diff --git a/maca_common/include/mcd.h b/maca_common/include/mcd.h
index aa4ade656c36921b030a52c3cd8314c1b7daa4d9..560fd6cb4baa6edcd911d9eee32f9b054f37513e 100644
--- a/maca_common/include/mcd.h
+++ b/maca_common/include/mcd.h
@@ -61,9 +61,9 @@
#define MCD_WF_Person 45
#define MCD_WF_Tense 46
-#define MCD_WF_FILE 47
-#define MCD_WF_DIRECTORY 48
-#define MCD_WF_SPEAKER 49
+#define MCD_WF_FILE 48
+#define MCD_WF_DIRECTORY 49
+#define MCD_WF_SPEAKER 50
diff --git a/maca_common/include/word.h b/maca_common/include/word.h
index 7e8f30bdbb048ac2767d692d1cbc182fc3ed0c99..dac612deb263dd47a31fe4865187e5a6929eab52 100644
--- a/maca_common/include/word.h
+++ b/maca_common/include/word.h
@@ -63,6 +63,11 @@ typedef struct _word {
#define word_get_label(w) (((w) == NULL) ? -1 : (w)->wf_array[MCD_WF_LABEL])
#define word_get_stag(w) (((w) == NULL) ? -1 : (w)->wf_array[MCD_WF_STAG])
#define word_get_sent_seg(w) (((w) == NULL) ? 0 : (w)->wf_array[MCD_WF_SENT_SEG])
+
+#define word_get_file(w) (((w) == NULL) ? 0 : (w)->wf_array[MCD_WF_FILE])
+#define word_get_directory(w) (((w) == NULL) ? 0 : (w)->wf_array[MCD_WF_DIRECTORY])
+#define word_get_speaker(w) (((w) == NULL) ? 0 : (w)->wf_array[MCD_WF_SPEAKER])
+
#define word_get_A(w) (((w) == NULL) ? -1 : (w)->wf_array[MCD_WF_A])
#define word_get_B(w) (((w) == NULL) ? -1 : (w)->wf_array[MCD_WF_B])
#define word_get_C(w) (((w) == NULL) ? -1 : (w)->wf_array[MCD_WF_C])
diff --git a/maca_tools/src/mcf2json.c b/maca_tools/src/mcf2json.c
index 4381177cfd388d8012380286f515dc1ae5fd19a8..a194d9bfe29340a82d24a540db1f8e08786bd873 100644
--- a/maca_tools/src/mcf2json.c
+++ b/maca_tools/src/mcf2json.c
@@ -2,6 +2,9 @@
#include<stdlib.h>
#include<string.h>
#include<getopt.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
#include"mcd.h"
#include"util.h"
@@ -16,6 +19,7 @@ typedef struct {
char *mcf_filename;
char *mcd_filename;
mcd *mcd_struct;
+ char *root_dir;
} context;
void mcf2json_context_free(context *ctx)
@@ -31,6 +35,8 @@ void mcf2json_context_free(context *ctx)
free(ctx->mcd_filename);
if(ctx->mcd_struct)
mcd_free(ctx->mcd_struct);
+ if(ctx->root_dir)
+ free(ctx->root_dir);
free(ctx);
}
}
@@ -47,6 +53,7 @@ context *mcf2json_context_new(void)
ctx->mcf_filename = NULL;
ctx->mcd_filename = NULL;
ctx->mcd_struct = NULL;
+ ctx->root_dir = NULL;
return ctx;
}
@@ -58,7 +65,7 @@ void mcf2json_context_general_help_message(context *ctx)
fprintf(stderr, "\t-v --verbose : activate verbose mode\n");
fprintf(stderr, "\t-C --mcd : mcd filename\n");
fprintf(stderr, "\t-i --mcf : mcf filename (read from stdin if absent)\n");
- fprintf(stderr, "\t-o --conll : conll filename (write to stdout if absent)\n");
+ fprintf(stderr, "\t-r --root : root directory of the json files\n");
}
void mcf2json_check_options(context *ctx){
@@ -81,14 +88,14 @@ context *mcf2json_context_read_options(int argc, char *argv[])
{"help", no_argument, 0, 'h'},
{"verbose", no_argument, 0, 'v'},
{"debug", no_argument, 0, 'd'},
- {"conll", required_argument, 0, 'o'},
{"mcd", required_argument, 0, 'C'},
{"mcf", required_argument, 0, 'i'},
+ {"root", required_argument, 0, 'r'},
};
optind = 0;
opterr = 0;
- while ((c = getopt_long (argc, argv, "hvdo:C:i:", long_options, &option_index)) != -1){
+ while ((c = getopt_long (argc, argv, "hvdC:i:r:", long_options, &option_index)) != -1){
switch (c)
{
case 'd':
@@ -100,15 +107,15 @@ context *mcf2json_context_read_options(int argc, char *argv[])
case 'v':
ctx->verbose = 1;
break;
- case 'o':
- ctx->conll_filename = strdup(optarg);
- break;
case 'i':
ctx->mcf_filename = strdup(optarg);
break;
case 'C':
ctx->mcd_filename = strdup(optarg);
break;
+ case 'r':
+ ctx->root_dir = strdup(optarg);
+ break;
}
}
@@ -118,7 +125,6 @@ context *mcf2json_context_read_options(int argc, char *argv[])
else{
ctx->mcd_struct = mcd_build_wpmlgfs();
}
-
return ctx;
}
@@ -317,7 +323,7 @@ void print_sentence(FILE *output_file, int sentence_nb, word_buffer *wb, int ind
int main(int argc, char *argv[])
{
- FILE *output_file;
+ FILE *output_file = NULL;
context *ctx = mcf2json_context_read_options(argc, argv);
word_buffer *wb = NULL;
word *w = NULL;
@@ -326,41 +332,101 @@ int main(int argc, char *argv[])
int index_first_word;
int index_last_word;
int sentence_nb = 0;
-
+ char current_directory[1000];
+ char current_file[1000];
+ char previous_directory[1000];
+ char previous_file[1000];
+ char *root_directory = NULL;
+ char destination_file[1000];
+ char destination_dir[1000];
+ struct stat st = {0};
+
mcf2json_check_options(ctx);
mcd_extract_dico_from_corpus(ctx->mcd_struct, ctx->mcf_filename);
-
- output_file = (ctx->conll_filename)? myfopen_no_exit(ctx->conll_filename, "w"): stdout;
wb = word_buffer_load_mcf(ctx->mcf_filename, ctx->mcd_struct);
- print_header(output_file, ctx->mcd_struct);
- do{
- w = word_buffer_b0(wb);
- if(new_sentence){
- new_sentence = 0;
- sentence_nb++;
- index_first_word = word_buffer_get_current_index(wb);
- }
- if(word_get_sent_seg(w)){
- index_last_word = word_buffer_get_current_index(wb);
- new_sentence = 1;
-
- if(first_sentence == 1)
- first_sentence = 0;
- else
- fprintf(output_file, ",");
- fprintf(output_file, "\n");
- print_sentence(output_file, sentence_nb, wb, index_first_word, index_last_word);
+ if(ctx->root_dir){
+ if(stat(ctx->root_dir, &st) == -1) {
+ mkdir(ctx->root_dir, 0700);
+ fprintf(stderr, "creating directory %s\n", ctx->root_dir);
}
- } while(word_buffer_move_right(wb));
-
- print_footer(output_file);
- if(ctx->conll_filename)
+ do{
+ w = word_buffer_b0(wb);
+ if(w == NULL) break;
+ word_sprint_col_n(current_directory, w, ctx->mcd_struct->wf2col[MCD_WF_DIRECTORY]);
+ word_sprint_col_n(current_file, w, ctx->mcd_struct->wf2col[MCD_WF_FILE]);
+
+ if(strcmp(current_directory, previous_directory)){
+ strcpy(destination_dir, ctx->root_dir);
+ strcat(destination_dir, "/");
+ strcat(destination_dir, current_directory);
+ if (stat(destination_dir, &st) == -1) {
+ mkdir(destination_dir, 0700);
+ fprintf(stderr, "creating directory %s\n", destination_dir);
+ }
+ }
+ if(strcmp(current_file, previous_file)){
+ strcpy(destination_file, destination_dir);
+ strcat(destination_file, "/");
+ strcat(destination_file, current_file);
+ strcat(destination_file, ".json");
+ fprintf(stderr, "creating file %s\n", destination_file);
+ if(output_file){
+ print_footer(output_file);
+ fclose(output_file);
+ }
+ output_file = myfopen_no_exit(destination_file, "w");
+ print_header(output_file, ctx->mcd_struct);
+ }
+ if(new_sentence){
+ new_sentence = 0;
+ sentence_nb++;
+ index_first_word = word_buffer_get_current_index(wb);
+ }
+ if(word_get_sent_seg(w)){
+ index_last_word = word_buffer_get_current_index(wb);
+ new_sentence = 1;
+
+ if(first_sentence == 1)
+ first_sentence = 0;
+ else
+ fprintf(output_file, ",");
+ fprintf(output_file, "\n");
+ print_sentence(output_file, sentence_nb, wb, index_first_word, index_last_word);
+ }
+ strcpy(previous_file, current_file);
+ strcpy(previous_directory, current_directory);
+ } while(word_buffer_move_right(wb));
+ print_footer(output_file);
fclose(output_file);
- mcf2json_context_free(ctx);
-
-
+ }
+ else{ //ctx->root_dir is NULL dump everything to stdout
+ output_file = stdout;
+ print_header(output_file, ctx->mcd_struct);
+ do{
+ w = word_buffer_b0(wb);
+ if(new_sentence){
+ new_sentence = 0;
+ sentence_nb++;
+ index_first_word = word_buffer_get_current_index(wb);
+ }
+ if(word_get_sent_seg(w)){
+ index_last_word = word_buffer_get_current_index(wb);
+ new_sentence = 1;
+
+ if(first_sentence == 1)
+ first_sentence = 0;
+ else
+ fprintf(output_file, ",");
+ fprintf(output_file, "\n");
+ print_sentence(output_file, sentence_nb, wb, index_first_word, index_last_word);
+ }
+ } while(word_buffer_move_right(wb));
+ print_footer(output_file);
+ }
+
+ mcf2json_context_free(ctx);
return 0;
}