def extend_write_eds(fig2parents, e2types, outfile, e2freq):
e2types_parents = {}
for mye in e2types:
tt = e2types[mye]
extended_tt = addhighleveltypes(tt, fig2parents)
e2types_parents[mye] = extended_tt
write_ds(e2types_parents, e2freq, outfile)
return e2types_parents
f = open(outdir + '/ds_lines_figer', 'w')
logger.info('write test lines in %s', outdir + 'ds_lines_figer')
for myline in lines:
f.write(myline.strip())
f.write('\n')
f.close()
if __name__ == '__main__':
fbname2figer = laod_figermapping('/nfs/data3/yadollah/nlptools_resources/figer/config/types.map')
fig2parents = load_parents('/nfs/data3/yadollah/nlptools_resources/figer/config/yy_type2parents')
mid2name = load_type_names('/nfs/data1/proj/yadollah/cluewebwork/nlu/dataForImport/type.name')
logger.info('size types with names: %d', len(mid2name))
dsfile = sys.argv[1]
ds_linesfile = sys.argv[2]
outdir = sys.argv[3]
(e2types, t2ents, e2freq) = load_dataset(dsfile, logger)
logger.info(len(e2types))
newConvertedDs = filter_write_ds(e2types, fbname2figer, mid2name, outdir, fig2parents)
logger.info('size of dataset after filtering to figer type: %d', len(newConvertedDs))
(new_dslines, e2freq) = filter_ds_lines(ds_linesfile, newConvertedDs, outdir)
logger.info('#lines after filtering to figer type: %d', len(new_dslines))
write_ds(newConvertedDs, e2freq, outdir + 'Eds_figer')
write_lines(new_dslines, e2freq, outdir)
