lexibank_bowernpny.py

from csv import QUOTE_NONE
from pathlib import Path

from clldutils.misc import slug
import pylexibank


class Dataset(pylexibank.Dataset):
    dir = Path(__file__).parent
    id = "bowernpny"
    writer_options=dict(keep_languages=False, keep_parameters=False)

    form_spec = pylexibank.FormSpec(
        separators=";,~/",
        missing_data=("-", "?", "—"),
        brackets={"(": ")", "{": "}", "[": "]"},
        strip_inside_brackets=True,
        replacements=[(" ", "_")],
        first_form_only=True,
    )

    def cmd_makecldf(self, args):
        # corrections to the data in concepticon
        concept_replacements = {
            "ashes": "ash",
            "bad": "bad/evil",
            "blowfly": "fly",
            "pimple": "boil/pimple",
            "boomerang": "boomerang/throwing stick",
            "true": "correct/true",
            "cut": "cut/hack",
            "die": "die/be dead",
            "dingo": "dingo/wolf",
            "blunt": "dull/blunt",
            "earth": "earth/soil",
            "faeces": "feces",
            "fat (n)": "fat/grease",
            "hair of head": "hair",
            "1sg": "I",
            "2sg": "you.SG",
            "3sg": "he/she",
            "1pl.excl": "we.EXCL.PL",
            "1pl.incl": "we.INCL.PL",
            "2pl": "thou",
            "3pl": "they",
            "hit": "hit (with hand)",
            "how?": "how",
            "what?": "what",
            "when?": "when",
            "where?": "where",
            "who?": "who",
            "inside": "in/inside",
            "kangaroo": "kangaroo/deer",
            "know": "know/be knowledgeable",
            "be alive": "live/be alive",
            "man": "man/male",
            "meat": "meat/flesh",
            "near": "near/close",
            "open": "open/uncover",
            "person": "person/human being",
            "road": "road/path",
            "ashamed": "shy/ashamed",
            "smell (tr/intr)": "sniff/smell",
            "stab": "stab/pierce",
            "tie up": "tie up/fasten",
            "woman": "woman/female",
            "sick": "painful/sick",
        }

        args.writer.add_sources()
        language_lookup = args.writer.add_languages(
            id_factory=lambda l: slug(l["Name"], lowercase=False), lookup_factory="Name"
        )
        concept_lookup = args.writer.add_concepts(
            id_factory=lambda x: x.id.split("-")[-1] + "_" + slug(x.english), lookup_factory="Name"
        )
        for s, t in concept_replacements.items():
            concept_lookup[s] = concept_lookup[t]

        singletons = 1
        for entry in pylexibank.progressbar(
            self.raw_dir.read_csv("bowernpny.tsv", delimiter="\t", quoting=QUOTE_NONE, dicts=True)
        ):
            if entry["Code"] == "0":
                cogid = singletons
                singletons += 1
            else:
                cogid = entry["Code"] + "-" + slug(entry["Gloss"])
            for form in args.writer.add_forms_from_value(
                Language_ID=language_lookup[entry["Language"]],
                Parameter_ID=concept_lookup[entry["Gloss"]],
                Value=entry["Form"],
                Source=["Bowern2012"],
            ):
                args.writer.add_cognate(lexeme=form, Cognateset_ID=cogid, Source=["Bowern2012"])

    def cmd_download(self, args):
        self.raw_dir.download(
            "https://zenodo.org/record/1005671/files/Pny2012Codes.tsv", "bowernpny.tsv"
        )
        self.raw_dir.write("sources.bib", pylexibank.getEvoBibAsBibtex("Bowern2012"))