| Strong Number | Greek/Hebrew | Pronounciation | Description
Here is an xample:
H1 | אב | 'âb | awb | A primitive word; father in a literal and
immediate, or figurative and remote application: - chief, (fore-)
father ([-less]), X patrimony, principal. Compare names in “Abi-”|
I need help on how I can convert the files in OpenScripture GitHub
into that format
I'm happy with PHP, Python or C++ in converting
Thanks!
I can take a look at writing some Python to do that conversion, but it
will probably be over the weekend.
--
Nathan Smith
http://nathan.smithfam.info/
On Dec 2, 8:21 am, Nathan Smith <nat...@smithfam.info> wrote:
OK, here is what I've come up with:
https://gitorious.org/biblical-studies/biblical-studies/blobs/raw/master/strongs-sqlite.py
You can checkout the git repo which contains it here:
git://gitorious.org/biblical-studies/biblical-studies.git
This was a bit more than I bargained for. :-) The
strongs-dictionary.xhtml from the Openscriptures strongs repo is good,
but there were two major drawbacks: 1. some of the unicode strings for
Hebrew entries are missing; 2. it lacks transliteration data.
So I decided to get this info from the two source XML files. They had
difference datatypes, so I had to write two different parsers using
xml.sax. The bottom line is that after filling in the proper variables
in the script and running it, you'll have a sqlite database matching
what you requested. I have done very limited QA on it, so let me know
if there are bugs. I put the code under an MIT license so you can
incorporate it if you like. There are actually more data fields than
just description (derivation/exegesis, definition/explanation,
kjv_def/translation in the Hebrew/Greek xml respectively, as well as
some outside those tags in Greek). You can modify the script to pick
those up separately into different database fields if you like (I did
that in one iteration of this script).
Here's the readme:
Requirements:
- Python v2.6 or later
Notes:
- Fill in the appropriate paths for hebrew_xml, greek_xml, db_file, and log_file
- Some strongsref tags in the Greek XML make reference to Strongs
numbers which are not included in the XML file. In those cases the
strongs number is listed in place of the unicode for the reference
lemma (e.g. "G5689").
- Hebrew when viewed in the console may appear LTR, but in text files
and other uses should appear RTL.
http://nathan.smithfam.info/
--
You received this message because you are subscribed to the Google Groups "Open Scriptures" group.
To post to this group, send email to openscr...@googlegroups.com.
To unsubscribe from this group, send email to openscripture...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/openscriptures?hl=en.
And yes, I will report back any bug!
On Dec 5, 2:57 am, Nathan Smith <nat...@smithfam.info> wrote:
> On Sat, Dec 3, 2011 at 6:40 AM, evstevemd <mwinjili...@gmail.com> wrote:
> > Thanks a Million Nathan,
> > I will appreciate that!
>
> OK, here is what I've come up with:
>
> https://gitorious.org/biblical-studies/biblical-studies/blobs/raw/mas...
I cannot make sense of this, would you please explain what it means?
Is something missing?
Yes, basically. I think it may be a mistake in the original Strong's.
The data for G25 has a <strongsref> tag which points to Greek 5689,
but the list only goes to 5624. I checked another online Strong's and
it had the same reference to the missing entry, so the error lies
somewhere "upstream" from our XML source. I don't have a print
Strong's anymore so I could not check up on it. Someone might even be
able to figure out which # is supposed to be referenced there.
As for the acknowledgements, just my name is fine. Thanks.
In my printed Strong's I found the same listing. The 5689 is in normal
type, referring to Hebrew, rather than italics for Greek. The
corresponding Hebrew words seems to fit.
Peace,
David
Thanks. I'll check to see if this is correct in version 1.5 and submit
a bug report to Ulrik if not.
On Dec 5, 9:25 pm, Nathan Smith <nat...@smithfam.info> wrote:
As far as I know, yes. It is corrected in the 1.5 version, so I'll try
running the script against that version.
I updated the script so that is uses the latest Strongs Greek XML
version 1.5, which has the aforementioned error corrected. Also, I
made the script automatically download the source files, so you don't
have to checkout the OpenScriptures git repo to run it. New code is
here: https://gitorious.org/biblical-studies
On Dec 11, 12:33 am, Nathan Smith <nat...@smithfam.info> wrote:
> On Wed, Dec 7, 2011 at 7:23 PM, Nathan Smith <nat...@smithfam.info> wrote: