update copyright

[iramuteq] / corpus.py

diff --git a/corpus.py b/corpus.py
index 5fe448a..555a034 100644 (file)
--- a/corpus.py
+++ b/corpus.py
@@ -20,6 +20,9 @@ from operator import itemgetter
 from uuid import uuid4
 import datetime
 from copy import copy
+#------test spacy------------
+#import spacy
+#nlp = spacy.load("fr_core_news_lg")
 
 #------------------------------------
 # import des fichiers du projet
@@ -255,14 +258,14 @@ class Corpus :
 
     def getucisize(self) :
         ucesize = self.getucesize()
-        return [sum(ucesize[uci.uces[0].ident:(uci.uces[-1].ident + 1)]) for uci in self.ucis]
+        return [sum(ucesize[uci.uces[0].ident:(uci.uces[-1].ident + 1)]) for uci in self.ucis if len(uci.uces) != 0]
 
     def getucesize(self) :
         res = self.getalluces()
         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([repr(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]
@@ -290,6 +293,11 @@ class Corpus :
     def getucesfrometoile(self, etoile) :
         return [uce.ident for uci in self.ucis for uce in uci.uces if etoile in uci.etoiles]
 
+    def getucisfrometoile(self, etoile):
+        uces = [uce.ident for uci in self.ucis for uce in uci.uces if etoile in uci.etoiles]
+        return list(set([self.getucefromid(val).uci for val in uces]))
+
+
     def getetoileuces(self) :
         log.info('get uces etoiles')
         etoileuces = {}
@@ -875,7 +883,7 @@ class Corpus :
             etoileuces = self.getetoileuces()
         else :
             etoileuces = self.getetoileucis()
-        etoileuces = dict([[et, etoileuces[et]] for et in etoileuces if len(etoileuces[et]) > 1])
+        etoileuces = dict([[et, etoileuces[et]] for et in etoileuces if len(etoileuces[et]) > 0])
         with open(fileout, 'w', encoding='utf8') as f :
             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()
@@ -1502,7 +1510,7 @@ class BuildSubCorpus(BuildCorpus):
                 newuces = []
                 newpara = []
                 for et in uci.paras :
-                    keepuces = [CopyUce(uce) for uce in uci.uces if uce.ident in dictucekeep]
+                    keepuces = [CopyUce(uce) for uce in uci.uces if uce.ident in dictucekeepand and uce.para == idpara]
                     idpara += 1
                     if keepuces != [] :
                         newuces += keepuces
@@ -1546,6 +1554,7 @@ class BuildSubCorpus(BuildCorpus):
 class BuildFromAlceste(BuildCorpus) :
 
     def read_corpus(self, infile) :
+
         if self.dlg is not None :
             self.dlg.Pulse('textes : 0 - segments : 0')
         self.limitshow = 0
@@ -1565,6 +1574,8 @@ class BuildFromAlceste(BuildCorpus) :
                     if self.testuci(line) :
                         iduci += 1
                         if txt != [] :
+                            #doc = nlp(' '.join(txt))
+                            #print([[word, word.pos_, word.lemma_] for word in doc])
                             iduce, idpara = self.treattxt(txt, iduce, idpara, iduci - 1)
                             txt = []
                             self.corpus.ucis.append(Uci(iduci, line))