initialization

README.txt

@@ -0,0 +1,17 @@
+1. lefff --> lemmafile
+	lefff2lemmm.py
+	produces lemmm.txt
+
+2. build c++ file:
+	decide on SEP_PREF "<w><a ref=\" in buildlexicon.h
+	g++ -o buildlexicon buildlexicon.cc
+	produces executable buildlexicon
+
+3. build tree:
+	./buildlexicon -d . -p lemmdico build < lemmm.txt
+	produces lemmdico.tbl lemmdico.fsa
+
+4. test result:
+	./buildlexicon -d . -p lemmdico consult
+
+

analyste.py

@@ -0,0 +1,224 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+
+####
+# Copyright (C) 2006-2015 Kim Gerdes
+# kim AT gerdes.fr
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License
+# as published by the Free Software Foundation; either version 2
+# of the License, or (at your option) any later version.
+#
+# This script is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE
+# See the GNU General Public License (www.gnu.org) for more details.
+#
+# You can retrieve a copy of the GNU General Public License
+# from http://www.gnu.org/.  For a copy via US Mail, write to the
+#     Free Software Foundation, Inc.
+#     59 Temple Place - Suite 330,
+#     Boston, MA  02111-1307
+#     USA
+####
+
+import time, re, cgitb, cgi,sha, Cookie, os,sys
+cgitb.enable()
+import session
+from random import choice
+from frenchLinguist import FrenchLinguist
+from lookup import lookup
+from rogers import Rogers
+
+
+class Elizia:
+
+
+    listeInterro=["pourquoi","quand","comment"]
+    fini = False
+    dicoDejaDitElizia = {}
+
+    cas = None # contient le mot clé trouvé   
+
+
+
+    def __init__(self,session):
+
+	    self.sess = session
+	    self.memory = session.data.get('memory',None)
+	    #if not self.memory: self.memory=[([],[],[],[])]
+	    if self.memory:self.memoProblems,self.memoAnswers,self.memoCases,self.memoTimes = zip(*self.memory)
+	    else:self.memoProblems,self.memoAnswers,self.memoCases,self.memoTimes=[],[],[],[]
+	    #session.data.get('memoireinputsPatient',[])
+	    #self.avantCas = session.data.get('memoireCas',[])
+	    self.dicoDejaDitElizia = session.data.get('dejaDitelizia',{})
+	    self.sex = session.data.get('sex',"unknown")
+	    self.rogers = Rogers()
+	    self.goodCases=[c for c in self.memoCases if c not in self.rogers.noRecall]
+	    self.linguist = FrenchLinguist()
+	    self.problem = "???" # devrait être écrasé par le vrai input
+
+
+    def analyse(self,problem):
+	""" ici se fait tout le boulot. """
+
+	self.problemBase=problem.strip()
+
+	#first letter of sentence lower key:
+	if len(self.problemBase)>1 : self.problemBase = self.problemBase[0].lower()+self.problemBase[1:]
+	self.special=self.problemBase.split(":")[0]
+	#print "***",self.problemBase.split(":"),self.special
+	if self.special in ["words","lemmas","all"]:self.problemBase=" ".join(self.problemBase.split(":")[1:])
+	else:self.special=None
+	#self.problem = self.problemBase
+	self.problem = self.linguist.decontracte(self.problemBase)
+	if self.special=="words":return "words: "+self.problem
+
+	self.cas = [] # contains (quality,case) couples
+	inputPropre = self.linguist.nettoyerTexte(self.problem)
+	inputSplit = inputPropre.split()
+
+	#print inputSplit
+
+	if self.special=="lemmas":	return "lemmas: "+" / ".join(lookup(inputPropre))
+	lemmas = lookup(inputPropre) # list of lemmas (that are different from the form)
+	sex = self.linguist.guesssex(self.problem)
+	if sex != "unknown":
+		self.sess.data['sex'] = sex
+	 	if self.sex != "unknown" and sex != self.sex:
+			c="votre identité sexuelle"
+			self.cas += [(self.rogers.quality[c],c)]
+			#return self.repo()
+		self.sex = sex
+	#for lemma in lemmas:
+		#print lemma,"-"
+	allKeys = list(set(inputSplit + lemmas))
+	allKeys.sort()
+	if self.special=="all": return "all: "+" / ".join(allKeys)
+
+
+	# special cases:
+	if len(self.problem) > 0 and self.cleTrouvee(self.rogers.keywords['vos SMS'],self.problem,self.problem.split()):
+		c="vos SMS"
+		self.cas += [(self.rogers.quality.get(c,0),c)]
+	#voir si le patient a dit qqch
+	if len(inputSplit)==0:
+	#if len(self.patientInput) > 0:
+		c="Vide"
+		self.cas += [(self.rogers.quality.get(c,0),c)]
+		return self.repo()
+	# voir si le patient a pê posé une question :
+	# donc si ça commence avec un mot interrogatif ou termine sur un point d'interrogation
+	if inputSplit[0] in self.listeInterro or self.problem[-1]=="?":
+		c="Question"
+		self.cas += [(self.rogers.quality.get(c,0),c)]
+	elif self.problemBase in self.memoProblems :
+		c="Répétition"
+		self.cas += [(self.rogers.quality.get(c,0),c)]
+
+	# regular cases:
+	for nom in self.rogers.keywords.keys(): # pour chaque clé		
+		if self.cleTrouvee(self.rogers.keywords[nom],inputPropre,allKeys):
+			self.cas += [(self.rogers.quality[nom],nom)]
+	#voir si le patient est très bref
+	if len(inputSplit)==1:
+		c="Bref"
+		self.cas += [(self.rogers.quality.get(c,0),c)]				
+	c="Retour"
+	self.cas += [(self.rogers.quality.get(c,0),c)]
+
+	return self.repo()
+
+    def cleTrouvee(self,cles,texte,liste):
+        """ 
+	cherche une liste cles dans un self.texte (string) et dans la liste de mots correspondante
+	deux cas : clé est un mot : il doit apparaître entouré de blancs 
+	ou clé contient des espaces : peu importe, tant qu'il apparaît
+	"""
+
+        for c in cles :
+		if " " in c and c in texte:
+			return True
+		elif c in liste :
+			return True
+        return False
+
+
+
+    def repo(self):
+	""" fonction importante qui rend la réponse et gêre les mémoires """
+
+	ordered,reponse,cases,maxiq = False,"",[],0
+	for q,c in self.cas:
+		if q>maxiq:
+			cases=[c]
+			maxiq=q
+		elif q==maxiq:cases+=[c]
+	#print cases
+	nomcas=choice(cases) # take one of the maxi cases by chance
+	#print nomcas
+
+	memoire =  self.dicoDejaDitElizia.get(nomcas,[])
+	#if nomcas in self.dicoDejaDitElizia.keys() : memoire =  self.dicoDejaDitElizia[nomcas]
+	#else : memoire = []	
+
+
+
+	ordered = nomcas in self.rogers.ordered  #== "Insulte" ,...
+	# if (ordered, no more answers and no cycle) or end case: end it all
+	self.fini = (nomcas in self.rogers.ordered and len(memoire) >= len(self.rogers.answers[nomcas])-2 and nomcas in self.rogers.noCycle ) or nomcas == "Fin"
+
+	#if case is worse than 1, we do have good cases, come back to old cases from time to time    (nomcas=="Retour" )
+	if maxiq<1 and self.goodCases and 0 == choice([0,1]) :  
+		nomcas = "Reprise"
+		casRepris = self.goodCases[-1]
+		#print self.goodCases,self.memoCases
+		if 0 == choice([0,1,2]): # parfois (1/3) on ajoute une réponse de l'ancien cas
+			if casRepris in self.dicoDejaDitElizia.keys() : memoReprise =  self.dicoDejaDitElizia[casRepris]
+			else : memoReprise = []
+			reponse = " \n "+self.choisirReponse(casRepris,memoReprise,False)
+
+	reponse = self.choisirReponse(nomcas,memoire,ordered) + reponse
+
+	if nomcas == "Reprise" : reponse = reponse.replace("xxx",casRepris)
+
+	if nomcas == "Retour" and " xxx" in reponse:
+		reponse = reponse.replace("xxx","")
+		reponse = reponse + self.linguist.phraseEnchassee(self.problem,self.sex)
+	# se souvenir de ce qu'on a dit
+	try:
+		self.sess.data['memory']  = (self.sess.data.get('memory') or [])+[( self.problemBase,reponse,nomcas,repr(time.time())  )]
+	except:
+		print "Putain, j'ai un problème de mémoire !"#.encode("utf-8")
+
+	return self.linguist.beautifier(reponse,self.sex)
+
+    def choisirReponse(self,casActuel,memoireActuel,ordered):
+	# le [:] permet de faire une copie
+        #       - pour qu'on ne touche pas à la liste originale
+        listeReponseLocale=self.rogers.answers[casActuel][:]
+
+        for rep in memoireActuel:
+		if rep in listeReponseLocale: # normalement, ça devrait tjrs être dedans
+			listeReponseLocale.remove(rep) # on enlève ce qu'on a déjà dit
+        if ordered : 	reponse = listeReponseLocale[0]
+	else : 		reponse = choice(listeReponseLocale) # choix par hasard parmi le reste
+
+	memoireCas=memoireActuel+[reponse] # on retient la réponse (sans phrase du patient en cas Retour)
+
+	# vider le mémoire sauf la dernière réponse au cas où on a déjà	utilisé toutes les réponses proposées
+
+	if len(memoireCas) >= len(self.rogers.answers[casActuel]):
+            while len(memoireCas)>1:
+                memoireCas.pop(0)
+
+	self.dicoDejaDitElizia[casActuel]=memoireCas
+
+
+
+	self.sess.data['dejaDitelizia'] = self.dicoDejaDitElizia
+
+	return reponse
+
+