附加在 for 循环中不适用于存储令牌列表

Question

In the for loop below, I'm reading .dat files from a folder and parsing each file to extract the token list and then storing it in a list. My code does this, but for individual files. I have 1187 files, but the ud_file.append() just adds the tokens from the latest file, and ignores the tokens it appended in the earlier iteration. So, the list contains only the latest tokens and not all the tokens from the 1187 files. How should I fix this?

from io import open
from conllu import parse_incr
import os
import glob
import pandas as pd

#create a dict to store the results
word_lemma_dict = {}
ud_files = []
dat_files = []

#open the files and load the sentences to a list

datfolder = "Lemma/venv/Hindi corpus 2/CoNLL/utf" #Folder where all the .dat files are stored.

datfiles = glob.glob(os.path.join(datfolder, '*.dat'))

for file in datfiles:
    data_file = open(file, "r", encoding = "utf-8")
    for tokenlist in parse_incr(data_file):
         ud_files.append(tokenlist). #Only stores tokens from the latest file. Should ideally stores tokens from all the files it read in the for loop.

Here's the sample .dat file. I have 1187 such files:

 sent_id = dev-s1
# text = रामायण काल में भगवान राम के पुत्र कुश की राजधानी कुशावती को 483 ईसा पूर्व बुद्ध ने अपने अंतिम विश्राम के लिए चुना ।
1   रामायण  रामायण  PROPN   NNPC    Case=Nom|Gender=Masc|Number=Sing|Person=3   2   compound    _   Vib=0|Tam=0|ChunkId=NP|ChunkType=child|Translit=rāmāyaṇa
2   काल काल PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   23  obl _   Vib=0_में|Tam=0|ChunkId=NP|ChunkType=head|Translit=kāla
3   में में ADP PSP AdpType=Post    2   case    _   ChunkId=NP|ChunkType=child|Translit=meṁ
4   भगवान   भगवान   NOUN    NNC Case=Nom|Gender=Masc|Number=Sing|Person=3   5   compound    _   Vib=0|Tam=0|ChunkId=NP2|ChunkType=child|Translit=bhagavāna
5   राम राम PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   7   nmod    _   Vib=0_का|Tam=0|ChunkId=NP2|ChunkType=head|Translit=rāma
6   के  का  ADP PSP AdpType=Post|Case=Acc|Gender=Masc|Number=Sing   5   case    _   ChunkId=NP2|ChunkType=child|Translit=ke
7   पुत्र   पुत्र   NOUN    NN  Case=Acc|Gender=Masc|Number=Sing|Person=3   8   nmod    _   Vib=0|Tam=0|ChunkId=NP3|ChunkType=head|Translit=putra
8   कुश कुश PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   10  nmod    _   Vib=0_का|Tam=0|ChunkId=NP4|ChunkType=head|Translit=kuśa
9   की  का  ADP PSP AdpType=Post|Case=Acc|Gender=Fem|Number=Sing    8   case    _   ChunkId=NP4|ChunkType=child|Translit=kī
10  राजधानी राजधानी NOUN    NN  Case=Acc|Gender=Fem|Number=Sing|Person=3    11  nmod    _   Vib=0|Tam=0|ChunkId=NP5|ChunkType=head|Translit=rājadhānī
11  कुशावती कुशावती PROPN   NNP Case=Acc|Gender=Fem|Number=Sing|Person=3    23  obj _   Vib=0_को|Tam=0|ChunkId=NP6|ChunkType=head|Translit=kuśāvatī
12  को  को  ADP PSP AdpType=Post    11  case    _   ChunkId=NP6|ChunkType=child|Translit=ko
13  483 483 PROPN   NNPC    Case=Nom|Gender=Masc|Number=Sing|Person=3   15  compound    _   Vib=0|Tam=0|ChunkId=NP7|ChunkType=child|Translit=483
14  ईसा ईसा PROPN   NNPC    Case=Nom|Gender=Masc|Number=Sing|Person=3   15  compound    _   Vib=0|Tam=0|ChunkId=NP7|ChunkType=child|Translit=īsā
15  पूर्व   पूर्व   PROPN   NNP Case=Nom|Gender=Masc|Number=Sing|Person=3   23  obl _   Vib=0|Tam=0|ChunkId=NP7|ChunkType=head|Translit=pūrva
16  बुद्ध   बुद्ध   PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   23  nsubj   _   Vib=0_ने|Tam=0|ChunkId=NP8|ChunkType=head|Translit=buddha
17  ने  ने  ADP PSP AdpType=Post    16  case    _   ChunkId=NP8|ChunkType=child|Translit=ne
18  अपने    अपना    PRON    PRP Case=Acc|Gender=Masc|PronType=Prs   20  nmod    _   Vib=0|Tam=0|ChunkId=NP9|ChunkType=head|Translit=apane
19  अंतिम   अंतिम   ADJ JJ  Case=Acc    20  amod    _   ChunkId=NP10|ChunkType=child|Translit=aṁtima
20  विश्राम विश्राम NOUN    NN  Case=Acc|Gender=Masc|Number=Sing|Person=3   23  obl _   Vib=0_के_लिए|Tam=0|ChunkId=NP10|ChunkType=head|Translit=viśrāma
21  के  के  ADP PSP AdpType=Post    20  case    _   ChunkId=NP10|ChunkType=child|Translit=ke
22  लिए लिए ADP PSP AdpType=Post    20  case    _   ChunkId=NP10|ChunkType=child|Translit=lie
23  चुना    चुन VERB    VM  Aspect=Perf|Gender=Masc|Number=Sing|VerbForm=Part|Voice=Act 0   root    _   Vib=या|Tam=yA|ChunkId=VGF|ChunkType=head|Stype=declarative|Translit=cunā
24  ।   ।   PUNCT   SYM _   23  punct   _   ChunkId=BLK|ChunkType=head|Translit=.

# sent_id = dev-s2
# text = मल्‍लों की राजधानी होने के कारण प्राचीनकाल में इस स्‍थान का अत्‍यंत महत्‍व था ।
1   मल्‍लों मल्ला   NOUN    NN  Case=Acc|Gender=Masc|Number=Plur|Person=3   3   nmod    _   Vib=0_का|Tam=0|ChunkId=NP|ChunkType=head|Translit=malloṁ
2   की  का  ADP PSP AdpType=Post|Case=Nom|Gender=Fem|Number=Sing    1   case    _   ChunkId=NP|ChunkType=child|Translit=kī
3   राजधानी राजधानी NOUN    NN  Case=Nom|Gender=Fem|Number=Sing|Person=3    4   nsubj   _   Vib=0|Tam=0|ChunkId=NP2|ChunkType=head|Translit=rājadhānī
4   होने    हो  VERB    VM  Case=Acc|Gender=Masc|VerbForm=Inf   14  advcl   _   Vib=ना_के_कारण|Tam=nA|ChunkId=VGNN|ChunkType=head|Translit=hone
5   के  के  ADP PSP AdpType=Post|Case=Acc|Gender=Masc   4   mark    _   ChunkId=VGNN|ChunkType=child|Translit=ke
6   कारण    कारण    ADP PSP Case=Acc|Gender=Masc    4   mark    _   ChunkId=VGNN|ChunkType=child|Translit=kāraṇa
7   प्राचीनकाल  प्राचीनकाल  NOUN    NN  Case=Acc|Gender=Masc|Number=Sing|Person=3   14  obl _   Vib=0_में|Tam=0|ChunkId=NP3|ChunkType=head|Translit=prācīnakāla
8   में में ADP PSP AdpType=Post    7   case    _   ChunkId=NP3|ChunkType=child|Translit=meṁ
9   इस  यह  DET DEM Case=Acc|Number=Sing|Person=3|PronType=Dem  10  det _   ChunkId=NP4|ChunkType=child|Translit=isa
10  स्‍थान  स्थान   NOUN    NN  Case=Acc|Gender=Masc|Number=Sing|Person=3   13  nmod    _   Vib=0_का|Tam=0|ChunkId=NP4|ChunkType=head|Translit=sthāna
11  का  का  ADP PSP AdpType=Post|Case=Nom|Gender=Masc|Number=Sing   10  case    _   ChunkId=NP4|ChunkType=child|Translit=kā
12  अत्‍यंत अत्यंत  ADJ JJ  Case=Nom    13  amod    _   ChunkId=NP5|ChunkType=child|Translit=atyaṁta
13  महत्‍व  महत्व   NOUN    NN  Case=Nom|Gender=Masc|Number=Sing|Person=3   14  nsubj   _   Vib=0|Tam=0|ChunkId=NP5|ChunkType=head|Translit=mahatva
14  था  था  VERB    VM  Gender=Masc|Mood=Ind|Number=Sing|Tense=Past|VerbForm=Fin|Voice=Act  0   root    _   Vib=था|Tam=WA|ChunkId=VGF|ChunkType=head|Stype=declarative|Translit=thā
15  ।   ।   PUNCT   SYM _   14  punct   _   ChunkId=BLK|ChunkType=head|Translit=.

# sent_id = dev-s3
# text = बौद्ध धर्मावलंबियों के अनुसार लुंबनी, बोधगया और सारनाथ के साथ ही इस स्‍थान का विशद् महत्‍व है ।
1   बौद्ध   बौद्ध   PROPN   NNP Case=Nom|Gender=Masc|Number=Sing|Person=3   2   nmod    _   Vib=0|Tam=0|ChunkId=NP|ChunkType=child|Translit=bauddha
2   धर्मावलंबियों   धर्मावलंबी  NOUN    NN  Case=Acc|Gender=Masc|Number=Plur|Person=3   17  nmod    _   Vib=0_के_अनुसार|Tam=0|ChunkId=NP|ChunkType=head|Translit=dharmāvalaṁbiyoṁ
3   के  के  ADP PSP AdpType=Post    2   case    _   ChunkId=NP|ChunkType=child|Translit=ke
4   अनुसार  अनुसार  ADP PSP AdpType=Post    2   case    _   ChunkId=NP|ChunkType=child|Translit=anusāra
5   लुंबनी  लुंबनी  PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   17  nmod    _   SpaceAfter=No|Vib=0|Tam=0|ChunkId=NP2|ChunkType=head|Translit=luṁbanī
6   ,   COMMA   PUNCT   SYM _   7   punct   _   ChunkId=NP2|ChunkType=child|Translit=,
7   बोधगया  बोधगया  PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   5   conj    _   Vib=0|Tam=0|ChunkId=NP3|ChunkType=head|Translit=bodhagayā
8   और  और  CCONJ   CC  _   9   cc  _   ChunkId=CCP|ChunkType=head|Translit=aura
9   सारनाथ  सारनाथ  PROPN   NNP Case=Acc|Gender=Masc|Number=Sing|Person=3   5   conj    _   Vib=0_के_साथ|Tam=0|ChunkId=NP4|ChunkType=head|Translit=sāranātha
10  के  के  ADP PSP AdpType=Post    9   case    _   ChunkId=NP4|ChunkType=child|Translit=ke
11  साथ साथ ADP NST AdpType=Post|Case=Nom|Gender=Masc|Number=Sing|Person=3  9   case    _   AltTag=ADP-NOUN|ChunkId=NP4|ChunkType=child|Translit=sātha
12  ही  ही  PART    RP  _   9   dep _   ChunkId=NP4|ChunkType=child|Translit=hī
13  इस  यह  DET DEM Case=Acc|Number=Sing|Person=3|PronType=Dem  14  det _   ChunkId=NP5|ChunkType=child|Translit=isa
14  स्‍थान  स्थान   NOUN    NN  Case=Acc|Gender=Masc|Number=Sing|Person=3   17  nmod    _   Vib=0_का|Tam=0|ChunkId=NP5|ChunkType=head|Translit=sthāna
15  का  का  ADP PSP AdpType=Post|Case=Nom|Gender=Masc|Number=Sing   14  case    _   ChunkId=NP5|ChunkType=child|Translit=kā
16  विशद्   विशद्   ADJ JJ  Case=Nom    17  amod    _   ChunkId=NP6|ChunkType=child|Translit=viśad
17  महत्‍व  महत्व   NOUN    NN  Case=Nom|Gender=Masc|Number=Sing|Person=3   0   root    _   Vib=0|Tam=0|ChunkId=NP6|ChunkType=head|Translit=mahatva
18  है  है  AUX VM  Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin|Voice=Act 17  cop _   Vib=है|Tam=hE|ChunkId=VGF|ChunkType=head|Stype=declarative|Translit=hai
19  ।   ।   PUNCT   SYM _   17  punct   _   ChunkId=BLK|ChunkType=head|Translit=.

Answer 1

Use the debugger and watch your datfiles variable. Are there really all file paths in? glob.glob does not work recursively by default unless you explicitly specify. You my want to give a shot for this:

datfiles = glob.glob(os.path.join(datfolder, '**/*.dat'), recursive=True)

I was filing up a sample with only two text files in a test dir. And I got it to work. I'd recommend to start over with a new venv, beside that put your python script and 2 test files. Then run your code. It should do, mine did also.

Just a note: check your indentation and the '.' on the last line (before the comment).

tst.txt:

1  sifasf  ncadasfdv
2  asfdias  askfnhoas

tst1.txt:

1  ddsds
2  asfdgasfg
3  asgas

the script:

#! /path/to/your/venv/python/interprter
from io import open
from conllu import parse_incr
import os
import glob

#create a dict to store the results
word_lemma_dict = {}
ud_files = []
dat_files = []

#open the files and load the sentences to a list

datfolder = "./" #Folder where all the .txt files are stored.

datfiles = glob.glob(os.path.join(datfolder, '*.txt'))
print(datfiles)

for file in datfiles:
  data_file = open(file, "r", encoding = "utf-8")
  for tokenlist in parse_incr(data_file):
    ud_files.append(tokenlist)

print(ud_files)

and the output:

['./tst1.txt', './tst.txt']
[TokenList<ddsds, asfdgasfg, asgas>, TokenList<sifasf, asfdias>]

I bet you can add more files and it will do...

I am guessing it's a path / join or grammar to the conllu parser issue.

You might post some contents of your different *.dat files to be parsed to your expectation.