This is a repository for things, which are related to the systematic extraction of information from grammatical descriptions. This means linguistic descriptions of languages are the target; not to be confused with the computational term "grammar mining".
A couple of text (markdown) files describe the workflow used in the project. The basic outline is described in workflow.md. Specific steps of the workflow are explained in more depth in the files Creating_bibliographies.md, duplicate_reduction.md and Acquire_sources.md.
The R script SA_scope_languoids.R is used for the filtering of languoids used in the sample (using input from Glottolog), cf. workflow.md.
As input for the download script, we need a list of filenames (stored as csv). This list is extracted from a bib file (cf. Creating_bibliographies.md) using filename_extraction_glottolog_v3.R.
One version of the output of the filename extraction can be found here as well (pigritiae causa vel for reasons of convenience): SA_alldoctypes_bestfn.csv. Note that this is not necessarily the output of the current version of the script and it doesn't claim to be; it's only there for testing the script below without having to create a file yourself.
THe file downloading_script_v2.2.4_flexible.py is a python3 script, which takes as input a csv file (cf. above) and downloads the specified items from the (not publically accessible!) website gramfinder. If you want to customize the script: The input file needs to specify the latter part of the URL (i.e. the folder), in the format "/macro-area/author_keywordYear(_suffix).pdf" like "eurasia\zydenbos_kannada2011_o.pdf" (can be found in the bib-files from Glottolog). (Note: this is my first python script; it's rather basic, and you may receive a warning about the security of the session - however, as I have been told by the owner of the website, this is at current unavoidable. You may, of course, report any issues via an issue, DM or pull request.)
Note: In case that there is really someone out there who wants to make use of those scripts and does not only see them as the documentation of the scientific work of our project, let me say that a) there is some information on how to use the scripts / what to be aware of in the scripts themselves (comment lines), and feel free to ask me or point me towards any inconsistencies, problems etc.