ajout de fonctions pour éliminer ou ajouter un mot dans un mcf, utlisation de...

ajout de fonctions pour éliminer ou ajouter un mot dans un mcf, utlisation de ces fonctions pour la conversion json2mcf

CMakeLists.txt

@@ -15,7 +15,7 @@ SET( CMAKE_EXE_LINKER_FLAGS  "${CMAKE_EXE_LINKER_FLAGS} -lm -lopenblas" )
 
 
 if (${CMAKE_C_COMPILER_VERSION} VERSION_LESS 5.3)
-	set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=gnu11")
+	set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=gnu11 -ggdb")
 	# better, but needs CMake >= 3.0
 	#set_property(GLOBAL PROPERTY CXX_STANDARD 11)
 	#set_property(GLOBAL PROPERTY C_STANDARD 11)

maca_common/include/word_buffer.h

@@ -45,6 +45,8 @@ typedef struct {
 word_buffer *word_buffer_new(FILE *input_file, mcd *mcd_struct, int lookahead);
 void         word_buffer_free(word_buffer *wb);
 int          word_buffer_add(word_buffer *wb, word *w);
+void         word_buffer_insert(word_buffer *wb, word *w, int index);
+void         word_buffer_rm(word_buffer *wb, int index);
 word*        word_buffer_get_word_relative(word_buffer *wb, int dist);
 word*        word_buffer_get_word_n(word_buffer *wb, int n);
 int          word_buffer_read_next_word(word_buffer *wb);

maca_common/src/json_parser.c

@@ -207,7 +207,7 @@ void consommer(json_parser_ctx *ctx, int c ) {
     ctx->uc = yylex(ctx); /* consommer le caractère */
   }
   else
-    erreur(ctx,  (char *) "erreure lexicale" );
+    erreur(ctx,  (char *) "erreur lexicale" );
 }
 
 /*-------------------------------------------------------------------------*/

maca_common/src/word_buffer.c

@@ -72,6 +72,66 @@ void word_buffer_free(word_buffer *wb)
   free(wb);
 }
 
+/* remove word at position index */
+void word_buffer_rm(word_buffer *wb, int index)
+{
+  int i;
+  if((index < 0) || (index >= wb->nbelem)) {
+    fprintf(stderr, "cannot remove word %d, index out of range\n", index);
+    return;
+  }
+  /* check if word at index has daughters */
+  for(i=0; i < wb->nbelem; i++){
+    if(word_get_gov_index(word_buffer_get_word_n(wb, i)) == index){
+      fprintf(stderr, "cannot remove word %d, it has at least one daughter", index);
+      return;
+    }
+  }
+  /* decrease dependencies length whenever gov and dep are on different sides of index */
+  for(int dep_index=0; dep_index < wb->nbelem; dep_index++){
+    word *dep = word_buffer_get_word_n(wb, dep_index);
+    int gov_index = word_get_gov_index(dep);
+    if((dep_index < index && gov_index > index)
+       || (dep_index > index && gov_index < index)){
+	word_set_gov(dep, word_get_gov(dep) - 1); 
+      }
+  }
+  word_free(wb->array[index]);
+  wb->array[index] = NULL;
+  for(i=index+1; i < wb->nbelem; i++){
+    wb->array[i-1] = wb->array[i];
+  }
+  wb->nbelem--;
+}
+
+/* insert word w at position index */
+void word_buffer_insert(word_buffer *wb, word *w, int index)
+{
+  if(wb->nbelem == wb->size -1){
+    wb->size = 2 * (wb->size + 1);
+    wb->array = (word **)realloc(wb->array, wb->size * sizeof(word *));
+  }
+
+  /* increase dependencies length whenever gov and dep are on different sides of index */
+  for(int dep_index=0; dep_index < wb->nbelem; dep_index++){
+    word *dep = word_buffer_get_word_n(wb, dep_index);
+    int gov_index = word_get_gov_index(dep);
+    if((dep_index < index && gov_index >= index)
+       || (dep_index >= index && gov_index < index)){
+	word_set_gov(dep, word_get_gov(dep) + 1); 
+      }
+  }
+  
+  for(int i=wb->nbelem; i >= index; i--){
+    wb->array[i] = wb->array[i-1];
+  }
+  wb->array[index] = w;
+  wb->nbelem++;
+  
+}
+
+
+
 int word_buffer_add(word_buffer *wb, word *w)
 {
   if(wb->nbelem == wb->size -1){

maca_tools/src/json2mcf.c

@@ -238,7 +238,7 @@ void update_segment(word_buffer *wb, int start, int end, char *label, char *stat
 
   mcd_struct = word_buffer_get_mcd(wb); 
   d = mcd_struct->dico_array[mcd_get_pos_col(mcd_struct)];
-  
+
   if(status_lab && !strcmp(status_lab, "G")){
     fprintf(stderr, "updating label of segment [%d-%d] with \"%s\"\n", start, end, label);
     w = word_buffer_get_word_n(wb, offset + start);
@@ -409,14 +409,44 @@ void check_token(json_attr_val *avl, word_buffer *wb, int offset)
   int id;
   char *form_json = NULL;
   char *form_mcf = NULL;
+  char *status = NULL;
   json_attr_val *av;
   word *w = NULL;
   for(av = avl; av != NULL; av = av->next){
     //      printf("attr = %s\n", av->attr);
     if(!strcmp(av->attr, "id")){id = (int)(av->val->u.number); continue;}
     if(!strcmp(av->attr, "word")){form_json = av->val->u.string; continue;}
+    if(!strcmp(av->attr, "status")){status = av->val->u.string; continue;}
+
+  }
+  /* ajouté le 24 juillet 2020 par Alexis */
+  if(status != NULL && !strcmp(status, "inserted")){
+    int form_column = wb->mcd_struct->wf2col[MCD_WF_FORM];
+    fprintf(stderr, "inserting token at position %d\n", id);
+    word *w = word_new(NULL);
+    //    word_set_form(w, form_json);
+    int code = dico_add(wb->mcd_struct->dico_array[form_column], form_json);
+    w->wf_array[MCD_WF_FORM] = dico_add(wb->mcd_struct->dico_array[form_column], form_json);
+    w->form = strdup(form_json);
+    word_buffer_insert(wb, w, id);
+  }
+
+  /* ajouté le 24 juillet 2020 par Alexis */
+  if(status != NULL && !strcmp(status, "deleted")){
+    fprintf(stderr, "deleting token at position %d\n", id);
+    word_buffer_rm(wb, id);
+  }
 
+  /* ajouté le 24 juillet 2020 par Alexis */
+  if(status != NULL && !strcmp(status, "modified")){
+    fprintf(stderr, "modifying token at position %d\n", id);
+    int form_column = wb->mcd_struct->wf2col[MCD_WF_FORM];
+    int code = dico_add(wb->mcd_struct->dico_array[form_column], form_json);
+    word *w = word_buffer_get_word_n(wb, id);
+    w->wf_array[MCD_WF_FORM] = dico_add(wb->mcd_struct->dico_array[form_column], form_json);
+    w->form = strdup(form_json);
   }
+  
   w = word_buffer_get_word_n(wb, id);
   form_mcf = w->form;
   fprintf(stderr, "id : %d \t json : %s \t mcf : %s\n", id, form_json, form_mcf);