"late-pluperfect-singulars",
"sigma-loss-pmd.2s",
"HGrk",
])
c = form.count("/") + 1
stem = ginflexion.find_stems(lemma, key, tags)
generated = ginflexion.generate(lemma, key, tags)
if strip_length(form) in [
strip_length(w) for w in sorted(generated)
]:
correct = "✓"
stem_guess = None
else:
correct = "✕"
incorrect_count += 1
possible_stems = [(key_to_part(a), b, a)
for a, b in ginflexion.possible_stems(form)]
likely_stems = [(key_to_part(a), b)
for a, b in ginflexion.possible_stems(
form, "^" + key + "$")]
possible_parses = ginflexion.parse(form)
if debug or correct == "✕":
output_item(lemma, key, key_to_part(key), form, None, stem,
possible_stems, likely_stems, possible_parses,
generated, correct)
print("{}/{} incorrect".format(incorrect_count, total_count))
if stem:
stem_guess = None
else:
stem_guess = [
stem for key, stem in ginflexion.possible_stems(
form, "^" + key + "$")
]
if [strip_length(w) for w in sorted(generated)] == \
[strip_length(w) for w in sorted(form.split("/"))]:
correct = "✓"
else:
correct = "✕"
if correct == "✕":
if stem_guess:
STEM_GUESSES[lemma][key_to_part(key)].add(
frozenset(stem_guess))
for lemma, parts in sorted(STEM_GUESSES.items()):
print()
print("{}:".format(lemma))
print(" stems:".format(lemma))
for part, stem_sets in sorted(parts.items()):
stem = set.intersection(*(set(s) for s in stem_sets))
if len(stem) == 0:
print(" {}: {} # @0".format(part, stem_sets))
elif len(stem) == 1:
print(" {}: {} # @1".format(part, stem.pop()))
else:
print(" {}: {} # @m".format(part, stem))
"sigma-loss-pmd.2s",
"HGrk",
])
c = form.count("/") + 1
stem = ginflexion.find_stems(lemma, key, tags)
generated = ginflexion.generate(lemma, key, tags)
segmented_lemma = ginflexion.segmented_lemmas.get(lemma)
if strip_length(form) in [
strip_length(w) for w in sorted(generated)
]:
correct = "✓"
stem_guess = None
else:
correct = "✕"
incorrect_count += 1
possible_stems = [(key_to_part(a), b, a)
for a, b in ginflexion.possible_stems(form)]
likely_stems = [(key_to_part(a), b)
for a, b in ginflexion.possible_stems(
form, "^" + key + "$")]
possible_parses = ginflexion.parse(form)
if debug or correct == "✕":
output_item(lemma, segmented_lemma, key, key_to_part(key),
form, None, stem, possible_stems, likely_stems,
possible_parses, generated, correct)
print("{}/{} incorrect".format(incorrect_count, total_count))
if stem:
stem_guess = None
else:
stem_guess = [
stem
for key, stem in ginflexion.possible_stems2(form, "^" + key + "$")
]
if [strip_length(w) for w in sorted(generated)] == \
[strip_length(w) for w in sorted(form.split("/"))]:
correct = "✓"
else:
correct = "✕"
if correct == "✕":
if stem_guess:
STEM_GUESSES[lemma][key_to_part(key)].add(frozenset(stem_guess))
for lemma, parts in sorted(STEM_GUESSES.items()):
print()
print("{}:".format(lemma))
print(" stems:".format(lemma))
for part, stem_sets in sorted(parts.items(),
key=lambda x: (x[0][0], {
"-": 0,
"+": 1
}[x[0][1]])):
stem = set.intersection(*(set(s) for s in stem_sets))
if len(stem) == 0:
print(" {}: {} # @0".format(part, stem_sets))
elif len(stem) == 1:
print(" {}: {} # @1".format(part, stem.pop()))
"HGrk",
])
c = form.count("/") + 1
stem = ginflexion.find_stems(lemma, key, tags)
generated = ginflexion.generate(lemma, key, tags)
if strip_length(form) in [
strip_length(w) for w in sorted(generated)]:
correct = "✓"
else:
correct = "✕"
incorrect_count += 1
summary_by_lemma[lemma].add(key)
possible_stems = [
(key_to_part(a), b, a)
for a, b in ginflexion.possible_stems(form)
]
likely_stems = [
(key_to_part(a), b)
for a, b in ginflexion.possible_stems(form, "^" + key + "$")
]
possible_parses = []
for plemma, pparse in ginflexion.parse(form):
possible_parses.append((
plemma,
pparse,
set(ginflexion.generate(plemma, pparse, tags))
))
if debug or correct == "✕":