# -*- coding: utf-8 -*-
#Author: Pierre Ratinaud
+#Copyright (c) 2008-2020 Pierre Ratinaud
+#modification pour python 3 : Laurent Mérat, 6x7 - mai 2020
+#License: GNU/GPL
+#------------------------------------
+# import des modules python
+#------------------------------------
import codecs
import os
-import gettext
-_ = gettext.gettext
import locale
import sys
from time import time
-from functions import decoupercharact, ReadDicoAsDico, DoConf, ReadLexique, progressbar
import re
import sqlite3
import itertools
import logging
from operator import itemgetter
from uuid import uuid4
+import datetime
+from copy import copy
+
+#------------------------------------
+# import des fichiers du projet
+#------------------------------------
+from functions import decoupercharact, ReadDicoAsDico, DoConf, ReadLexique, progressbar
from chemins import PathOut
from dialog import CorpusPref, SubTextFromMetaDial, MergeClusterFrame
-from copy import copy
from colors import colors
-import datetime
+
+import langue
+langue.run()
log = logging.getLogger('iramuteq.corpus')
def CopyUce(uce) :
return Uce(uce.ident, uce.para, uce.uci)
-
def CopyUci(uci):
nuci = Uci(uci.ident, '')
nuci.etoiles = copy(uci.etoiles)
return nuci
-
class Corpus :
+
"""Corpus class
list of text
"""
gramtype = self.parent.lexique[word][1]
lem = self.parent.lexique[word][0]
elif word.isdigit() :
- gramtype = u'num'
+ gramtype = 'num'
lem = word
else :
- gramtype = u'nr'
+ gramtype = 'nr'
lem = word
self.formes[word] = Word(word, gramtype, len(self.formes), lem)
self.idformesuces[self.formes[word].ident] = {self.ucis[-1].uces[-1].ident : 1}
res = self.ccorpus.execute('SELECT * FROM etoiles;')
for row in res :
self.ucis.append(Uci(row[0], row[1], row[2]))
- uces = self.conncorpus.cursor().execute('SELECT * FROM luces where uci=?;',(`self.ucis[-1].ident`,))
+ uces = self.conncorpus.cursor().execute('SELECT * FROM luces where uci=?;',(repr(self.ucis[-1].ident),))
for uce in uces:
self.ucis[-1].uces.append(Uce(uce[2], uce[1], uce[0]))
res = self.ccorpus.execute('SELECT * FROM formes;')
self.ccorpus.close()
def getworduces(self, wordid) :
- if isinstance(wordid, basestring) :
+ if isinstance(wordid, str) :
wordid = self.formes[wordid].ident
- res = self.cformes.execute('SELECT uces FROM uces where id=? ORDER BY id;', (`wordid`,))
+ res = self.cformes.execute('SELECT uces FROM uces where id=? ORDER BY id;', (repr(wordid),))
return list(itertools.chain(*[[int(val) for val in row[0].split()] if not isinstance(row[0], int) else [row[0]] for row in res]))
def getworducis(self, wordid) :
return list(set([self.getucefromid(uce).uci for uce in res]))
def getformeuceseff(self, formeid) :
- if isinstance(formeid, basestring) :
+ if isinstance(formeid, str) :
formeid = self.formes[formeid].ident
- res = self.cformes.execute('SELECT uces FROM uces where id=? ORDER BY id;', (`formeid`,))
+ res = self.cformes.execute('SELECT uces FROM uces where id=? ORDER BY id;', (repr(formeid),))
uces = list(itertools.chain(*[[int(val) for val in row[0].split()] if not isinstance(row[0], int) else [row[0]] for row in res]))
query = 'SELECT eff FROM eff where id=%i ORDER BY id' % formeid
res = self.cformes.execute(query)
return formeuceeff
def getlemuces(self, lem) :
- formesid = ', '.join([`val` for val in self.lems[lem].formes])
+ formesid = ', '.join([repr(val) for val in self.lems[lem].formes])
query = 'SELECT uces FROM uces where id IN (%s) ORDER BY id' % formesid
res = self.cformes.execute(query)
return list(set(list(itertools.chain(*[[int(val) for val in row[0].split()] if not isinstance(row[0], int) else [row[0]] for row in res]))))
if lem in self.lems :
formesid += self.lems[lem].formes
else :
- print 'abscent : %s' % lem
+ print('abscent : %s' % lem)
query = 'SELECT uces FROM uces where id IN %s ORDER BY id' % str(tuple(formesid))
res = self.cformes.execute(query)
return list(set(list(itertools.chain(*[[int(val) for val in row[0].split()] if not isinstance(row[0], int) else [row[0]] for row in res]))))
self.tgenlem[lem] = [0] * clnb
self.tgenlem[lem][i] = len(set(lemst).intersection(classe))
else :
- print 'abscent: ',lem
+ print('abscent: ',lem)
return list(set(tgenst))
def gettgentxt(self, tgen):
return list(set([self.getucefromid(val).uci for val in uces]))
def getlemuceseff(self, lem, luces = None) :
- formesid = ', '.join([`val` for val in self.lems[lem].formes])
+ formesid = ', '.join([repr(val) for val in self.lems[lem].formes])
query = 'SELECT uces FROM uces where id IN (%s) ORDER BY id' % formesid
res = self.cformes.execute(query)
uces = list(itertools.chain(*[[int(val) for val in row[0].split()] if not isinstance(row[0], int) else [row[0]] for row in res]))
return [len(uce[1].split()) for uce in res]
def getconcorde(self, uces) :
- return self.cuces.execute('select * from uces where id IN (%s) ORDER BY id;' % ', '.join([`i` for i in uces]))
+ return self.cuces.execute('select * from uces where id IN (%s) ORDER BY id;' % ', '.join([repr(i) for i in uces]))
def getuciconcorde(self, ucis) :
uces = [[val,[uce.ident for uce in self.ucis[val].uces]] for val in ucis]
def getactivesnb(self, key) :
return len([lem for lem in self.lems if self.lems[lem].act == key])
+
+# fonction inactive mais avec une incertitude concernant l'indentation sur le dernier else
# def make_lems(self, lem = True) :
# log.info('make lems')
# self.lems = {}
lem = dictionnaire[forme][0]
gram = dictionnaire[forme][1]
elif forme.isdigit() :
- gram = u'num'
+ gram = 'num'
lem = forme
else :
- gram = u'nr'
+ gram = 'nr'
lem = forme
self.formes[forme].lem = lem
self.formes[forme].gram = gram
self.iduces = dict([[uce.ident, uce] for uci in self.ucis for uce in uci.uces])
def make_lexitable(self, mineff, etoiles, gram = 0) :
- log.info('making lexical table...')
if gram == 0 :
grams = {1:'', 2:''}
else :
tab = []
for lem in tokeep :
deff = self.getlemuceseff(lem)
- ucesk = deff.keys()
+ ucesk = list(deff.keys())
line = [lem] + [sum([deff[uce] for uce in et.intersection(ucesk)]) for et in etuces]
if sum(line[1:]) >= mineff :
tab.append(line)
for forme in self.formes :
formeuceeff = self.getformeuceseff(forme)
for i, classe in enumerate(lclasses) :
- concern = sets[i].intersection(formeuceeff.keys())
+ concern = sets[i].intersection(list(formeuceeff.keys()))
if len(concern) :
totoccurrences[etoiles[i]] += sum([formeuceeff[uce] for uce in concern])
#tgenoccurrences = dict([[val, 0] for val in etoiles])
for lem in tgen[t] :
lemuceeff = self.getlemuceseff(lem)
for i, classe in enumerate(lclasses) :
- concern = sets[i].intersection(lemuceeff.keys())
+ concern = sets[i].intersection(list(lemuceeff.keys()))
if len(concern) :
tgenoccurrences[t][etoiles[i]] += sum([lemuceeff[uce] for uce in concern])
return tgenoccurrences, totoccurrences
-
+
def make_tgen_profile(self, tgen, ucecl, uci = False) :
log.info('tgen/classes')
self.tgenlem = {}
tab = [[lem] + [len(set(self.gettgenstprof(tgen[lem], classe, i, clnb)).intersection(classe)) for i, classe in enumerate(ucecl)] for lem in tgen]
tab = [[line[0]] + [val for val in line[1:]] for line in tab if sum(line[1:]) >= 3]
return tab
+
#i = 0
#nam = 'total'
#while nam + `i` in tgen :
#tab = [line0] + tab
#with open(fileout, 'w') as f :
# f.write('\n'.join(['\t'.join(line) for line in tab]).encode(self.parametres['syscoding']))
-
+
def make_efftype_from_etoiles(self, etoiles) :
dtype = {}
etuces = [[] for et in etoiles]
etuces = [set(val) for val in etuces]
for lem in self.lems :
deff = self.getlemuceseff(lem)
- ucesk = deff.keys()
+ ucesk = list(deff.keys())
gram = self.lems[lem].gram
if gram in dtype :
dtype[gram] = [i + j for i, j in zip(dtype[gram], [sum([deff[uce] for uce in et.intersection(ucesk)]) for et in etuces])]
log.info('taille uc1 : %i - taille uc2 : %i' % (len(uc1), len(uc2)))
self.write_ucmatrix(uc1, actives, uc1out)
self.write_ucmatrix(uc2, actives, uc2out)
- listuce1 = [['uce', 'uc']] + [[`uce`, `i`] for i, ucl in enumerate(uc1) for uce in ucl]
- listuce2 = [['uce', 'uc']] + [[`uce`, `i`] for i, ucl in enumerate(uc2) for uce in ucl]
+ listuce1 = [['uce', 'uc']] + [[repr(uce), repr(i)] for i, ucl in enumerate(uc1) for uce in ucl]
+ listuce2 = [['uce', 'uc']] + [[repr(uce), repr(i)] for i, ucl in enumerate(uc2) for uce in ucl]
with open(listuce1out, 'w') as f :
f.write('\n'.join([';'.join(line) for line in listuce1]))
with open(listuce2out, 'w') as f :
for uce in self.getlemuces(lem):
if (uces_uc[uce], i) not in deja_la :
nbl += 1
- f.write(''.join([' '.join([`uces_uc[uce]+1`,`i+1`,`1`]),'\n']))
+ f.write(''.join([' '.join([repr(uces_uc[uce]+1),repr(i+1),repr(1)]),'\n']))
deja_la[(uces_uc[uce], i)] = 0
f.seek(0)
with open(fileout, 'w') as ffin :
with open(outf,'w') as f :
for uce in res :
if self.iduces[uce[0]].uci == actuci and self.iduces[uce[0]].para == actpara :
- f.write(uce[1].encode(self.parametres['syscoding']) + '\n')
+ f.write(uce[1] + '\n')
elif self.iduces[uce[0]].uci != actuci :
actuci = self.iduces[uce[0]].uci
if self.ucis[self.iduces[uce[0]].uci].paras == [] :
actpara = self.iduces[uce[0]].para
- f.write('\n' + ' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles).encode(self.parametres['syscoding']) + '\n' + uce[1].encode(self.parametres['syscoding']) + '\n')
+ f.write('\n' + ' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles) + '\n' + uce[1] + '\n')
else :
ident = 0
actpara = self.iduces[uce[0]].para
- f.write('\n'.join([' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles).encode(self.parametres['syscoding']), self.ucis[self.iduces[uce[0]].uci].paras[ident].encode(self.parametres['syscoding']), uce[1].encode(self.parametres['syscoding'])]) + '\n')
+ f.write('\n'.join([' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles), self.ucis[self.iduces[uce[0]].uci].paras[ident], uce[1]] + '\n'))
elif self.iduces[uce[0]].para != actpara :
actpara = self.iduces[uce[0]].para
ident += 1
- f.write('\n'.join([self.ucis[self.iduces[uce[0]].uci].paras[ident].encode(self.parametres['syscoding']), uce[1].encode(self.parametres['syscoding'])]) + '\n')
-
+ f.write('\n'.join([self.ucis[self.iduces[uce[0]].uci].paras[ident], uce[1]]) + '\n')
+
def export_meta_table(self, outf) :
- metas = [[`i`] + text.etoiles[1:] for i, text in enumerate(self.ucis)]
+ metas = [[repr(i)] + text.etoiles[1:] for i, text in enumerate(self.ucis)]
longueur_max = max([len(val) for val in metas])
first = ['column_%i' % i for i in range(longueur_max)]
metas.insert(0, first)
with open(outf, 'w') as f :
- f.write('\n'.join(['\t'.join(line) for line in metas]).encode(self.parametres['syscoding']))
-
+ f.write('\n'.join(['\t'.join(line) for line in metas]))
+
def export_corpus_classes(self, outf, alc = True, lem = False, uci = False) :
ucecl = {}
for i, lc in enumerate(self.lc) :
etline = ' '.join(self.ucis[actuci].etoiles + ['*classe_%i' % ucecl[uce[0]]])
else :
etline = ' '.join(['<' + '='.join(et.split('_')) + '>' for et in self.ucis[actuci].etoiles[1:]])
- f.write(etline.encode(self.parametres['syscoding']) + '\n')
- f.write(guce.encode(self.parametres['syscoding']) + '\n\n')
+ f.write(etline + '\n')
+ f.write(guce + '\n\n')
def export_classe(self, outf, classe, lem = False, uci = False) :
sts = self.lc[classe - 1]
for uce in res :
guce = uce[1]
if not uci :
- f.write(' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles).encode(self.parametres['syscoding']) + '\n')
+ f.write(' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles) + '\n')
else :
- f.write(' '.join(self.ucis[uce[0]].etoiles).encode(self.parametres['syscoding']) + '\n')
+ f.write(' '.join(self.ucis[uce[0]].etoiles) + '\n')
if lem :
guce = ' '.join([self.formes[forme].lem for forme in guce.split()])
- f.write(guce.encode(self.parametres['syscoding']) + '\n\n')
-
+ f.write(guce + '\n\n')
+
def export_owledge(self, rep, classe, lem = False, uci = False) :
sts = self.lc[classe - 1]
if not uci :
for uce in res :
ident = uce[0]
guce = uce[1]
- outf = '.'.join([`ident`, 'txt'])
+ outf = '.'.join([repr(ident), 'txt'])
outf = os.path.join(rep, outf)
if lem :
guce = ' '.join([self.formes[forme].lem for forme in guce.split()])
with open(outf, 'w') as f :
- f.write(guce.encode('cp1252', errors = 'replace'))
+ f.write(guce) #.encode('cp1252', errors = 'replace'))
def export_tropes(self, fileout, classe, lem = False, uci = False) :
sts = self.lc[classe - 1]
guce = uce[1]
if lem :
guce = ' '.join([self.formes[forme].lem for forme in guce.split()])
- f.write(guce.encode('cp1252', errors = 'replace'))
+ f.write(guce) #.encode('cp1252', errors = 'replace'))
f.write('\n')
def make_and_write_sparse_matrix_from_uces(self, actives, outfile, listuce = False) :
for i, lem in enumerate(actives) :
for uce in sorted(self.getlemuces(lem)) :
nbl += 1
- f.write(''.join([' '.join([`uce+1`, `i+1`,`1`]),'\n']))
+ f.write(''.join([' '.join([repr(uce+1), repr(i+1),repr(1)]),'\n']))
f.seek(0)
with open(outfile, 'w') as ffin :
ffin.write("%%%%MatrixMarket matrix coordinate integer general\n%i %i %i\n" % (self.getucenb(), len(actives), nbl))
os.remove(outfile + '~')
if listuce :
with open(listuce, 'w') as f :
- f.write('\n'.join(['uce;uc'] + [';'.join([`i`,`i`]) for i in range(0, self.getucenb())]))
+ f.write('\n'.join(['uce;uc'] + [';'.join([repr(i),repr(i)]) for i in range(0, self.getucenb())]))
def make_and_write_sparse_matrix_from_uci(self, actives, outfile, listuci = False) :
log.info('make_and_write_sparse_matrix_from_ucis %s' % outfile)
for i, lem in enumerate(actives) :
for uci in sorted(self.getlemucis(lem)) :
nbl += 1
- f.write(''.join([' '.join([`uci+1`, `i+1`,`1`]),'\n']))
+ f.write(''.join([' '.join([repr(uci+1), repr(i+1),repr(1)]),'\n']))
f.seek(0)
with open(outfile, 'w') as ffin :
ffin.write("%%%%MatrixMarket matrix coordinate integer general\n%i %i %i\n" % (self.getucinb(), len(actives), nbl))
os.remove(outfile + '~')
if listuci :
with open(listuci, 'w') as f :
- f.write('\n'.join(['uci;uc'] + [';'.join([`i`,`i`]) for i in range(0, self.getucinb())]))
+ f.write('\n'.join(['uci;uc'] + [';'.join([repr(i),repr(i)]) for i in range(0, self.getucinb())]))
def make_and_write_sparse_matrix_from_classe(self, actives, uces, outfile) :
log.info('make_and_write_sparse_matrix_from_classe %s' % outfile)
for i, lem in enumerate(actives) :
uces_ok = list(set(self.getlemuces(lem)).intersection(uces))
for uce in uces_ok :
- f.write(''.join([' '.join([`duces[uce]+1`,`i+1`,`1`]),'\n']))
+ f.write(''.join([' '.join([repr(duces[uce]+1),repr(i+1),repr(1)]),'\n']))
f.seek(0)
with open(outfile, 'w') as ffin :
ffin.write("%%%%MatrixMarket matrix coordinate integer general\n%i %i %i\n" % (len(uces), len(actives), nbl))
tab = [[lem] + [len(set(self.getlemucis(lem)).intersection(classe)) for classe in ucecl] for lem in actives]
else :
tab = [[lem] + [len(set(self.getlemuces(lem)).intersection(classe)) for classe in ucecl] for lem in actives]
- tab = [[line[0]] + [`val` for val in line[1:]] for line in tab if sum(line[1:]) >= 3]
+ tab = [[line[0]] + [repr(val) for val in line[1:]] for line in tab if sum(line[1:]) >= 3]
with open(fileout, 'w') as f :
- f.write('\n'.join([';'.join(line) for line in tab]).encode(self.parametres['syscoding']))
- f.write('\n')
+ f.write('\n'.join([';'.join(line) for line in tab]))
def make_etoiles(self) :
etoiles = set([])
for uci in self.ucis :
etoiles.update(uci.etoiles[1:])
return list(etoiles)
-
+
def make_themes(self):
themes = set([])
for uci in self.ucis :
except IndexError :
det[et[0]] = 1
return det
-
+
def make_theme_dict(self):
themes = [val for uci in self.ucis for val in uci.paras]
det = {}
etoileuces = self.getetoileuces()
else :
etoileuces = self.getetoileucis()
- print 'etoilesuces ok'
- etoileuces = dict([[et, etoileuces[et]] for et in etoileuces if
- len(etoileuces[et]) > 1 ]) #and not et.startswith(u'*reference_')
- print len(etoileuces)
- print 'etoilesuces ok2'
+ etoileuces = dict([[et, etoileuces[et]] for et in etoileuces if len(etoileuces[et]) > 1])
with open(fileout, 'w') as f :
- print 'write...'
- f.write('\n'.join([';'.join([et] + [`len(set(etoileuces[et]).intersection(classe))` for classe in ucecl]) for et in etoileuces]).encode(self.parametres['syscoding']))
+ f.write('\n'.join([';'.join([et] + [repr(len(set(etoileuces[et]).intersection(classe))) for classe in ucecl]) for et in etoileuces])) #.encode(self.parametres['syscoding'])
#etoiles = self.make_etoiles()
#with open(fileout, 'w') as f :
# f.write('\n'.join([';'.join([etoile] + [`len(set(self.getucesfrometoile(etoile)).intersection(classe))` for classe in ucecl]) for etoile in etoiles]).encode(self.parametres['syscoding']))
return txt + '\n</body></html>'
def make_cut_corpus(self, uci = False) :
- txt = u''
+ txt = ''
if not uci :
res = self.getalluces()
self.make_iduces()
for uce in res :
if self.iduces[uce[0]].uci != actuci :
actuci = self.iduces[uce[0]].uci
- txt += u'\n' + ' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles) + u'\n'
- txt += ''.join([u'\n',uce[1],u'\n'])
+ txt += '\n' + ' '.join(self.ucis[self.iduces[uce[0]].uci].etoiles) + '\n'
+ txt += ''.join(['\n',uce[1],'\n'])
else :
- txt += ''.join([u'\n',uce[1],u'\n'])
+ txt += ''.join(['\n',uce[1],'\n'])
else :
res = self.getallucis()
actuci = ''
for uce in res :
if self.ucis[uce[0]].ident != actuci :
actuci = self.ucis[uce[0]].ident
- txt += u'\n' + ' '.join(self.ucis[self.ucis[uce[0]].ident].etoiles) + u'\n'
- txt += ''.join([u'\n',uce[1],u'\n'])
+ txt += '\n' + ' '.join(self.ucis[self.ucis[uce[0]].ident].etoiles) + '\n'
+ txt += ''.join(['\n',uce[1],'\n'])
else :
- txt += ''.join([u'\n',uce[1],u'\n'])
+ txt += ''.join(['\n',uce[1],'\n'])
return txt
def count_from_list(self, l, d) :
res[gram] = [0 for val in self.lc]
lemuceeff = self.getlemuceseff(lem)
for i, classe in enumerate(self.lc) :
- concern = set(classe).intersection(lemuceeff.keys())
+ concern = set(classe).intersection(list(lemuceeff.keys()))
res[gram][i] += sum([lemuceeff[uce] for uce in concern])
- res = [[gram] + [`val` for val in res[gram]] for gram in res]
+ res = [[gram] + [repr(val) for val in res[gram]] for gram in res]
res.sort()
with open(outf, 'w') as f :
- f.write('\n'.join([';'.join(line) for line in res]).encode(self.parametres['syscoding']))
-
+ f.write('\n'.join([';'.join(line) for line in res]))
def make_ucecl_from_R(self, filein) :
- with open(filein, 'rU') as f :
+ with open(filein, 'r') as f :
c = f.readlines()
c.pop(0)
self.lc = []
for forme in self.formes :
formeuceeff = self.getformeuceseff(forme)
for i, classe in enumerate(lclasses) :
- concern = sets[i].intersection(formeuceeff.keys())
+ concern = sets[i].intersection(list(formeuceeff.keys()))
if len(concern) :
occurrences[i+1] += sum([formeuceeff[uce] for uce in concern])
formescl[i+1] += 1
hapaxcl[i+1] += 1
log.info('%f' % (time() - t1))
if outf is not None :
- toprint = '\n'.join([';'.join([`i`, `occurrences[i]`, `formescl[i]`, `hapaxcl[i]`, `lenclasses[i]`, `float(hapaxcl[i])/float(formescl[i])`]) for i in occurrences])
+ toprint = '\n'.join([';'.join([repr(i), repr(occurrences[i]), repr(formescl[i]), repr(hapaxcl[i]), repr(lenclasses[i]), repr(float(hapaxcl[i])/float(formescl[i]))]) for i in occurrences])
with open(outf, 'w') as f :
f.write(toprint)
else :
- return [[`occurrences[i]`, `formescl[i]`, `hapaxcl[i]`, `lenclasses[i]`, `float(hapaxcl[i])/float(formescl[i])`] for i in occurrences]
+ return [[repr(occurrences[i]), repr(formescl[i]), repr(hapaxcl[i]), repr(lenclasses[i]), repr(float(hapaxcl[i])/float(formescl[i]))] for i in occurrences]
def get_stat_by_et(self, outf, etoiles) :
lclasses = [self.getucesfrometoile(etoile) for etoile in etoiles]
huces[hucesdict[uce][0]].append(uce)
else :
huces[hucesdict[uce][0]] = [uce]
- huces = zip(huces, huces.values())
+ huces = list(zip(huces, list(huces.values())))
huces.sort(reverse=True)
txt = """
<html><body>
def export_dictionary(self, fileout, syscoding) :
listformes = [[self.formes[forme].freq, forme, self.formes[forme].lem, self.formes[forme].gram] for forme in self.formes]
listformes.sort(reverse = True)
- listformes = [forme[1:] + [`forme[0]`] for forme in listformes]
+ listformes = [forme[1:] + [repr(forme[0])] for forme in listformes]
with open(fileout, 'w') as f :
- f.write('\n'.join(['\t'.join(forme) for forme in listformes]).encode(syscoding))
+ f.write('\n'.join(['\t'.join(forme) for forme in listformes]))
def export_lems(self, fileout, syscoding) :
self.make_idformes()
- listlem = [[lem, '\t'.join(['\t'.join([self.idformes[forme].forme, `self.lems[lem].formes[forme]`]) for forme in self.lems[lem].formes])] for lem in self.lems]
+ listlem = [[lem, '\t'.join(['\t'.join([self.idformes[forme].forme, repr(self.lems[lem].formes[forme])]) for forme in self.lems[lem].formes])] for lem in self.lems]
listlem.sort()
with open(fileout, 'w') as f :
- f.write('\n'.join(['\t'.join(lem) for lem in listlem]).encode(syscoding))
-
+ f.write('\n'.join(['\t'.join(lem) for lem in listlem]))
class MakeUciStat :
self.formes[forme.ident] = forme.freq
self.freq += forme.freq
+
def decouperlist(chaine, longueur, longueurOptimale) :
"""
on part du dernier caractère, et on recule jusqu'au début de la chaîne.
Si on trouve un '$', c'est fini.
Sinon, on cherche le meilleur candidat. C'est-à-dire le rapport poids/distance le plus important.
"""
- separateurs = [[u'.', 6.0], [u'?', 6.0], [u'!', 6.0], [u'£$£', 6.0], [u':', 5.0], [u';', 4.0], [u',', 1.0], [u' ', 0.01]]
+ separateurs = [['.', 6.0], ['?', 6.0], ['!', 6.0], ['£$£', 6.0], [':', 5.0], [';', 4.0], [',', 1.0], [' ', 0.01]]
dsep = dict([[val[0],val[1]] for val in separateurs])
trouve = False # si on a trouvé un bon séparateur
iDecoupe = 0 # indice du caractere ou il faut decouper
-
longueur = min(longueur, len(chaine) - 1)
chaineTravail = chaine[:longueur + 1]
nbCar = longueur
meilleur = ['', 0, 0] # type, poids et position du meilleur separateur
-
try :
- indice = chaineTravail.index(u'$')
+ indice = chaineTravail.index('$')
trouve = True
iDecoupe = indice - 1
except ValueError :
nbCar = nbCar - 1
# si on a trouvé
if trouve:
+
#if meilleur[0] != ' ' :
# fin = chaine[iDecoupe + 1:]
# retour = chaineTravail[:iDecoupe]
#else :
+
fin = chaine[iDecoupe + 1:]
retour = chaineTravail[:iDecoupe + 1]
return len(retour) > 0, retour, fin
return False, chaine, ''
def testetoile(line) :
- return line.startswith(u'****')
+ return line.startswith('****')
def testint(line) :
- return line[0:4].isdigit() and u'*' in line
+ return line[0:4].isdigit() and '*' in line
def prep_txtlist(txt) :
- return txt.split() + [u'$']
+ return txt.split() + ['$']
def prep_txtcharact(txt) :
- return txt + u'$'
+ return txt + '$'
+
class BuildCorpus :
"""
Class for building a corpus
"""
+
def __init__(self, infile, parametres_corpus, lexique = None, expressions = None, dlg = None) :
log.info('begin building corpus...')
self.lexique = lexique
self.corpus.pathout = PathOut(filename = parametres_corpus['originalpath'], dirout = parametres_corpus['pathout'])
self.corpus.pathout.createdir(parametres_corpus['pathout'])
self.corpus.parametres['uuid'] = str(uuid4())
- self.corpus.parametres['corpus_name'] = parametres_corpus['corpus_name']#os.path.split(self.corpus.parametres['pathout'])[1]
+ self.corpus.parametres['corpus_name'] = parametres_corpus['corpus_name'] #os.path.split(self.corpus.parametres['pathout'])[1]
self.corpus.parametres['type'] = 'corpus'
if self.corpus.parametres['keep_ponct'] :
self.ponctuation_espace = [' ', '']
else :
- self.ponctuation_espace = [' ','.', u'£$£', ';', '?', '!', ',', ':','']
+ self.ponctuation_espace = [' ','.', '£$£', ';', '?', '!', ',', ':','']
self.cleans = []
self.tolist = self.corpus.parametres.get('tolist', 0)
self.buildcleans()
t1 = time()
try :
self.read_corpus(self.infile)
- except Warning, args :
+ except Warning as args :
log.info('pas kool %s' % args)
raise Warning
else :
if self.corpus.parametres.get('firstclean', 1) :
self.cleans.append(self.firstclean)
if self.corpus.parametres['charact'] :
- self.rule = self.corpus.parametres.get('keep_caract', u"^a-zA-Z0-9àÀâÂäÄáÁéÉèÈêÊëËìÌîÎïÏòÒôÔöÖùÙûÛüÜçÇßœŒ’ñ.:,;!?*'_")
+ self.rule = self.corpus.parametres.get('keep_caract', "^a-zA-Z0-9àÀâÂäÄáÁéÉèÈêÊëËìÌîÎïÏòÒôÔöÖùÙûÛüÜçÇßœŒ’ñ.:,;!?*'_")
self.cleans.append(self.docharact)
if self.corpus.parametres.get('expressions', 1) :
self.cleans.append(self.make_expression)
self.cleans.append(self.dotiret)
def make_expression(self,txt) :
- exp = self.expressions.keys()
+ exp = list(self.expressions.keys())
exp.sort(reverse=True)
for expression in exp :
if expression in txt :
def docharact(self, txt) :
#rule = u"^a-zA-Z0-9àÀâÂäÄáÁéÉèÈêÊëËìÌîÎïÏòÒôÔöÖùÙûÛüÜçÇßœŒ’ñ.:,;!?*'_-"
- list_keep = u"[" + self.rule + "]+"
+ list_keep = "[" + self.rule + "]+"
return re.sub(list_keep, ' ', txt)
def doapos(self, txt) :
- return txt.replace(u'\'', u' ')
+ return txt.replace('\'', ' ')
def dotiret(self, txt) :
- return txt.replace(u'-', u' ')
+ return txt.replace('-', ' ')
def firstclean(self, txt) :
- txt = txt.replace(u'’',"'")
- txt = txt.replace(u'œ', u'oe')
- return txt.replace('...',u' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace(u'…', u' £$£ ')
+ txt = txt.replace('’',"'")
+ txt = txt.replace('œ', 'oe')
+ return txt.replace('...',' £$£ ').replace('?',' ? ').replace('.',' . ').replace('!', ' ! ').replace(',',' , ').replace(';', ' ; ').replace(':',' : ').replace('…', ' £$£ ')
def make_cleans(self, txt) :
for clean in self.cleans :
def backup_uce(self) :
if self.corpus.idformesuces != {} :
log.info('backup %i' % len(self.corpus.idformesuces))
- touce = [(`forme`, ' '.join([`val` for val in self.corpus.idformesuces[forme].keys()])) for forme in self.corpus.idformesuces]
- toeff = [(`forme`, ' '.join([`val` for val in self.corpus.idformesuces[forme].values()])) for forme in self.corpus.idformesuces]
+ touce = [(repr(forme), ' '.join([repr(val) for val in list(self.corpus.idformesuces[forme].keys())])) for forme in self.corpus.idformesuces]
+ toeff = [(repr(forme), ' '.join([repr(val) for val in list(self.corpus.idformesuces[forme].values())])) for forme in self.corpus.idformesuces]
self.cf.executemany('INSERT INTO uces VALUES (?,?);', touce)
self.cf.executemany('INSERT INTO eff VALUES (?,?);', toeff)
self.corpus.idformesuces = {}
for uci in self.corpus.ucis :
self.ccorpus.execute('INSERT INTO etoiles VALUES (?,?,?);' ,(uci.ident,' '.join(uci.etoiles), ' '.join(uci.paras,)))
for uce in uci.uces :
- self.ccorpus.execute('INSERT INTO luces VALUES (?,?,?);',(`uci.ident`,`uce.para`,`uce.ident`,))
+ self.ccorpus.execute('INSERT INTO luces VALUES (?,?,?);',(repr(uci.ident),repr(uce.para),repr(uce.ident),))
for forme in self.corpus.formes :
- self.ccorpus.execute('INSERT INTO formes VALUES (?,?,?,?,?);', (`self.corpus.formes[forme].ident`, forme, self.corpus.formes[forme].lem, self.corpus.formes[forme].gram, `self.corpus.formes[forme].freq`,))
+ self.ccorpus.execute('INSERT INTO formes VALUES (?,?,?,?,?);', (repr(self.corpus.formes[forme].ident), forme, self.corpus.formes[forme].lem, self.corpus.formes[forme].gram, repr(self.corpus.formes[forme].freq),))
log.info('%f' % (time() - t))
def dofinish(self) :
self.corpus.parametres['hapax'] = '%i - %.2f %% des formes - %.2f %% des occurrences' % (hapaxnb, pourhapaxf, pourhapaxocc)
class BuildSubCorpus(BuildCorpus):
+
def __init__(self, corpus, parametres, dlg = None) :
log.info('begin subcorpus...')
self.dlg = dlg
self.corpus.parametres['meta'] = parametres.get('meta', False)
self.corpus.parametres['uuid'] = str(uuid4())
if parametres.get('frommeta', False) :
- print 'make subtexts'
+ print('make subtexts')
self.corpus.ucis = [CopyUci(uci) for uci in self.ori.ucis if set(parametres['meta']).intersection(uci.etoiles) != set()]
elif parametres.get('fromtheme', False) :
- print 'make subtexts from theme'
+ print('make subtexts from theme')
idpara = 0
for uci in self.ori.ucis :
if uci.paras != [] :
self.dobuild()
def fromuceids(self):
- print 'fromuceids'
- dictucekeep = dict(zip(self.parametres['uceids'], self.parametres['uceids']))
+ print('fromuceids')
+ dictucekeep = dict(list(zip(self.parametres['uceids'], self.parametres['uceids'])))
idpara = 0
for uci in self.ori.ucis :
if uci.paras == [] :
newuces = []
newpara = []
for et in uci.paras :
- keepuces = [CopyUce(uce) for uce in uci.uces if uce.ident in dictucekeep and uce.para == idpara]
+ keepuces = [CopyUce(uce) for uce in uci.uces if uce.ident in dictucekeep]
idpara += 1
if keepuces != [] :
newuces += keepuces
ident_para = -1
lastpara = -1
newuceident = {}
- print 'redo text, para and st ident'
+ print('redo text, para and st ident')
for uci in self.corpus.ucis :
uci.ident = ident_uci
ident_uci += 1
newuceident[uce.ident] = ident_uce
uce.ident = ident_uce
ident_uce += 1
- print 'backup st text and forms'
+ print('backup st text and forms')
for row in self.ori.getconcorde(self.olduceid) :
- self.c.execute('INSERT INTO uces VALUES(?,?);', (`newuceident[row[0]]`, row[1]))
+ self.c.execute('INSERT INTO uces VALUES(?,?);', (repr(newuceident[row[0]]), row[1]))
for word in row[1].split() :
self.corpus.add_word_from_forme(self.ori.formes[word], newuceident[row[0]])
self.backup_uce()
- print 'done'
+ print('done')
class BuildFromAlceste(BuildCorpus) :
+
def read_corpus(self, infile) :
if self.dlg is not None :
self.dlg.Pulse('textes : 0 - segments : 0')
else :
if iduci > 0 :
if self.corpus.ucis[-1].uces == [] :
- log.info(u'Empty text : %i' % linenb)
+ log.info('Empty text : %i' % linenb)
iduci -= 1
self.corpus.ucis.pop()
self.corpus.ucis.append(Uci(iduci, line))
if self.dlg is not None :
if not (iduci + 1) % 10 :
self.dlg.Pulse('textes : %i - segments : %i' % (iduci + 1, iduce +1))
- elif line.startswith(u'-*') :
+ elif line.startswith('-*') :
if iduci != -1 :
if txt != [] :
iduce, idpara = self.treattxt(txt, iduce, idpara, iduci)
if iduci != -1 and iduce != -1:
self.backup_uce()
else :
- log.info(_(u"No Text in corpus. Are you sure of the formatting ?"))
+ log.info(_("No Text in corpus. Are you sure of the formatting ?"))
raise Exception('TextBeforeTextMark %i' % linenb)
except UnicodeDecodeError :
raise Exception("CorpusEncoding")
for uce in ucetxt :
iduce += 1
self.corpus.ucis[-1].uces.append(Uce(iduce, idpara, iduci))
- self.c.execute('INSERT INTO uces VALUES(?,?);', (`iduce`,uce))
+ self.c.execute('INSERT INTO uces VALUES(?,?);', (repr(iduce),uce))
if not self.tolist :
uce = uce.split()
else :
for word in uce :
self.last += 1
self.corpus.add_word(word)
- log.debug(' '.join([`iduci`,`idpara`,`iduce`]))
+ log.debug(' '.join([repr(iduci),repr(idpara),repr(iduce)]))
if self.last > self.lim :
self.backup_uce()
self.last = 0
#read (treat_txt)
class Builder :
+
def __init__(self, parent, dlg = None) :
self.parent = parent
self.dlg = dlg
-
parametres = DoConf(os.path.join(self.parent.UserConfigPath,'corpus.cfg')).getoptions('corpus')
parametres['pathout'] = PathOut(parent.filename, 'corpus').mkdirout()
parametres['corpus_name'] = os.path.split(parametres['pathout'])[1]
return BuildFromAlceste(self.parent.filename, self.parametres, self.parent.lexique, self.parent.expressions, dlg = self.dlg).corpus
class SubBuilder :
+
def __init__(self, parent, corpus, parametres = None, dlg = None):
self.parent = parent
self.ori = corpus
self.dlg = dlg
corpus_name = 'Sub' + corpus.parametres['corpus_name']
if dlg is not None :
- busy = wx.BusyInfo(_("Please wait...").decode('utf8'), self)
+ busy = wx.BusyInfo(_("Please wait..."), self)
wx.SafeYield()
parametres['corpus_name'] = corpus_name
if parametres.get('frommeta', False) :
elif parametres.get('fromtheme', False) :
parametres['meta'] = corpus.make_themes()
elif parametres.get('fromclusters', False) :
- parametres['meta'] = [' '.join(['classe', `i`]) for i in range(1,parametres['clnb'] + 1)]
+ parametres['meta'] = [' '.join(['classe', repr(i)]) for i in range(1,parametres['clnb'] + 1)]
else :
parametres['meta'] = []
if 'fromclusters' not in parametres :
return BuildSubCorpus(self.ori, parametres = self.parametres, dlg = self.dlg).corpus
class BuildMergeFromClusters(BuildCorpus):
+
def __init__(self, analyses, parametres, dlg = None) :
log.info('begin subcorpus...')
self.dlg = dlg
self.dobuild()
def fromuceids(self):
- print 'fromuceids'
- dictucekeep = dict(zip(self.parametres['uceids'], self.parametres['uceids']))
+ print('fromuceids')
+ dictucekeep = dict(list(zip(self.parametres['uceids'], self.parametres['uceids'])))
idpara = 0
for uci in self.ori.ucis :
if uci.paras == [] :
ident_para = -1
lastpara = -1
newuceident = {}
- print 'redo text, para and st ident'
+ print('redo text, para and st ident')
for uci in self.corpus.ucis :
#print uci.ident, ident_uci, [uce.ident for uce in uci.uces], uci.etoiles
uci.ident = ident_uci
uce.ident = ident_uce
#print uce.ident
ident_uce += 1
- print 'backup st text and forms'
+ print('backup st text and forms')
rowid = 0
for i, analyse in enumerate(self.analyses) :
#print analyse, self.parametres['corpusira']
self.corpus.add_word_from_forme(old.formes[word], newuceident['%i-%i' % (i,row[0])])
rowid += 1
self.backup_uce()
- print 'done'
+ print('done')
class MergeClusters :
+
def __init__(self, parent, parametres = None, dlg = None):
self.parent = parent
#self.ori = corpus
self.dlg = dlg
corpus_name = 'MergeFromClusters'
if dlg is not None :
- busy = wx.BusyInfo(_("Please wait...").decode('utf8'), self)
+ busy = wx.BusyInfo(_("Please wait..."), self)
wx.SafeYield()
parametres['corpus_name'] = corpus_name
if dlg is not None :
else :
self.clusters[cl[0]].append(cl[2])
self.newet[cl[0]].append(dial.selected[cl])
-
-
analyses = [val for val in self.clusters]
clusters = [self.clusters[val] for val in analyses]
self.newet = [self.newet[val] for val in analyses]
analyses = [self.analyses[val] for val in analyses]
pathout = os.path.dirname(os.path.dirname(analyses[0]['pathout']))
self.analyses = analyses
-
pathout = os.path.join(pathout, parametres['corpus_name'])
i = 1
while os.path.exists(pathout + '_%i' % i) :
projects / iramuteq / blobdiff