I'm testing a template [:it:s:Template:Pg]] (calling a Lua module) that fastly converts book page numbers, passed as the unique parameter to the template, into links to right djvu pages, without any need to add delta parameters and linking any format for page number (arabic, roman, other...) since page numbers are seen as strings.
The Lua script reads a data page, containing tables for conversion book page->djvu page and reverse, using mw.loadData() function for max efficiency (often there are dozens, sometimes hundreds of links to pages into a single book page; ity happens into analytical indexes, glossaries and so on).
The Lua data page is written in a eyeblink by a js script, which reads and parses html of Index: page, produced by pagelist tag; so it's very comfortable to fill such data page and to update it, if pagelist parameters are updated.
Is this "beginner's Lua exercise" someway interesting/inspiring in your opinion?
Alex brollo
Hi!
It looks very good! In my opinion the template will be more interesting (but also more heavy) if you work directly with the Index page by loading its content (local page = mw.title.new( 'Index:XXX.djvu' ); local text = page:getContent()) and parsing the <pagelist> tag that will give you the displayed number <-> number in the djvu conversion.
Thomas
Date: Fri, 31 May 2013 16:10:47 +0200 From: alex.brollo@gmail.com To: wikisource-l@lists.wikimedia.org Subject: [Wikisource-l] Playing with Lua, javascript and pagelist tag
I'm testing a template [:it:s:Template:Pg]] (calling a Lua module) that fastly converts book page numbers, passed as the unique parameter to the template, into links to right djvu pages, without any need to add delta parameters and linking any format for page number (arabic, roman, other...) since page numbers are seen as strings.
The Lua script reads a data page, containing tables for conversion book page->djvu page and reverse, using mw.loadData() function for max efficiency (often there are dozens, sometimes hundreds of links to pages into a single book page; ity happens into analytical indexes, glossaries and so on).
The Lua data page is written in a eyeblink by a js script, which reads and parses html of Index: page, produced by pagelist tag; so it's very comfortable to fill such data page and to update it, if pagelist parameters are updated.
Is this "beginner's Lua exercise" someway interesting/inspiring in your opinion? Alex brollo
_______________________________________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Thank you Thomas. There are two reasons to load "structural metadata" into a Lua data page, to be imported with mw.loadData(): 1. often the tl|Pg - which reads data page by mw.loadData() and converts a book page number/mark into a running link to djvu page - is called dozens, sometime hundreds of times by a single Page: this happens, i.e., into glossaries, indexes of people, indexes of lemmas; so, it's important IMHO that it's not server expensive; 2. I'm an absolute Lua beginner, and I'm searching for opportunities to test most interesting Lua features. ;-)
I'm using javascript to build data Lua data page since I find it comfortable for tests (much more comfortable when coupled with Chrome Shortcut Manager); data pages can be built with a python bot too, but I think that a "manual opportunity" by javascript should be saved for updates/tests and so on. And.... it's so exciting to see how code appears into a eyeblink.... :-)
I'm going to expand the data page adding ns0 names of pages and subpages and ranges of pages which they transclude (all the needed data are wrapped into Index: wiki code in it.source); so I hope that tl|Pg link will point djvu page in nsIndex and nsPage, and will point to ns0 page/subpage when transcluded into ns0.
Alex
2013/6/1 Thomas PT thomaspt@hotmail.fr
Hi!
It looks very good! In my opinion the template will be more interesting (but also more heavy) if you work directly with the Index page by loading its content (local page = mw.title.new( 'Index:XXX.djvu' ); local text = page:getContent()) and parsing the <pagelist> tag that will give you the displayed number <-> number in the djvu conversion.
Thomas
Date: Fri, 31 May 2013 16:10:47 +0200 From: alex.brollo@gmail.com To: wikisource-l@lists.wikimedia.org Subject: [Wikisource-l] Playing with Lua, javascript and pagelist tag
I'm testing a template [:it:s:Template:Pg]] (calling a Lua module) that fastly converts book page numbers, passed as the unique parameter to the template, into links to right djvu pages, without any need to add delta parameters and linking any format for page number (arabic, roman, other...) since page numbers are seen as strings.
The Lua script reads a data page, containing tables for conversion book page->djvu page and reverse, using mw.loadData() function for max efficiency (often there are dozens, sometimes hundreds of links to pages into a single book page; ity happens into analytical indexes, glossaries and so on).
The Lua data page is written in a eyeblink by a js script, which reads and parses html of Index: page, produced by pagelist tag; so it's very comfortable to fill such data page and to update it, if pagelist parameters are updated.
Is this "beginner's Lua exercise" someway interesting/inspiring in your opinion?
Alex brollo
_______________________________________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
I got it :-)
Take a look to this pagehttp://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go herehttp://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg http://it.wikisource.org/wiki/Template:Pg, Modulo:Pg http://it.wikisource.org/wiki/Modulo:Pg, and this Modulo:Dati/Manuale....http://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Scripts are rough, they don't consider any exception, they are not documented, I tested them on a couple of works only.... but they show that it can be done.
Alex
2013/6/1 Alex Brollo alex.brollo@gmail.com
Thank you Thomas. There are two reasons to load "structural metadata" into a Lua data page, to be imported with mw.loadData():
- often the tl|Pg - which reads data page by mw.loadData() and converts a
book page number/mark into a running link to djvu page - is called dozens, sometime hundreds of times by a single Page: this happens, i.e., into glossaries, indexes of people, indexes of lemmas; so, it's important IMHO that it's not server expensive; 2. I'm an absolute Lua beginner, and I'm searching for opportunities to test most interesting Lua features. ;-)
I'm using javascript to build data Lua data page since I find it comfortable for tests (much more comfortable when coupled with Chrome Shortcut Manager); data pages can be built with a python bot too, but I think that a "manual opportunity" by javascript should be saved for updates/tests and so on. And.... it's so exciting to see how code appears into a eyeblink.... :-)
I'm going to expand the data page adding ns0 names of pages and subpages and ranges of pages which they transclude (all the needed data are wrapped into Index: wiki code in it.source); so I hope that tl|Pg link will point djvu page in nsIndex and nsPage, and will point to ns0 page/subpage when transcluded into ns0.
Alex
2013/6/1 Thomas PT thomaspt@hotmail.fr
Hi!
It looks very good! In my opinion the template will be more interesting (but also more heavy) if you work directly with the Index page by loading its content (local page = mw.title.new( 'Index:XXX.djvu' ); local text = page:getContent()) and parsing the <pagelist> tag that will give you the displayed number <-> number in the djvu conversion.
Thomas
Date: Fri, 31 May 2013 16:10:47 +0200 From: alex.brollo@gmail.com To: wikisource-l@lists.wikimedia.org Subject: [Wikisource-l] Playing with Lua, javascript and pagelist tag
I'm testing a template [:it:s:Template:Pg]] (calling a Lua module) that fastly converts book page numbers, passed as the unique parameter to the template, into links to right djvu pages, without any need to add delta parameters and linking any format for page number (arabic, roman, other...) since page numbers are seen as strings.
The Lua script reads a data page, containing tables for conversion book page->djvu page and reverse, using mw.loadData() function for max efficiency (often there are dozens, sometimes hundreds of links to pages into a single book page; ity happens into analytical indexes, glossaries and so on).
The Lua data page is written in a eyeblink by a js script, which reads and parses html of Index: page, produced by pagelist tag; so it's very comfortable to fill such data page and to update it, if pagelist parameters are updated.
Is this "beginner's Lua exercise" someway interesting/inspiring in your opinion?
Alex brollo
_______________________________________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg http://it.wikisource.org/wiki/Template:Pg, Modulo:Pg http://it.wikisource.org/wiki/Modulo:Pg, and this Modulo:Dati/Manuale.... http://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
Thanks Lars: OK. :-)
2013/6/3 Lars Aronsson lars@aronsson.se
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page <http://it.wikisource.org/** wiki/Pagina:Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvu/583http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583>. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here <http://it.wikisource.org/**wiki/Manuale_di_economia_** politica_con_una_introduzione_**alla_scienza_sociale/Indice_** dei_nomi_di_autori#pagename583http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 **> where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg <http://it.wikisource.org/**wiki/Template:Pghttp://it.wikisource.org/wiki/Template:Pg>, Modulo:Pg <http://it.wikisource.org/**wiki/Modulo:Pghttp://it.wikisource.org/wiki/Modulo:Pg>, and this Modulo:Dati/Manuale.... <http://it.wikisource.org/** wiki/Modulo:Dati/Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvuhttp://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu>; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
-- Lars Aronsson (lars@aronsson.se) Project Runeberg - free Nordic literature - http://runeberg.org/
______________________________**_________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.**org Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/**mailman/listinfo/wikisource-lhttps://lists.wikimedia.org/mailman/listinfo/wikisource-l
Done Lars; I edited Modulo:Pg to keep Template:Pg simpler.
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Thanks Lars: OK. :-)
2013/6/3 Lars Aronsson lars@aronsson.se
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page <http://it.wikisource.org/** wiki/Pagina:Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvu/583http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583>. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here <http://it.wikisource.org/**wiki/Manuale_di_economia_** politica_con_una_introduzione_**alla_scienza_sociale/Indice_** dei_nomi_di_autori#pagename583http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 **> where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg <http://it.wikisource.org/** wiki/Template:Pg http://it.wikisource.org/wiki/Template:Pg>, Modulo:Pg <http://it.wikisource.org/**wiki/Modulo:Pghttp://it.wikisource.org/wiki/Modulo:Pg>, and this Modulo:Dati/Manuale.... <http://it.wikisource.org/** wiki/Modulo:Dati/Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvuhttp://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu>; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
-- Lars Aronsson (lars@aronsson.se) Project Runeberg - free Nordic literature - http://runeberg.org/
______________________________**_________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.**org Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/**mailman/listinfo/wikisource-lhttps://lists.wikimedia.org/mailman/listinfo/wikisource-l
I'm testing tl|Pg into a "hard case", transcluded page will contain more than 8000 Pg template and Pg module calls; no expensive parser function is added by Lua.
This is the page I'm working about: http://it.wikisource.org/wiki/Annali_del_Principato_ecclesiastico_di_Trento_...
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Done Lars; I edited Modulo:Pg to keep Template:Pg simpler.
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Thanks Lars: OK. :-)
2013/6/3 Lars Aronsson lars@aronsson.se
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page <http://it.wikisource.org/** wiki/Pagina:Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvu/583http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583>. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here <http://it.wikisource.org/**wiki/Manuale_di_economia_** politica_con_una_introduzione_**alla_scienza_sociale/Indice_** dei_nomi_di_autori#pagename583http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 **> where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg <http://it.wikisource.org/** wiki/Template:Pg http://it.wikisource.org/wiki/Template:Pg>, Modulo:Pg <http://it.wikisource.org/**wiki/Modulo:Pghttp://it.wikisource.org/wiki/Modulo:Pg>, and this Modulo:Dati/Manuale.... <http://it.wikisource.org/** wiki/Modulo:Dati/Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvuhttp://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu>; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
-- Lars Aronsson (lars@aronsson.se) Project Runeberg - free Nordic literature - http://runeberg.org/
______________________________**_________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.**org Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/**mailman/listinfo/wikisource-lhttps://lists.wikimedia.org/mailman/listinfo/wikisource-l
I found a limit at approx 4000 calls of template|Pg; so I splitted the whole page http://it.wikisource.org/wiki/Annali_del_Principato_ecclesiastico_di_Trento_... into smaller subpages A....Z and now the script runs happily.
In the meantime, I'm translating js script which builds Modulo:Dati/.... in python, so that Alebot (running into TS willow and listening #it.wikisource) could write/update Modulo:Dati/... page as soon as Index:page has been edited. Index page edits are not so frequent IMHO, they are relatively "stable" pages. Global number of source edits would not been flooded IMHO by such additive edits of Modulo:Dati subpages.
I'm thinking too about loading other useful book-related data which could be uploaded into Modulo:Dati to be used into any nsPage and ns0 page related with rootIndex: page. I'm dreaming about an almost integral automation of ns0 transclusion.
But I absolutely need Aubrey's suggestions..... :-)
Alex
2013/6/5 Alex Brollo alex.brollo@gmail.com
I'm testing tl|Pg into a "hard case", transcluded page will contain more than 8000 Pg template and Pg module calls; no expensive parser function is added by Lua.
This is the page I'm working about: http://it.wikisource.org/wiki/Annali_del_Principato_ecclesiastico_di_Trento_...
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Done Lars; I edited Modulo:Pg to keep Template:Pg simpler.
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Thanks Lars: OK. :-)
2013/6/3 Lars Aronsson lars@aronsson.se
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page <http://it.wikisource.org/** wiki/Pagina:Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvu/583http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583>. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here <http://it.wikisource.org/**wiki/Manuale_di_economia_** politica_con_una_introduzione_**alla_scienza_sociale/Indice_** dei_nomi_di_autori#pagename583http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 **> where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg <http://it.wikisource.org/** wiki/Template:Pg http://it.wikisource.org/wiki/Template:Pg>, Modulo:Pg <http://it.wikisource.org/**wiki/Modulo:Pghttp://it.wikisource.org/wiki/Modulo:Pg>, and this Modulo:Dati/Manuale.... <http://it.wikisource.org/** wiki/Modulo:Dati/Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvuhttp://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu>; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
-- Lars Aronsson (lars@aronsson.se) Project Runeberg - free Nordic literature - http://runeberg.org/
______________________________**_________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.**org Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/**mailman/listinfo/wikisource-lhttps://lists.wikimedia.org/mailman/listinfo/wikisource-l
@Alex: one thing I always would have liked to have is an /all page that would display all the ocr text and that would replace it by proofread text as soon as each page would be created. It would be the equivalent of the text dump by internet archive and it would allow search engines to find text in the books even if they haven't been proofread yet. Automatic creation of page transclusion is nice but also dangerous... too many structures to have an easy solution. It could be easier to have a parser that would read the proofread pages, detect "section" marks and create pages accordingly. IIRC, Phe did already something like that, perhaps you could port it to Lua?
By the way, do you think it would make sense that the Modulo:Dati is generated automatically by the Index: page? Maybe each time that it is saved?
Micru
On Thu, Jun 6, 2013 at 6:30 PM, Alex Brollo alex.brollo@gmail.com wrote:
I found a limit at approx 4000 calls of template|Pg; so I splitted the whole page http://it.wikisource.org/wiki/Annali_del_Principato_ecclesiastico_di_Trento_... into smaller subpages A....Z and now the script runs happily.
In the meantime, I'm translating js script which builds Modulo:Dati/.... in python, so that Alebot (running into TS willow and listening #it.wikisource) could write/update Modulo:Dati/... page as soon as Index:page has been edited. Index page edits are not so frequent IMHO, they are relatively "stable" pages. Global number of source edits would not been flooded IMHO by such additive edits of Modulo:Dati subpages.
I'm thinking too about loading other useful book-related data which could be uploaded into Modulo:Dati to be used into any nsPage and ns0 page related with rootIndex: page. I'm dreaming about an almost integral automation of ns0 transclusion.
But I absolutely need Aubrey's suggestions..... :-)
Alex
2013/6/5 Alex Brollo alex.brollo@gmail.com
I'm testing tl|Pg into a "hard case", transcluded page will contain more than 8000 Pg template and Pg module calls; no expensive parser function is added by Lua.
This is the page I'm working about: http://it.wikisource.org/wiki/Annali_del_Principato_ecclesiastico_di_Trento_...
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Done Lars; I edited Modulo:Pg to keep Template:Pg simpler.
Alex
2013/6/3 Alex Brollo alex.brollo@gmail.com
Thanks Lars: OK. :-)
2013/6/3 Lars Aronsson lars@aronsson.se
On 06/03/2013 12:10 AM, Alex Brollo wrote:
Take a look to this page <http://it.wikisource.org/** wiki/Pagina:Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvu/583http://it.wikisource.org/wiki/Pagina:Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu/583>. As you see, page numbers are simply wrapped into a tl|Pg. No other parameters. Nevertheless, they are transformed into active links do right djvu pages.
Now, go here <http://it.wikisource.org/**wiki/Manuale_di_economia_** politica_con_una_introduzione_**alla_scienza_sociale/Indice_** dei_nomi_di_autori#pagename583http://it.wikisource.org/wiki/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale/Indice_dei_nomi_di_autori#pagename583 **> where the same page is transcluded in ns0. Page numbers now are active links to the right subpage/chapter.
This trick uses: Template:Pg <http://it.wikisource.org/** wiki/Template:Pg http://it.wikisource.org/wiki/Template:Pg>, Modulo:Pg <http://it.wikisource.org/**wiki/Modulo:Pghttp://it.wikisource.org/wiki/Modulo:Pg>, and this Modulo:Dati/Manuale.... <http://it.wikisource.org/** wiki/Modulo:Dati/Manuale_di_**economia_politica_con_una_** introduzione_alla_scienza_**sociale.djvuhttp://it.wikisource.org/wiki/Modulo:Dati/Manuale_di_economia_politica_con_una_introduzione_alla_scienza_sociale.djvu>; nothing more than this, no javascript and no AJAX, so I presume (I didn't test by now) that links are running too into any html-based export of ns0 pages as server builds them.
Excellent! However, on the transcluded page, the link goes only to the chapter URL. It should also add #pag137 or #pagename147, since all page links in the margin of the transcluded chapter are marked like this:... id="pag137"><span id="pagename147">...
-- Lars Aronsson (lars@aronsson.se) Project Runeberg - free Nordic literature - http://runeberg.org/
______________________________**_________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.**org Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/**mailman/listinfo/wikisource-lhttps://lists.wikimedia.org/mailman/listinfo/wikisource-l
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous... too many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when you work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could suggest users to focus just on nsPage and Indexes. All the difficult transclusion part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user: * define the table of content once for all in the Index page * define the table of content once in the book Toc (there is often one, if not always, when needed) * define the table of content just putting templates thorough the book, as the reader goes through the book.
What do you all think?
Aubrey
@Micru: I like semi-automated, user-driven procedures much more than bot-automated procedures. I.e.: our it.source Index pages use a tl|Sommario for any ns0 section, with following parameters: nome= full name of ns0 page titolo= title for ns0 page (to be displayed only) from= number of djvu page where ns0 begins delta= delta between djvu page and book page
These data are almost sufficient (with some ajax) to self-compililing all the needed code when creating a new ns0 subpage, compiling both our header template (tl|IncludiIntestazione) and pages tag. Needed to= parameter of pages tag is calculated (it is simply the number of djvu page of following tl|Sommario -1) and navigation templates are calculated from preceeding and following tl|Sommario. But creation of the page is not fully automated: ns0auto() script is invoked by the user in edit mode, so that the user can verify and test resulting code and fix it if needed.
What I'm thinking about is, to so something similar by a Lua script, reading Modulo:Dati which is aligned with Index: page data (yes, I presume by a #irc bot and I presently working to write the needed code). So, this procedure will be semi-automated, since page creation would be a user-driven procedure.
Alex
2013/6/7 Andrea Zanni zanni.andrea84@gmail.com
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous... too many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when you work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could suggest users to focus just on nsPage and Indexes. All the difficult transclusion part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one, if
not always, when needed)
- define the table of content just putting templates thorough the book, as
the reader goes through the book.
What do you all think?
Aubrey
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
I have been wondering the same thing for years. When I upload prints to Wikimedia Commons, I am generally in a hurry and just use the default uploader to get it out there. Weeks or months or sometimes years later I will add in the detailed metadata like the book it was first published in, alternate sources for the print from the one I used, the publisher if that is a different person than the artist, etc. What I don't bother with is page numbers, because this is often unknown and changes from edition to edition. You can get around this problem by naming specific editions held in specific libraries with specific page numbers, which I have done occasionally. Some prints are so well known they go by their own titles, and the Wikimedia Commons artwork template even has a field "Original title" to deal with this issue.
When you go through an index of plates in any older book, generally there are some mistakes, such as blank pages that are indexed because the plate didn't make it to the printer, some plates the printer added that didn't make it into the index, and of course the really confusing one, the prints that a reprinter added that neither the original author nor the original publisher ever saw.
One reason I have not spent much time on Wikisource is because I feel I have to decide up front what the structure of the book will be with page numbering (which sometimes does not count the plates), so I need to base this on the original index or original list of chapters. Sometimes a book becomes famous just for one passage, and that passage may not even be indexed in the original version. How do you add these links? On Wikimedia Commons you can keep on adding values to fields, and change the "Information" template to "Artwork" to get more fields. You can even add annotations to files and then put links to other files in the annotations, so that through the "Global usage" property you can see where such prints have been "quoted" or re-used. How do you do this with books?
I would like to see a flexible way to set this up that makes it easy to come back and make corrections or additions to the published information in both indexes and ToC's based on later discovery. This book of prints for example shows a page order based on one edition that was reproduced in facsimile version, but other versions exist with different plates: http://commons.wikimedia.org/wiki/Category:32_afbeeldinge_der_Graven_van_HOL... How do you set up page numbers for this, because there weren't any to start with? Jane
2013/6/7, Andrea Zanni zanni.andrea84@gmail.com:
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous... too many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when you work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could suggest users to focus just on nsPage and Indexes. All the difficult transclusion part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one, if
not always, when needed)
- define the table of content just putting templates thorough the book, as
the reader goes through the book.
What do you all think?
Aubrey
Thanks for suggestions.... I can only promise, I'll think about them. The question by Micru is particularly hard. :-(
@ Jane: I've to read your mail again and again; nevertheless a well compiled pagelist tag can really identify into a unique way any page of the book, even if they have no page number, and tl|Pg manages djvu page/book page relationship easily even if book page is identified by something like "Figure 1, Figure 2....". I'll take a look at your book.
Alex
2013/6/7 Jane Darnell jane023@gmail.com
I have been wondering the same thing for years. When I upload prints to Wikimedia Commons, I am generally in a hurry and just use the default uploader to get it out there. Weeks or months or sometimes years later I will add in the detailed metadata like the book it was first published in, alternate sources for the print from the one I used, the publisher if that is a different person than the artist, etc. What I don't bother with is page numbers, because this is often unknown and changes from edition to edition. You can get around this problem by naming specific editions held in specific libraries with specific page numbers, which I have done occasionally. Some prints are so well known they go by their own titles, and the Wikimedia Commons artwork template even has a field "Original title" to deal with this issue.
When you go through an index of plates in any older book, generally there are some mistakes, such as blank pages that are indexed because the plate didn't make it to the printer, some plates the printer added that didn't make it into the index, and of course the really confusing one, the prints that a reprinter added that neither the original author nor the original publisher ever saw.
One reason I have not spent much time on Wikisource is because I feel I have to decide up front what the structure of the book will be with page numbering (which sometimes does not count the plates), so I need to base this on the original index or original list of chapters. Sometimes a book becomes famous just for one passage, and that passage may not even be indexed in the original version. How do you add these links? On Wikimedia Commons you can keep on adding values to fields, and change the "Information" template to "Artwork" to get more fields. You can even add annotations to files and then put links to other files in the annotations, so that through the "Global usage" property you can see where such prints have been "quoted" or re-used. How do you do this with books?
I would like to see a flexible way to set this up that makes it easy to come back and make corrections or additions to the published information in both indexes and ToC's based on later discovery. This book of prints for example shows a page order based on one edition that was reproduced in facsimile version, but other versions exist with different plates:
http://commons.wikimedia.org/wiki/Category:32_afbeeldinge_der_Graven_van_HOL... How do you set up page numbers for this, because there weren't any to start with? Jane
2013/6/7, Andrea Zanni zanni.andrea84@gmail.com:
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous...
too
many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when
you
work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could suggest users to focus just on nsPage and Indexes. All the difficult transclusion part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one,
if
not always, when needed)
- define the table of content just putting templates thorough the book,
as
the reader goes through the book.
What do you all think?
Aubrey
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
@ Jane again: I'd better look before, and talk after.... I see a collection jpg's from scans, not a djvu file or a Index: page into a wikisource project. :-)
So I presume that I can't find any pagelist tag..... :-)
Did you personally scan those pages? Did you scan all the pages of the book (if it's a book...)? Do you know if any complete scan of the book has been published previously (into Internet Archive, Google books, or other digital libraries)? Next time if you scan or take pictures to all the pages of a book, and load all the images to Commons, some willing wikisourcian could mount them into a multipage djvu file and to open an Index: page to proofread it into a wikisource project.
2013/6/7 Alex Brollo alex.brollo@gmail.com
Thanks for suggestions.... I can only promise, I'll think about them. The question by Micru is particularly hard. :-(
@ Jane: I've to read your mail again and again; nevertheless a well compiled pagelist tag can really identify into a unique way any page of the book, even if they have no page number, and tl|Pg manages djvu page/book page relationship easily even if book page is identified by something like "Figure 1, Figure 2....". I'll take a look at your book.
Alex
2013/6/7 Jane Darnell jane023@gmail.com
I have been wondering the same thing for years. When I upload prints to Wikimedia Commons, I am generally in a hurry and just use the default uploader to get it out there. Weeks or months or sometimes years later I will add in the detailed metadata like the book it was first published in, alternate sources for the print from the one I used, the publisher if that is a different person than the artist, etc. What I don't bother with is page numbers, because this is often unknown and changes from edition to edition. You can get around this problem by naming specific editions held in specific libraries with specific page numbers, which I have done occasionally. Some prints are so well known they go by their own titles, and the Wikimedia Commons artwork template even has a field "Original title" to deal with this issue.
When you go through an index of plates in any older book, generally there are some mistakes, such as blank pages that are indexed because the plate didn't make it to the printer, some plates the printer added that didn't make it into the index, and of course the really confusing one, the prints that a reprinter added that neither the original author nor the original publisher ever saw.
One reason I have not spent much time on Wikisource is because I feel I have to decide up front what the structure of the book will be with page numbering (which sometimes does not count the plates), so I need to base this on the original index or original list of chapters. Sometimes a book becomes famous just for one passage, and that passage may not even be indexed in the original version. How do you add these links? On Wikimedia Commons you can keep on adding values to fields, and change the "Information" template to "Artwork" to get more fields. You can even add annotations to files and then put links to other files in the annotations, so that through the "Global usage" property you can see where such prints have been "quoted" or re-used. How do you do this with books?
I would like to see a flexible way to set this up that makes it easy to come back and make corrections or additions to the published information in both indexes and ToC's based on later discovery. This book of prints for example shows a page order based on one edition that was reproduced in facsimile version, but other versions exist with different plates:
http://commons.wikimedia.org/wiki/Category:32_afbeeldinge_der_Graven_van_HOL... How do you set up page numbers for this, because there weren't any to start with? Jane
2013/6/7, Andrea Zanni zanni.andrea84@gmail.com:
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous...
too
many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when
you
work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could
suggest
users to focus just on nsPage and Indexes. All the difficult
transclusion
part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one,
if
not always, when needed)
- define the table of content just putting templates thorough the book,
as
the reader goes through the book.
What do you all think?
Aubrey
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Hi Alex, No the book I want to do is still on my computer. That link you looked at is a book that was printed about 20 years ago as a facsimile of one printed in the 17th century. It does happen to be a book that has been massively reused, and it's not even the original!
The book is just plates, and the only text on them is in the description section of the files. I don't see the point of having it on Wikisource because you can use it more easily on Commons (and each page can be linked to any language-pedia.
I will try to upload the part of the book I mean so you can take a look at my specific problem. It's 3 volumes but there is a section that is particularly problematic.
Jane
2013/6/7, Alex Brollo alex.brollo@gmail.com:
@ Jane again: I'd better look before, and talk after.... I see a collection jpg's from scans, not a djvu file or a Index: page into a wikisource project. :-)
So I presume that I can't find any pagelist tag..... :-)
Did you personally scan those pages? Did you scan all the pages of the book (if it's a book...)? Do you know if any complete scan of the book has been published previously (into Internet Archive, Google books, or other digital libraries)? Next time if you scan or take pictures to all the pages of a book, and load all the images to Commons, some willing wikisourcian could mount them into a multipage djvu file and to open an Index: page to proofread it into a wikisource project.
2013/6/7 Alex Brollo alex.brollo@gmail.com
Thanks for suggestions.... I can only promise, I'll think about them. The question by Micru is particularly hard. :-(
@ Jane: I've to read your mail again and again; nevertheless a well compiled pagelist tag can really identify into a unique way any page of the book, even if they have no page number, and tl|Pg manages djvu page/book page relationship easily even if book page is identified by something like "Figure 1, Figure 2....". I'll take a look at your book.
Alex
2013/6/7 Jane Darnell jane023@gmail.com
I have been wondering the same thing for years. When I upload prints to Wikimedia Commons, I am generally in a hurry and just use the default uploader to get it out there. Weeks or months or sometimes years later I will add in the detailed metadata like the book it was first published in, alternate sources for the print from the one I used, the publisher if that is a different person than the artist, etc. What I don't bother with is page numbers, because this is often unknown and changes from edition to edition. You can get around this problem by naming specific editions held in specific libraries with specific page numbers, which I have done occasionally. Some prints are so well known they go by their own titles, and the Wikimedia Commons artwork template even has a field "Original title" to deal with this issue.
When you go through an index of plates in any older book, generally there are some mistakes, such as blank pages that are indexed because the plate didn't make it to the printer, some plates the printer added that didn't make it into the index, and of course the really confusing one, the prints that a reprinter added that neither the original author nor the original publisher ever saw.
One reason I have not spent much time on Wikisource is because I feel I have to decide up front what the structure of the book will be with page numbering (which sometimes does not count the plates), so I need to base this on the original index or original list of chapters. Sometimes a book becomes famous just for one passage, and that passage may not even be indexed in the original version. How do you add these links? On Wikimedia Commons you can keep on adding values to fields, and change the "Information" template to "Artwork" to get more fields. You can even add annotations to files and then put links to other files in the annotations, so that through the "Global usage" property you can see where such prints have been "quoted" or re-used. How do you do this with books?
I would like to see a flexible way to set this up that makes it easy to come back and make corrections or additions to the published information in both indexes and ToC's based on later discovery. This book of prints for example shows a page order based on one edition that was reproduced in facsimile version, but other versions exist with different plates:
http://commons.wikimedia.org/wiki/Category:32_afbeeldinge_der_Graven_van_HOL... How do you set up page numbers for this, because there weren't any to start with? Jane
2013/6/7, Andrea Zanni zanni.andrea84@gmail.com:
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous...
too
many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when
you
work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could
suggest
users to focus just on nsPage and Indexes. All the difficult
transclusion
part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often
one,
if
not always, when needed)
- define the table of content just putting templates thorough the
book,
as
the reader goes through the book.
What do you all think?
Aubrey
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
On Fri, Jun 7, 2013 at 9:40 AM, Jane Darnell jane023@gmail.com wrote:
Sometimes a book becomes famous just for one passage, and that passage may not even be indexed in the original version. How do you add these links? On Wikimedia Commons you can keep on adding values to fields, and change the "Information" template to "Artwork" to get more fields. You can even add annotations to files and then put links to other files in the annotations, so that through the "Global usage" property you can see where such prints have been "quoted" or re-used. How do you do this with books?
I think this is complex and should be addressed with a new vision of what quotes, citations and annotations are, and how to exploit and manage them better. Right now, we don't have the (software) infrastructure to do that. If we could integrate textus or other annotation tool with MediaWiki, probably things could change.
Aubrey
I don't agree that it should be fully automated by any stretch of the imagination. I can see that it is an option that some may wish to use, but I dislike the limitations, and do not see it working as the only means to use.
Regards, Billinghurst
On Fri, 7 Jun 2013 08:52:12 +0200, Andrea Zanni zanni.andrea84@gmail.com wrote:
On Fri, Jun 7, 2013 at 1:36 AM, David Cuenca dacuetu@gmail.com wrote:
Automatic creation of page transclusion is nice but also dangerous...
too
many structures to have an easy solution.
What Alex is thinking, if I understand his work correctly, is that when
you
work on a new book in nsPage, you "define" what the structure is (his work right now is wrapping titles/chapters in {{title}} templates, to give the book a logic structure), and then a bot runs, create corrispondent ns0 chapters and transclude pages.
I think that ns0 automation is something long needed, as we could
suggest
users to focus just on nsPage and Indexes. All the difficult
transclusion
part would be automatic (or semi-automatic).
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one,
if
not always, when needed)
- define the table of content just putting templates thorough the book,
as
the reader goes through the book.
What do you all think?
Aubrey
On Fri, Jun 7, 2013 at 2:01 PM, billinghurst billinghurst@gmail.com wrote:
I don't agree that it should be fully automated by any stretch of the imagination. I can see that it is an option that some may wish to use, but I dislike the limitations, and do not see it working as the only means to use.
Well, I think that the bot/Lua/extension/whatever could show to the user a window with the structure of the book as it would be created. The user could change some things, or Cancel the automatic transclusion.
I don't know: to me, avoiding the burden of taking care of different namesapces, with diffirent templates needed and strange tags (eg pagelist) should be an aim for us, if we want the layman to understand Wikisource and contribute. I myself can't upload and create a whole book from the scan to the ns0 transclusion without mistakes or forgetting important things. Wikisource *is* difficult, Too much, IMHO.
Aubrey
On Fri, 7 Jun 2013 14:44:06 +0200, Andrea Zanni zanni.andrea84@gmail.com wrote:
On Fri, Jun 7, 2013 at 2:01 PM, billinghurst billinghurst@gmail.com wrote:
I don't agree that it should be fully automated by any stretch of the imagination. I can see that it is an option that some may wish to use, but I dislike the limitations, and do not see it working as the only means
to
use.
Well, I think that the bot/Lua/extension/whatever could show to the user
a
window with the structure of the book as it would be created. The user could change some things, or Cancel the automatic transclusion.
I don't know: to me, avoiding the burden of taking care of different namesapces, with diffirent templates needed and strange tags (eg
pagelist)
should be an aim for us, if we want the layman to understand Wikisource
and
contribute. I myself can't upload and create a whole book from the scan to the ns0 transclusion without mistakes or forgetting important things. Wikisource *is* difficult, Too much, IMHO.
Aubrey
I am not saying that it isn't part of the choice, I am just saying that it should not be enforced. I am explaining choice, not commenting on the development of the proposed tool and its availability. At a point of time, I may use it. Of course you make mistakes, we all do, and they are not just in the <pages> stuff. I make more mistakes in Page: ns than I do in main. I see mistakes in the published books, including mistakes in ToC. Humans while they make mistakes, also are able to error resolve.
English Wikisource has more components in its headers, and is able to adapt its {{header}} components more dynamically. Having the ability to tweak enables presentation to how it suits a work, and its readibility; to this point of time, I find the automated process too restricting.
Regards, Billinghurst
I'm going ahead with Modulo:Pg and Modulo:Dati/... idea. A test p.indice function into :it:s:Modulo:Dati now builds a nice list of links to subpages into any supported ns0 page - simply browsing one of Modulo/Dati:.... tables and selecting cap objects which are subpages of the calling page and adding a little bit of wiki code.
But I found a problem while trying to emulate fully our it.source, powerful tl|Testo, currently used to build links to subpages. A transclusion call by #lst tag is not considered "server expensive", while a Lua emulation of a transclusion call by mw.title.new() (to retrieve its content and select text into a section) is a server expensive function. Is there some other method to emulate #lst in Lua using not expensive server functions?
Alex
2013/6/8 billinghurst billinghurst@gmail.com
On Fri, 7 Jun 2013 14:44:06 +0200, Andrea Zanni zanni.andrea84@gmail.com wrote:
On Fri, Jun 7, 2013 at 2:01 PM, billinghurst billinghurst@gmail.com wrote:
I don't agree that it should be fully automated by any stretch of the imagination. I can see that it is an option that some may wish to use, but I dislike the limitations, and do not see it working as the only means
to
use.
Well, I think that the bot/Lua/extension/whatever could show to the user
a
window with the structure of the book as it would be created. The user could change some things, or Cancel the automatic transclusion.
I don't know: to me, avoiding the burden of taking care of different namesapces, with diffirent templates needed and strange tags (eg
pagelist)
should be an aim for us, if we want the layman to understand Wikisource
and
contribute. I myself can't upload and create a whole book from the scan to the ns0 transclusion without mistakes or forgetting important things. Wikisource *is* difficult, Too much, IMHO.
Aubrey
I am not saying that it isn't part of the choice, I am just saying that it should not be enforced. I am explaining choice, not commenting on the development of the proposed tool and its availability. At a point of time, I may use it. Of course you make mistakes, we all do, and they are not just in the <pages> stuff. I make more mistakes in Page: ns than I do in main. I see mistakes in the published books, including mistakes in ToC. Humans while they make mistakes, also are able to error resolve.
English Wikisource has more components in its headers, and is able to adapt its {{header}} components more dynamically. Having the ability to tweak enables presentation to how it suits a work, and its readibility; to this point of time, I find the automated process too restricting.
Regards, Billinghurst
Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
Strategy 3 is extremely interesting, from the "if you are repeating yourself, you are going wrong" point of view. Nevertheless such an approach needs a well-designed and complete "chapters tree" as step one; the best would be that titles of sections/subsections could be wrote one only to avoid any possible mistake.
A simple Excel page (or similar) would be probably the simplest way to produce the template code to be seeded into pages containing the start of chapters and ignoring other pages.
With very few simple conventions, anything - but pagelist - could be automatized, if a thorough seeding of such templates could be done as proofreading step 1, since a script could add too needed section tags, then build pages tags and fill chapter list into Index page, and all the needed code for ns0 pages/subpages.
2013/6/7 Andrea Zanni zanni.andrea84@gmail.com
.....
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one, if
not always, when needed)
- define the table of content just putting templates thorough the book, as
the reader goes through the book.
What do you all think?
Aubrey
First version of a python script which builds a Modulo:Dati page/... parsing Index and produces a running Lua script (first test using it:Indice:Georgiche.djvu producing Modulo:Dati/Georgiche.djvu). This means that Modulo:Dati/... pages could be built and updated both automagically (by an #irc bot) and manually (by javascript, or by a interactive bot).
Alex
2013/6/7 Alex Brollo alex.brollo@gmail.com
Strategy 3 is extremely interesting, from the "if you are repeating yourself, you are going wrong" point of view. Nevertheless such an approach needs a well-designed and complete "chapters tree" as step one; the best would be that titles of sections/subsections could be wrote one only to avoid any possible mistake.
A simple Excel page (or similar) would be probably the simplest way to produce the template code to be seeded into pages containing the start of chapters and ignoring other pages.
With very few simple conventions, anything - but pagelist - could be automatized, if a thorough seeding of such templates could be done as proofreading step 1, since a script could add too needed section tags, then build pages tags and fill chapter list into Index page, and all the needed code for ns0 pages/subpages.
2013/6/7 Andrea Zanni zanni.andrea84@gmail.com
.....
I wonder if there is a better way to define the logic structure of our book, maybe directly in the Index page. I don't know what would be easier for the user:
- define the table of content once for all in the Index page
- define the table of content once in the book Toc (there is often one,
if not always, when needed)
- define the table of content just putting templates thorough the book,
as the reader goes through the book.
What do you all think?
Aubrey
wikisource-l@lists.wikimedia.org