diff --git a/CMakeLists.txt b/CMakeLists.txt
index 903cc582a74196aabbb3bc58fbf6b9d50fbffb1d..009441f83bf81a0615aba3c9c42a665d1de53e58 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -10,8 +10,9 @@ SET(CMAKE_C_COMPILER g++)
SET(CMAKE_CXX_COMPILER g++)
-SET( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Ofast -DUSE_CBLAS")
+SET( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Ofast -DUSE_CBLAS -ggdb")
SET( CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -lm -lopenblas" )
+set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=gnu11 -ggdb")
if (${CMAKE_C_COMPILER_VERSION} VERSION_LESS 5.3)
diff --git a/maca_common/include/mcd.h b/maca_common/include/mcd.h
index 560fd6cb4baa6edcd911d9eee32f9b054f37513e..9bc126fe0f9902838a2953472a17936be92bb187 100644
--- a/maca_common/include/mcd.h
+++ b/maca_common/include/mcd.h
@@ -8,7 +8,7 @@
#define MCD_INVALID_VALUE -1
-#define MCD_WF_NB 51
+#define MCD_WF_NB 55
#define MCD_WF_ID 0
#define MCD_WF_OFFSET 0 /* ID and OFFSET are synonymous */
@@ -65,6 +65,11 @@
#define MCD_WF_DIRECTORY 49
#define MCD_WF_SPEAKER 50
+#define MCD_WF_WORD_SPAN_STATUS 51
+#define MCD_WF_POS_LABEL_STATUS 52
+#define MCD_WF_DEP_SPAN_STATUS 53
+#define MCD_WF_DEP_LABEL_STATUS 54
+
/*Abbr
@@ -148,6 +153,13 @@ Xtra*/
#define mcd_get_label_col(m) (m)->wf2col[MCD_WF_LABEL]
#define mcd_get_stag_col(m) (m)->wf2col[MCD_WF_STAG]
#define mcd_get_sent_seg_col(m) (m)->wf2col[MCD_WF_SENT_SEG]
+
+#define mcd_get_word_span_status_col(m) (m)->wf2col[MCD_WF_WORD_SPAN_STATUS]
+#define mcd_get_pos_label_status_col(m) (m)->wf2col[MCD_WF_POS_LABEL_STATUS]
+#define mcd_get_dep_span_status_col(m) (m)->wf2col[MCD_WF_DEP_SPAN_STATUS]
+#define mcd_get_dep_label_status_col(m) (m)->wf2col[MCD_WF_DEP_LABEL_STATUS]
+
+
#define mcd_get_a_col(m) (m)->wf2col[MCD_WF_A]
#define mcd_get_b_col(m) (m)->wf2col[MCD_WF_B]
#define mcd_get_c_col(m) (m)->wf2col[MCD_WF_C]
diff --git a/maca_common/include/word.h b/maca_common/include/word.h
index dcf18dc38b52b93a22209ccb73733a5eddd91b08..c0ee708daaced5b0d1dd74eda52b1c4d27fae34d 100644
--- a/maca_common/include/word.h
+++ b/maca_common/include/word.h
@@ -109,6 +109,15 @@ typedef struct _word {
#define word_set_label(w, val) ((w)->wf_array[MCD_WF_LABEL] = (val))
#define word_set_stag(w, val) ((w)->wf_array[MCD_WF_STAG] = (val))
#define word_set_sent_seg(w, val) ((w)->wf_array[MCD_WF_SENT_SEG] = (val))
+
+
+#define word_set_word_span_status(w, val) ((w)->wf_array[MCD_WF_WORD_SPAN_STATUS] = (val))
+#define word_set_pos_label_status(w, val) ((w)->wf_array[MCD_WF_POS_LABEL_STATUS] = (val))
+#define word_set_dep_span_status(w, val) ((w)->wf_array[MCD_WF_DEP_SPAN_STATUS] = (val))
+#define word_set_dep_label_status(w, val) ((w)->wf_array[MCD_WF_DEP_LABEL_STATUS] = (val))
+
+
+
#define word_set_A(w, val) ((w)->wf_array[MCD_WF_A] = (val))
#define word_set_B(w, val) ((w)->wf_array[MCD_WF_B] = (val))
#define word_set_C(w, val) ((w)->wf_array[MCD_WF_C] = (val))
diff --git a/maca_common/src/mcd.c b/maca_common/src/mcd.c
index 34fd53b7bfbae62d856189c29183af4f735cd3c3..4eff7b1a147438c15f1385a118d71640ebc38f68 100644
--- a/maca_common/src/mcd.c
+++ b/maca_common/src/mcd.c
@@ -517,6 +517,11 @@ int mcd_wf_code(char *wf)
if(!strcmp(wf, "DIRECTORY")) return MCD_WF_DIRECTORY;
if(!strcmp(wf, "SPEAKER")) return MCD_WF_SPEAKER;
+ if(!strcmp(wf, "WORD_SPAN_STATUS")) return MCD_WF_WORD_SPAN_STATUS;
+ if(!strcmp(wf, "POS_LABEL_STATUS")) return MCD_WF_POS_LABEL_STATUS;
+ if(!strcmp(wf, "DEP_SPAN_STATUS")) return MCD_WF_DEP_SPAN_STATUS;
+ if(!strcmp(wf, "DEP_LABEL_STATUS")) return MCD_WF_DEP_LABEL_STATUS;
+
if(!strcmp(wf, "A")) return MCD_WF_A;
if(!strcmp(wf, "B")) return MCD_WF_B;
diff --git a/maca_tools/src/json2mcf.c b/maca_tools/src/json2mcf.c
index 21e481c6a8a8b27a609f8ec456c9751999831ec1..738a6ac451db1f48db98372453e282ab8935ef14 100644
--- a/maca_tools/src/json2mcf.c
+++ b/maca_tools/src/json2mcf.c
@@ -239,25 +239,52 @@ 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)];
+ w = word_buffer_get_word_n(wb, offset + start);
+ if(w == NULL){
+ fprintf(stderr, "WARNING cannot access segment %d\n", offset + start);
+ return;
+ }
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);
+
/* -------------------------------------*/
- /* added by alexis 210/07/18 for datcha */
- word_set_F(w, 1);
+ /* added by alexis 21/07/18 for datcha */
+ // word_set_F(w, 1);
/* -------------------------------------*/
- if(w == NULL){
- fprintf(stderr, "WARNING cannot access segment %d\n", offset + start);
- }
- else{
- if(d)
- label_code = dico_string2int(d, label);
+ if(d)
+ label_code = dico_string2int(d, label);
+ if(label_code == -1)
+ fprintf(stderr, "label %s unknown\n", label);
+ else
+ word_set_pos(w, label_code);
+ }
+
+
+ /* -------------------------------------*/
+ /* added by alexis 04/05/2021 for orfeo */
+ if(status_seg){
+ dico *d1 = mcd_struct->dico_array[mcd_get_word_span_status_col(mcd_struct)];
+ if(d1){
+ label_code = dico_string2int(d1, status_seg);
if(label_code == -1)
fprintf(stderr, "label %s unknown\n", label);
else
- word_set_pos(w, label_code);
+ word_set_word_span_status(w, label_code);
+
}
}
+ if(status_lab){
+ dico *d1 = mcd_struct->dico_array[mcd_get_pos_label_status_col(mcd_struct)];
+ if(d1){
+ label_code = dico_string2int(d1, status_lab);
+ if(label_code == -1)
+ fprintf(stderr, "label %s unknown\n", label);
+ else
+ word_set_pos_label_status(w, label_code);
+
+ }
+ }
+ /* -------------------------------------*/
}
void process_segment(json_attr_val *avl, word_buffer *wb, int offset)
@@ -301,41 +328,57 @@ void update_link(word_buffer *wb, int orig, int dest, char *label, char *status
mcd_struct = word_buffer_get_mcd(wb);
d = mcd_struct->dico_array[mcd_get_label_col(mcd_struct)];
+
+ w = word_buffer_get_word_n(wb, offset + orig);
+
+ if(w == NULL){
+ fprintf(stderr, "WARNING cannot access segment %d\n", offset + orig);
+ return;
+ }
if(status_lab && !strcmp(status_lab, "G")){
fprintf(stderr, "updating label of link %d -> %d with \"%s\"\n", orig, dest, label);
- w = word_buffer_get_word_n(wb, offset + orig);
-
- if(w == NULL){
- fprintf(stderr, "WARNING cannot access segment %d\n", offset + orig);
- }
-
- else{
if(d)
label_code = dico_string2int(d, label);
if(label_code == -1)
fprintf(stderr, "WARNING : label %s unknown\n", label);
// else
word_set_label(w, label_code);
- }
}
if(status_link && !strcmp(status_link, "G")){
fprintf(stderr, "updating governor of token %d with %d\n", orig, dest);
- w = word_buffer_get_word_n(wb, offset + orig);
-
- if(w == NULL){
- fprintf(stderr, "WARNING cannot access segment %d\n", offset + orig);
- }
-
- else{
-
if(dest == -1) /* -1 is for root */
word_set_gov(w, 0);
else
word_set_gov(w, dest - orig);
+ }
+ /* -------------------------------------*/
+ /* added by alexis 04/05/2021 for orfeo */
+ if(status_link){
+ dico *d1 = mcd_struct->dico_array[mcd_get_dep_span_status_col(mcd_struct)];
+ if(d1){
+ label_code = dico_string2int(d1, status_link);
+ if(label_code == -1)
+ fprintf(stderr, "label %s unknown\n", status_link);
+ else
+ word_set_dep_span_status(w, label_code);
+
}
}
+ if(status_lab){
+ dico *d1 = mcd_struct->dico_array[mcd_get_dep_label_status_col(mcd_struct)];
+ if(d1){
+ label_code = dico_string2int(d1, status_lab);
+ if(label_code == -1)
+ fprintf(stderr, "label %s unknown\n", status_lab);
+ else
+ word_set_dep_label_status(w, label_code);
+
+ }
+ }
+ /* -------------------------------------*/
+
}
@@ -344,7 +387,10 @@ void update_link(word_buffer *wb, int orig, int dest, char *label, char *status
void process_link(json_attr_val *avl, word_buffer *wb, int offset)
{
int orig, dest;
- char *label, *status_link, *status_lab;
+ char *label = NULL;
+ char *status_link = NULL;
+ char *status_lab = NULL;
+ char *status_label = NULL;
json_attr_val *av;
for(av = avl; av != NULL; av = av->next){
@@ -353,11 +399,15 @@ void process_link(json_attr_val *avl, word_buffer *wb, int offset)
if(!strcmp(av->attr, "dest")){dest = (int)(av->val->u.number); continue;}
if(!strcmp(av->attr, "label")){label = av->val->u.string; continue;}
if(!strcmp(av->attr, "status_link")){status_link = av->val->u.string; continue;}
+ if(!strcmp(av->attr, "status_label")){status_label = av->val->u.string; continue;}
if(!strcmp(av->attr, "status_lab")){status_lab = av->val->u.string; continue;}
}
// fprintf(stderr, "link : orig = %d dest = %d label = %s status_link = %s status_lab = %s\n", orig, dest, label, status_link, status_lab);
- update_link(wb, orig, dest, label, status_link, status_lab, offset);
+ if(status_label == NULL)
+ update_link(wb, orig, dest, label, status_link, status_lab, offset);
+ else
+ update_link(wb, orig, dest, label, status_link, status_label, offset);
}
@@ -448,6 +498,10 @@ void check_token(json_attr_val *avl, word_buffer *wb, int offset)
}
w = word_buffer_get_word_n(wb, id);
+ if(w == NULL){
+ fprintf(stderr, "attention, w vaut nul, id = %d form = %s", id, form_json);
+ exit(1);
+ }
form_mcf = w->form;
fprintf(stderr, "id : %d \t json : %s \t mcf : %s\n", id, form_json, form_mcf);
if(!string_equal(form_json, form_mcf)){
@@ -506,7 +560,45 @@ int main(int argc, char *argv[])
json2mcf_check_options(ctx);
mcd_extract_dico_from_corpus(ctx->mcd_struct, ctx->mcf_filename);
+
+
+ /*---------------------------------------------------*/
+ /* ajouté par Alexis le 04/05/2021 pour prendre en compte la notion de statut dans les fichiers mcf*/
+ /* add (eventually) missing labels to status dictionnaries */
+
+ dico *d;
+ d = ctx->mcd_struct->dico_array[mcd_get_word_span_status_col(ctx->mcd_struct)];
+ if(d){
+ dico_add(d, "_");
+ dico_add(d, "G");
+ dico_add(d, "C");
+ }
+
+ d = ctx->mcd_struct->dico_array[mcd_get_pos_label_status_col(ctx->mcd_struct)];
+ if(d){
+ dico_add(d, "_");
+ dico_add(d, "G");
+ dico_add(d, "C");
+ }
+
+ d = ctx->mcd_struct->dico_array[mcd_get_dep_span_status_col(ctx->mcd_struct)];
+ if(d){
+ dico_add(d, "_");
+ dico_add(d, "G");
+ dico_add(d, "C");
+ }
+
+ d = ctx->mcd_struct->dico_array[mcd_get_dep_label_status_col(ctx->mcd_struct)];
+ if(d){
+ dico_add(d, "_");
+ dico_add(d, "G");
+ dico_add(d, "C");
+ }
+
+ /*---------------------------------------------------*/
+
wb = word_buffer_load_mcf(ctx->mcf_filename, ctx->mcd_struct);
+
root = json_parse_full(ctx->json_filename, 0);
if(root->type != JSON_AVL){
diff --git a/maca_tools/src/mcf2json.c b/maca_tools/src/mcf2json.c
index 08e523b22570a704ff91d75def8d6a85d6036d5e..0182a52a6ce387ae5b504fc1a0bcde90e558d530 100644
--- a/maca_tools/src/mcf2json.c
+++ b/maca_tools/src/mcf2json.c
@@ -181,6 +181,11 @@ void print_link(FILE *output_file, word_buffer *wb, int index_first_word, int in
int gov_col = mcd_get_gov_col(word_buffer_get_mcd(wb));
int label_col = mcd_get_label_col(word_buffer_get_mcd(wb));
int index_col = mcd_get_index_col(word_buffer_get_mcd(wb));
+
+ int dep_span_status_col = mcd_get_dep_span_status_col(word_buffer_get_mcd(wb));
+ int dep_label_status_col = mcd_get_dep_label_status_col(word_buffer_get_mcd(wb));
+
+
word *w = word_buffer_get_word_n(wb, index);
fprintf(output_file, "{");
@@ -206,8 +211,23 @@ void print_link(FILE *output_file, word_buffer *wb, int index_first_word, int in
fprintf(output_file, "_");
fprintf(output_file, "\", ");
- fprintf(output_file, "\"status_link\": \"\", ");
- fprintf(output_file, "\"status_lab\": \"\", ");
+ /* ajoute le 04/05/2021 par Alexis pour Orfeo */
+
+ fprintf(output_file, "\"status_link\": \"");
+ if(dep_span_status_col != -1)
+ word_print_col_n(output_file, w, dep_span_status_col);
+ else
+ fprintf(output_file, "_");
+ fprintf(output_file, "\", ");
+
+ fprintf(output_file, "\"status_lab\": \"");
+
+ if(dep_label_status_col != -1)
+ word_print_col_n(output_file, w, dep_label_status_col);
+ else
+ fprintf(output_file, "_");
+ fprintf(output_file, "\", ");
+ /* ------------------------------------------- */
fprintf(output_file, "\"timestamp\": \"\", ");
fprintf(output_file, "\"author\": \"\", ");
fprintf(output_file, "\"target\": \"\"");
@@ -234,6 +254,8 @@ void print_links(FILE *output_file, word_buffer *wb, int index_first_word, int i
void print_segment(FILE *output_file, word_buffer *wb, int index_first_word, int index)
{
int pos_col = mcd_get_pos_col(word_buffer_get_mcd(wb));
+ int word_span_status_col = mcd_get_word_span_status_col(word_buffer_get_mcd(wb));
+ int pos_label_status_col = mcd_get_pos_label_status_col(word_buffer_get_mcd(wb));
word *w = word_buffer_get_word_n(wb, index);
fprintf(output_file, "{ ");
@@ -247,8 +269,21 @@ void print_segment(FILE *output_file, word_buffer *wb, int index_first_word, int
fprintf(output_file, "_");
fprintf(output_file, "\", ");
- fprintf(output_file, "\"status_seg\": \"\", ");
- fprintf(output_file, "\"status_lab\": \"\", ");
+ fprintf(output_file, "\"status_seg\": \"");
+ if(word_span_status_col != -1)
+ word_print_col_n(output_file, w, word_span_status_col);
+ else
+ fprintf(output_file, "_");
+ fprintf(output_file, "\", ");
+
+ fprintf(output_file, "\"status_lab\": \"");
+ if(pos_label_status_col != -1)
+ word_print_col_n(output_file, w, pos_label_status_col);
+ else
+ fprintf(output_file, "_");
+ fprintf(output_file, "\", ");
+
+
fprintf(output_file, "\"timestamp\": \"\", ");
fprintf(output_file, "\"author\": \"\", ");
fprintf(output_file, "\"target\": \"\", ");