Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
C
corpus2
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Redmine
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
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
Analysers
corpus2
Commits
dd9a3136
Commit
dd9a3136
authored
May 24, 2011
by
ilor
Browse files
Options
Downloads
Patches
Plain Diff
IOB-CHAN~~~~
parent
cbf8a71a
No related branches found
No related tags found
No related merge requests found
Changes
3
Show whitespace changes
Inline
Side-by-side
Showing
3 changed files
libcorpus2/CMakeLists.txt
+1
-0
1 addition, 0 deletions
libcorpus2/CMakeLists.txt
libcorpus2/io/iob-chan.cpp
+216
-0
216 additions, 0 deletions
libcorpus2/io/iob-chan.cpp
libcorpus2/io/iob-chan.h
+88
-0
88 additions, 0 deletions
libcorpus2/io/iob-chan.h
with
305 additions
and
0 deletions
libcorpus2/CMakeLists.txt
+
1
−
0
View file @
dd9a3136
...
...
@@ -59,6 +59,7 @@ SET(libcorpus2_STAT_SRC
io/cclwriter.cpp
io/helpers.cpp
io/fastxces.cpp
io/iob-chan.cpp
io/nonewriter.cpp
io/orthwriter.cpp
io/pathwriter.cpp
...
...
This diff is collapsed.
Click to expand it.
libcorpus2/io/iob-chan.cpp
0 → 100644
+
216
−
0
View file @
dd9a3136
/*
Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
Part of the libcorpus2 project
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 3 of the License, or (at your option)
any later version.
This program 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 LICENSE and COPYING files for more details.
*/
#include
<libcorpus2/io/iob-chan.h>
#include
<libpwrutils/foreach.h>
#include
<libcorpus2/ann/annotatedsentence.h>
#include
<boost/algorithm/string.hpp>
#include
<boost/make_shared.hpp>
#include
<fstream>
#include
<boost/algorithm/string/split.hpp>
namespace
Corpus2
{
bool
IobChanWriter
::
registered
=
TokenWriter
::
register_writer
<
IobChanWriter
>
(
"iob-chan"
,
"nowarn,noforce"
);
IobChanWriter
::
IobChanWriter
(
std
::
ostream
&
os
,
const
Tagset
&
tagset
,
const
string_range_vector
&
params
)
:
TokenWriter
(
os
,
tagset
,
params
),
warn_on_no_lexemes_
(
true
),
force_
(
true
)
{
foreach
(
const
string_range
&
param
,
params
)
{
std
::
string
p
=
boost
::
copy_range
<
std
::
string
>
(
param
);
if
(
p
==
"nowarn"
)
{
warn_on_no_lexemes_
=
false
;
}
else
if
(
p
==
"noforce"
)
{
force_
=
false
;
}
}
}
void
IobChanWriter
::
write_token
(
const
Token
&
t
)
{
os
()
<<
t
.
orth_utf8
();
if
(
t
.
lexemes
().
empty
())
{
if
(
warn_on_no_lexemes_
)
{
std
::
cerr
<<
"No lexemes for token!"
;
}
}
else
{
const
Lexeme
&
pref
=
t
.
get_preferred_lexeme
(
tagset
());
os
()
<<
"
\t
"
;
write_tag
(
pref
.
tag
());
}
os
()
<<
"
\n
"
;
}
void
IobChanWriter
::
write_sentence
(
const
Sentence
&
s
)
{
const
AnnotatedSentence
*
ann
=
dynamic_cast
<
const
AnnotatedSentence
*>
(
&
s
);
if
(
force_
)
{
// I sincerely apologize
AnnotatedSentence
*
hax
=
const_cast
<
AnnotatedSentence
*>
(
ann
);
foreach
(
const
AnnotatedSentence
::
chan_map_t
::
value_type
&
v
,
hax
->
all_channels
())
{
hax
->
get_channel
(
v
.
first
).
make_iob_from_segments
();
}
}
for
(
size_t
idx
=
0
;
idx
<
s
.
size
();
++
idx
)
{
const
Token
*
t
=
s
.
tokens
()[
idx
];
os
()
<<
t
->
orth_utf8
();
if
(
t
->
lexemes
().
empty
())
{
if
(
warn_on_no_lexemes_
)
{
std
::
cerr
<<
"No lexemes for token!"
;
}
}
else
{
const
Lexeme
&
pref
=
t
->
get_preferred_lexeme
(
tagset
());
os
()
<<
"
\t
"
;
os
()
<<
pref
.
lemma_utf8
();
os
()
<<
"
\t
"
;
write_tag
(
pref
.
tag
());
os
()
<<
"
\t
"
;
}
if
(
ann
)
{
bool
first
=
true
;
foreach
(
const
AnnotatedSentence
::
chan_map_t
::
value_type
&
v
,
ann
->
all_channels
())
{
if
(
!
first
)
{
os
()
<<
","
;
}
os
()
<<
v
.
first
<<
"-"
;
os
()
<<
Corpus2
::
IOB
::
to_string
(
v
.
second
.
get_iob_at
(
idx
));
first
=
false
;
}
}
os
()
<<
"
\n
"
;
}
os
()
<<
"
\n
"
;
}
void
IobChanWriter
::
write_chunk
(
const
Chunk
&
c
)
{
foreach
(
const
Sentence
::
ConstPtr
&
s
,
c
.
sentences
())
{
write_sentence
(
*
s
);
}
}
void
IobChanWriter
::
write_tag
(
const
Tag
&
tag
)
{
os
()
<<
tagset
().
tag_to_string
(
tag
);
}
bool
IobChanReader
::
registered
=
TokenReader
::
register_reader
<
IobChanReader
>
(
"iob-chan"
,
"ign,loose,strict,no_set_disamb"
);
IobChanReader
::
IobChanReader
(
const
Tagset
&
tagset
,
std
::
istream
&
is
)
:
BufferedSentenceReader
(
tagset
),
is_
(
&
is
),
disamb_
(
true
)
{
}
IobChanReader
::
IobChanReader
(
const
Tagset
&
tagset
,
const
std
::
string
&
filename
)
:
BufferedSentenceReader
(
tagset
),
is_
(),
disamb_
(
true
)
{
is_owned_
.
reset
(
new
std
::
ifstream
(
filename
.
c_str
(),
std
::
ifstream
::
in
));
if
(
!
this
->
is_owned_
->
good
())
{
throw
Corpus2Error
(
"File not found!"
);
}
else
{
this
->
is_
=
is_owned_
.
get
();
}
}
Sentence
::
Ptr
IobChanReader
::
actual_next_sentence
()
{
std
::
string
line
;
boost
::
shared_ptr
<
AnnotatedSentence
>
s
;
typedef
boost
::
split_iterator
<
std
::
string
::
const_iterator
>
string_split_iterator
;
while
(
is
().
good
())
{
std
::
getline
(
is
(),
line
);
if
(
line
.
empty
())
{
break
;
}
std
::
vector
<
std
::
string
>
spl
;
boost
::
algorithm
::
split
(
spl
,
line
,
boost
::
is_any_of
(
"
\t
"
));
if
(
spl
.
size
()
!=
4
)
{
std
::
cerr
<<
"Invalid line: "
<<
line
<<
"("
<<
spl
.
size
()
<<
")
\n
"
;
}
else
{
const
std
::
string
&
orth
=
spl
[
0
];
const
std
::
string
&
lemma
=
spl
[
1
];
const
std
::
string
&
tag_string
=
spl
[
2
];
const
std
::
string
&
anns
=
spl
[
3
];
Tag
tag
=
parse_tag
(
tag_string
);
Token
*
t
=
new
Token
();
t
->
set_orth
(
UnicodeString
::
fromUTF8
(
orth
));
t
->
set_wa
(
PwrNlp
::
Whitespace
::
Space
);
t
->
add_lexeme
(
Lexeme
(
UnicodeString
::
fromUTF8
(
lemma
),
tag
));
if
(
disamb_
)
{
t
->
lexemes
().
back
().
set_disamb
(
true
);
}
if
(
!
s
)
{
s
=
boost
::
make_shared
<
AnnotatedSentence
>
();
t
->
set_wa
(
PwrNlp
::
Whitespace
::
Newline
);
}
s
->
append
(
t
);
std
::
vector
<
std
::
string
>
annsplit
;
boost
::
algorithm
::
split
(
annsplit
,
anns
,
boost
::
is_any_of
(
","
));
foreach
(
const
std
::
string
&
a
,
annsplit
)
{
std
::
vector
<
std
::
string
>
one_ann_split
;
boost
::
algorithm
::
split
(
one_ann_split
,
a
,
boost
::
is_any_of
(
"-"
));
if
(
one_ann_split
.
size
()
!=
2
)
{
std
::
cerr
<<
"Invalid annotation:"
<<
a
<<
"
\n
"
;
}
else
{
const
std
::
string
&
aname
=
one_ann_split
[
0
];
const
std
::
string
&
aiob
=
one_ann_split
[
1
];
Corpus2
::
IOB
::
Enum
iob
=
Corpus2
::
IOB
::
from_string
(
aiob
);
if
(
iob
==
Corpus2
::
IOB
::
PostLast
)
{
std
::
cerr
<<
"Invalid IOB tag: "
<<
aiob
<<
"
\n
"
;
}
else
{
if
(
!
s
->
has_channel
(
aname
))
{
s
->
create_channel
(
aname
);
}
s
->
get_channel
(
aname
).
set_iob_at
(
s
->
size
()
-
1
,
iob
);
}
}
}
}
}
if
(
s
)
{
foreach
(
const
AnnotatedSentence
::
chan_map_t
::
value_type
&
v
,
s
->
all_channels
())
{
s
->
get_channel
(
v
.
first
).
make_segments_from_iob
();
}
}
return
s
;
}
void
IobChanReader
::
set_option
(
const
std
::
string
&
option
)
{
if
(
option
==
"no_set_disamb"
)
{
disamb_
=
false
;
}
else
{
BufferedSentenceReader
::
set_option
(
option
);
}
}
std
::
string
IobChanReader
::
get_option
(
const
std
::
string
&
option
)
const
{
if
(
option
==
"no_set_disamb"
)
{
return
!
disamb_
?
option
:
""
;
}
return
BufferedSentenceReader
::
get_option
(
option
);
}
}
/* end ns Corpus2 */
This diff is collapsed.
Click to expand it.
libcorpus2/io/iob-chan.h
0 → 100644
+
88
−
0
View file @
dd9a3136
/*
Copyright (C) 2010 Tomasz Śniatowski, Adam Radziszewski
Part of the libcorpus2 project
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 3 of the License, or (at your option)
any later version.
This program 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 LICENSE and COPYING files for more details.
*/
#ifndef LIBSORPUS2_IO_IOB_CHAN_H
#define LIBCORPUS2_IO_IOB_CHAN_H
#include
<libcorpus2/io/reader.h>
#include
<libcorpus2/io/writer.h>
#include
<boost/scoped_ptr.hpp>
namespace
Corpus2
{
/**
* Simple writer class to output token in RFTagger-compatible corpora form.
*
* One token per line, token line consists of the orth, followed by the
* tag, followed by newline (one tag per token only). Each sentence
* is followed by a blank line.
*
* The first lexeme is used. No-lexeme tokens trigger a warning unless
* nowarn is passed.
*/
class
IobChanWriter
:
public
TokenWriter
{
public:
IobChanWriter
(
std
::
ostream
&
os
,
const
Tagset
&
tagset
,
const
string_range_vector
&
params
);
void
write_token
(
const
Token
&
t
);
void
write_sentence
(
const
Sentence
&
s
);
void
write_chunk
(
const
Chunk
&
p
);
void
write_tag
(
const
Tag
&
tag
);
static
bool
registered
;
private:
bool
warn_on_no_lexemes_
;
bool
force_
;
};
class
IobChanReader
:
public
BufferedSentenceReader
{
public:
IobChanReader
(
const
Tagset
&
tagset
,
std
::
istream
&
is
);
IobChanReader
(
const
Tagset
&
tagset
,
const
std
::
string
&
filename
);
std
::
istream
&
is
()
{
return
*
is_
;
}
void
set_option
(
const
std
::
string
&
option
);
std
::
string
get_option
(
const
std
::
string
&
option
)
const
;
static
bool
registered
;
protected
:
/// BufferedSentenceReader override
Sentence
::
Ptr
actual_next_sentence
();
std
::
istream
*
is_
;
boost
::
scoped_ptr
<
std
::
istream
>
is_owned_
;
/// Whether to mark all incoming tags as disambiguated
bool
disamb_
;
};
}
/* end ns Corpus2 */
#endif // LIBCORPUS2_IO_IOB_CHAN_H
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