Python 3 tool for generating (initially Biblical) Greek readers

Example from John 18:1-11. The steps to produce this are listed below under A More Extended Example.

MorphGNT and my Morphological Lexicon aren't quite rich enough yet to produce the kind of readers I've long wanted to (much less the larger vision of a New Kind of Graded Reader) but I've been inspired by Brian Renshaw's (presumably manually produced) Greek Readers (e.g. A Good Friday Greek Reader) to at least put together a tool to show what's possible now and then build on it.

Of course, this isn't the first time I've written code to generate documents from my Greek New Testament databases. This year marks the 20th anniversary of my Index to the Greek New Testament which was the first major project I undertook based on MorphGNT.

What I'm initially putting together here is a Python 3 library and command-line script driven by text files. Eventually, I'll make a website out of this so the majority of the target audience can actually use it :-)

For other Greek projects of mine, see http://jktauber.com/.

As well as Python 3, you'll need to install the packages in requirements.txt via pip (preferably in a virtualenv).

XeTeX is required as the current output of my scripts is LaTeX with Unicode (although I do plan to support other backends eventually). On OS X, I use the MacTeX distribution.

Assuming you've installed the requirements, you can just type:

./reader.py "John 18:1-11" > reader.tex

You can then run:

xelatex reader.tex
xelatex reader.tex

The two rendering passes ensure that the footnotes are properly numbered.

Note that the reader.pdf PDF that results will footnote every word with the lemma from MorphGNT and, in the case of verbs will include parsing codes. No glosses will be included.

If you want to exclude certain words (for example, very common words) from being annotated, you can pass an --exclude option to reader.py, giving the name of a file which simply lists the lemmas to exclude. For example:

αὐτός
καί
ὁ

You can easily generate such a file for any words occurring more than N times by running frequency_exclusion.py with N as an argument. For example, to create an exclusion file with any words occurring 31 times or more, run:

./frequency_exclusion.py 31 > exclude31.txt

and then run ./reader.py with --exclude exclude31.txt.

Note that you can make edits to the file after running frequency_exclusion.py to tailor the exclusion list to your needs.

If you want to provide glosses, you can pass a --glosses option to reader.py with the name of a YAML file that maps each lemma to a default gloss and possibly per-verse overrides. A file with just default (i.e. global) glosses might look like this:

ἀποκόπτω:
    default: cut off
ἕλκω:
    default: draw
θήκη:
    default: sheath

You can auto-generate an initial gloss file based on John Jeffrey Dodson's public domain lexicon (via lexemes.yaml in this repo) using make_glosses.py which takes a verse range argument just like reader.py as well as an --exclude option.

If you want to extend an existing glosses file you can pass its name in using the --existing option. This is useful if you've already made edits to the file and you don't want to lose them when expanding the coverage of the file to more verses (or fewer exclusions).

When typesetting a reader with glosses, the gloss language should be specified when you generate the reader file. If no value is specified it will default to English, but if your gloss words are in another language it should be specified with the --language option. Languages should be specified using three letter ISO-639-3 codes (e.g. --language rus for Russion or --language spa for Spanish).

If you want to provide more detailed headwords (such as the article or adjective endings) you can pass a --headwords option to reader.py with the name of a YAML file that maps each lemma you want to override with the full headword you want to use instead. For example:

θήκη: θήκη, ης, ἡ
Κεδρών: Κεδρών, ὁ

You can run make_headwords.py to generate headword overrides for nouns and adjectives based on Danker's Concise Lexicon (via the lexemes.yaml file). make_headwords.py takes a verse range argument just like reader.py as well as an --exclude option.

If you want to extend an existing headword file you can pass its name in using the --existing option. This is useful if you've already made edits to the file and you don't want to lose them when expanding the coverage of the file to more verses (or fewer exclusions).

The default typeface is now Times New Roman but you can change this by passing a --typeface option to reader.py.

Here is how you might typically use the tools:

./frequency_exclusion.py 31 > example/exclude.txt
./make_glosses.py \
    --exclude example/exclude.txt \
    "John 18:1-11" > example/glosses.yaml
# edit example/glosses.yaml to your liking
./make_headwords.py \
    --exclude example/exclude.txt \
    "John 18:1-11" > example/headwords.yaml
./reader.py \
    --headwords example/headwords.yaml \
    --glosses example/glosses.yaml \
    --language eng \
    --exclude example/exclude.txt \
    --typeface "Skolar PE" \
    --backend backends.LaTeX \
    "John 18:1-11" > example/reader.tex
cd example
xelatex reader.tex
xelatex reader.tex
open reader.pdf

You can see the results of this in the examples directory.

A --backend option can be provided to reader.py to use an alternative backend. This option takes a module-qualified Python class name. As well as the default backends.LaTeX, there is an experimental backends.SILE included for the SILE Typesetter and backends.MARKDOWN for Markdown, most useful for Markdown processors that support footnotes, for example GitHub (for example).

./reader.py --backend backends.SILE "John 18:1-11" > reader.sil
sile reader.sil