[sword-devel] French Bible Segond 1910 with Strong's number

Wed Nov 6 08:32:20 EST 2024

Hi all,

I support Cyrille's suggestion of using wikisource as the main source for
public domain bibles when we have a good facsimile available.

For those not familiar with how wikisource works, the idea is that one can
upload a facsimile (in djvu format), and then users can transcribe it into
text and correct it page by page, making it easy to compare and potentially
fix.
See for example this view, on one specific page of the Segond 1910, with
the facsimile on the right and the transcribed text on the left :
https://fr.wikisource.org/wiki/Page:La_Sainte_Bible,_trad._Segond,_ed._1910.djvu/8
Each page can be edited, then reviewed, corrected, and so on - and the wiki
setup makes it easy to collaborate.
There is even the possibility with dedicated markup to annotate corrections
to the original text : when the original text contains an obvious typo that
you want to fix, you can edit your fix in while preserving the original
text with specific markup, so that the document still keeps track of the
original form and what was corrected.

Then, proper markup while transcribing the text allows wikisource to group
it as proper books and chapters, format the text, integrate notes, etc.
See the full bible index here :
https://fr.wikisource.org/wiki/Livre:La_Sainte_Bible,_trad._Segond,_ed._1910.djvu
With the index page by page on the left, and the reconstructed text index
by book on the right.

And then, as long as you ensure that your markup is consistent throughout
your transcription of all pages of the same text (eg. titles of the same
level are transcribed with the same markup, etc.), it is easy enough with
my bible scraper to fetch all the books from wikisource and convert them
into USFM or OSIS - as nicely advertised by Cyrille and Matěj above.
Which means that if anyone brought some corrections to the wikisource text
over time, you can just re-run the scraper to get an updated USFM/OSIS
document with these corrections - you don't need to maintain your own
document separately.

The only question is how you intend to handle Strong's numbers, since they
are not present in the original facsimile (at least not the one present on
wikisource), and therefore not stored in wikisource.
This means my scraper can only output a document without Strong's numbers,
and you'd have to inject them in post processing with another tool - but
from your message above it seems you already have what you need for this.

If you want to try going into this direction, I can easily write a small
scraper configuration file for the Segond version currently available on
wikisource so that you can try it out.

Regards,

Arnaud

Le mer. 6 nov. 2024 à 09:27, Fr Cyrille <fr.cyrille at tiberiade.be> a écrit :

>
>
> Le 06/11/2024 à 00:48, yvand a écrit :
>
> Thank you all for your answers!
>
> I think if we can get the text (which is in public domain) from a
> well-known bible editor, it is interesting and easier. Currently the
> version Segond 1910 exists in different versions (differences in
> notes/titles, maybe also in accents Ésaü/Esaü ?) and it is interesting to
> know to what it corresponds. We can still share typos upstream or/and use a
> patch to fix typos, and update wikisource text.
>
> If wikisource is the source, we need to know if and what paper edition it
> matches exactly. If the reader can refer to a paper edition, it is
> interesting. From where will come the text and notes in wikisource? Which
> choices for notes/titles?
>
> Personally, I don't know, but you can check for yourself if the facsimile
> <https://fr.wikisource.org/wiki/Fichier:La_Sainte_Bible,_trad._Segond,_ed._1910.djvu>
> is correct. I know that a contributor has searched long and hard for this
> facsimile. I sincerely think that wikisource is the best place. We no
> longer need to ask anyone for the right to use the text (even in the public
> domain). What's more, Matej beat me to it: Arnaud's tool is absolutely
> incredible in its ability to convert to usfm or osis. I invite you to try
> it out. What's more, Arnaud is very responsive and saved me a lot of work.
> I now have 3 bibles in French on wikisource, the Glaire, the Giguet and the
> Crampon.
>
> Now the text on wikisource only has the titles for some books. It remains
> to add the notes/crossreferences. But if you have a text with notes, it can
> be quite easy to download the wiki text and add the notes.
>
>
> https://fr.wikisource.org/wiki/Livre:La_Sainte_Bible,_trad._Segond,_ed._1910.djvu
>
> Personally, I found domcox's proposition interesting.
>
>
> I connected to DBL and found 2 different projects :
>
> - "Français Louis Segond 1910" (last archive 2020-12-15)
>
> - "Louis Segond 1910 Original Corrected (Public Domain version)" (last
> archive 2024-08-01) ("La Sainte Bible par Louis Segond 1910. Société
> Biblique britannique et étrangere. Texte historique, références corrigées
> en 2019." "Ce qu'il y a de nouveau dans l'édition actuelle, ce sont surtout
> les parallèles. Là où ces parallèles rendaient superflues les notes de M.
> Segond, celles-ci ont été supprimées. De meme, les notes qui exprimaient
> une opinion théologique, ou qui présentaient un commentairs sur le texte,
> ont été laissés de cote, conformement aux statuts de la Société, qui
> excluent de ses éditions de Écritures des notes de ce genre.")
>
> Both are published by British & Foreign Bible Society
>
> But the access seem not public. As this Bible is in public domain, I guess
> they will share with us if we ask.
>
>
> Maybe Michael can help us to how the source of his module and how he did
> it? Is it based on DBL? If so which project and what tool do you use? from
> USX or USFM export?
>
>
> For versification, I agree with domcox's remark, we can keep the original
> versification. I can convert from our versification (WLC for OT / NA for
> NT) to the segond versification with a script.
>
>
> If we go that way, I guess we need to do those steps:
>
> 1. Access DBL, export data (text, notes, titles if present) in USX or USFM
> from Paratext
>
> 2. Convert to OSIS (better for sword and required because USX and USFM do
> not support Strong's number, isn't it ?)
>
> 3. Add Strong's numbers (= add <w> tags on words) : for this, I can work
> to match exactly the text from the previous OSIS in our database, and
> develop a tool to add the last version of our Strong's numbers on it.
>
> 4. Update sword module
>
> The last two steps (3. and 4.) will be regularly relaunched, because we do
> corrections on Strong's numbers regurlarly. It is still a work in progress,
> but I think the work is already not so not too bad.
>
>
> @Fr Cyrille : We need to exchange by email (not here) if and how we can
> help you to add Strong's numbers on your néo-crampon version. Our initial
> assignment tool (for Strong's numbers) has gathered dust and requires
> french lemmatization and proper versification versions with the WLC or NA
> to be effective. Can you send me a mail in French, with more details (what
> is this new version? is it close to other French versions other than
> Crampon? what is the actual format you use? which versification?…).
>
>
> yvand
>
>
>
> Le 05/11/2024 à 19:38, domcox a écrit :
>
> yvand <yvand.sword at gmail.com> <yvand.sword at gmail.com> writes:
>
>
> For Segond 1910, the text is in public domain. I used mod2imp to get the
> text from the FreSegond (Crosswire) module months
> ago. The only changes we made are about versification. Indeed all our
> texts use the same versification (based on WLC, for OT =
> German in sword), so the text matches the original (hebrew WLC or greek
> MT/WH) one and we can easily compare
> translations.
>
>
> As lot of people still read the printed French Segond 1910 or derivatives,
> I think it's very important to match the printed
> versification. Regular readers generally view Bible software with not
> matching versifications as second-rate software.
> We have a matching Bible Segond versification, let's keep it that way.
>
>
> In the past, the FreLSG module had Strong's numbers. If I am correct (see
> discussion "Strong numbers for Louis Segond" in
> 2013/07), we do not have the authorization (but still available in STEP)
> so this module has been removed or replaced by
> FreSegond module (renamed ?).
>
>
> In the past, the FreLSG had Strong's numbers. It was upgraded later to a
> new version that was horrible, filled with errors
> and even containing deuteros that never existed in the Segond Bible.
>
> This is why I made the FreSegond module, it was not an easy task. The
> Segond Bible is everywhere on the Web, but it looks like
> all sites use the same source, and it's not accurate.
>
>
> The fraLSG1910eb (F10) module is a more recent one, available in e-bible
> repo with headings, WoC and Strong's numbers. It
> seems to be a work in progress for Strong's numbers. They are not
> available for all books. I compare Strong's numbers
> affectations for the first verses of Genesis 1 (see attached screenshot)
> but there are errors in this module. On the left this
> module, on the right our data (https://concordance.bible/Sg1910/Gen/1/).
>
>
> If I remember well, Michael have access to the DBL, so his module is
> certainly based on the official Public Domain version,
> edited by the Alliance Biblique Française (
> https://lire.la-bible.net/bible/LSG/GEN.1)
>
>
> So I contact all of you (especially actual maintainers of Segond 1910
> modules and French users) what would be the
> best way of giving our Strong's numbers for Segond 1910 in a sword module.
> I don't know who is concerned or interested
> that's why I send to the mailing list.
>
> As there is already many existing Segond 1910 sword module, it guess it
> would not be a good idea of creating a new one. On
> one side, as our text is based from FreSegond module, it might be a good
> idea of updating this module with Strong's numbers
> (the only difference is versification). On the other hand, the
> fraLSG1910eb module offers Strong's numbers but our data seems
> more accurate and complete, so it could be interesting of updating this
> module. In any case, it would be important to submit
> updates regularly for Strong's numbers.
>
>
> I think the best option for this module should be:
>
> 1. Have access to the DBL version
> 2. Update the current CrossWire module from that source with footnotes,
> comments and cross-references, keeping the original versification
> 3. Add Strong's numbers
>
>
>
>
> _______________________________________________
> sword-devel mailing list: sword-devel at crosswire.orghttp://crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page
>
>
> --
> Vous aimez la Bible ? Vous êtes étudiant en théologie ? Utilisez
> l'application libre Xiphos <https://xiphos.org/> ou Andbible
> <https://andbible.github.io/> et accédez aux textes sources, à des
> commentaires, des dictionnaires et beaucoup d'autres fonctionnalités... Me
> contacter pour des traductions en français.
> _______________________________________________
> sword-devel mailing list: sword-devel at crosswire.org
> http://crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://crosswire.org/pipermail/sword-devel/attachments/20241106/1175739d/attachment-0001.htm>