-
Notifications
You must be signed in to change notification settings - Fork 0
/
00_preproc_plenar.py
446 lines (363 loc) · 14.7 KB
/
00_preproc_plenar.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
# %%
from pprint import pprint
import json
import pandas as pd
import glob
from bs4 import BeautifulSoup
# %%
def parse_xml(file):
'''parse xml-files of plenary protocols of WP18 (2013-2017) for date and protocol-number'''
infile = open(file, "r", encoding='utf-8')
contents = infile.read()
soup = BeautifulSoup(contents, 'xml')
datum = soup.find('DATUM')
datum = datum.text
nummer = soup.find('NR')
nummer = nummer.text
return datum, nummer
# combine text from csv + date & number from xml
files_csv = glob.glob('app/data/raw/plenar/WP18/*.csv')
files_xml = glob.glob('app/data/raw/plenar/WP18/WP18_xml/*.xml')
files_csv.sort()
files_xml.sort()
#%%
dfs = []
for i, (file_csv, file_xml) in enumerate(zip(files_csv, files_xml)):
df = pd.read_csv(file_csv)
datum, nummer = parse_xml(file_xml)
df['datum'] = datum
df['nummer'] = nummer
df['filenum'] = str(file_xml)
dfs.append(df)
df = pd.concat(dfs)
# %%
# filter plenary speeches (exclude questions, zwischenrufe, etc.) and combine to list
df_filter = df.loc[df['type'] == 'speech', :]
# df_text = df_filter.groupby(['speaker_key', 'top_id'], as_index=False)['text'].apply(list)
g = df_filter.groupby(['speaker_key', 'top_id'])
df_text = g['text'].apply(list)
#%%
df_wp18 = pd.merge(pd.DataFrame(df_text), df_filter, on=[
'speaker_key', 'top_id'], how='inner')
#%%
# speeches as contineous string
df_wp18['text_clean'] = (
df_wp18['text_x']
.map(lambda x: [str(y) for y in x])
.map(lambda x: ' '.join(x))
)
df_wp18.drop_duplicates(subset='text_clean', keep="first", inplace=True)
# check number of speeches of mdb Volker Beck, should be 137 according to offenesparlament.de
df_wp18.loc[df_wp18['speaker_fp'] == 'volker-beck', :].shape
# df_wp18.to_csv('../../data/raw/plenar/plenar_WP18.csv', mode='w', index=False, sep='|')
# %%
df_wp18.loc[df_wp18['datum'].isna()]
# %%
# drop unnecessary columns, convert datum to datetime
# df_wp18.drop(labels=['Unnamed: 0', 'text', 0], axis=1, inplace=True)
df_wp18['datum'] = pd.to_datetime(df_wp18['datum'])
# df_wp18.sort_values(by='datum').iloc[[0, -1]]
# %%
def parse_xml(file):
print(file)
infile = open(file, "r", encoding='utf-8')
contents = infile.read()
soup = BeautifulSoup(contents, 'lxml')
reden = soup.find_all('rede')
daten = soup.find('veranstaltungsdaten')
datum = daten.find('datum')
date = datum['date']
plenar_file = []
for rede in reden:
# plenar_file.append(rede)
plenar = dict()
plenar['rede_id'] = rede['id']
# print(plenar['rede_id'])
vorname = rede.find('vorname')
nachname = rede.find('nachname')
namenszusatz = rede.find('namenszusatz')
if vorname == None:
vorname = 'NA'
else:
vorname = vorname.text
if nachname == None:
nachname = 'NA'
else:
nachname = nachname.text
if namenszusatz != None:
plenar['name'] = str(str(vorname) + ' ' +
str(namenszusatz.text) + ' ' + str(nachname))
else:
plenar['name'] = str(str(vorname) + ' ' + str(nachname))
rolle = rede.find('rolle_lang')
if rolle != None:
plenar['rolle'] = rolle.text
else:
plenar['rolle'] = 'NA'
fraktion = rede.find('fraktion')
if fraktion != None:
plenar['fraktion'] = fraktion.text
else:
plenar['fraktion'] = 'NA'
id_redner = rede.find('redner')
plenar['id_redner'] = id_redner['id']
plenar['id_rede'] = rede['id']
plenar['date'] = date
plenar['filenum'] = str(file)
# Clear: Kommentare, Namen und Text von (Vize-)Präsidenten
kommentare = rede.find_all('kommentar')
for kommentar in kommentare:
kommentar.clear()
rede_res = []
name_edits = rede.find_all('name')
for name_edit in name_edits:
next_p = name_edit.find_next_sibling("p")
while True:
if next_p == None:
break
try:
if next_p['klasse'] == 'redner':
check_redner = next_p.find('redner')
# print(check_redner)
if check_redner['id'] == plenar['id_redner']:
break
else:
next_p = next_p.find_next_sibling('p')
except:
break
else:
delete_p = next_p
next_p = next_p.find_next_sibling("p")
delete_p.clear()
name_edit.clear()
relevant = True
for p in rede.find_all('p'):
try:
if p['klasse'] == 'redner':
red = p.find('redner')
if red['id'] != plenar['id_redner']:
relevant = False
else:
relevant = True
else:
if relevant:
rede_res.append(p.text)
# print(p.text)
except:
if relevant:
rede_res.append(p.text)
plenar['text'] = " ".join(rede_res)
plenar_file.append(plenar)
return plenar_file
files = glob.glob('app/data/raw/plenar/WP19/*.xml')
# files = glob.glob('data/raw/plenar/WP19/19007-data.xml')
plenar_all = []
for file in files:
plenar_all.extend(parse_xml(file))
df_wp19 = pd.DataFrame(plenar_all)
df_wp19.shape
# %%
# RENAME
df_wp19[['name', 'rolle']] = df_wp19[['name', 'rolle']].fillna('')
df_wp19['speaker_cleaned'] = df_wp19.apply(
lambda x: x['name'].replace(x['rolle'], ''), axis=1)
df_wp19.rename(columns={'fraktion': 'speaker_party', 'date': 'datum',
'name': 'speaker', 'id_redner': 'speaker_key', 'text': 'text_clean'}, inplace=True)
df_wp19['wahlperiode'] = 19
df_wp19['datum'] = pd.to_datetime(df_wp19['datum'])
# df_wp19.sort_values(by='datum')
df_wp19
# %%
# concatenate wp18 & wp19
df_all = pd.concat([df_wp18, df_wp19], axis=0, ignore_index=True)
len(df_all)
# %%
df_all.dropna(subset=['text_clean'], inplace=True)
# df_all.sort_values(by='datum', inplace=True)
len(df_all)
# %%
# - merge with metadata
import os
# df = pd.read_csv('data/mdbs_metadata.csv')
DIR_META = os.path.join('app/data/', "mdbs_metadata.pkl")
df = pd.read_pickle(DIR_META)
df_plenar = df_all
# %%
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('Dr.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('Prof.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('Dr.-Ing.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('h.c.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('h. c.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('med.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('med')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].split('-Ing.')[-1], axis=1)
df_plenar['speaker_cleaned'] = df_plenar.apply(
lambda row: row['speaker_cleaned'].strip(), axis=1)
df_plenar['speaker_cleaned_'] = df_plenar['speaker_cleaned']
# %%
# to check
dfp = df_plenar
l1 = dfp.speaker_cleaned.unique()
# l2 = df.name_res.unique()
l2 = df.name.unique()
x = [y for y in l1 if y not in l2]
# %%
dfp.loc[dfp['speaker_cleaned'] == 'Angela Merkel,',
'speaker_cleaned'] = 'Angela Merkel'
dfp.loc[dfp['speaker_cleaned'] == 'Dagmar G. Wöhrl',
'speaker_cleaned'] = 'Dagmar Wöhrl'
dfp.loc[dfp['speaker_cleaned'] == 'Andreas G. Lämmel',
'speaker_cleaned'] = 'Andreas Lämmel'
dfp.loc[dfp['speaker_cleaned'] == 'Franz Josef Jung',
'speaker_cleaned'] = 'Franz-Josef Jung'
dfp.loc[dfp['speaker_cleaned'] == 'Aydan Özoğuz',
'speaker_cleaned'] = 'Aydan Özoguz'
dfp.loc[dfp['speaker_cleaned'] == 'Sevim Dağdelen',
'speaker_cleaned'] = 'Sevim Dagdelen'
dfp.loc[dfp['speaker_cleaned'] == 'Jan Ralf Nolte',
'speaker_cleaned'] = 'Jan Nolte'
dfp.loc[dfp['speaker_cleaned'] == 'Alexander Graf Graf Lambsdorff',
'speaker_cleaned'] = 'Alexander Graf Lambsdorff'
dfp.loc[dfp['speaker_cleaned'] == 'Michael Georg Link',
'speaker_cleaned'] = 'Michael Link'
dfp.loc[dfp['speaker_cleaned'] == 'Eberhardt Alexander Gauland',
'speaker_cleaned'] = 'Alexander Gauland'
dfp.loc[dfp['speaker_cleaned'] == 'Fabio De Masi',
'speaker_cleaned'] = 'Fabio de Masi'
dfp.loc[dfp['speaker_cleaned'] == 'Ulrich Oehme',
'speaker_cleaned'] = 'Ulrich Öhme'
dfp.loc[dfp['speaker_cleaned'] == 'Armin-Paulus Hampel',
'speaker_cleaned'] = 'Armin Paul Hampel'
dfp.loc[dfp['speaker_cleaned'] == 'Johann David Wadephul',
'speaker_cleaned'] = 'Johann Wadephul'
dfp.loc[dfp['speaker_cleaned'] == 'Joana Eleonora Cotar',
'speaker_cleaned'] = 'Joana Cotar'
dfp.loc[dfp['speaker_cleaned'] == 'Sonja Amalie Steffen',
'speaker_cleaned'] = 'Sonja Steffen'
dfp.loc[dfp['speaker_cleaned'] == 'Konstantin Elias Kuhle',
'speaker_cleaned'] = 'Konstantin Kuhle'
dfp.loc[dfp['speaker_cleaned'] == 'Roman Johannes Reusch',
'speaker_cleaned'] = 'Roman Reusch'
dfp.loc[dfp['speaker_cleaned'] == 'Gero Clemens Hocker',
'speaker_cleaned'] = 'Gero Hocker'
dfp.loc[dfp['speaker_cleaned'] == 'Ali',
'speaker_cleaned'] = 'Amira Mohamed Ali'
dfp.loc[dfp['speaker_cleaned'] == 'Christian Freiherr von Freiherr Stetten',
'speaker_cleaned'] = 'Christian Freiherr von Stetten'
dfp.loc[dfp['speaker_cleaned'] == 'Tobias Matthias Peterka',
'speaker_cleaned'] = 'Tobias Peterka'
dfp.loc[dfp['speaker_cleaned'] == 'Mariana Iris Harder-Kühnel',
'speaker_cleaned'] = 'Mariana Harder-Kühnel'
dfp.loc[dfp['speaker_cleaned'] == 'Johannes Graf Schraps',
'speaker_cleaned'] = 'Johannes Schraps'
dfp.loc[dfp['speaker_cleaned'] == 'Siegbert Droese',
'speaker_cleaned'] = 'Siegbert Dröse'
dfp.loc[dfp['speaker_cleaned'] == 'Martin Erwin Renner',
'speaker_cleaned'] = 'Martin E. Renner'
dfp.loc[dfp['speaker_cleaned'] == 'Bettina Margarethe Wiesmann',
'speaker_cleaned'] = 'Bettina Wiesmann '
dfp.loc[dfp['speaker_cleaned'] == 'Jan Ralf Graf Nolte',
'speaker_cleaned'] = 'Jan Nolte'
dfp.loc[dfp['speaker_cleaned'] == 'Gerd Graf Müller',
'speaker_cleaned'] = 'Gerd Müller'
dfp.loc[dfp['speaker_cleaned'] == 'Helin Evrim Sommer',
'speaker_cleaned'] = 'Evrim Sommer'
dfp.loc[dfp['speaker_cleaned'] == 'Udo Theodor Hemmelgarn',
'speaker_cleaned'] = 'Udo Hemmelgarn'
dfp.loc[dfp['speaker_cleaned'] == 'Eva-Maria Elisabeth Schreiber',
'speaker_cleaned'] = 'Eva Schreiber'
dfp.loc[dfp['speaker_cleaned'] == 'Norbert Maria Altenkamp',
'speaker_cleaned'] = 'Norbert Altenkamp'
dfp.loc[dfp['speaker_cleaned'] == 'Katharina Graf Dröge',
'speaker_cleaned'] = 'Katharina Dröge'
dfp.loc[dfp['speaker_cleaned'] == 'Britta Katharina Dassler',
'speaker_cleaned'] = 'Britta Dassler'
dfp.loc[dfp['speaker_cleaned'] == 'Michael Graf Leutert',
'speaker_cleaned'] = 'Michael Leutert'
dfp.loc[dfp['speaker_cleaned'] == 'Eva-Maria Schreiber',
'speaker_cleaned'] = 'Eva Schreiber'
dfp.loc[dfp['speaker_cleaned'] == 'Jens Graf Spahn',
'speaker_cleaned'] = 'Jens Spahn'
dfp.loc[dfp['speaker_cleaned'] == 'Rolf Graf Mützenich',
'speaker_cleaned'] = 'Rolf Mützenich'
dfp.loc[dfp['speaker_cleaned'] == 'Paul Viktor Podolay',
'speaker_cleaned'] = 'Paul Podolay'
dfp.loc[dfp['speaker_cleaned'] == 'Martin Graf Hebner',
'speaker_cleaned'] = 'Martin Hebner'
dfp.loc[dfp['speaker_cleaned'] == 'Albert H. Weiler',
'speaker_cleaned'] = 'Albert Weiler'
dfp.loc[dfp['speaker_cleaned'] == 'Jens Graf Kestner',
'speaker_cleaned'] = 'Jens Kestner'
dfp.loc[dfp['speaker_cleaned'] == 'Heidrun Bluhm-Förster',
'speaker_cleaned'] = 'Heidrun Bluhm'
dfp.loc[dfp['speaker_cleaned'] == 'Elvan Korkmaz-Emre',
'speaker_cleaned'] = 'Elvan Korkmaz'
dfp.loc[dfp['speaker_cleaned'] == 'Katharina Kloke',
'speaker_cleaned'] = 'katharina willkomm'
dfp.loc[dfp['speaker_cleaned'] == 'in der beek',
'speaker_cleaned'] = 'olaf in der beek'
#dfp.loc[dfp['speaker_cleaned'] == 'aaa'] = 'bbb'
# %%
dfp.speaker_cleaned = dfp.speaker_cleaned.apply(lambda x: x.lower().strip())
# %%
# check again
l1 = dfp.speaker_cleaned.unique()
l2 = df.name.unique()
remove = [y for y in l1 if y not in l2]
# %%
# df['is_add'] = 0
# df_add = pd.DataFrame({'name_res': ['Katharina Kloke', 'Hans-Peter Bartels'], 'is_add': [1, 1]})
# df = df.append(df_add, ignore_index = True)
dfp_drop = dfp.drop(dfp[dfp.speaker_cleaned.isin(remove)].index)
# %%
df_res = pd.merge(df, dfp_drop, left_on='name',
right_on='speaker_cleaned', how='inner', suffixes=('_left', '_right'))
df_res['typ'] = 'plenar'
# %%
df_res_rename = df_res.rename(columns={"name": "name_res", "filenum": "plenar_file", "id_rede": "plenar_id_rede", "wahlperiode": "plenar_wahlperiode",
"profile_url_left": "aw_profil_url", "profile": "social_media_profile", "text_clean": "text"})
# %%
df_res_rename.columns
# %%
df_clean = df_res_rename[['name_res', 'party', 'id_party', 'agw_18', 'agw_19', 'education', 'election_list', 'gender', 'social_media_profile',
'facebook', 'twitter', 'youtube', 'instagram', 'flickr', 'typ', 'datum', 'text', 'plenar_file', 'plenar_id_rede', 'plenar_wahlperiode']]
# %%
# set id for each speech and write textfiles to disc
# texts = [str(x) for x in df_clean['text']]
texts = df_clean['text'].tolist()
filenames = []
for i, text in enumerate(texts):
filename = 'plenar_{:0>6}.txt'.format(i)
filenames.append(filename)
with open('app/data/corpus/plenar/{}'.format(filename), "w") as text_file:
# try:
text_file.write(text)
# except:
# pass
df_clean['filename'] = filenames
df_clean['file_id'] = filenames
print(len(filenames))
print(df_clean.shape)
# %%
# save metadata as JSON
df_clean.rename(columns={'datum': 'date'}, inplace=True)
df_clean = df_clean.set_index('filename')
df_clean.loc[:, df_clean.columns != 'text'].to_json(
'plenar_meta.json', orient='index')
# pprint(data)
# print(data[]) "Dimension: ", data['cubes'][cube]['dim']
# %%
# check
json_file = 'app/data/corpus/plenar_meta.json'
with open(json_file) as json_data:
data = json.load(json_data)
for i in range(10):
pprint(data['plenar_00{}000.txt'.format(i)])