There are a few reasons for it. Mainly, the extension doesn't handle parsing; it only fetches the page, extracts the section, and lets the parser do its thing, just as it would with an ordinary transclusion. So it's not a simple matter of changing the order of things, since the extension doesn't control that.
The closest we can come in faking that would be for the extension to do part of the parsing out of the normal order to substitute templates, then let the parser repeat that same stage in the normal order. This out of order parsing would likely cause some things to parse differently than they would if parsed in the order, to it could cause some breakage. It would certainly have performance issues, since getting a small section from a large page would potentially mean getting a lot of templates from the database, substituting their text, then discarding the result.
It may be useful to have some way to explicitly request an out of order parse in certain circumstances, but I don't know how difficult that would be (I haven't had time to digest all the changes in the new preprocessor) and I don't think it would be desirable in the general case. Anyway, I hope that helps clarify things.-Steve Sanbeg (talk) 21:15, 24 September 2008 (UTC)
Thanks very much. Another question for you. Do you think that using sections to identify "costants" or "fields" into a page is so critically resource-consuming, to discourage such a trick? --Alex brollo (talk) 04:44, 25 September 2008 (UTC)
It shouldn't be. The section tags are (to the parser) a no-op, so they will require some resources to run the hook, but not as much as tag that does something, or some formatting. Extracting a lot of small sections could take more processing power than a simple template substitution, since it will have to scan through an entire long page to find the section for each call, while the templates would be smaller; but each of those templates would require a database call to retrieve, while the #lst calls (after the first) would just get the page from a local cache. -Steve Sanbeg (talk) 22:15, 26 September 2008 (UTC)
Thanks very much. I quoted your answer into my presentation of the whole idea in it.s. --Alex brollo (talk) 08:49, 29 September 2008 (UTC)

LST using {{#tag:section}}Edit

Hello Sanbeg. Could you add a self-contained version of the section tag to the labeled section transclusion extension for short text (like <section name="foo">text here</section> or <section name="foo" text="text here" />)? We could then use it as {{#tag:section}} in templates, which might allow us to transclude parsed automatically-delimited portions of the page. For example, headers on subpages could transclude the author value from the header on the top page. —{admin} Pathoschild 13:25:29, 03 October 2008 (UTC)

I'll have to think about whether there is a way to get the behavior you're looking for; it comes up from time to time, but I suspect that it won't be simple to do. I don't think that #tag would be helpful, since the hook function for section is a noop anyway; those markers are interpreted by the extension, not the parser. There is some more discussion in the thread above. I think what's needed is an alternate form of #lst that will at least substitute templates prior to looking for section markers. -Steve Sanbeg (talk) 16:43, 3 October 2008 (UTC)
Could the extension run the page through parser::parse() before retrieving sections? —{admin} Pathoschild 23:49:11, 03 October 2008 (UTC)
I don't know if there's any way to have that function pass the section markers through, but calling something like (IIRC, it's been awhile) parser::braceSubstitution() may be workable. -Steve Sanbeg (talk) 16:45, 6 October 2008 (UTC)
Just wondering whether there had been any progress on this issue. I have found that empty section tags work fine, eg.
<section begin=whatever></section>
This text to be transcluded
<section end=whatever></section>
I am trying to get Template:Footnote end to function, and it seems to need the use of #tag, yet it fails (and quietly) when bibs and bobs are done to it. billinghurst (talk) 10:29, 7 December 2009 (UTC)
Sorry, I'm afraid trying to use #tag with a section marker is doomed to fail, so we'll need to find another way. I guess I should explain that better, though :). As I'm sure you know, markup like <thing>text</thing> is treated by the parser as an extension tag, which calls the extension on the specified text; and #tag is alternative way to call the same extension, to do what the parser what do with that tag. In this case, the section is bound to a simple callback function that just ignores its input and always returns an empty string. Not too useful by itself, it's just done that way so that #lst will have some kind of marker that it can look for, but that the parser can easily remove when rendering the text. So with #tag:section you should be able to get the noop tag behavior, but not create the marker, which is what you really want. -Steve Sanbeg (talk) 00:54, 8 December 2009 (UTC)

nested transclusionEdit

On Bible/Leviticus/1/1, the entry for JPS1917 comes from Bible (Jewish Publication Society 1917)/Leviticus, and that text was recently moved to Page:JPS1917-Torah.djvu/130. I think the only way to make the comparison text work currently is to do this. Ideally the labels that are transcluded together in Bible (Jewish Publication Society 1917)/Leviticus become visible as defined sections. If that is too much magic, I could keep doing these and moaning at the loss of the [edit] button on Bible (Jewish Publication Society 1917)/Leviticus taking the reader to the pagescans, or I could create Bible (Jewish Publication Society 1917)/Leviticus/LST and put the voodoo there to be included into Bible/Leviticus/1/1. John Vandenberg (chat) 09:30, 7 October 2008 (UTC)

It looks like there's a lot to think about there; I'm not even 100% clear on what you're looking for. It may be useful at some point to implement some kind recursive version of #lst, for use in these cases. For what you're doing now, it may be useful to transclude the chapter headings the same way you're transcluding the text, to keep the [edit] links consistent. -Steve Sanbeg (talk) 16:29, 7 October 2008 (UTC)

interaction between poem extension and transclusionEdit

Hi, could you opine on an issue I've run into? See;

Currently, it's all falling apart. What I'm trying to do is have the two blocks of trancluded text wrapped in a single table and poem tag set; the opening of the table and poem occurs in the first page and the close occurs in the second. Each of the pages has half of the markup in the either the header or the footer to hide it from the transclusion so that the tags are properly nested when the fragments are considered as standalone pages. I've successfully done the table aspect of this elsewhere, but it's not working out for the poem pseudo-tag. I expect that the extension simply doesn't work correctly across multiple block of transcluded text.

Another thing to look at is;

Here, the transclusion is to the left-side and I'm only hiding bits of the table, not the poem-tag. I tried to hide the poem here, too, but it's undone; in the histories.

The reason I want to do this with the poem is that when there are multiple poem blocks, MediaWiki is generating paragraph elements differently and this results in the left side of The Grass so little has to do — being built as two paragraphs instead of one and the spacing gets messed up; on that page, "For such a noticing." is the last bit of a paragraph and "And even when it dies, to pass" is the beginning of the next; this is instead of there being two br-elements generated. FWIW, I really think it would be better if this extension were generating opening/closing sets of p-tags for all cases where it's generating the 2-br sequence.

nb, the single-cell table is in this picture simply to get the poem centered and the padding-left is just a tweak for 'apparent-centering' due to the typical white-space at the end of each line.

Also, I see that this extension is marked as obsolete, so it may well be that I should be jumping to something entirely new; pointers welcome ;)

Cheers, Jack Merridew 06:55, 6 November 2008 (UTC)

The obsolete tag looks fishy, so I've reverted that. XML-style wiki tags generally don't work quite the same as other wiki syntax; for example they tend to be hidden during part of the parsing, and unhidden at a later stage. There was a compact poem that should do this, but that was broken by some change to the parser, and I haven't had a chance to look into why. Another option may be to wrap the body of the poem in <onlyinclude>s and repeat the markup in the transcluding page, or try to get something working with LST. -Steve Sanbeg (talk) 18:46, 7 November 2008 (UTC)
It looks like someone did copy the poem code into the code, so eventually the extension will go away, but there should be no noticeable changes. -Steve Sanbeg (talk) 23:15, 7 November 2008 (UTC)
Looking at how we should generally handle books of poetry. I would reccomend using LST inside the poem tags on every page. Generally LST will already be neccessary to seperate poems on a good portion of pages of any given book of poetry. So I think we should just use LST regardless of if it neccessary for seperation so it may also handle the issue with the poem tag. Rather than using a seperate system for each issue.--BirgitteSB 19:15, 7 November 2008 (UTC)
I tried to make an example but the LST is not working. User:BirgitteSB/Sandbox Any thoughts?--BirgitteSB 20:28, 7 November 2008 (UTC)
Maybe you can't nest LST in a poem? I'll check if that was broken when poem was moved into the core. -Steve Sanbeg (talk) 23:28, 7 November 2008 (UTC)
  • I've been aways awhile and wanted to check in on if you had a chance to look into this at all. Thanks. --BirgitteSB 20:14, 9 January 2009 (UTC)
    • I've given it some thought, but I haven't gotten a chance to implement anything yet. I think the problem is that we've been looking for some trick that will happen to work, but even if we do find one it may not be stable, so it would only help for awhile; I think we really need a feature specifically for this. My best idea so far is to do something like <section name=example render-tag=poem>this is part of a poem</section> and have LST add the the poem tags itself on the page; then it may be transcluded with something like {{#tag:poem| {{#include:page:something|example}} ... }} -Steve Sanbeg (talk) 19:14, 12 January 2009 (UTC)

Thanks, both; I'm going to review the ws:lst/mw:Extension:Labeled Section Transclusion doc and try and get a grip on this stuff. Thanks for the help.

Birgitte, I just noticed that your sandbox test has "section=The First Chantey" re page 201, but that page names it w/"begin=First"; I previewed a tweak that didn't seem to make any difference; possibly it needs to match the h2 id. Off to read about it.

Steve, the compact option, broken or not, would take things in the opposite direction than I intend. The point of merging the two poem-tags on the discreet pages into one when they are transcluded to a combined page is to get all the hard newlines to expand to the 2 br-element sequence, including at the seam between two transcluded pages. Changing the behavior (<br /> <br /> to </p> <p>) of this now non-extension bit of MediaWiki probably would have issues, but it might be something to consider as an option such as compact or some cousin-xml-taq. "Poem" is really a poor name for this as it really doesn't have much to do with poems besides that being, presumably, the initial intended use for it; I see this as much more akin to the true pre-element.

It does seem that the issues I've encountered and you both have mentioned above are odd interactions between different bits of code. I've written an awful lot of code and expect that things get pretty complex deep inside MediaWiki. I'm liking the more robust seeming LST approach, and see onlyinclude as a fallback. Cheers, Jack Merridew 05:31, 8 November 2008 (UTC)

I just tried an onlyinclude-test at 86 and 87; diffs: diff86 diff87 — which caused The Grass so little has to do — to fall apart completely; I undid the test.

Then I tried wrapping only the raw text with the onlyinclude tags; diff86 diff87 — which required that I duplicate the table and poem-tags in the presentation layer page that's pulling these in: diff:The Grass…. This produces the rendering I'm after, although I don't like having to duplicate the code. This approach is rather fragile in that it needs to have the newlines omitted in the transcluding page; at least no newline between the two transclusions of 86 and 87; I expect that this could be 'moved' to the pagespace by tweaking newlines there, instead; all too fussy for non-technical folk. Off to review LST, next. Cheers, Jack Merridew 06:32, 8 November 2008 (UTC)

I was able to add some newlines to unscrunch things a bit in The Grass so little has to do — page; the newline between the transclusions still must be omitted. Cheers, Jack Merridew 06:42, 8 November 2008 (UTC)

Two more tweaks diff87 diff:The Grass… and the newlines are reasonable; I'm not keen on the <onlyinclude> having to be on the same line as the first line of prose, but could live with this. re LST; for the poems at hand, I believe there will only be one section per pagespace page and I would label it 'prose' or something like it. There is, I expect, an order to all the parsing MediaWiki does, so it may be that there are a different set of issues here. I'll run a test somewhere of this technique, too. Cheers, Jack Merridew 07:03, 8 November 2008 (UTC)

I've realized another issue; when page 31 was validated, a bit of script rewrote the header, adding some newlines. This is going to mess with any attempts to put stuff up in the header, so we're certainly moving in the right direction. I've done that trick in a few spots in this work and they'll have to be revisited. Cheers, Jack Merridew 07:48, 8 November 2008 (UTC)

Hmm, I may have still spoken too soon. Transcluding poems in pieces may be doable as simply as the only slightly ugly example at user:sanbeg/pbc. Unfortunately, the ugliness can't be hidden in a template, but it should work without code changes, and not be as fragile as other things we've tried. -Steve Sanbeg (talk) 02:15, 17 January 2009 (UTC)
Sorry, I missed this reply. I'll look into the example and the state of The Belle's works. There were, I recall some loose ends. I big piece of it was just how to mix the rather longstanding versions with the versions from that specific printing; i.e all her friends tweaks and made-up titles. Cheers, Jack Merridew 09:16, 6 May 2009 (UTC)

Scripting skills required {{Sidenotes begin}} or similarEdit

Hi Sanbeg. Working on a book that has sidenotes left and right. Looks good in the Page: space, however, when brought to main namespace it needs some intervention. Discussion took place at Template_talk:Sidenotes begin. I would either like to have some complexity to the template or a separate template that can be used in the main namespace. I would rather work on works, rather than try to learn to script it, so I am here seeking advice/help. Thx. -- billinghurst (talk) 00:40, 4 January 2009 (UTC)

If I'm reading it correctly, the issue isn't namespaces, but that sidenotes and page both put text in the left margin, and having two left margins looks bad. We may need to find a way to put them into the same margin, or at least make it look a little neater. -Steve Sanbeg (talk) 18:31, 12 January 2009 (UTC)
Not quite. In the Page namespace, I am wanting the rendition to show as per the book, which is left and right sidenotes depending on the page spread respectively. However, when the pages are transcluded to the main ns, the rendering of L R L R L R looks messy, and there it would look better if they were all forced to the one side. -- billinghurst (talk) 00:26, 13 January 2009 (UTC)
That should be simpler, then; just put all the sidenotes on the right, so they won't interfere. Sure, we could have a template that switches sides based on the namespace or pagename; i.e. user:sanbeg/Sidenote tries to guess which side a note would go on based on pagename, and currently displays left or right. We could integrate that logic into the sidenote display, if the pages tend to follow the right name format, so it can guess right most of the time. -Steve Sanbeg (talk) 17:28, 13 January 2009 (UTC)
I've also implemented template:sidenote which is quite experimental; it calls that userspace template and the appropriate sidenote template. -Steve Sanbeg (talk) 18:47, 13 January 2009 (UTC)
Thanks. I finally got back to this work, and while my scripting is pretty rustic, I have written {{tl:Force sidenote}}. It inhales the existing left and right sidenote and allows me to make most left sidenotes when taken out of the Page: namespace. If you think that it needs tidying or improving, feel free to modify.   -- billinghurst (talk) 14:43, 2 May 2009 (UTC)

a technical reviewEdit

See Template talk:Dropinitial-span‎ and the template implementation. I also just left a note at User talk:Jayvdb#a technical review. I'm pretty sure this is a good direction to go, but would appreciate more eyes on the details before I make this the norm.

Cheers, Jack Merridew 09:12, 6 May 2009 (UTC)

lst and overflowing referencesEdit


I have been trying to make lst work with overflowing references such as you can see at the bottom of Page:Miscellaneousbot01brow.djvu/29 and Page:Miscellaneousbot01brow.djvu/30. My idea was to break page 30 into separate sections for text and overflow; then tranclude the overflow section into the page 29 reference, but within an includeonly block. I've tried a few things and I am not even close to figuring it out. Do you know whether or not it can be done?

Hesperian 04:28, 8 June 2009 (UTC)

OK, I think I've gotten it. It seems like a straightforward combination of lst & includeonly. Although I'm not sure if it's useful to nest any of that inside of a ref tag, which may have been what was giving you trouble. -Steve Sanbeg (talk) 17:35, 10 June 2009 (UTC)
Thanks for trying—an interesting approach—but there's two problems. 1. The whole point for me was to shift the overflow text onto the page where it appears; whereas you have retained it at the other page. 2. More importantly, it has to work when the pages are transcluded at The miscellaneous botanical works of Robert Brown/Volume 1/General remarks on the botany of Terra Australis; and it don't. e.g. the includeonly affects both the overflow transclusion and the page transclusion.
If it is not possible to call an lst from inside a ref, then I guess it can't be done. :-(
Hesperian 23:24, 10 June 2009 (UTC)
Oh, I guess I'm not as clear as I thought about what you want to do. Calling lst from a ref shouldn't be a problem, as long as the section markers aren't nested in an extension tag. -Steve Sanbeg (talk) 18:37, 11 June 2009 (UTC)

After further investigation, I think the root cause of my difficulties is the fact that <includeonly> tags do not work inside <ref> tags. e.g. if you transclude a page with content


the result does not contain the B, which is an error in my opinion. Hesperian 00:16, 12 June 2009 (UTC)

I've submitted a bug report. Hesperian 00:32, 12 June 2009 (UTC)
Yeah, the issues with nested XML tags have been around for awhile; i.e. the parser doesn't know which extensions contain wiki text and which don't, so it assumes they don't. I'd expect this is related. But you could probably get the same affect if you duplicate the content, i.e.

-Steve Sanbeg (talk) 15:23, 12 June 2009 (UTC)

I think something like {{#tag:ref|A<includeonly>B</includeonly>}} should work as a workaround for the problem with includeonly tags in ref tags. /EnDumEn (talk) 09:38, 21 June 2009 (UTC)

Hebrew pointingEdit

Hi, again. Could you please peek at this and the edit I describe there?

I see that you made some subsequent tweaks to that code. I think the span-close I added is fine; I suspect that a lang="rtl" may be needed in there; maybe in each span. I've run into this issue elsewhere recently, too; see:

Cheers, Jack Merridew 05:13, 14 June 2009 (UTC)

p.s. I'll get back to Emily and the lst of poems; promise.

I actually don't know much about RTL myself. My edits there were just to add a missing piece that the javascript was looking for (IIRC Zhaladshar ported the rest for Hebrew wikisource, but missed that bit) and to try getting the same functionality without using javascript. That last one seemed OK to me, but someone else claimed to have problems with it, so I didn't do anything more with it. -Steve Sanbeg (talk) 21:45, 15 June 2009 (UTC)
The changes to Edittools seem stable and I've not seen any complaints. My attention was drawn there by the Firefox HTML Validator plug-in. Cheers, Jack Merridew 10:56, 16 June 2009 (UTC)


User:Fandalbot, probable returning vandal. Charles Matthews (talk) 09:50, 30 June 2009 (UTC)

Now active as User:Return of the Fandalbot. Charles Matthews (talk) 15:09, 30 June 2009 (UTC)

Vandal againEdit

User:Fandal - A New Beginning. Charles Matthews (talk) 10:33, 2 July 2009 (UTC)


One of your subpages is found in Category:Wikisource; I left it there for Justin Case's benefit, it may be some maintenance thing beyond my ken. Cygnis insignis (talk) 07:45, 3 October 2009 (UTC)

OK, it's deleted now. It was an old test page, but that bot is pretty stable now. -Steve Sanbeg (talk) 15:06, 5 October 2009 (UTC)
Thanks Steve. Nice to meet you. Cygnis insignis (talk) 19:40, 5 October 2009 (UTC)

Applying your greater geniusEdit

Gday Steve, I was wondering if we could borrow that Developer component of your brain to help us nut out some of the issues around Wikisource:Scriptorium#PediaPress and validated texts. While it is not a highly used component of the site, it probably should be. Thanks if you can assist. — billinghurst sDrewth 13:27, 4 May 2010 (UTC)

Wikisource:Scriptorium#mw:Extension:LilyPond and BugzillaEdit

Hi Steve. Tim Starling has done some responses, and I wondered whether you might be able to give some opinion here. Thanks. — billinghurst sDrewth 13:04, 11 May 2010 (UTC)

Bot flagsEdit

I am contacting everyone who operates an active bot on en.WS. When granting bot-flags I am heavily reliant on the consensus of people like you to ensure that prospective bots approriate and should be flagged. I am not at all competant to evaluate bots independantly and this leads to people waiting a rather long time to find out if their bots will be santioned. Could you please help me out and take the time to examine one of the two current requests at the Scriptorium?[2] [3] Thank you.--BirgitteSB 20:21, 11 May 2010 (UTC)

Bible/2 John/1/6 and the likeEdit

With progress in other tools that allow side by side, I am wondering on the status of pieces like Bible/2 John/1/6. It would seem that we would be able to compare a lot of works side by side now, and pages like this that are not actual works but constructs, should be reviewed for their original purpose. Thoughts? — billinghurst sDrewth 01:25, 4 July 2010 (UTC)

I mostly did the technical aspects of that project, implementation of the extension & bot. Birgette or Dovi, who were on the project before I was could probably give you a better idea of the original purpose. But I don't think the side-by-side comparison really replaces the ability to compare parts of 6 similar works. -Steve Sanbeg (talk) 23:50, 13 July 2010 (UTC)


With the implentation of {{plain sister}} into {{header}}. I was wondering whether you thought that the template was still needed? Currently it is unused. — billinghurst sDrewth 09:23, 15 January 2011 (UTC)

It's been along time since I've looked at it, but I don't think it was ever widely used; more a proof of concept that we could use one template instead of different templates with slightly different display formats for each site. If none of those are used anymore, than this probably isn't needed either. -Steve Sanbeg (talk) 01:04, 31 January 2011 (UTC)
Sweet, plain sister probably belongs here in that case, rather than with its extended name. — billinghurst sDrewth 04:03, 31 January 2011 (UTC)

Archiving Scriptorium among othersEdit


Just a friendly reminder asking to please archive Scriptorium as soon as possible (or by the end of the month) It didn't get done in March and it is getting rather long at this point. TIA. -- George Orwell III (talk) 21:10, 24 April 2012 (UTC)

Wikisource User GroupEdit

Wikisource, the free digital library is moving towards better implementation of book management, proofreading and uploading. All language communities are very important in Wikisource. We would like to propose a Wikisource User Group, which would be a loose, volunteer organization to facilitate outreach and foster technical development, join if you feel like helping out. This would also give a better way to share and improve the tools used in the local Wikisources. You are invited to join the mailing list 'wikisource-l' (English), the IRC channel #wikisource, the facebook page or the Wikisource twitter. As a part of the Google Summer of Code 2013, there are four projects related to Wikisource. To get the best results out of these projects, we would like your comments about them. The projects are listed at Wikisource across projects. You can find the midpoint report for developmental work done during the IEG on Wikisource here.

Global message delivery, 23:22, 24 July 2013 (UTC)

Work indexEdit

Hello Sanberg, I have just noticed—making me happy—that you created Wikisource:Works-A. Thank you very much. If you give me the script you used and a detailed explanation of how to use it (I've never done things like that before), I can create and update the pages on my own without bothering you :)

The script only needs a few adjustments:

  • the titles would look better between <span style="background-color:#F0F0F0"> and </span>, as the light gray background would make the title more easily distinguishable from the categories,
  • the categories would look better if separated by spaced em dashes ( – ).

Thanks in advance, Erasmo Barresi (talk) 17:32, 2 September 2013 (UTC)

You're welcome. That's a quick perl script I threw together a little while before I went on a vacation, so it's probably not in the best state, but you can see how much sense it makes. So I pushed the script to https://github.com/sanbeg/wikisource/blob/master/work-index.pl, and some of the common code that I use in several of my scripts is on CPAN http://search.cpan.org/~sanbeg/MediaWiki-EditFramework/. I didn't set the script up to edit yet; it currently just dumps the text to the standard output; I just looked it over then copied it in by hand. So if you're reasonably competent with perl, you should be able to install my module, either create a 'passwd.pm' file with your username and password or remove the long part, which shouldn't really be necessary at this point anyway, and see if you can get the expected output. -Steve Sanbeg (talk) 03:11, 6 September 2013 (UTC)
I've never done things like that before :) It looks more difficult than I had imagined. It would be probably better to use your bot to create and then update the pages (maybe three times a year, for example in January, May, and September). Again, thanks for all your previous and future work.--Erasmo Barresi (talk) 17:10, 12 September 2013 (UTC)

removal of admin rightsEdit

Hi Sanbeg,

You've lost the admin bit at your annual confirmation due to inactivity. If you're gone for good, then thank you for your years of service to the project, and good luck for whatever you do next. Should you return some time in the future, then of course you'll be welcomed back and reinstatement of the admin bit is a routine matter.

Hesperian 00:59, 1 October 2014 (UTC)

Sorry to hear that. I do still pop in as time permits, although I've mostly been sticking with a little archiving, so my edits tend to be credited to the bot account. -Steve Sanbeg (talk) 04:14, 30 December 2014 (UTC)

Archiving WS:SEdit

Hi. Are you planning on running Sanbeg (bot) through WS:S? If we are needing to look to alternate archiving means, please let the community know. Thanks. — billinghurst sDrewth 23:33, 19 July 2016 (UTC)

Encoding problemEdit

Hi Sanbeg, your bot made this edit which broke the original wikilink de:Seite:Ludwig Bechstein - Thüringer Sagenbuch - Erster Band.pdf/19 due to a German umlaut. I guess it it related to an encoding problem. Could you fix this. Thank you for cooperation, --Aschroet (talk) 04:37, 10 September 2016 (UTC)

Oh, that's strange; yeah, it definitely looks like the encoding got messed up. I'll try and track down what went wrong there. -Steve Sanbeg (talk) 02:25, 14 September 2016 (UTC)

Bot unexpectedly archiving WS:SEdit

Hi, your bot has suddenly archived a couple of chunks from WS:S. This included one of my recently added comments, which is why I noticed it. We've got SpBot currently doing the archiving, so is there a way of stopping your bot from doing it as well? Thanks, Beeswaxcandle (talk) 06:37, 7 May 2019 (UTC)

Ah sorry. Things have been so busy lately I haven't paid too much attention to the wiki, so I didn't realize there was another bot now. I've backfilled the index in Wikisource:Scriptorium/Archives through the end of 2018, since it seems like it was being updated off & on before it was stopped in mid year. I guess I'll leave it at that, then. Steve Sanbeg (talk) 03:10, 26 June 2019 (UTC)