Updating for lemma representation of word_form. Also cleaning code, adding tqdm,...
This commit is contained in:
parent
3c669c7901
commit
4c2b5f2b13
185
wani.py
185
wani.py
|
@ -1,7 +1,7 @@
|
|||
from xml.etree import ElementTree
|
||||
import re
|
||||
from enum import Enum
|
||||
from collections import defaultdict
|
||||
from collections import defaultdict, namedtuple
|
||||
import sys
|
||||
import logging
|
||||
import argparse
|
||||
|
@ -12,6 +12,7 @@ import concurrent.futures
|
|||
import tempfile
|
||||
|
||||
from msd_translate import MSD_TRANSLATE
|
||||
from tqdm import tqdm
|
||||
|
||||
|
||||
MAX_NUM_COMPONENTS = 5
|
||||
|
@ -217,7 +218,7 @@ class ComponentRendition:
|
|||
return self.rendition is rendition
|
||||
|
||||
@staticmethod
|
||||
def set_representations(matches, structure, word_renderer):
|
||||
def set_representations(matches, structure, word_renderer, lemma_msds):
|
||||
representations = {
|
||||
c.idx: [[], None] if c.representation.isit(Rendition.WordForm) else [True, ""]
|
||||
for c in structure.components
|
||||
|
@ -225,24 +226,34 @@ class ComponentRendition:
|
|||
found_agreements = {}
|
||||
word_component_id = {}
|
||||
|
||||
# doprint = structure.id == '1' and matches[0]['1'].text.startswith('evrop') and matches[0]['2'].text.startswith('prv')
|
||||
doprint = False
|
||||
|
||||
def render_all(component_id, lst):
|
||||
matches.representations[component_id] = "/".join([w.text for w in set(lst)])
|
||||
def render_all(component_id, lst, _bw):
|
||||
rep = "/".join([w.text for w in set(lst)]) if len(lst) > 0 else None
|
||||
matches.representations[component_id] = rep
|
||||
|
||||
def render_form(component_id, lst):
|
||||
sorted_lst = sorted(set(lst), key=lst.count)
|
||||
for word in sorted_lst:
|
||||
def render_form(component_id, lst, backup_word):
|
||||
if backup_word is not None:
|
||||
lst.append(backup_word)
|
||||
|
||||
lst_ctr = []
|
||||
for word in lst:
|
||||
lst_ctr.append((word.msd, word.lemma))
|
||||
sorted_lst = sorted(set(lst_ctr), key=lst.count)
|
||||
|
||||
if len(lst) > 3:
|
||||
a = 3
|
||||
|
||||
for word_msd, word_lemma in sorted_lst:
|
||||
if component_id in found_agreements:
|
||||
other_component_id, other_word, agreements = found_agreements[component_id]
|
||||
print(word.lemma, other_word.lemma, component_id, other_component_id, word.msd, word.msd)
|
||||
agr = are_agreements_ok(word.msd, other_word.lemma, other_word.msd, agreements)
|
||||
agr = are_agreements_ok(word_msd, other_word.lemma, other_word.msd, agreements)
|
||||
if agr is None:
|
||||
continue
|
||||
|
||||
matches.representations[other_component_id] = agr
|
||||
|
||||
matches.representations[word_component_id[word.id]] = word.most_frequent_text(word_renderer)
|
||||
if word_lemma is not None:
|
||||
matches.representations[component_id] = word_renderer.render(word_lemma, word_msd)
|
||||
|
||||
break
|
||||
|
||||
# othw = are_agreements_ok(word, found_agreements)
|
||||
|
@ -255,10 +266,7 @@ class ComponentRendition:
|
|||
if ow_msd[0] != w2_msd[0]:
|
||||
continue
|
||||
|
||||
print(w1_msd, w2_msd)
|
||||
if check_agreement(w1_msd, w2_msd, agreements):
|
||||
if doprint:
|
||||
print("GOOD :)")
|
||||
return w2_txt
|
||||
|
||||
def check_msd(word, selectors):
|
||||
|
@ -279,7 +287,6 @@ class ComponentRendition:
|
|||
# if not in msd, some strange msd was tries, skipping...
|
||||
if agr_case not in TAGSET[t1]:
|
||||
logging.warning("Cannot do agreement: {} for msd {} not found!".format(agr_case, msd1))
|
||||
print("BAAAD")
|
||||
return False
|
||||
|
||||
v1 = TAGSET[t1].index(agr_case)
|
||||
|
@ -293,7 +300,6 @@ class ComponentRendition:
|
|||
t2 = msd2[0]
|
||||
if agr_case not in TAGSET[t2]:
|
||||
logging.warning("Cannot do agreement: {} for msd {} not found!".format(agr_case, msd2))
|
||||
print("BAAAD")
|
||||
return False
|
||||
v2 = TAGSET[t2].index(agr_case)
|
||||
if v2 + 1 >= len(msd2):
|
||||
|
@ -328,44 +334,26 @@ class ComponentRendition:
|
|||
assert(rep.isit(Rendition.WordForm))
|
||||
wf_type, more = rep.more
|
||||
|
||||
if wf_type is WordFormSelection.All:
|
||||
add = True
|
||||
func = render_all
|
||||
elif wf_type is WordFormSelection.Msd:
|
||||
if wf_type is WordFormSelection.Msd:
|
||||
add = check_msd(w, more)
|
||||
func = render_form
|
||||
elif wf_type is WordFormSelection.All:
|
||||
add = True
|
||||
func = render_all
|
||||
elif wf_type is WordFormSelection.Any:
|
||||
add = True
|
||||
func = render_form
|
||||
else:
|
||||
assert(wf_type is WordFormSelection.Agreement)
|
||||
other_w, agreements = more
|
||||
found_agreements[other_w] = (w_id, w.lemma, agreements)
|
||||
found_agreements[other_w] = (w_id, w, agreements)
|
||||
|
||||
add = True
|
||||
func = lambda *x: None
|
||||
|
||||
representations[w_id][1] = func
|
||||
if add:
|
||||
representations[w_id][0].append(w)
|
||||
representations[w_id][1] = func
|
||||
|
||||
if doprint:
|
||||
print(len(matches), len(found_agreements))
|
||||
|
||||
# for w1i, w2i, agreements in found_agreements:
|
||||
# w1, w2 = words[w1i], words[w2i]
|
||||
# if doprint:
|
||||
# print("? ", w1.msd, w2.msd, end="")
|
||||
|
||||
# if w2i not in bad_words:
|
||||
#
|
||||
# if check_agreement(w1, w2, agreements):
|
||||
# representations[w1i][0].append(w1.text)
|
||||
# if doprint:
|
||||
# print(" :)")
|
||||
# elif doprint:
|
||||
# print(" :(")
|
||||
# elif doprint:
|
||||
# print(" :((")
|
||||
|
||||
# just need to set representation to first group,
|
||||
# but in correct order, agreements last!
|
||||
|
@ -379,20 +367,12 @@ class ComponentRendition:
|
|||
|
||||
for w_id, w in representation_sorted_words:
|
||||
data = representations[w_id]
|
||||
if doprint:
|
||||
print([(r.text, r.lemma, r.msd) for r in data[0]])
|
||||
|
||||
if type(data[1]) is str:
|
||||
matches.representations[w_id] = None if data[0] else data[1]
|
||||
elif len(data[0]) == 0:
|
||||
matches.representations[w_id] = None
|
||||
else:
|
||||
data[1](str(w_id), data[0])
|
||||
backup_word = lemma_only_word(lemma_msds[w.msd[0]]) if w.msd[0] in lemma_msds else None
|
||||
data[1](str(w_id), data[0], backup_word)
|
||||
|
||||
if doprint:
|
||||
print(matches.representations)
|
||||
print('--')
|
||||
|
||||
def __str__(self):
|
||||
return str(self.rendition)
|
||||
|
||||
|
@ -468,7 +448,7 @@ def build_morphology_regex(restriction):
|
|||
return False
|
||||
return True
|
||||
|
||||
return " ".join(rgx), matcher
|
||||
return rgx, matcher
|
||||
|
||||
|
||||
def build_lexis_regex(restriction):
|
||||
|
@ -493,7 +473,8 @@ class Restriction:
|
|||
restriction_type = restriction_tag.get('type')
|
||||
if restriction_type == "morphology":
|
||||
self.type = RestrictionType.Morphology
|
||||
self.present, self.matcher = build_morphology_regex(list(restriction_tag))
|
||||
present, self.matcher = build_morphology_regex(list(restriction_tag))
|
||||
self.present = " ".join(present)
|
||||
elif restriction_type == "lexis":
|
||||
self.type = RestrictionType.Lexis
|
||||
self.present, self.matcher = build_lexis_regex(list(restriction_tag))
|
||||
|
@ -822,17 +803,44 @@ class SyntacticStructure:
|
|||
# to_ret.append((m, self.check_agreements(m)))
|
||||
|
||||
|
||||
def build_structures(filename):
|
||||
structures = []
|
||||
def load_structures(filename):
|
||||
with open(filename, 'r') as fp:
|
||||
et = ElementTree.XML(fp.read())
|
||||
for structure in et.iter('syntactic_structure'):
|
||||
to_append = SyntacticStructure.from_xml(structure)
|
||||
if to_append is None:
|
||||
continue
|
||||
structures.append(to_append)
|
||||
|
||||
return build_structures(et), get_lemma_features(et)
|
||||
|
||||
def build_structures(et):
|
||||
structures = []
|
||||
for structure in et.iter('syntactic_structure'):
|
||||
to_append = SyntacticStructure.from_xml(structure)
|
||||
if to_append is None:
|
||||
continue
|
||||
structures.append(to_append)
|
||||
return structures
|
||||
|
||||
def get_lemma_features(et):
|
||||
lf = et.find('lemma_features')
|
||||
if lf is None:
|
||||
return {}
|
||||
|
||||
result = {}
|
||||
for pos in lf.iter('POS'):
|
||||
rgx_list, _ = build_morphology_regex(pos)
|
||||
rgx_str = ""
|
||||
for position in rgx_list:
|
||||
if position == ".":
|
||||
rgx_str += " "
|
||||
elif len(position) == 1:
|
||||
rgx_str += position
|
||||
elif len(position) == 3 and position[0] == "[" and position[2] == "]":
|
||||
rgx_str += position[1]
|
||||
else:
|
||||
raise RuntimeError("Strange rgx for lemma_feature...")
|
||||
|
||||
assert(rgx_str[0].isupper())
|
||||
result[rgx_str[0]] = rgx_str.strip().replace(' ', '-')
|
||||
|
||||
return result
|
||||
|
||||
def get_msd(comp):
|
||||
d = dict(comp.items())
|
||||
|
@ -844,6 +852,10 @@ def get_msd(comp):
|
|||
logging.error(d, file=sys.stderr)
|
||||
raise NotImplementedError("MSD?")
|
||||
|
||||
def lemma_only_word(msd):
|
||||
WordLemma = namedtuple('WordLemmaOnly', 'msd most_frequent_text lemma')
|
||||
return WordLemma(msd=msd, most_frequent_text=lambda *x: None, lemma=None)
|
||||
|
||||
class Word:
|
||||
def __init__(self, xml, do_msd_translate):
|
||||
self.lemma = xml.get('lemma')
|
||||
|
@ -882,6 +894,7 @@ class WordMsdRenderer:
|
|||
def __init__(self):
|
||||
self.all_words = []
|
||||
self.rendered_words = {}
|
||||
self.frequent_words = {}
|
||||
|
||||
def add_words(self, words):
|
||||
self.all_words.extend(words)
|
||||
|
@ -893,19 +906,30 @@ class WordMsdRenderer:
|
|||
|
||||
for lemma, ld in data.items():
|
||||
self.rendered_words[lemma] = {}
|
||||
freq_words = defaultdict(int)
|
||||
|
||||
for msd, texts in ld.items():
|
||||
rep = max(set(texts), key=texts.count)
|
||||
self.rendered_words[lemma][msd] = rep
|
||||
|
||||
self.rendered_words[lemma][msd] = (rep, len(texts))
|
||||
|
||||
for txt in texts:
|
||||
freq_words[(msd, txt)] += 1
|
||||
|
||||
self.frequent_words[lemma] = []
|
||||
for (msd, txt), n in sorted(freq_words.items(), key=lambda x: -x[1]):
|
||||
self.frequent_words[lemma].append((msd, txt, n))
|
||||
|
||||
def render(self, lemma, msd):
|
||||
if lemma in self.rendered_words:
|
||||
if msd in self.rendered_words[lemma]:
|
||||
return self.rendered_words[lemma][msd]
|
||||
return self.rendered_words[lemma][msd][0]
|
||||
|
||||
def available_words(self, lemma):
|
||||
if lemma in self.rendered_words:
|
||||
for msd in self.rendered_words[lemma].keys():
|
||||
yield (msd, self.rendered_words[lemma][msd])
|
||||
if lemma in self.frequent_words:
|
||||
# print("--")
|
||||
for msd, text, _ in self.frequent_words[lemma]:
|
||||
# print(lemma, msd, text, _)
|
||||
yield (msd, text)
|
||||
|
||||
def is_root_id(id_):
|
||||
return len(id_.split('.')) == 3
|
||||
|
@ -1010,7 +1034,6 @@ class Writer:
|
|||
elif self.all:
|
||||
return [word.id, word.text, word.lemma, word.msd]
|
||||
else:
|
||||
# print("1", word)
|
||||
if representation is None:
|
||||
return [word.lemma, word.lemma, "lemma_fallback"]
|
||||
else:
|
||||
|
@ -1137,17 +1160,21 @@ class ColocationIds:
|
|||
if group:
|
||||
break
|
||||
|
||||
def set_representations(self, structures, word_renderer):
|
||||
def set_representations(self, structures, word_renderer, lemma_msds):
|
||||
components_dict = {structure.id: structure for structure in structures}
|
||||
for _1, sm in self.data.items():
|
||||
ComponentRendition.set_representations(sm, components_dict[sm.structure_id], word_renderer)
|
||||
idx = 1
|
||||
for _1, sm in tqdm(self.data.items()):
|
||||
if idx == 120:
|
||||
a = 3
|
||||
ComponentRendition.set_representations(sm, components_dict[sm.structure_id], word_renderer, lemma_msds)
|
||||
idx += 1
|
||||
|
||||
|
||||
def match_file(words, structures):
|
||||
matches = {s.id: [] for s in structures}
|
||||
|
||||
for idx, s in enumerate(structures):
|
||||
logging.info("{}/{}: {:7s}".format(idx, len(structures), s.id))
|
||||
for idx, s in tqdm(list(enumerate(structures))):
|
||||
# logging.info("{}/{}: {:7s}".format(idx, len(structures), s.id))
|
||||
for w in words:
|
||||
mhere = s.match(w)
|
||||
logging.debug(" GOT: {}".format(len(mhere)))
|
||||
|
@ -1162,13 +1189,19 @@ def match_file(words, structures):
|
|||
|
||||
|
||||
def main(input_file, structures_file, args):
|
||||
structures = build_structures(structures_file)
|
||||
structures, lemma_msds = load_structures(structures_file)
|
||||
for s in structures:
|
||||
logging.debug(str(s))
|
||||
|
||||
colocation_ids = ColocationIds()
|
||||
word_renderer = WordMsdRenderer()
|
||||
|
||||
# if True:
|
||||
# with open("match_word.p", "rb") as fp:
|
||||
# words, matches = pickle.load(fp)
|
||||
# colocation_ids.add_matches(matches)
|
||||
# word_renderer.add_words(words)
|
||||
|
||||
if args.parallel:
|
||||
num_parallel = int(args.parallel)
|
||||
|
||||
|
@ -1215,7 +1248,7 @@ def main(input_file, structures_file, args):
|
|||
# get word renders for lemma/msd
|
||||
word_renderer.generate_renders()
|
||||
# figure out representations!
|
||||
colocation_ids.set_representations(structures, word_renderer)
|
||||
colocation_ids.set_representations(structures, word_renderer, lemma_msds)
|
||||
|
||||
if args.all:
|
||||
Writer.make_all_writer(args).write_out(structures, colocation_ids)
|
||||
|
|
Loading…
Reference in New Issue
Block a user