gen_context.py

"""\
Parse a CSV namespace and generate the corresponding JSON-LD context.
"""

import argparse
import csv
import json
import sys
from pathlib import Path


RO_TERMS_PREFIX = "https://w3id.org/ro/terms"
VOCAB_FN = "vocabulary.csv"


def build_context(namespace, vocab_path):
    add_terms = {}
    with open(vocab_path, newline="") as f:
        reader = csv.DictReader(f)
        for row in reader:
            if not "term" in row:
                raise RuntimeError("Unexpected keys in CSV %s, try: term,type,label,description,domain,range" % row.keys())
            k = row["term"]
            if not k:
                continue # empty line!
            add_terms[k] = f"{RO_TERMS_PREFIX}/{namespace}#{k}"
    return {"@context": add_terms}


def main(args):
    ns_dir = Path(args.ns_dir)
    vocab_path = ns_dir / VOCAB_FN
    if not vocab_path.is_file():
        raise RuntimeError(f"{vocab_path} not found")
    context = build_context(ns_dir.name, vocab_path)
    if not args.output:
        args.output = [ns_dir / f"context.{ext}" for ext in ("json", "jsonld")]
    else:
        args.output = [args.output]
    for out_fn in args.output:
        with open(out_fn, "wt", encoding="utf8") as f:
            json.dump(context, f, ensure_ascii=False, indent=4, sort_keys=False)


if __name__ == "__main__":
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("ns_dir", metavar="NAMESPACE", help="namespace dir")
    parser.add_argument('-o', '--output', metavar="FILE")
    main(parser.parse_args(sys.argv[1:]))