fixParadigms.py

# -*- coding: utf-8 -*-
import codecs
import re
import argparse
import sys

# gather arguments
parser = argparse.ArgumentParser(description="Clean out paradigms with errors.")
parser.add_argument("-language", action="store", dest="language", help="Language.")
args = parser.parse_args()

fin = codecs.open(
    "./tabular_results/" + args.language + "_tabular_paradigms.txt", "rb", "utf-8"
)  # CHANGE
fout = codecs.open(
    "./tabular_results/" + args.language + "_tabular_paradigms_norm.txt", "wb", "utf-8"
)  # CHANGE

paradigms = fin.read().split("\n\n")

vcount = 0
ncount = 0
acount = 0

for p in paradigms:
    # Exclude paradigms that caused exceptions during parsing.
    if "----" not in p:
        if "\tV;" in p:
            vcount += 1
        if "\tN;" in p:
            ncount += 1
        if "\tADJ;" in p:
            acount += 1
        fout.write(p + "\n\n")

# print language, number of verbs, adjectives, and nouns
print(args.language, vcount, acount, ncount)

# clean up
fin.close()
fout.close()