Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
M
macaon2
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container registry
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
GitLab community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Alexis Nasr
macaon2
Commits
fa7d3b1f
Commit
fa7d3b1f
authored
7 years ago
by
Alexis Nasr
Browse files
Options
Downloads
Patches
Plain Diff
added maca_trans_attach_punct a post parsing program that attaches punctuation
parent
ad724099
No related branches found
No related tags found
No related merge requests found
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
maca_trans_parser/src/maca_trans_attach_punct.c
+209
-0
209 additions, 0 deletions
maca_trans_parser/src/maca_trans_attach_punct.c
with
209 additions
and
0 deletions
maca_trans_parser/src/maca_trans_attach_punct.c
0 → 100644
+
209
−
0
View file @
fa7d3b1f
#include
<stdio.h>
#include
<stdlib.h>
#include
<string.h>
#include
<unistd.h>
#include
<getopt.h>
#include
"context.h"
#include
"feat_vec.h"
#include
"dico_vec.h"
#include
"word_emb.h"
#include
"config2feat_vec.h"
void
maca_trans_attach_punct_help_message
(
context
*
ctx
)
{
context_general_help_message
(
ctx
);
context_mode_help_message
(
ctx
);
context_sent_nb_help_message
(
ctx
);
fprintf
(
stderr
,
"INPUT
\n
"
);
context_conll_help_message
(
ctx
);
fprintf
(
stderr
,
"IN TEST MODE
\n
"
);
context_vocabs_help_message
(
ctx
);
fprintf
(
stderr
,
"OUTPUT
\n
"
);
context_cff_help_message
(
ctx
);
fprintf
(
stderr
,
"IN TRAIN MODE
\n
"
);
context_vocabs_help_message
(
ctx
);
context_root_label_help_message
(
ctx
);
context_punct_label_help_message
(
ctx
);
}
void
maca_trans_attach_punct_check_options
(
context
*
ctx
)
{
if
(
!
ctx
->
input_filename
||
ctx
->
help
/* || !ctx->mcd_filename */
/* || !(ctx->cff_filename || ctx->fann_filename) */
){
maca_trans_attach_punct_help_message
(
ctx
);
exit
(
1
);
}
}
int
look_left_for_target
(
word_buffer
*
wb
,
int
target
)
{
int
position
=
word_buffer_get_current_index
(
wb
)
-
1
;
int
gov
;
int
step
=
0
;
do
{
step
--
;
if
(
position
==
target
)
return
step
;
gov
=
word_get_gov
(
word_buffer_get_word_n
(
wb
,
position
));
position
+=
gov
;
// printf("position = %d gov = %d\n", position, gov);
}
while
((
position
>=
0
)
&&
(
gov
<
0
));
return
0
;
}
int
look_right_for_target
(
word_buffer
*
wb
,
int
target
)
{
int
position
=
word_buffer_get_current_index
(
wb
)
+
1
;
int
gov
;
int
step
=
0
;
do
{
step
++
;
if
(
position
==
target
)
return
step
;
gov
=
word_get_gov
(
word_buffer_get_word_n
(
wb
,
position
));
position
+=
gov
;
// printf("position = %d gov = %d\n", position, gov);
}
while
((
position
<
word_buffer_get_nbelem
(
wb
))
&&
(
gov
>
0
));
return
0
;
}
int
get_left_attachement_site
(
word_buffer
*
wb
)
{
int
position
=
word_buffer_get_current_index
(
wb
)
-
1
;
word
*
w
;
int
gov
=
0
;
do
{
position
+=
gov
;
w
=
word_buffer_get_word_n
(
wb
,
position
);
gov
=
word_get_gov
(
w
);
// printf("position = %d gov = %d\n", position, gov);
}
while
((
position
>=
0
)
&&
(
gov
<
0
));
return
position
;
}
int
get_right_attachement_site
(
word_buffer
*
wb
)
{
int
position
=
word_buffer_get_current_index
(
wb
)
+
1
;
word
*
w
;
int
gov
=
0
;
do
{
position
+=
gov
;
w
=
word_buffer_get_word_n
(
wb
,
position
);
gov
=
word_get_gov
(
w
);
// printf("position = %d gov = %d\n", position, gov);
}
while
((
position
<
word_buffer_get_nbelem
(
wb
))
&&
(
gov
>
0
));
return
position
;
}
void
generate_training_file
(
FILE
*
output_file
,
context
*
ctx
)
{
config
*
c
;
int
mvt_code
;
char
mvt_type
;
int
mvt_label
;
feat_vec
*
fv
=
feat_vec_new
(
feature_types_nb
);
int
sentence_nb
=
0
;
int
root_label
=
dico_string2int
(
ctx
->
dico_labels
,
(
char
*
)
ctx
->
root_label
);
int
punct_label
=
dico_string2int
(
ctx
->
dico_labels
,
(
char
*
)
ctx
->
punct_label
);
word_buffer
*
ref
=
word_buffer_load_mcf
(
ctx
->
input_filename
,
ctx
->
mcd_struct
);
FILE
*
mcf_file
=
myfopen
(
ctx
->
input_filename
,
"r"
);
word
*
b0
;
int
left_attachement_site
;
int
right_attachement_site
;
int
target
;
int
cla
;
int
highest_left
,
highest_right
;
// c = config_new(mcf_file, mcd_struct_hyp, 5);
while
(
!
word_buffer_end
(
ref
)
&&
(
sentence_nb
<
ctx
->
sent_nb
)){
b0
=
word_buffer_b0
(
ref
);
if
(
word_get_label
(
b0
)
==
punct_label
){
left_attachement_site
=
get_left_attachement_site
(
ref
);
right_attachement_site
=
get_right_attachement_site
(
ref
);
highest_left
=
(
word_get_index
(
b0
)
+
word_get_gov
(
b0
)
==
left_attachement_site
)
?
1
:
0
;
highest_right
=
(
word_get_index
(
b0
)
+
word_get_gov
(
b0
)
==
right_attachement_site
)
?
1
:
0
;
if
(
highest_left
)
printf
(
"class = HL
\n
"
);
else
if
(
highest_right
)
printf
(
"class = HR
\n
"
);
else
{
target
=
word_get_index
(
b0
)
+
word_get_gov
(
b0
);
if
(
word_get_gov
(
b0
)
<
0
){
cla
=
look_left_for_target
(
ref
,
target
);
printf
(
"class = %d"
,
cla
);
// if(highest_left) printf("*");
printf
(
"
\n
"
);
}
if
(
word_get_gov
(
b0
)
>
0
){
cla
=
look_right_for_target
(
ref
,
target
);
printf
(
"class = %d"
,
cla
);
// if(highest_right) printf("*");
printf
(
"
\n
"
);
}
}
}
word_buffer_move_right
(
ref
);
}
}
int
main
(
int
argc
,
char
*
argv
[])
{
context
*
ctx
;
FILE
*
output_file
;
ctx
=
context_read_options
(
argc
,
argv
);
maca_trans_attach_punct_check_options
(
ctx
);
// ctx->features_model = feat_model_read(ctx->features_model_filename, feat_lib_build(), ctx->verbose);
if
(
ctx
->
mode
==
TRAIN_MODE
){
mcd_extract_dico_from_corpus
(
ctx
->
mcd_struct
,
ctx
->
input_filename
);
ctx
->
vocabs
=
mcd_build_dico_vec
(
ctx
->
mcd_struct
);
}
else
if
(
ctx
->
mode
==
TEST_MODE
){
ctx
->
vocabs
=
dico_vec_read
(
ctx
->
vocabs_filename
,
ctx
->
hash_ratio
);
mcd_link_to_dico
(
ctx
->
mcd_struct
,
ctx
->
vocabs
,
ctx
->
verbose
);
}
ctx
->
dico_labels
=
dico_vec_get_dico
(
ctx
->
vocabs
,
(
char
*
)
"LABEL"
);
if
(
ctx
->
dico_labels
==
NULL
){
fprintf
(
stderr
,
"cannot find label names
\n
"
);
return
1
;
}
/* in train mode create feature dictionnary for perceptron */
if
(
ctx
->
mode
==
TRAIN_MODE
)
ctx
->
d_perceptron_features
=
dico_new
((
char
*
)
"d_perceptron_features"
,
10000000
);
/* in test mode read feature dictionnary for perceptron */
if
(
ctx
->
mode
==
TEST_MODE
)
ctx
->
d_perceptron_features
=
dico_vec_get_dico
(
ctx
->
vocabs
,
(
char
*
)
"d_perceptron_features"
);
/* add the feature dictionnary to the dico vector */
// dico_vec_add(ctx->vocabs, ctx->d_perceptron_features);
/* open output file */
output_file
=
(
ctx
->
cff_filename
)
?
myfopen_no_exit
(
ctx
->
cff_filename
,
"w"
)
:
stdout
;
generate_training_file
(
output_file
,
ctx
);
/* if(ctx->mode == TRAIN_MODE)
dico_vec_print(ctx->vocabs_filename, ctx->vocabs);*/
if
(
ctx
->
cff_filename
)
fclose
(
output_file
);
context_free
(
ctx
);
return
0
;
}
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment