line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
######################################################################## |
2
|
|
|
|
|
|
|
# Author: Patrik Lambert (lambert@talp.ucp.es) |
3
|
|
|
|
|
|
|
# Description: Tools library to manage an Alignment Sets, i.e. a set of |
4
|
|
|
|
|
|
|
# sentences aligned at the word (or phrase) level. |
5
|
|
|
|
|
|
|
#----------------------------------------------------------------------- |
6
|
|
|
|
|
|
|
# Copyright 2004 by Patrik Lambert |
7
|
|
|
|
|
|
|
# |
8
|
|
|
|
|
|
|
# This program is free software; you can redistribute it and/or modify |
9
|
|
|
|
|
|
|
# it under the terms of the GNU General Public License as published by |
10
|
|
|
|
|
|
|
# the Free Software Foundation; either version 2 of the License, or |
11
|
|
|
|
|
|
|
# (at your option) any later version. |
12
|
|
|
|
|
|
|
# |
13
|
|
|
|
|
|
|
# This program is distributed in the hope that it will be useful, |
14
|
|
|
|
|
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of |
15
|
|
|
|
|
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
16
|
|
|
|
|
|
|
# GNU General Public License for more details. |
17
|
|
|
|
|
|
|
# |
18
|
|
|
|
|
|
|
# You should have received a copy of the GNU General Public License |
19
|
|
|
|
|
|
|
# along with this program; if not, write to the Free Software |
20
|
|
|
|
|
|
|
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. |
21
|
|
|
|
|
|
|
######################################################################## |
22
|
|
|
|
|
|
|
|
23
|
|
|
|
|
|
|
package Lingua::AlignmentSet; |
24
|
1
|
|
|
1
|
|
25122
|
use 5.005; |
|
1
|
|
|
|
|
4
|
|
|
1
|
|
|
|
|
68
|
|
25
|
1
|
|
|
1
|
|
7
|
use vars qw($VERSION); |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
58
|
|
26
|
1
|
|
|
1
|
|
5
|
use strict; |
|
1
|
|
|
|
|
7
|
|
|
1
|
|
|
|
|
62
|
|
27
|
|
|
|
|
|
|
$VERSION = 1.1; |
28
|
1
|
|
|
1
|
|
661
|
use Lingua::AlSetLib 1.1; |
|
1
|
|
|
|
|
37
|
|
|
1
|
|
|
|
|
40
|
|
29
|
1
|
|
|
1
|
|
927
|
use Lingua::Alignment 1.1; |
|
1
|
|
|
|
|
40
|
|
|
1
|
|
|
|
|
48
|
|
30
|
1
|
|
|
1
|
|
706
|
use Lingua::WriteLatexFile; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
30
|
|
31
|
1
|
|
|
1
|
|
552
|
use Lingua::AlignmentEval; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
34
|
|
32
|
|
|
|
|
|
|
|
33
|
1
|
|
|
1
|
|
7
|
use Dumpvalue; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
23
|
|
34
|
1
|
|
|
1
|
|
1128
|
use IO::File; |
|
1
|
|
|
|
|
12986
|
|
|
1
|
|
|
|
|
14449
|
|
35
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
my $dumper=new Dumpvalue; |
37
|
|
|
|
|
|
|
my $true = 1; |
38
|
|
|
|
|
|
|
my $false = 0; |
39
|
|
|
|
|
|
|
|
40
|
|
|
|
|
|
|
sub new { |
41
|
2
|
|
|
2
|
0
|
33
|
my ($pkg,$refToFileSets) = @_; |
42
|
2
|
|
|
|
|
9
|
my $refToLocation = readLocation($refToFileSets->[0][0]); |
43
|
2
|
|
|
|
|
5
|
my $format = $refToFileSets->[0][1]; |
44
|
2
|
|
|
|
|
3
|
my $range = $refToFileSets->[0][2]; |
45
|
2
|
|
|
|
|
4
|
my $alSet = {}; |
46
|
|
|
|
|
|
|
#default values: |
47
|
2
|
50
|
|
|
|
5
|
if (!defined($format)){$format="TALP"} |
|
0
|
|
|
|
|
0
|
|
|
2
|
|
|
|
|
4
|
|
48
|
|
|
|
|
|
|
else {$format = uc $format}; |
49
|
2
|
100
|
|
|
|
7
|
if (!defined($range)){$range="1-"}; |
|
1
|
|
|
|
|
2
|
|
50
|
2
|
100
|
|
|
|
6
|
if ($format eq "BLINKER"){ |
51
|
|
|
|
|
|
|
#for future ease we save detailed infos contained in the source sample path |
52
|
1
|
|
|
|
|
4
|
completeBlinkerLocation($refToLocation); |
53
|
|
|
|
|
|
|
} |
54
|
|
|
|
|
|
|
|
55
|
2
|
|
|
|
|
5
|
$alSet->{location}=$refToLocation; |
56
|
2
|
|
|
|
|
5
|
$alSet->{format}=$format; |
57
|
|
|
|
|
|
|
|
58
|
2
|
|
|
|
|
6
|
setRange($alSet,$range); |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
#checking the data: |
61
|
2
|
50
|
|
|
|
16
|
if ($format eq "GIZA"){ |
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
62
|
|
|
|
|
|
|
# if ($ambiguity || $confidence){die "GIZA format not compatible with ambiguity or confidence features"} |
63
|
|
|
|
|
|
|
} elsif ($format eq "BLINKER"){ |
64
|
|
|
|
|
|
|
} elsif ($format eq "NAACL"){ |
65
|
|
|
|
|
|
|
} elsif ($format eq "TALP"){ |
66
|
|
|
|
|
|
|
} else { |
67
|
0
|
|
|
|
|
0
|
die "Unknown format $format. Can't create alignment set object"; |
68
|
|
|
|
|
|
|
} |
69
|
2
|
|
|
|
|
9
|
return bless $alSet,$pkg; |
70
|
|
|
|
|
|
|
} |
71
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
# create a new AlignmentSet that contains the same data of an already existing alignment set (without copying the addresses) |
73
|
|
|
|
|
|
|
sub copy { |
74
|
0
|
|
|
0
|
0
|
0
|
my $alSet = shift; |
75
|
|
|
|
|
|
|
|
76
|
0
|
|
|
|
|
0
|
my $cloneLocation={}; |
77
|
0
|
|
|
|
|
0
|
my ($field,$value); |
78
|
0
|
|
|
|
|
0
|
while (($field,$value)=each (%{$alSet->{location}})){ |
|
0
|
|
|
|
|
0
|
|
79
|
0
|
|
|
|
|
0
|
$cloneLocation->{$field}=$value; |
80
|
|
|
|
|
|
|
} |
81
|
0
|
|
|
|
|
0
|
return Lingua::AlignmentSet->new([[$cloneLocation,$alSet->{format},$alSet->{firstSentPair}."-".$alSet->{lastSentPair}]]); |
82
|
|
|
|
|
|
|
} |
83
|
|
|
|
|
|
|
|
84
|
|
|
|
|
|
|
sub setWordFiles{ |
85
|
1
|
|
|
1
|
0
|
10
|
my ($alSet,$sourcePath,$targetPath) = @_; |
86
|
|
|
|
|
|
|
|
87
|
1
|
|
|
|
|
7
|
$alSet->{location}->{source}=$sourcePath; |
88
|
1
|
|
|
|
|
4
|
$alSet->{location}->{target}=$targetPath; |
89
|
|
|
|
|
|
|
} |
90
|
|
|
|
|
|
|
sub setSourceFile{ |
91
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$sourcePath) = @_; |
92
|
|
|
|
|
|
|
|
93
|
0
|
|
|
|
|
0
|
$alSet->{location}->{source}=$sourcePath; |
94
|
|
|
|
|
|
|
} |
95
|
|
|
|
|
|
|
sub setTargetFile{ |
96
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$targetPath) = @_; |
97
|
|
|
|
|
|
|
|
98
|
0
|
|
|
|
|
0
|
$alSet->{location}->{target}=$targetPath; |
99
|
|
|
|
|
|
|
} |
100
|
|
|
|
|
|
|
|
101
|
|
|
|
|
|
|
sub setTargetToSourceFile{ |
102
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$targetToSourcePath) = @_; |
103
|
|
|
|
|
|
|
|
104
|
0
|
|
|
|
|
0
|
$alSet->{location}->{targetToSource}=$targetToSourcePath; |
105
|
|
|
|
|
|
|
} |
106
|
|
|
|
|
|
|
|
107
|
|
|
|
|
|
|
sub chFormat { |
108
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat,$alignMode)=@_; |
109
|
|
|
|
|
|
|
|
110
|
0
|
|
|
|
|
0
|
$alSet->convert($newLocation,$newFormat,$alignMode); |
111
|
|
|
|
|
|
|
} |
112
|
|
|
|
|
|
|
|
113
|
|
|
|
|
|
|
|
114
|
|
|
|
|
|
|
# Won't work if the sentence files are not specified |
115
|
|
|
|
|
|
|
sub visualise { |
116
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$representation,$format,$outputFH,$mark,$alignMode,$maxRows,$maxCols)=@_; |
117
|
0
|
|
|
|
|
0
|
$representation = lc $representation; |
118
|
0
|
|
|
|
|
0
|
$format = lc $format; |
119
|
0
|
0
|
|
|
|
0
|
if (!defined($outputFH)){$outputFH=*STDOUT} |
|
0
|
|
|
|
|
0
|
|
120
|
0
|
0
|
|
|
|
0
|
if ($representation eq "matrix"){ |
121
|
0
|
0
|
|
|
|
0
|
if (!defined($mark)){$mark = "cross"} |
|
0
|
|
|
|
|
0
|
|
122
|
0
|
0
|
|
|
|
0
|
if (!defined($maxRows)){$maxRows = 53} #default maxRows value |
|
0
|
|
|
|
|
0
|
|
123
|
0
|
0
|
|
|
|
0
|
if (!defined($maxCols)){$maxCols = 35} #default maxRows value |
|
0
|
|
|
|
|
0
|
|
124
|
0
|
|
|
|
|
0
|
$format="latex"; |
125
|
|
|
|
|
|
|
} |
126
|
0
|
|
|
|
|
0
|
my $latex = Lingua::Latex->new; |
127
|
0
|
0
|
|
|
|
0
|
if ($format eq "latex"){ |
128
|
0
|
|
|
|
|
0
|
print $outputFH $latex->startFile; |
129
|
0
|
|
|
|
|
0
|
print $outputFH $latex->setTabcolsep("0.5mm"); |
130
|
|
|
|
|
|
|
} |
131
|
0
|
|
|
|
|
0
|
my $output = ""; |
132
|
0
|
|
|
|
|
0
|
my $inputSentPairNum = $alSet->{firstSentPair}; |
133
|
0
|
|
|
|
|
0
|
my $i; |
134
|
0
|
|
|
|
|
0
|
my ($al,$alSetChunk); |
135
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
136
|
0
|
0
|
0
|
|
|
0
|
if (($alSet->{format} ne "GIZA") && (!$FH->{source} || !$FH->{target})){ |
|
|
|
0
|
|
|
|
|
137
|
0
|
|
|
|
|
0
|
die "To use the 'visualise' function, you must specify the sentence (words) files.\n"; |
138
|
|
|
|
|
|
|
} |
139
|
0
|
|
|
|
|
0
|
while ($alSetChunk = $alSet->loadChunk($FH,$inputSentPairNum,$alignMode)){ # returns 0 if eof or last sentence pair |
140
|
0
|
|
|
|
|
0
|
$output = ""; |
141
|
0
|
|
|
|
|
0
|
for ($i=0;$i<@$alSetChunk;$i++){ |
142
|
0
|
|
|
|
|
0
|
$al = $$alSetChunk[$i]; |
143
|
|
|
|
|
|
|
# print main::Dumper($al); |
144
|
0
|
0
|
|
|
|
0
|
if ($representation eq "matrix"){ |
|
|
0
|
|
|
|
|
|
145
|
0
|
|
|
|
|
0
|
$output.= "\n$inputSentPairNum\n".$al->displayAsMatrix($latex,$mark,$maxRows,$maxCols); |
146
|
|
|
|
|
|
|
}elsif ($representation eq "enumlinks"){ |
147
|
0
|
|
|
|
|
0
|
$output.= "\n$inputSentPairNum\n".$al->displayAsLinkEnumeration($format,$latex); |
148
|
|
|
|
|
|
|
} #elsif |
149
|
|
|
|
|
|
|
}#for |
150
|
0
|
|
|
|
|
0
|
print $outputFH $output; |
151
|
0
|
|
|
|
|
0
|
$inputSentPairNum++; |
152
|
|
|
|
|
|
|
} |
153
|
0
|
0
|
|
|
|
0
|
if ($format eq "latex"){print $outputFH $latex->endFile}; |
|
0
|
|
|
|
|
0
|
|
154
|
|
|
|
|
|
|
} |
155
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
#only work if the text files are given (not only the alignment files). |
157
|
|
|
|
|
|
|
sub getSize { |
158
|
0
|
|
|
0
|
0
|
0
|
my $alSet = shift; |
159
|
0
|
|
|
|
|
0
|
my ($file,$factor); |
160
|
0
|
|
|
|
|
0
|
my $size; |
161
|
|
|
|
|
|
|
|
162
|
0
|
0
|
0
|
|
|
0
|
if ($alSet->{format} eq "GIZA"){ |
|
|
0
|
0
|
|
|
|
|
163
|
0
|
|
|
|
|
0
|
$file = $alSet->{location}->{sourceToTarget}; |
164
|
0
|
|
|
|
|
0
|
$factor = 3; |
165
|
|
|
|
|
|
|
}elsif ($alSet->{format} eq "NAACL" || $alSet->{format} eq "BLINKER" || $alSet->{format} eq "TALP"){ |
166
|
0
|
0
|
|
|
|
0
|
if (!$alSet->{location}->{source}){ |
167
|
0
|
|
|
|
|
0
|
die "One of the functions your are using requires you specify the sentence files (source and target)\n"; |
168
|
|
|
|
|
|
|
} |
169
|
0
|
|
|
|
|
0
|
$file = $alSet->{location}->{source}; |
170
|
0
|
|
|
|
|
0
|
$factor = 1; |
171
|
|
|
|
|
|
|
} |
172
|
0
|
|
|
|
|
0
|
open (FILE,"<$file"); |
173
|
0
|
|
|
|
|
0
|
$size += tr/\n/\n/ while sysread(FILE, $_, 2 ** 16); |
174
|
0
|
|
|
|
|
0
|
close(FILE); |
175
|
0
|
|
|
|
|
0
|
$size = $size / $factor; |
176
|
0
|
|
|
|
|
0
|
return $size; |
177
|
|
|
|
|
|
|
} |
178
|
|
|
|
|
|
|
|
179
|
|
|
|
|
|
|
# returns a list (in random order) of lineNumbers |
180
|
|
|
|
|
|
|
# to sort this list, do: my @sortedSelection = sort { $a <=> $b; } @selection; |
181
|
|
|
|
|
|
|
|
182
|
|
|
|
|
|
|
sub chooseSubsets { |
183
|
|
|
|
|
|
|
#TO DO: possibility of percentage input for the size |
184
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$seed,$size) = @_; |
185
|
0
|
|
|
|
|
0
|
my $alSetSize = $alSet->getSize(); |
186
|
0
|
|
|
|
|
0
|
my $count; |
187
|
0
|
|
|
|
|
0
|
my @selected=(); |
188
|
0
|
|
|
|
|
0
|
my @notSelected = (); |
189
|
0
|
|
|
|
|
0
|
my ($ind,$elt); |
190
|
|
|
|
|
|
|
|
191
|
0
|
|
|
|
|
0
|
for ($count=1;$count<$alSetSize;$count++){ |
192
|
0
|
|
|
|
|
0
|
push @notSelected,$count; |
193
|
|
|
|
|
|
|
} |
194
|
0
|
|
|
|
|
0
|
srand $seed; |
195
|
0
|
|
|
|
|
0
|
for ($count=0;$count<$size;$count++){ |
196
|
0
|
|
|
|
|
0
|
$ind = rand @notSelected; |
197
|
0
|
|
|
|
|
0
|
$elt = $notSelected[$ind]; |
198
|
0
|
|
|
|
|
0
|
splice @notSelected, $ind, 1; |
199
|
0
|
|
|
|
|
0
|
push @selected,$elt; |
200
|
|
|
|
|
|
|
} |
201
|
0
|
|
|
|
|
0
|
return \@selected; |
202
|
|
|
|
|
|
|
} |
203
|
|
|
|
|
|
|
################################################################### |
204
|
|
|
|
|
|
|
### EVALUATION ### |
205
|
|
|
|
|
|
|
################################################################### |
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
#code adapted from Rada Mihalcea's wa_eval_align.pl, rada@cs.unt.edu |
208
|
|
|
|
|
|
|
# Evaluation is performed using: |
209
|
|
|
|
|
|
|
# - Standard Precision, Recall, F-measure, separate for S (Sure) and P (Possible) cases |
210
|
|
|
|
|
|
|
# - AER measure, defined as |
211
|
|
|
|
|
|
|
# AER = 1 - ( |A & S| + |A & P| ) / ( |A| + |S| ) |
212
|
|
|
|
|
|
|
# [where A represents the alignment, S and P represent the S (Sure) and P (Possible) gold standard alignments] |
213
|
|
|
|
|
|
|
|
214
|
|
|
|
|
|
|
sub evaluate { |
215
|
1
|
|
|
1
|
0
|
8
|
my ($submissionAlSet,$answerAlSet,$alignMode,$weighted)=@_; |
216
|
1
|
50
|
|
|
|
4
|
if (!defined($weighted)){$weighted=0} |
|
1
|
|
|
|
|
1
|
|
217
|
1
|
|
|
|
|
2
|
my ($line,$alignment); |
218
|
0
|
|
|
|
|
0
|
my ($FH,$alSetChunk,$i,$al,$fhPos); |
219
|
0
|
|
|
|
|
0
|
my ($inputSentPairNum,$internalSentPairNum,$sentPairNum); |
220
|
0
|
|
|
|
|
0
|
my ($sureMatch,$possibleMatch,$possibleMatchSure); |
221
|
0
|
|
|
|
|
0
|
my ($surePrecision,$sureRecall,$possiblePrecision,$possibleRecall,$sureFMeasure,$possibleFMeasure,$AER); |
222
|
|
|
|
|
|
|
# 1 READ ANSWER AND SUBMISSION FILES |
223
|
|
|
|
|
|
|
# (in the case of NAACL format file it's more efficient to treat it directly, otherwise load to internal structure) |
224
|
|
|
|
|
|
|
# answer file |
225
|
0
|
|
|
|
|
0
|
my %sureAnswer; |
226
|
0
|
|
|
|
|
0
|
my %possibleAnswer; |
227
|
1
|
|
|
|
|
2
|
my $INFINITY = 9999999999; |
228
|
1
|
|
|
|
|
2
|
$inputSentPairNum = $answerAlSet->{firstSentPair}; |
229
|
1
|
|
|
|
|
2
|
$internalSentPairNum = 1; |
230
|
1
|
50
|
33
|
|
|
7
|
if ( $answerAlSet->{format} eq "NAACL" && $alignMode eq "as-is" && $answerAlSet->{firstSentPair} == 1){ |
|
|
|
33
|
|
|
|
|
231
|
0
|
0
|
|
|
|
0
|
my $answerFH = IO::File->new("<".$answerAlSet->{location}{sourceToTarget}) or die "Answer alignment file opening error:$!"; |
232
|
|
|
|
|
|
|
#go to first sentence pair: |
233
|
0
|
|
|
|
|
0
|
$fhPos = $answerFH->getpos; |
234
|
0
|
|
0
|
|
|
0
|
while ($answerFH->getline() !~ m/^0*$inputSentPairNum .*/o && !$answerFH->eof()) { |
235
|
0
|
|
|
|
|
0
|
$fhPos = $answerFH->getpos; |
236
|
|
|
|
|
|
|
} |
237
|
0
|
0
|
|
|
|
0
|
if ($answerFH->eof()){ |
238
|
0
|
|
|
|
|
0
|
die "First sentence pair of range not found in ".$answerAlSet->{location}{sourceToTarget}; |
239
|
|
|
|
|
|
|
} |
240
|
0
|
|
|
|
|
0
|
$answerFH->setpos($fhPos); #if we changed sentences, we read the first line of next sentence=>go back one line |
241
|
|
|
|
|
|
|
|
242
|
|
|
|
|
|
|
#read file: |
243
|
0
|
0
|
|
|
|
0
|
if ($answerAlSet->{lastSentPair} eq "eof"){ |
244
|
0
|
|
|
|
|
0
|
$inputSentPairNum = $INFINITY; |
245
|
|
|
|
|
|
|
}else{ |
246
|
0
|
|
|
|
|
0
|
$inputSentPairNum = $answerAlSet->{lastSentPair}+1; |
247
|
|
|
|
|
|
|
} |
248
|
0
|
|
0
|
|
|
0
|
while(!$answerFH->eof() && ( ($line=$answerFH->getline()) !~ m/^0*$inputSentPairNum .*/o )) { |
249
|
0
|
|
|
|
|
0
|
chomp $line; |
250
|
0
|
|
|
|
|
0
|
$line =~ s/^\s+|\s+$//g; |
251
|
0
|
|
|
|
|
0
|
identifySurePossible($line,\%sureAnswer,\%possibleAnswer); |
252
|
|
|
|
|
|
|
} |
253
|
0
|
|
|
|
|
0
|
$answerFH->close(); |
254
|
|
|
|
|
|
|
}else{ |
255
|
1
|
|
|
|
|
7
|
$FH = $answerAlSet->openFiles(); |
256
|
1
|
|
|
|
|
12
|
while ($alSetChunk = $answerAlSet->loadChunk($FH,$inputSentPairNum,$alignMode)){ # returns 0 if eof or last sentence pair |
257
|
10
|
|
|
|
|
52
|
for ($i=0;$i<@$alSetChunk;$i++){ |
258
|
10
|
|
|
|
|
15
|
$al = $$alSetChunk[$i]; |
259
|
|
|
|
|
|
|
# print "EVALUATE: answer al:\n"; |
260
|
|
|
|
|
|
|
# print main::Dumper($al); |
261
|
10
|
|
|
|
|
52
|
foreach $line (@{$al->writeToBlinker()}){ |
|
10
|
|
|
|
|
32
|
|
262
|
90
|
|
|
|
|
142
|
$line = "$internalSentPairNum ".$line; |
263
|
90
|
|
|
|
|
179
|
identifySurePossible($line,\%sureAnswer,\%possibleAnswer); |
264
|
|
|
|
|
|
|
} |
265
|
|
|
|
|
|
|
} |
266
|
10
|
|
|
|
|
12
|
$inputSentPairNum++; |
267
|
10
|
|
|
|
|
27
|
$internalSentPairNum++; |
268
|
|
|
|
|
|
|
} |
269
|
1
|
|
|
|
|
4
|
closeFiles($FH,$answerAlSet->{format}); |
270
|
|
|
|
|
|
|
}# if format - else |
271
|
|
|
|
|
|
|
|
272
|
|
|
|
|
|
|
# submission file |
273
|
1
|
|
|
|
|
2
|
my %sureSubmission; |
274
|
|
|
|
|
|
|
my %possibleSubmission; |
275
|
1
|
|
|
|
|
2
|
$inputSentPairNum = $submissionAlSet->{firstSentPair}; |
276
|
1
|
|
|
|
|
2
|
$internalSentPairNum = 1; |
277
|
|
|
|
|
|
|
|
278
|
1
|
50
|
33
|
|
|
8
|
if ($submissionAlSet->{format} eq "NAACL" && $alignMode eq "as-is" && $submissionAlSet->{firstSentPair}==1){ |
|
|
|
33
|
|
|
|
|
279
|
0
|
0
|
|
|
|
0
|
my $submissionFH = IO::File->new("<".$submissionAlSet->{location}{sourceToTarget}) or die "Submission alignment file opening error:$!"; |
280
|
|
|
|
|
|
|
|
281
|
|
|
|
|
|
|
#go to first sentence pair: |
282
|
0
|
|
|
|
|
0
|
$fhPos = $submissionFH->getpos; |
283
|
0
|
|
0
|
|
|
0
|
while ($submissionFH->getline() !~ m/^0*$inputSentPairNum .*/o && !$submissionFH->eof()) { |
284
|
0
|
|
|
|
|
0
|
$fhPos = $submissionFH->getpos; |
285
|
|
|
|
|
|
|
} |
286
|
0
|
0
|
|
|
|
0
|
if ($submissionFH->eof()){ |
287
|
0
|
|
|
|
|
0
|
die "First sentence pair of range not found in ".$submissionAlSet->{location}{sourceToTarget}; |
288
|
|
|
|
|
|
|
} |
289
|
0
|
|
|
|
|
0
|
$submissionFH->setpos($fhPos); #if we changed sentences, we read the first line of next sentence=>go back one line |
290
|
|
|
|
|
|
|
|
291
|
|
|
|
|
|
|
#read file: |
292
|
0
|
0
|
|
|
|
0
|
if ($submissionAlSet->{lastSentPair} eq "eof"){ |
293
|
0
|
|
|
|
|
0
|
$inputSentPairNum = $INFINITY; |
294
|
|
|
|
|
|
|
}else{ |
295
|
0
|
|
|
|
|
0
|
$inputSentPairNum = $submissionAlSet->{lastSentPair}+1; |
296
|
|
|
|
|
|
|
} |
297
|
0
|
|
0
|
|
|
0
|
while(!$submissionFH->eof() && (($line = $submissionFH->getline()) !~ m/^0*$inputSentPairNum .*/o )) { |
298
|
0
|
|
|
|
|
0
|
chomp $line; |
299
|
0
|
|
|
|
|
0
|
$line =~ s/^\s+|\s+$//g; |
300
|
0
|
|
|
|
|
0
|
identifySurePossible($line,\%sureSubmission,\%possibleSubmission); |
301
|
|
|
|
|
|
|
} |
302
|
0
|
|
|
|
|
0
|
$submissionFH->close(); |
303
|
|
|
|
|
|
|
}else{ |
304
|
1
|
|
|
|
|
4
|
$FH = $submissionAlSet->openFiles(); |
305
|
1
|
|
|
|
|
4
|
while ($alSetChunk = $submissionAlSet->loadChunk($FH,$inputSentPairNum,$alignMode)){ # returns 0 if eof or last sentence pair |
306
|
10
|
|
|
|
|
27
|
for ($i=0;$i<@$alSetChunk;$i++){ |
307
|
10
|
|
|
|
|
12
|
$al = $$alSetChunk[$i]; |
308
|
|
|
|
|
|
|
# print "submission al:\n"; |
309
|
|
|
|
|
|
|
# $dumper->dumpValue($al); |
310
|
10
|
|
|
|
|
56
|
foreach $line (@{$al->writeToBlinker()}){ |
|
10
|
|
|
|
|
38
|
|
311
|
59
|
|
|
|
|
100
|
$line = "$internalSentPairNum ".$line; |
312
|
59
|
|
|
|
|
120
|
identifySurePossible($line,\%sureSubmission,\%possibleSubmission); |
313
|
|
|
|
|
|
|
} |
314
|
|
|
|
|
|
|
} |
315
|
10
|
|
|
|
|
11
|
$inputSentPairNum++; |
316
|
10
|
|
|
|
|
29
|
$internalSentPairNum++; |
317
|
|
|
|
|
|
|
} |
318
|
1
|
|
|
|
|
6
|
closeFiles($FH,$submissionAlSet->{format}); |
319
|
|
|
|
|
|
|
}# if format=NAACL else |
320
|
|
|
|
|
|
|
|
321
|
|
|
|
|
|
|
|
322
|
|
|
|
|
|
|
# print "weighted:$weighted\n"; |
323
|
|
|
|
|
|
|
# print "SA:".join("-",keys %sureAnswer),"\nSS:".join(" - ",keys %sureSubmission),"\nPA:".join(" - ",keys %possibleAnswer),"\nPS:".join(" - ",keys %possibleSubmission)."\n"; |
324
|
|
|
|
|
|
|
|
325
|
|
|
|
|
|
|
# 2 WEIGHT LINKS |
326
|
|
|
|
|
|
|
# It is a kind of "normalization" of multiple links: each link (j i) is weighted according to |
327
|
|
|
|
|
|
|
# the number of links in which j and i are involved: weight(j,i)=0.5*(1/numLinks(j)+1/numLinks(i)). |
328
|
|
|
|
|
|
|
|
329
|
1
|
|
|
|
|
3
|
my ($link,$j,$hash,$value); |
330
|
0
|
|
|
|
|
0
|
my %weightsSure; |
331
|
0
|
|
|
|
|
0
|
my %linksSure; |
332
|
0
|
|
|
|
|
0
|
my @linksSureInSentence; |
333
|
0
|
|
|
|
|
0
|
my %linksPossible; |
334
|
0
|
|
|
|
|
0
|
my @linksPossibleInSentence; |
335
|
0
|
|
|
|
|
0
|
my %weightsPossible; |
336
|
|
|
|
|
|
|
|
337
|
1
|
50
|
|
|
|
6
|
if ($weighted){ |
338
|
|
|
|
|
|
|
# When only sure links are considered (calculation of Ps and Rs), they are weighted with respect to the union of both sure sets |
339
|
|
|
|
|
|
|
# take union |
340
|
0
|
|
|
|
|
0
|
foreach $hash ( \%sureSubmission, \%sureAnswer ) { |
341
|
0
|
|
|
|
|
0
|
while (($link, $value) = each %$hash) { |
342
|
0
|
|
|
|
|
0
|
($sentPairNum,$j,$i)=split(" ",$link); |
343
|
0
|
|
|
|
|
0
|
$linksSure{$sentPairNum}{"$j $i"} = $value; |
344
|
|
|
|
|
|
|
} |
345
|
|
|
|
|
|
|
} |
346
|
|
|
|
|
|
|
# calculate weight of each link |
347
|
0
|
|
|
|
|
0
|
foreach $sentPairNum (keys %linksSure){ |
348
|
0
|
|
|
|
|
0
|
@linksSureInSentence =keys %{$linksSure{$sentPairNum}}; |
|
0
|
|
|
|
|
0
|
|
349
|
0
|
|
|
|
|
0
|
foreach $link (@linksSureInSentence){ |
350
|
0
|
|
|
|
|
0
|
($j,$i)=split(" ",$link); |
351
|
0
|
|
|
|
|
0
|
$weightsSure{"$sentPairNum $link"}=0.5*( 1/grep(/^$j /,@linksSureInSentence)+1/grep(/ $i$/,@linksSureInSentence) ); |
352
|
|
|
|
|
|
|
} |
353
|
|
|
|
|
|
|
} |
354
|
|
|
|
|
|
|
|
355
|
|
|
|
|
|
|
# When all links are considered (calculation of Pp and Rp, AER), possible AND sure links are weighted with respect to the union of all sets. |
356
|
0
|
|
|
|
|
0
|
%linksPossible=%linksSure; |
357
|
|
|
|
|
|
|
# add union of possible links |
358
|
0
|
|
|
|
|
0
|
foreach $hash (\%possibleSubmission, \%possibleAnswer ) { |
359
|
0
|
|
|
|
|
0
|
while (($link, $value) = each %$hash) { |
360
|
0
|
|
|
|
|
0
|
($sentPairNum,$j,$i)=split(" ",$link); |
361
|
0
|
|
|
|
|
0
|
$linksPossible{$sentPairNum}{"$j $i"} = $value; |
362
|
|
|
|
|
|
|
} |
363
|
|
|
|
|
|
|
} |
364
|
|
|
|
|
|
|
# calculate weight of each link |
365
|
0
|
|
|
|
|
0
|
foreach $sentPairNum (keys %linksPossible){ |
366
|
0
|
|
|
|
|
0
|
@linksPossibleInSentence =keys %{$linksPossible{$sentPairNum}}; |
|
0
|
|
|
|
|
0
|
|
367
|
0
|
|
|
|
|
0
|
foreach $link (@linksPossibleInSentence){ |
368
|
0
|
|
|
|
|
0
|
($j,$i)=split(" ",$link); |
369
|
0
|
|
|
|
|
0
|
$weightsPossible{"$sentPairNum $link"}=0.5*( 1/grep(/^$j /,@linksPossibleInSentence)+1/grep(/ $i$/,@linksPossibleInSentence) ); |
370
|
|
|
|
|
|
|
} |
371
|
|
|
|
|
|
|
} |
372
|
|
|
|
|
|
|
} |
373
|
|
|
|
|
|
|
|
374
|
|
|
|
|
|
|
# 3 SUM UP LINKS |
375
|
|
|
|
|
|
|
# in case of weights distinct from 1: sum of %possibleAnswer and %possibleSubmission is always with %weightsPossible. |
376
|
|
|
|
|
|
|
# however the sum of %sureAnswer and %sureSubmission is with %weightsSure to calculate Ps and Rs, %weightsPossible for Pp, Rp and AER. |
377
|
1
|
|
|
|
|
4
|
my ($totalPossibleAnswer,$totalPossibleSubmission)=(0,0); |
378
|
1
|
|
|
|
|
2
|
my ($totalSureAnswer_weightsSure,$totalSureSubmission_weightsSure,$totalSureAnswer_weightsPossible,$totalSureSubmission_weightsPossible)=(0,0,0,0); |
379
|
1
|
50
|
|
|
|
9
|
if ($weighted){ |
380
|
0
|
|
|
|
|
0
|
foreach $link (keys %sureAnswer){ |
381
|
0
|
|
|
|
|
0
|
$totalSureAnswer_weightsSure+=$weightsSure{$link}; |
382
|
0
|
|
|
|
|
0
|
$totalSureAnswer_weightsPossible+=$weightsPossible{$link}; |
383
|
|
|
|
|
|
|
} |
384
|
0
|
|
|
|
|
0
|
foreach $link (keys %sureSubmission){ |
385
|
0
|
|
|
|
|
0
|
$totalSureSubmission_weightsSure+=$weightsSure{$link}; |
386
|
0
|
|
|
|
|
0
|
$totalSureSubmission_weightsPossible+=$weightsPossible{$link}; |
387
|
|
|
|
|
|
|
} |
388
|
0
|
|
|
|
|
0
|
foreach $link (keys %possibleAnswer){ |
389
|
0
|
|
|
|
|
0
|
$totalPossibleAnswer+=$weightsPossible{$link}; |
390
|
|
|
|
|
|
|
} |
391
|
0
|
|
|
|
|
0
|
foreach $link (keys %possibleSubmission){ |
392
|
0
|
|
|
|
|
0
|
$totalPossibleSubmission+=$weightsPossible{$link}; |
393
|
|
|
|
|
|
|
} |
394
|
|
|
|
|
|
|
}else{ #every link has a weight 1 |
395
|
1
|
|
|
|
|
8
|
$totalSureAnswer_weightsSure=scalar(keys %sureAnswer); |
396
|
1
|
|
|
|
|
2
|
$totalSureAnswer_weightsPossible= $totalSureAnswer_weightsSure; |
397
|
1
|
|
|
|
|
3
|
$totalSureSubmission_weightsSure=scalar(keys %sureSubmission); |
398
|
1
|
|
|
|
|
3
|
$totalSureSubmission_weightsPossible=$totalSureSubmission_weightsSure; |
399
|
1
|
|
|
|
|
2
|
$totalPossibleAnswer=scalar(keys %possibleAnswer); |
400
|
1
|
|
|
|
|
3
|
$totalPossibleSubmission=scalar(keys %possibleSubmission); |
401
|
|
|
|
|
|
|
} |
402
|
|
|
|
|
|
|
|
403
|
|
|
|
|
|
|
# 4 COUNT MATCHES |
404
|
|
|
|
|
|
|
|
405
|
|
|
|
|
|
|
# print "sureSubmission:",join("|",keys %sureSubmission),"\n"; |
406
|
|
|
|
|
|
|
# print "possibleSubmission:",join("|",keys %possibleSubmission),"\n"; |
407
|
|
|
|
|
|
|
# print "sureAnswer:",join("|",keys %sureAnswer),"\n"; |
408
|
|
|
|
|
|
|
# print "possibleAnswer:",join("|",keys %possibleAnswer),"\n"; |
409
|
|
|
|
|
|
|
# print "\n"; |
410
|
|
|
|
|
|
|
# now determine the S[ure] matches |
411
|
1
|
|
|
|
|
2
|
$sureMatch = 0; |
412
|
1
|
|
|
|
|
13
|
foreach $alignment (keys %sureSubmission) { |
413
|
59
|
100
|
|
|
|
122
|
if(defined($sureAnswer{$alignment})) { |
414
|
54
|
50
|
|
|
|
80
|
if (!$weighted){$sureMatch++} |
|
54
|
|
|
|
|
68
|
|
|
0
|
|
|
|
|
0
|
|
415
|
|
|
|
|
|
|
else {$sureMatch += $weightsSure{$alignment}} |
416
|
|
|
|
|
|
|
} |
417
|
|
|
|
|
|
|
} |
418
|
|
|
|
|
|
|
# and the [P]robable matches |
419
|
|
|
|
|
|
|
# these are checked against both S[ure] and P[robable] correct alignments |
420
|
1
|
|
|
|
|
6
|
$possibleMatch = 0; |
421
|
1
|
|
|
|
|
10
|
foreach $alignment (keys %possibleSubmission, keys %sureSubmission) { |
422
|
59
|
100
|
66
|
|
|
153
|
if(defined($sureAnswer{$alignment}) || defined($possibleAnswer{$alignment})) { |
423
|
54
|
50
|
|
|
|
79
|
if (!$weighted){$possibleMatch++} |
|
54
|
|
|
|
|
66
|
|
|
0
|
|
|
|
|
0
|
|
424
|
|
|
|
|
|
|
else{$possibleMatch += $weightsPossible{$alignment}} |
425
|
|
|
|
|
|
|
} |
426
|
|
|
|
|
|
|
} |
427
|
|
|
|
|
|
|
# and also the intersection between all submitted alignments |
428
|
|
|
|
|
|
|
# and the S [Sure] correct alignments -- as needed by AER |
429
|
1
|
|
|
|
|
8
|
$possibleMatchSure = 0; |
430
|
1
|
|
|
|
|
9
|
foreach $alignment (keys %possibleSubmission, keys %sureSubmission) { |
431
|
59
|
100
|
|
|
|
120
|
if(defined($sureAnswer{$alignment})) { |
432
|
54
|
50
|
|
|
|
77
|
if (!$weighted){$possibleMatchSure++} |
|
54
|
|
|
|
|
67
|
|
|
0
|
|
|
|
|
0
|
|
433
|
|
|
|
|
|
|
else{$possibleMatchSure+= $weightsPossible{$alignment}} |
434
|
|
|
|
|
|
|
} |
435
|
|
|
|
|
|
|
} |
436
|
|
|
|
|
|
|
# print "sureMatch:$sureMatch possibleMatch:$possibleMatch possibleMatchSure:$possibleMatchSure\n"; |
437
|
|
|
|
|
|
|
|
438
|
|
|
|
|
|
|
# 5 COMPUTE EVALUATION MEASURES |
439
|
|
|
|
|
|
|
# now determine the precision, recall, and F-measure for [S]ure alignments |
440
|
1
|
50
|
|
|
|
8
|
if(scalar(keys %sureSubmission) != 0) { |
441
|
1
|
|
|
|
|
3
|
$surePrecision = $sureMatch / $totalSureSubmission_weightsSure; |
442
|
|
|
|
|
|
|
}else { |
443
|
0
|
|
|
|
|
0
|
$surePrecision = 0; |
444
|
|
|
|
|
|
|
} |
445
|
1
|
50
|
|
|
|
5
|
if(scalar(keys %sureAnswer) != 0) { |
446
|
1
|
|
|
|
|
4
|
$sureRecall = $sureMatch / $totalSureAnswer_weightsSure; |
447
|
|
|
|
|
|
|
}else { |
448
|
0
|
|
|
|
|
0
|
$sureRecall = 0; |
449
|
|
|
|
|
|
|
} |
450
|
1
|
50
|
33
|
|
|
9
|
if($sureRecall != 0 && $surePrecision != 0) { |
451
|
1
|
|
|
|
|
4
|
$sureFMeasure = 2 * $sureRecall * $surePrecision / ($sureRecall + $surePrecision); |
452
|
|
|
|
|
|
|
}else { |
453
|
0
|
|
|
|
|
0
|
$sureFMeasure = 0; |
454
|
|
|
|
|
|
|
} |
455
|
|
|
|
|
|
|
|
456
|
|
|
|
|
|
|
|
457
|
|
|
|
|
|
|
# and now determine the precision, recall, and F-measure for [P]robable alignments |
458
|
1
|
50
|
|
|
|
6
|
if(scalar(keys %sureSubmission) + scalar(keys %possibleSubmission) != 0) { |
459
|
1
|
|
|
|
|
2
|
$possiblePrecision = $possibleMatch / ($totalSureSubmission_weightsPossible+$totalPossibleSubmission); |
460
|
|
|
|
|
|
|
}else { |
461
|
0
|
|
|
|
|
0
|
$possiblePrecision = 0; |
462
|
|
|
|
|
|
|
} |
463
|
1
|
50
|
|
|
|
7
|
if(scalar(keys %sureAnswer) + scalar(keys %possibleAnswer)!= 0) { |
464
|
1
|
|
|
|
|
3
|
$possibleRecall = $possibleMatch / ($totalSureAnswer_weightsPossible+$totalPossibleAnswer); |
465
|
|
|
|
|
|
|
}else { |
466
|
0
|
|
|
|
|
0
|
$possibleRecall = 0; |
467
|
|
|
|
|
|
|
} |
468
|
1
|
50
|
33
|
|
|
8
|
if($possibleRecall != 0 && $possiblePrecision != 0) { |
469
|
1
|
|
|
|
|
2
|
$possibleFMeasure = 2 * $possibleRecall * $possiblePrecision / ($possibleRecall + $possiblePrecision); |
470
|
|
|
|
|
|
|
}else { |
471
|
0
|
|
|
|
|
0
|
$possibleFMeasure = 0; |
472
|
|
|
|
|
|
|
} |
473
|
|
|
|
|
|
|
|
474
|
|
|
|
|
|
|
# and determine the AER |
475
|
1
|
50
|
|
|
|
3
|
if(scalar(keys %sureSubmission) + scalar(keys %possibleSubmission) != 0) { |
476
|
1
|
|
|
|
|
3
|
$AER = 1 - ($possibleMatchSure + $possibleMatch) / ($totalSureSubmission_weightsPossible+$totalPossibleSubmission+$totalSureAnswer_weightsPossible); |
477
|
|
|
|
|
|
|
}else { |
478
|
0
|
|
|
|
|
0
|
$AER = 0; |
479
|
|
|
|
|
|
|
} |
480
|
1
|
|
|
|
|
11
|
return Lingua::AlignmentEval->new($surePrecision,$sureRecall,$sureFMeasure,$possiblePrecision,$possibleRecall,$possibleFMeasure,$AER); |
481
|
|
|
|
|
|
|
} |
482
|
|
|
|
|
|
|
################################################################### |
483
|
|
|
|
|
|
|
### PROCESSING ### |
484
|
|
|
|
|
|
|
################################################################### |
485
|
|
|
|
|
|
|
|
486
|
|
|
|
|
|
|
sub processAlignment{ |
487
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$AlignmentSub,$newLocation,$newFormat,$alignMode)=@_; |
488
|
0
|
|
|
|
|
0
|
my $newAlSet = $alSet->copy; |
489
|
0
|
0
|
|
|
|
0
|
if (ref($AlignmentSub) eq 'ARRAY'){ |
490
|
0
|
0
|
|
|
|
0
|
if ($AlignmentSub->[0] eq "Lingua::Alignment::eliminateWord"){ |
491
|
0
|
0
|
|
|
|
0
|
if (@$AlignmentSub<3){die "Missing parameters for Lingua::Alignment::eliminateWord\n"} |
|
0
|
|
|
|
|
0
|
|
492
|
|
|
|
|
|
|
else{ |
493
|
0
|
|
|
|
|
0
|
my $side = lc $AlignmentSub->[2]; |
494
|
0
|
0
|
0
|
|
|
0
|
if (!$alSet->{location}{$side} || !$newLocation->{$side}){die "Missing $side file for Lingua::Alignment::eliminateWord\n"} |
|
0
|
|
|
|
|
0
|
|
495
|
|
|
|
|
|
|
} |
496
|
|
|
|
|
|
|
} |
497
|
|
|
|
|
|
|
} |
498
|
0
|
|
|
|
|
0
|
$newAlSet->convert($newLocation,$newFormat,$alignMode,$AlignmentSub); |
499
|
0
|
|
|
|
|
0
|
return $newAlSet; |
500
|
|
|
|
|
|
|
} |
501
|
|
|
|
|
|
|
|
502
|
|
|
|
|
|
|
sub symmetrize { |
503
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat,$ENV,$selectSubgroups,$alignMode,$globals)=@_; |
504
|
|
|
|
|
|
|
#defaults |
505
|
0
|
0
|
|
|
|
0
|
if (!defined($selectSubgroups)){$selectSubgroups=0} |
|
0
|
|
|
|
|
0
|
|
506
|
0
|
0
|
|
|
|
0
|
if (!defined($alignMode)){$alignMode="no-null-align"} |
|
0
|
|
|
|
|
0
|
|
507
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"minPhraseFrequency"})){$globals->{"minPhraseFrequency"}=2}; |
|
0
|
|
|
|
|
0
|
|
508
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"extendGroups"})){$globals->{"extendGroups"}=0}; |
|
0
|
|
|
|
|
0
|
|
509
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"onlyGroups"})){$globals->{"onlyGroups"}=1}; |
|
0
|
|
|
|
|
0
|
|
510
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"defaultActionGrouping"})){$globals->{"defaultActionGrouping"}="Lingua::Alignment::getUnion"}; |
|
0
|
|
|
|
|
0
|
|
511
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"defaultActionGeneral"})){$globals->{"defaultActionGeneral"}="Lingua::Alignment::intersect"}; |
|
0
|
|
|
|
|
0
|
|
512
|
0
|
0
|
|
|
|
0
|
if (!defined($globals->{"verbose"})){$globals->{"verbose"}="0"}; |
|
0
|
|
|
|
|
0
|
|
513
|
0
|
|
|
|
|
0
|
my $verbose = $globals->{"verbose"}; |
514
|
0
|
|
|
|
|
0
|
my $al; # reference alignment- remains unchanged |
515
|
|
|
|
|
|
|
my $modAl; #reference alignment modified with the successive aplication of symRules |
516
|
|
|
|
|
|
|
#load in memory a chunk of the alignment set as a list |
517
|
|
|
|
|
|
|
#of references to (internal representation) alignment objects: |
518
|
0
|
|
|
|
|
0
|
my ($k,$alSetChunk); |
519
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
520
|
0
|
|
|
|
|
0
|
my $newFH; |
521
|
0
|
0
|
|
|
|
0
|
if ($selectSubgroups==0){ |
522
|
0
|
|
|
|
|
0
|
$newFH = openLocation($newLocation,$newFormat,">",$alSet->{location}); |
523
|
|
|
|
|
|
|
} |
524
|
0
|
|
|
|
|
0
|
my $internalSentPairNum = 1; |
525
|
0
|
|
|
|
|
0
|
my ($sentenceNum,$ruleApplied) = ($alSet->{firstSentPair},0); |
526
|
|
|
|
|
|
|
|
527
|
0
|
|
|
|
|
0
|
my $grSentPairNum=1; |
528
|
|
|
|
|
|
|
|
529
|
0
|
|
|
|
|
0
|
my ($j,$i); |
530
|
0
|
|
|
|
|
0
|
my ($lines,$line); |
531
|
0
|
|
|
|
|
0
|
my $groups = {}; |
532
|
0
|
|
|
|
|
0
|
my $groupsCurrentSentence = {}; |
533
|
0
|
|
|
|
|
0
|
my $groupKeys = []; |
534
|
0
|
|
|
|
|
0
|
my $subGroups={}; |
535
|
0
|
|
|
|
|
0
|
my $subGroupsCurrentSentence = {}; |
536
|
0
|
|
|
|
|
0
|
my $subGroupKeys=[]; |
537
|
0
|
|
|
|
|
0
|
my ($candidate,$count); |
538
|
|
|
|
|
|
|
|
539
|
0
|
0
|
|
|
|
0
|
if (!$selectSubgroups){ #load subgroup hash and array: |
540
|
0
|
|
|
|
|
0
|
open(GROUPS,"<$ENV/groups"); |
541
|
0
|
|
|
|
|
0
|
while (){ |
542
|
0
|
|
|
|
|
0
|
push @$groupKeys,$_; |
543
|
0
|
|
|
|
|
0
|
@$line = split " | ",$_,2; |
544
|
0
|
|
|
|
|
0
|
$groups->{$line->[1]}=$line->[0]; |
545
|
|
|
|
|
|
|
} |
546
|
0
|
0
|
|
|
|
0
|
if ($globals->{onlyGroups}==0){ |
547
|
0
|
|
|
|
|
0
|
open(SUBGROUPS,"<$ENV/subGroups"); |
548
|
0
|
|
|
|
|
0
|
while (){ |
549
|
0
|
|
|
|
|
0
|
push @$subGroupKeys,$_; |
550
|
0
|
|
|
|
|
0
|
@$line = split " | ",$_,2; |
551
|
0
|
|
|
|
|
0
|
$subGroups->{$line->[1]}=$line->[0]; |
552
|
|
|
|
|
|
|
} |
553
|
|
|
|
|
|
|
} |
554
|
|
|
|
|
|
|
} |
555
|
|
|
|
|
|
|
|
556
|
0
|
|
|
|
|
0
|
my %anchors; |
557
|
0
|
|
|
|
|
0
|
my %sourcePerturbed={}; |
558
|
0
|
|
|
|
|
0
|
my %targetPerturbed={}; # Perturbations must be distinct so we keep track of the already detected "Perturbed" $j's |
559
|
0
|
|
|
|
|
0
|
my ($perturbation,$perturbationNoMod); |
560
|
0
|
|
|
|
|
0
|
my ($lastAnchorSource,$lastAnchorTarget,$newAnchorSource,$newAnchorTarget); |
561
|
0
|
|
|
|
|
0
|
my ($ind,$newPerturbationDetected,$anchorsInTarget); |
562
|
0
|
|
|
|
|
0
|
my ($countPertubs,$countGrouping,$countOneToMany,$countElse,$countNoGroup)=(0,0,0,0,0); |
563
|
0
|
|
|
|
|
0
|
while ($alSetChunk = $alSet->loadChunk($FH,$sentenceNum,$alignMode)){ # returns 0 if eof or last sentence pair |
564
|
0
|
|
|
|
|
0
|
for ($k=0;$k<@$alSetChunk;$k++){ |
565
|
|
|
|
|
|
|
# print "\nsentence pair $sentenceNum\n"; |
566
|
0
|
|
|
|
|
0
|
$ruleApplied=0; |
567
|
0
|
|
|
|
|
0
|
$al = $$alSetChunk[$k]; |
568
|
0
|
0
|
|
|
|
0
|
if ($verbose >0){ |
569
|
0
|
|
|
|
|
0
|
print $sentenceNum."\n"; |
570
|
0
|
|
|
|
|
0
|
print $al->sourceSentence."\n"; |
571
|
0
|
|
|
|
|
0
|
print $al->targetSentence."\n"; |
572
|
|
|
|
|
|
|
} |
573
|
0
|
|
|
|
|
0
|
$modAl = $al->clone(); |
574
|
0
|
|
|
|
|
0
|
($lastAnchorSource,$lastAnchorTarget)=(0,0); |
575
|
0
|
|
|
|
|
0
|
%sourcePerturbed=(); |
576
|
0
|
|
|
|
|
0
|
%targetPerturbed=(); |
577
|
0
|
|
|
|
|
0
|
$j = 1; |
578
|
|
|
|
|
|
|
#detect "perturbations" in the anchor diagonal looping only over $j (to have less repeated zones). We can only miss those where $i is aligned only to NULL |
579
|
0
|
|
|
|
|
0
|
while ($j<@{$al->{sourceAl}}){ |
|
0
|
|
|
|
|
0
|
|
580
|
0
|
|
0
|
|
|
0
|
while ( !$al->isAnchor($j,"source") && $j<(@{$al->{sourceAl}})){ |
|
0
|
|
|
|
|
0
|
|
581
|
0
|
|
|
|
|
0
|
$j++; |
582
|
|
|
|
|
|
|
} |
583
|
0
|
0
|
|
|
|
0
|
if ($j<=@{$al->{sourceAl}}){ |
|
0
|
|
|
|
|
0
|
|
584
|
0
|
0
|
|
|
|
0
|
if ($j==@{$al->{sourceAl}}){ |
|
0
|
|
|
|
|
0
|
|
585
|
0
|
|
|
|
|
0
|
($newAnchorSource,$newAnchorTarget) = ($j,scalar(@{$al->{targetAl}})); |
|
0
|
|
|
|
|
0
|
|
586
|
|
|
|
|
|
|
}else{ |
587
|
0
|
|
|
|
|
0
|
($newAnchorSource,$newAnchorTarget) = ($j,$al->{sourceAl}[$j][0]); |
588
|
|
|
|
|
|
|
} |
589
|
0
|
|
|
|
|
0
|
$newPerturbationDetected=0; |
590
|
0
|
0
|
0
|
|
|
0
|
if (($newAnchorSource-$lastAnchorSource)!=1 && !$sourcePerturbed{$lastAnchorSource+1}){ |
|
|
0
|
0
|
|
|
|
|
591
|
0
|
|
|
|
|
0
|
$newPerturbationDetected = 1; |
592
|
|
|
|
|
|
|
} elsif (($newAnchorTarget-$lastAnchorTarget)!=1 && !$targetPerturbed{$lastAnchorTarget+1}){ |
593
|
0
|
|
|
|
|
0
|
$anchorsInTarget=1; |
594
|
0
|
|
|
|
|
0
|
for ($i=$lastAnchorTarget+1;$i<$newAnchorTarget;$i++){ |
595
|
0
|
0
|
|
|
|
0
|
if (!$al->isAnchor($i,"target")){$anchorsInTarget=0} |
|
0
|
|
|
|
|
0
|
|
596
|
|
|
|
|
|
|
} |
597
|
0
|
0
|
|
|
|
0
|
if (!$anchorsInTarget){$newPerturbationDetected=1}; |
|
0
|
|
|
|
|
0
|
|
598
|
|
|
|
|
|
|
} |
599
|
0
|
0
|
|
|
|
0
|
if ( $newPerturbationDetected ){ |
600
|
0
|
|
|
|
|
0
|
$countPertubs++; |
601
|
|
|
|
|
|
|
# print "\n($lastAnchorSource,$lastAnchorTarget,$newAnchorSource,$newAnchorTarget)\n"; |
602
|
|
|
|
|
|
|
|
603
|
0
|
|
|
|
|
0
|
$perturbation = $al->cut($lastAnchorSource,$lastAnchorTarget,$newAnchorSource,$newAnchorTarget); |
604
|
0
|
|
|
|
|
0
|
$perturbationNoMod = $al->cut($lastAnchorSource,$lastAnchorTarget,$newAnchorSource,$newAnchorTarget); |
605
|
|
|
|
|
|
|
|
606
|
|
|
|
|
|
|
# PRINT PERTURBATION TO FILE |
607
|
|
|
|
|
|
|
#if (exists($newFH->{source})){ |
608
|
|
|
|
|
|
|
# $newFH->{source}->print(" ".join(" ",@{$perturbation->{sourceWords}})." \n"); |
609
|
|
|
|
|
|
|
#} |
610
|
|
|
|
|
|
|
#if (exists($newFH->{target})){ |
611
|
|
|
|
|
|
|
# $newFH->{target}->print(" ".join(" ",@{$perturbation->{targetWords}})." \n"); |
612
|
|
|
|
|
|
|
#} |
613
|
|
|
|
|
|
|
#$perturbation->getUnion; |
614
|
|
|
|
|
|
|
#$lines = $perturbation->writeToBlinker; |
615
|
|
|
|
|
|
|
#foreach $line (@$lines){ |
616
|
|
|
|
|
|
|
# $newFH->{sourceToTarget}->print("$grSentPairNum $line\n"); |
617
|
|
|
|
|
|
|
#} |
618
|
|
|
|
|
|
|
#$grSentPairNum++; |
619
|
|
|
|
|
|
|
|
620
|
|
|
|
|
|
|
#END PRINT PERTURBATION TO FILE |
621
|
0
|
0
|
|
|
|
0
|
if ($selectSubgroups){ |
622
|
0
|
|
|
|
|
0
|
$perturbation->selectSubgroups($groupsCurrentSentence,$subGroupsCurrentSentence,$globals); |
623
|
|
|
|
|
|
|
}else{ |
624
|
0
|
0
|
|
|
|
0
|
if (1==0){ |
625
|
|
|
|
|
|
|
# if ($ruleApplied=$perturbation->applyOneToMany_2()){ |
626
|
|
|
|
|
|
|
$countOneToMany++; |
627
|
0
|
|
|
|
|
0
|
}elsif (($ruleApplied=$perturbation->applyGrouping($groupKeys,$subGroupKeys,$globals))>0){ |
628
|
0
|
|
|
|
|
0
|
$countGrouping++; |
629
|
|
|
|
|
|
|
}else{ |
630
|
0
|
|
|
|
|
0
|
my $defaultActionGen = $globals->{defaultActionGeneral}; |
631
|
0
|
|
|
|
|
0
|
$perturbation->$defaultActionGen(); |
632
|
0
|
0
|
|
|
|
0
|
if ($ruleApplied==-1){ |
633
|
0
|
|
|
|
|
0
|
$countNoGroup++; |
634
|
|
|
|
|
|
|
}else{ |
635
|
0
|
|
|
|
|
0
|
$perturbation->processNull(); |
636
|
0
|
|
|
|
|
0
|
$countElse++; |
637
|
|
|
|
|
|
|
} |
638
|
|
|
|
|
|
|
} |
639
|
0
|
|
|
|
|
0
|
$perturbation->paste($modAl); |
640
|
|
|
|
|
|
|
} |
641
|
|
|
|
|
|
|
# print "\ns indices:",join (" ",keys %{$perturbation->{sourceIndices}}),"\n"; |
642
|
|
|
|
|
|
|
# print "t indices:",join (" ",keys %{$perturbation->{targetIndices}}),"\n"; |
643
|
0
|
|
|
|
|
0
|
foreach $ind (keys %{$perturbation->{sourceIndices}}){ |
|
0
|
|
|
|
|
0
|
|
644
|
0
|
0
|
|
|
|
0
|
if ($ind>0){ |
645
|
0
|
|
|
|
|
0
|
$sourcePerturbed{$ind+$perturbation->{zeroSource}}=1; |
646
|
|
|
|
|
|
|
} |
647
|
|
|
|
|
|
|
} |
648
|
0
|
|
|
|
|
0
|
foreach $ind (keys %{$perturbation->{targetIndices}}){ |
|
0
|
|
|
|
|
0
|
|
649
|
0
|
0
|
|
|
|
0
|
if ($ind>0){ |
650
|
0
|
|
|
|
|
0
|
$targetPerturbed{$ind+$perturbation->{zeroTarget}}=1; |
651
|
|
|
|
|
|
|
} |
652
|
|
|
|
|
|
|
} |
653
|
|
|
|
|
|
|
# print "s perturbed:",join (" ",keys %sourcePerturbed),"\n"; |
654
|
|
|
|
|
|
|
# print "t perturbed:",join (" ",keys %targetPerturbed),"\n"; |
655
|
|
|
|
|
|
|
} #if perturbation |
656
|
|
|
|
|
|
|
|
657
|
0
|
|
|
|
|
0
|
$anchors{"$newAnchorSource $newAnchorTarget"}=1; |
658
|
0
|
|
|
|
|
0
|
($lastAnchorSource,$lastAnchorTarget) = ($newAnchorSource,$newAnchorTarget); |
659
|
0
|
|
|
|
|
0
|
$j++; |
660
|
|
|
|
|
|
|
} |
661
|
|
|
|
|
|
|
} #while j... |
662
|
0
|
0
|
0
|
|
|
0
|
if ($newFormat eq "NAACL" && !$selectSubgroups){ |
663
|
0
|
0
|
|
|
|
0
|
if (exists($newFH->{source})){ |
664
|
0
|
|
|
|
|
0
|
$newFH->{source}->print(" ".join(" ",@{$modAl->{sourceWords}})." \n"); |
|
0
|
|
|
|
|
0
|
|
665
|
|
|
|
|
|
|
} |
666
|
0
|
0
|
|
|
|
0
|
if (exists($newFH->{target})){ |
667
|
0
|
|
|
|
|
0
|
$newFH->{target}->print(" ".join(" ",@{$modAl->{targetWords}})." \n"); |
|
0
|
|
|
|
|
0
|
|
668
|
|
|
|
|
|
|
} |
669
|
0
|
|
|
|
|
0
|
$al->intersect(); |
670
|
0
|
|
|
|
|
0
|
$lines = $modAl->writeToBlinker; |
671
|
0
|
|
|
|
|
0
|
foreach $line (@$lines){ |
672
|
0
|
|
|
|
|
0
|
$newFH->{sourceToTarget}->print("$internalSentPairNum $line\n"); |
673
|
|
|
|
|
|
|
} |
674
|
|
|
|
|
|
|
} |
675
|
0
|
0
|
|
|
|
0
|
if (($internalSentPairNum % 1000)==0){print STDERR $internalSentPairNum} |
|
0
|
0
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
676
|
|
|
|
|
|
|
elsif (($internalSentPairNum % 100)==0){print STDERR "."} |
677
|
0
|
|
|
|
|
0
|
$sentenceNum++; |
678
|
0
|
|
|
|
|
0
|
$internalSentPairNum++; |
679
|
0
|
0
|
|
|
|
0
|
if ($verbose > 0){print "Candidates:\n";} |
|
0
|
|
|
|
|
0
|
|
680
|
0
|
0
|
|
|
|
0
|
if ($selectSubgroups){ |
681
|
0
|
|
|
|
|
0
|
foreach $candidate (keys %$groupsCurrentSentence){ |
682
|
0
|
0
|
|
|
|
0
|
if ($verbose > 0){print "$candidate\n";} |
|
0
|
|
|
|
|
0
|
|
683
|
0
|
|
|
|
|
0
|
$groups->{$candidate}=$groups->{$candidate}+1; |
684
|
|
|
|
|
|
|
} |
685
|
0
|
|
|
|
|
0
|
%$groupsCurrentSentence=(); |
686
|
0
|
0
|
|
|
|
0
|
if ($globals->{onlyGroups}==0){ |
687
|
0
|
|
|
|
|
0
|
foreach $candidate (keys %$subGroupsCurrentSentence){ |
688
|
0
|
|
|
|
|
0
|
$subGroups->{$candidate}=$subGroups->{$candidate}+1; |
689
|
|
|
|
|
|
|
} |
690
|
0
|
|
|
|
|
0
|
%$subGroupsCurrentSentence=(); |
691
|
|
|
|
|
|
|
} |
692
|
|
|
|
|
|
|
} |
693
|
|
|
|
|
|
|
}#for k<@alSetChunk |
694
|
|
|
|
|
|
|
} #while alsetchunk |
695
|
0
|
|
|
|
|
0
|
print STDERR "\n"; |
696
|
0
|
0
|
|
|
|
0
|
if ($selectSubgroups==0){ |
697
|
0
|
|
|
|
|
0
|
closeFiles($newFH,$newFormat); |
698
|
|
|
|
|
|
|
} |
699
|
0
|
|
|
|
|
0
|
closeFiles($FH,$alSet->{format}); |
700
|
0
|
0
|
|
|
|
0
|
if ($selectSubgroups){ |
701
|
0
|
0
|
|
|
|
0
|
if ($verbose>0){print "\ngroups:",scalar(keys(%$groups))," - subgroups:",scalar(keys(%$subGroups)),"\n";} |
|
0
|
|
|
|
|
0
|
|
702
|
0
|
0
|
|
|
|
0
|
open(GROUPS, ">$ENV/groups") or die "File opening error:$!";; |
703
|
0
|
|
|
|
|
0
|
while (($candidate,$count)=each(%$groups)){ |
704
|
|
|
|
|
|
|
# print "groups $count | $candidate\n"; |
705
|
0
|
0
|
|
|
|
0
|
if ($count >= $globals->{minPhraseFrequency}){ |
706
|
0
|
|
|
|
|
0
|
print GROUPS "$count | $candidate\n"; |
707
|
|
|
|
|
|
|
} |
708
|
|
|
|
|
|
|
} |
709
|
0
|
0
|
|
|
|
0
|
if ($globals->{onlyGroups}==0){ |
710
|
0
|
0
|
|
|
|
0
|
open(SUBGROUPS, ">$ENV/subGroups") or die "File opening error:$!";; |
711
|
0
|
|
|
|
|
0
|
while (($candidate,$count)=each(%$subGroups)){ |
712
|
|
|
|
|
|
|
# print "SUBGROUPS $count | $candidate\n"; |
713
|
0
|
0
|
|
|
|
0
|
if ($count >= $globals->{minPhraseFrequency}){ |
714
|
0
|
|
|
|
|
0
|
print SUBGROUPS "$count | $candidate\n"; |
715
|
|
|
|
|
|
|
} |
716
|
|
|
|
|
|
|
} |
717
|
|
|
|
|
|
|
} |
718
|
|
|
|
|
|
|
}else{ |
719
|
0
|
|
|
|
|
0
|
print STDERR "perturbations:$countPertubs (oneToMany:$countOneToMany grouped:$countGrouping not grouped:$countNoGroup others:$countElse)\n"; |
720
|
0
|
|
|
|
|
0
|
$alSet->{location}=$newLocation; |
721
|
0
|
|
|
|
|
0
|
$alSet->{format}=$newFormat; |
722
|
0
|
|
|
|
|
0
|
$alSet->{firstSentPair}=1; |
723
|
0
|
|
|
|
|
0
|
$alSet->{lastSentPair}="eof"; |
724
|
|
|
|
|
|
|
} |
725
|
|
|
|
|
|
|
} |
726
|
|
|
|
|
|
|
|
727
|
|
|
|
|
|
|
sub orderAsBilCorpus { |
728
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat,$alignMode,$corpSrc,$corpTrg,$verbose)=@_; |
729
|
0
|
0
|
|
|
|
0
|
if (!defined($newFormat)){$newFormat="TALP"} |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
730
|
|
|
|
|
|
|
else {$newFormat = uc $newFormat} |
731
|
0
|
|
|
|
|
0
|
$newLocation = readLocation($newLocation); |
732
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
733
|
0
|
|
|
|
|
0
|
my $newFH = openLocation($newLocation,$newFormat,">",$alSet->{location}); |
734
|
0
|
|
|
|
|
0
|
my $inputSentPairNum=$alSet->{firstSentPair}; |
735
|
0
|
|
|
|
|
0
|
my $internalSentPairNum = 1; |
736
|
0
|
0
|
|
|
|
0
|
if ($verbose >0){ |
737
|
0
|
|
|
|
|
0
|
select STDOUT; $| = 1; # enable autoflush (desactivate buffering) |
|
0
|
|
|
|
|
0
|
|
738
|
|
|
|
|
|
|
} |
739
|
0
|
0
|
|
|
|
0
|
open(CS,"<$corpSrc") || die "$corpSrc file opening error !"; |
740
|
0
|
0
|
|
|
|
0
|
open(CT,"<$corpTrg") || die "$corpTrg file opening error !"; |
741
|
|
|
|
|
|
|
|
742
|
|
|
|
|
|
|
# LOAD BILINGUAL CORPUS IN HASH |
743
|
0
|
|
|
|
|
0
|
my %newcorp; |
744
|
0
|
|
|
|
|
0
|
my $cntCorp=0; |
745
|
0
|
|
|
|
|
0
|
while (my $s=) { |
746
|
0
|
|
|
|
|
0
|
chomp $s; |
747
|
0
|
|
|
|
|
0
|
my $t=; |
748
|
0
|
|
|
|
|
0
|
chomp $t; |
749
|
0
|
|
|
|
|
0
|
$newcorp{"$s ||| $t"}=1; |
750
|
0
|
|
|
|
|
0
|
$cntCorp++; |
751
|
|
|
|
|
|
|
} |
752
|
0
|
0
|
|
|
|
0
|
if ($verbose > 0){ |
753
|
0
|
|
|
|
|
0
|
print "Number of different sentence pairs in new corpus:".scalar(keys %newcorp)."\n"; |
754
|
|
|
|
|
|
|
} |
755
|
|
|
|
|
|
|
|
756
|
|
|
|
|
|
|
# PARSE AL SET |
757
|
0
|
|
|
|
|
0
|
my $cntToFind=$cntCorp; |
758
|
0
|
|
|
|
|
0
|
my %found; |
759
|
0
|
|
|
|
|
0
|
while (my $alSetChunk = $alSet->loadChunk($FH,$inputSentPairNum,"as-is")){ # returns 0 if eof or last sentence pair |
760
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<@$alSetChunk;$i++){ |
761
|
0
|
|
|
|
|
0
|
my $al = $$alSetChunk[$i]; |
762
|
0
|
|
|
|
|
0
|
my $pair=$al->sourceSentence." ||| ".$al->targetSentence; |
763
|
0
|
0
|
|
|
|
0
|
if ($newcorp{$pair}){ |
764
|
0
|
|
|
|
|
0
|
$found{$pair}=$al; |
765
|
0
|
|
|
|
|
0
|
$cntToFind--; |
766
|
|
|
|
|
|
|
} |
767
|
|
|
|
|
|
|
} #for |
768
|
0
|
0
|
|
|
|
0
|
if ($cntToFind==0){ |
769
|
0
|
0
|
|
|
|
0
|
if ($verbose>0){print STDERR "Leaving loop of giza file at line:$inputSentPairNum\n";} |
|
0
|
|
|
|
|
0
|
|
770
|
0
|
|
|
|
|
0
|
last; |
771
|
|
|
|
|
|
|
} |
772
|
0
|
|
|
|
|
0
|
$inputSentPairNum++; |
773
|
0
|
0
|
|
|
|
0
|
if ($verbose>0){ |
774
|
0
|
0
|
|
|
|
0
|
if ($inputSentPairNum % 100000 ==0){print $inputSentPairNum;} |
|
0
|
|
|
|
|
0
|
|
775
|
0
|
0
|
|
|
|
0
|
if ($inputSentPairNum % 10000 ==0){print ".";} |
|
0
|
|
|
|
|
0
|
|
776
|
|
|
|
|
|
|
} |
777
|
|
|
|
|
|
|
} #while |
778
|
0
|
0
|
|
|
|
0
|
if ($verbose>0){print "\n";} |
|
0
|
|
|
|
|
0
|
|
779
|
|
|
|
|
|
|
# REORDER AL SET |
780
|
0
|
|
|
|
|
0
|
print "reordering Alignment set...\n"; |
781
|
0
|
|
|
|
|
0
|
seek CS,0,0; #go back to beginning of file |
782
|
0
|
|
|
|
|
0
|
seek CT,0,0; |
783
|
0
|
|
|
|
|
0
|
while (my $s=) { |
784
|
0
|
|
|
|
|
0
|
chomp $s; |
785
|
0
|
|
|
|
|
0
|
my $t=; |
786
|
0
|
|
|
|
|
0
|
chomp $t; |
787
|
0
|
0
|
|
|
|
0
|
if (exists($found{"$s ||| $t"})){ |
788
|
0
|
|
|
|
|
0
|
my $al= $found{"$s ||| $t"}; |
789
|
0
|
|
|
|
|
0
|
$al->output($FH,$newFormat,$newFH,$newLocation,$internalSentPairNum); |
790
|
|
|
|
|
|
|
}else{ |
791
|
0
|
|
|
|
|
0
|
die "ERROR: not found sentence pair $internalSentPairNum in Alignment Set\n"; |
792
|
|
|
|
|
|
|
} |
793
|
0
|
|
|
|
|
0
|
$internalSentPairNum++; |
794
|
|
|
|
|
|
|
} |
795
|
0
|
|
|
|
|
0
|
closeFiles($newFH,$newFormat); |
796
|
0
|
|
|
|
|
0
|
closeFiles($FH,$alSet->{format}); |
797
|
0
|
|
|
|
|
0
|
$alSet->updateObject($newFormat,$newLocation,$internalSentPairNum-1); |
798
|
|
|
|
|
|
|
} |
799
|
|
|
|
|
|
|
|
800
|
|
|
|
|
|
|
sub adaptToBilCorpus { |
801
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat,$alignMode,$corpSrc,$corpTrg,$restrictions,$verbose)=@_; |
802
|
0
|
|
|
|
|
0
|
my $pdiff=$restrictions->{allowedPercentWordDiff}; |
803
|
0
|
|
|
|
|
0
|
my $mindiff=$restrictions->{minAllowedNumWordDiff}; |
804
|
0
|
|
|
|
|
0
|
my $maxdiff=$restrictions->{maxAllowedNumWordDiff}; |
805
|
0
|
|
|
|
|
0
|
my $nfirst=$restrictions->{numWordsConsideredFirst}; |
806
|
0
|
|
|
|
|
0
|
my $dumper = new Dumpvalue; |
807
|
|
|
|
|
|
|
|
808
|
0
|
0
|
|
|
|
0
|
if (!defined($newFormat)){$newFormat="TALP"} |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
809
|
|
|
|
|
|
|
else {$newFormat = uc $newFormat} |
810
|
0
|
|
|
|
|
0
|
$newLocation = readLocation($newLocation); |
811
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
812
|
0
|
|
|
|
|
0
|
my $newFH = openLocation($newLocation,$newFormat,">",$alSet->{location}); |
813
|
0
|
|
|
|
|
0
|
my $inputSentPairNum=$alSet->{firstSentPair}; |
814
|
0
|
|
|
|
|
0
|
my $internalSentPairNum = 1; |
815
|
0
|
|
|
|
|
0
|
select STDOUT; $| = 1; # enable autoflush (desactivate buffering) |
|
0
|
|
|
|
|
0
|
|
816
|
0
|
0
|
|
|
|
0
|
open(CS,"<$corpSrc") || die "$corpSrc file opening error !"; |
817
|
0
|
0
|
|
|
|
0
|
open(CT,"<$corpTrg") || die "$corpTrg file opening error !"; |
818
|
|
|
|
|
|
|
|
819
|
|
|
|
|
|
|
# DETECT ALSET SENTENCES THAT ARE IN THE CORPUS |
820
|
0
|
|
|
|
|
0
|
my %newcorp; |
821
|
0
|
|
|
|
|
0
|
while (my $s=) { |
822
|
0
|
|
|
|
|
0
|
chomp $s; |
823
|
0
|
|
|
|
|
0
|
my $t=; |
824
|
0
|
|
|
|
|
0
|
chomp $t; |
825
|
0
|
|
|
|
|
0
|
$newcorp{"$s ||| $t"}=1; |
826
|
|
|
|
|
|
|
} |
827
|
0
|
0
|
|
|
|
0
|
if ($verbose > 0){ |
828
|
0
|
|
|
|
|
0
|
print "Number of different sentence pairs in new corpus:".scalar(keys %newcorp)."\n"; |
829
|
|
|
|
|
|
|
} |
830
|
0
|
|
|
|
|
0
|
my $count=0; |
831
|
0
|
|
|
|
|
0
|
while (my $alSetChunk = $alSet->loadChunk($FH,$inputSentPairNum,$alignMode)){ |
832
|
|
|
|
|
|
|
# print $inputSentPairNum."\n"; |
833
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<@$alSetChunk;$i++){ |
834
|
0
|
|
|
|
|
0
|
my $al = $$alSetChunk[$i]; |
835
|
0
|
|
|
|
|
0
|
my $s = $al->sourceSentence; |
836
|
0
|
|
|
|
|
0
|
my $t = $al->targetSentence; |
837
|
0
|
|
|
|
|
0
|
my @ws=split / /,$s; |
838
|
0
|
|
|
|
|
0
|
my @wt=split / /,$t; |
839
|
0
|
|
|
|
|
0
|
my ($nums,$numt)=(scalar(@ws),scalar(@wt)); |
840
|
0
|
0
|
|
|
|
0
|
if (!$newcorp{"$s ||| $t"}){ |
841
|
|
|
|
|
|
|
# DETECT CLOSEST SENTENCES IN NEW CORPUS AND MODIFY ALIGNMENT SET |
842
|
|
|
|
|
|
|
# calculate values for length test (to, later, skip lcs calculation) |
843
|
0
|
|
|
|
|
0
|
my $sAllowedDiff = Lingua::AlSetLib::max(Lingua::AlSetLib::min($nums*$pdiff/100,$maxdiff),$mindiff); |
844
|
0
|
|
|
|
|
0
|
my $tAllowedDiff = Lingua::AlSetLib::max(Lingua::AlSetLib::min($numt*$pdiff/100,$maxdiff),$mindiff); |
845
|
0
|
|
|
|
|
0
|
my $sMin=$nums-$sAllowedDiff; |
846
|
0
|
|
|
|
|
0
|
my $sMax=$nums+$sAllowedDiff; |
847
|
0
|
|
|
|
|
0
|
my $tMin=$numt-$tAllowedDiff; |
848
|
0
|
|
|
|
|
0
|
my $tMax=$numt+$tAllowedDiff; |
849
|
0
|
0
|
|
|
|
0
|
if ($verbose > 0){ |
850
|
0
|
|
|
|
|
0
|
print $s."\n"; |
851
|
0
|
|
|
|
|
0
|
print $t."\n"; |
852
|
0
|
|
|
|
|
0
|
print "ns:$nums nt:$numt allowed length diff: s:$sAllowedDiff t:$tAllowedDiff;\n\n"; |
853
|
|
|
|
|
|
|
} |
854
|
0
|
|
|
|
|
0
|
my $bestLcs = 0; |
855
|
0
|
|
|
|
|
0
|
my @bestSrc; |
856
|
|
|
|
|
|
|
my @bestTrg; |
857
|
0
|
|
|
|
|
0
|
my ($cntGoodLength,$cntPassedFirstLCS,$cnt2)=(0,0,0); |
858
|
|
|
|
|
|
|
# parse new corpus |
859
|
0
|
|
|
|
|
0
|
foreach my $pair (keys %newcorp){ |
860
|
0
|
|
|
|
|
0
|
my ($cs,$ct)=split / \|\|\| /,$pair; |
861
|
0
|
|
|
|
|
0
|
my @wcs = split / /,$cs; |
862
|
0
|
|
|
|
|
0
|
my @wct = split / /,$ct; |
863
|
0
|
|
|
|
|
0
|
my ($numcs,$numct)=(scalar(@wcs),scalar(@wct)); |
864
|
|
|
|
|
|
|
|
865
|
0
|
0
|
|
|
|
0
|
if ($verbose >2){print "ncs:$numcs nct:$numct\n";} |
|
0
|
|
|
|
|
0
|
|
866
|
0
|
0
|
|
|
|
0
|
if ($verbose >2){print "CORPUS:\n$cs\n$ct\n";} |
|
0
|
|
|
|
|
0
|
|
867
|
|
|
|
|
|
|
#length test: |
868
|
0
|
0
|
0
|
|
|
0
|
if ($numcs<$sMin || $numcs>$sMax || $numct<$tMin || $numct>$tMax){ |
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
869
|
|
|
|
|
|
|
# skip lcs ratio calculation |
870
|
|
|
|
|
|
|
}else{ |
871
|
0
|
|
|
|
|
0
|
$cntGoodLength++; |
872
|
|
|
|
|
|
|
#calculate LCS, but first looking at the nfirst first words |
873
|
0
|
|
|
|
|
0
|
my @fws; |
874
|
|
|
|
|
|
|
my @fwt; |
875
|
0
|
|
|
|
|
0
|
my @fwcs; |
876
|
0
|
|
|
|
|
0
|
my @fwct; |
877
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<$nfirst;$i++){ |
878
|
0
|
|
|
|
|
0
|
push @fws,$ws[$i]; |
879
|
0
|
|
|
|
|
0
|
push @fwt,$wt[$i]; |
880
|
0
|
|
|
|
|
0
|
push @fwcs,$wcs[$i]; |
881
|
0
|
|
|
|
|
0
|
push @fwct,$wct[$i]; |
882
|
|
|
|
|
|
|
} |
883
|
0
|
|
|
|
|
0
|
my $sLcs = Lingua::AlSetLib::LCS_ratio(\@fws,\@fwcs); |
884
|
0
|
|
|
|
|
0
|
my $tLcs = Lingua::AlSetLib::LCS_ratio(\@fwt,\@fwct); |
885
|
|
|
|
|
|
|
# first words LCS test: |
886
|
0
|
0
|
0
|
|
|
0
|
if ($sLcs == 0 || $tLcs == 0){ |
887
|
|
|
|
|
|
|
# skip lcs ratio calculation |
888
|
|
|
|
|
|
|
}else{ |
889
|
0
|
|
|
|
|
0
|
$cntPassedFirstLCS++; |
890
|
0
|
|
|
|
|
0
|
my $sLcs = Lingua::AlSetLib::LCS_ratio(\@ws,\@wcs); |
891
|
0
|
|
|
|
|
0
|
my $tLcs = Lingua::AlSetLib::LCS_ratio(\@wt,\@wct); |
892
|
0
|
0
|
|
|
|
0
|
if ($verbose >1){ |
893
|
0
|
|
|
|
|
0
|
print "ncs:$numcs nct:$numct\n"; |
894
|
0
|
|
|
|
|
0
|
print "CORPUS:\n$cs\n$ct\n"; |
895
|
0
|
|
|
|
|
0
|
print "chars src lcsr: $sLcs\t trg lcsr:$tLcs\n"; |
896
|
|
|
|
|
|
|
} |
897
|
0
|
|
|
|
|
0
|
my $lcs = $sLcs+$tLcs; |
898
|
0
|
0
|
|
|
|
0
|
if ($lcs > $bestLcs){ |
|
|
0
|
|
|
|
|
|
899
|
0
|
|
|
|
|
0
|
$bestLcs = $lcs; |
900
|
0
|
|
|
|
|
0
|
@bestSrc = ($cs); |
901
|
0
|
|
|
|
|
0
|
@bestTrg = ($ct); |
902
|
|
|
|
|
|
|
}elsif($lcs == $bestLcs){ |
903
|
0
|
|
|
|
|
0
|
push @bestSrc,$cs; |
904
|
0
|
|
|
|
|
0
|
push @bestTrg,$ct; |
905
|
|
|
|
|
|
|
} |
906
|
|
|
|
|
|
|
} |
907
|
|
|
|
|
|
|
} # if length test |
908
|
0
|
0
|
|
|
|
0
|
if ($verbose >1){ |
909
|
0
|
0
|
0
|
|
|
0
|
if ($cnt2>0 && ($cnt2 % 100000)==0){print "$cnt2";} |
|
0
|
0
|
0
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
910
|
|
|
|
|
|
|
elsif ($cnt2>0 && ($cnt2 % 1000)==0){print ".";} |
911
|
|
|
|
|
|
|
} |
912
|
0
|
|
|
|
|
0
|
$cnt2++; |
913
|
|
|
|
|
|
|
} # for each sent pair in corpus |
914
|
0
|
|
|
|
|
0
|
my $uniqBestSrc; |
915
|
|
|
|
|
|
|
my $uniqBestTrg; |
916
|
0
|
|
|
|
|
0
|
my $cntCharLevel=scalar(@bestSrc); |
917
|
0
|
0
|
|
|
|
0
|
if ($cntCharLevel>1){ |
918
|
0
|
|
|
|
|
0
|
my $bestLcs = 0; |
919
|
|
|
|
|
|
|
# calculate LCS at character level |
920
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<$cntCharLevel;$i++){ |
921
|
0
|
|
|
|
|
0
|
my $cs=$bestSrc[$i]; |
922
|
0
|
|
|
|
|
0
|
my $ct=$bestTrg[$i]; |
923
|
0
|
|
|
|
|
0
|
$cntCharLevel++; |
924
|
0
|
|
|
|
|
0
|
my @chars = split //,$s; |
925
|
0
|
|
|
|
|
0
|
my @charcs = split //,$cs; |
926
|
0
|
|
|
|
|
0
|
my $sLcs = Lingua::AlSetLib::LCS_ratio(\@chars,\@charcs); |
927
|
0
|
|
|
|
|
0
|
my @chart = split //,$t; |
928
|
0
|
|
|
|
|
0
|
my @charct = split //,$ct; |
929
|
0
|
|
|
|
|
0
|
my $tLcs = Lingua::AlSetLib::LCS_ratio(\@chart,\@charct); |
930
|
0
|
|
|
|
|
0
|
my $lcs = $sLcs+$tLcs; |
931
|
0
|
0
|
|
|
|
0
|
if ($lcs > $bestLcs){ |
932
|
0
|
|
|
|
|
0
|
$bestLcs = $lcs; |
933
|
0
|
|
|
|
|
0
|
$uniqBestSrc = $cs; |
934
|
0
|
|
|
|
|
0
|
$uniqBestTrg = $ct; |
935
|
|
|
|
|
|
|
} |
936
|
|
|
|
|
|
|
} |
937
|
|
|
|
|
|
|
}else{ |
938
|
0
|
|
|
|
|
0
|
$uniqBestSrc = $bestSrc[0]; |
939
|
0
|
|
|
|
|
0
|
$uniqBestTrg = $bestTrg[0]; |
940
|
|
|
|
|
|
|
} |
941
|
0
|
0
|
|
|
|
0
|
if ($cntGoodLength == 0){ |
942
|
0
|
|
|
|
|
0
|
print "WARNING: sentence pair $inputSentPairNum not found in corpus\n"; |
943
|
|
|
|
|
|
|
}else{ |
944
|
0
|
0
|
|
|
|
0
|
if ($verbose >0){ |
945
|
0
|
|
|
|
|
0
|
my $numCorp=scalar(keys %newcorp); |
946
|
0
|
|
|
|
|
0
|
print "Passed length test: $cntGoodLength / $numCorp\n"; |
947
|
0
|
|
|
|
|
0
|
print "Passed first words LCS test: $cntPassedFirstLCS\n"; |
948
|
0
|
0
|
|
|
|
0
|
print "LCS calculated at character level: "; if ($cntCharLevel==1){print "0"}else{print $cntCharLevel}; print "\n"; |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
949
|
0
|
|
|
|
|
0
|
print "\nbest lcsr: $bestLcs, best pair:\n$uniqBestSrc\n$uniqBestTrg\n"; |
950
|
|
|
|
|
|
|
} |
951
|
|
|
|
|
|
|
# detect edits to pass from alset sent pair to corpus sent pair |
952
|
0
|
|
|
|
|
0
|
my @bs=split / /,$uniqBestSrc; |
953
|
0
|
|
|
|
|
0
|
my @diffs = Lingua::AlSetLib::diff( \@ws, \@bs ); |
954
|
0
|
0
|
|
|
|
0
|
if ($verbose>2){print $dumper->dumpValue(\@diffs);} |
|
0
|
|
|
|
|
0
|
|
955
|
|
|
|
|
|
|
|
956
|
|
|
|
|
|
|
# parse output of diff function |
957
|
0
|
|
|
|
|
0
|
my @updatedPosi; #array: orig posis -> updated posis |
958
|
|
|
|
|
|
|
my %reversePosi; #hash: updated posis -> orig posis |
959
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<=$nums;$i++){ |
960
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]=$i; |
961
|
0
|
|
|
|
|
0
|
$reversePosi{$i}=$i; |
962
|
|
|
|
|
|
|
} |
963
|
|
|
|
|
|
|
|
964
|
0
|
|
|
|
|
0
|
foreach my $hunk (@diffs){ |
965
|
0
|
|
|
|
|
0
|
my @delPosi; |
966
|
|
|
|
|
|
|
my @del; |
967
|
0
|
|
|
|
|
0
|
my @addPosi; |
968
|
0
|
|
|
|
|
0
|
my @add; |
969
|
0
|
|
|
|
|
0
|
foreach my $change (@$hunk) { |
970
|
0
|
0
|
|
|
|
0
|
if ($change->[0] eq '-'){ |
971
|
0
|
|
|
|
|
0
|
push @delPosi,$change->[1]+1; |
972
|
0
|
|
|
|
|
0
|
push @del,$change->[2]; |
973
|
|
|
|
|
|
|
}else{ |
974
|
0
|
|
|
|
|
0
|
push @addPosi,$change->[1]+1; |
975
|
0
|
|
|
|
|
0
|
push @add,$change->[2]; |
976
|
|
|
|
|
|
|
} |
977
|
|
|
|
|
|
|
} |
978
|
|
|
|
|
|
|
# del posis are relative to first array (@ws) => update posis |
979
|
|
|
|
|
|
|
# add posis are relative to second array (@bs) => don't update posis |
980
|
0
|
|
|
|
|
0
|
my $numDel=scalar(@delPosi); |
981
|
0
|
|
|
|
|
0
|
my $numAdd=scalar(@addPosi); |
982
|
0
|
0
|
|
|
|
0
|
if ($numDel==0){ #insertion |
983
|
0
|
|
|
|
|
0
|
$al->splice("source",$addPosi[0],0,\@add); |
984
|
0
|
|
|
|
|
0
|
print "insert '",join(" ",@add),"' at position { ",$addPosi[0]," }\n"; |
985
|
|
|
|
|
|
|
#update updatedPosi array |
986
|
0
|
|
|
|
|
0
|
for (my $i=$reversePosi{"$addPosi[0]"};$i<=$nums;$i++){ |
987
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]+=$numAdd; |
988
|
0
|
|
|
|
|
0
|
$reversePosi{"$updatedPosi[$i]"}=$i; |
989
|
|
|
|
|
|
|
} |
990
|
|
|
|
|
|
|
}else{ # substitution or deletion |
991
|
0
|
|
|
|
|
0
|
$al->splice("source",$updatedPosi[$delPosi[0]],$numDel,\@add); |
992
|
0
|
|
|
|
|
0
|
print "substitute '",join(" ",@del),"' at positions { ",join(" ",@delPosi)," } by '",join(" ",@add),"'\n"; |
993
|
|
|
|
|
|
|
#update updatedPosi array |
994
|
0
|
|
|
|
|
0
|
for (my $i=$delPosi[0]+$numDel;$i<=$nums;$i++){ |
995
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]+=$numAdd-$numDel; |
996
|
0
|
|
|
|
|
0
|
$reversePosi{"$updatedPosi[$i]"}=$i; |
997
|
|
|
|
|
|
|
} |
998
|
|
|
|
|
|
|
} |
999
|
|
|
|
|
|
|
} |
1000
|
|
|
|
|
|
|
#target |
1001
|
0
|
|
|
|
|
0
|
my @bt=split / /,$uniqBestTrg; |
1002
|
0
|
|
|
|
|
0
|
my @diffs = Lingua::AlSetLib::diff( \@wt, \@bt ); |
1003
|
0
|
0
|
|
|
|
0
|
if ($verbose>2){print $dumper->dumpValue(\@diffs);} |
|
0
|
|
|
|
|
0
|
|
1004
|
|
|
|
|
|
|
# parse output of diff function |
1005
|
0
|
|
|
|
|
0
|
@updatedPosi=(); |
1006
|
0
|
|
|
|
|
0
|
%reversePosi=(); |
1007
|
0
|
|
|
|
|
0
|
for (my $i=0;$i<=$numt;$i++){ |
1008
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]=$i; |
1009
|
0
|
|
|
|
|
0
|
$reversePosi{$i}=$i; |
1010
|
|
|
|
|
|
|
} |
1011
|
0
|
|
|
|
|
0
|
foreach my $hunk (@diffs){ |
1012
|
0
|
|
|
|
|
0
|
my @delPosi; |
1013
|
|
|
|
|
|
|
my @del; |
1014
|
0
|
|
|
|
|
0
|
my @add; |
1015
|
0
|
|
|
|
|
0
|
my @addPosi; |
1016
|
0
|
|
|
|
|
0
|
foreach my $change (@$hunk) { |
1017
|
0
|
0
|
|
|
|
0
|
if ($change->[0] eq '-'){ |
1018
|
0
|
|
|
|
|
0
|
push @delPosi,$change->[1]+1; |
1019
|
0
|
|
|
|
|
0
|
push @del,$change->[2]; |
1020
|
|
|
|
|
|
|
}else{ |
1021
|
0
|
|
|
|
|
0
|
push @addPosi,$change->[1]+1; |
1022
|
0
|
|
|
|
|
0
|
push @add,$change->[2]; |
1023
|
|
|
|
|
|
|
} |
1024
|
|
|
|
|
|
|
} |
1025
|
|
|
|
|
|
|
#update updatedPosi array |
1026
|
0
|
|
|
|
|
0
|
my $numDel=scalar(@delPosi); |
1027
|
0
|
|
|
|
|
0
|
my $numAdd=scalar(@addPosi); |
1028
|
0
|
0
|
|
|
|
0
|
if ($numDel==0){ #insertion |
1029
|
0
|
|
|
|
|
0
|
$al->splice("target",$addPosi[0],0,\@add); |
1030
|
0
|
|
|
|
|
0
|
print "insert '",join(" ",@add),"' at position { ",$addPosi[0]," }\n"; |
1031
|
|
|
|
|
|
|
#update updatedPosi array |
1032
|
0
|
|
|
|
|
0
|
for (my $i=$reversePosi{"$addPosi[0]"};$i<=$numt;$i++){ |
1033
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]+=$numAdd; |
1034
|
0
|
|
|
|
|
0
|
$reversePosi{"$updatedPosi[$i]"}=$i; |
1035
|
|
|
|
|
|
|
} |
1036
|
|
|
|
|
|
|
}else{ # substitution or deletion |
1037
|
0
|
|
|
|
|
0
|
$al->splice("target",$updatedPosi[$delPosi[0]],$numDel,\@add); |
1038
|
0
|
|
|
|
|
0
|
print "substitute '",join(" ",@del),"' at positions { ",join(" ",@delPosi)," } by '",join(" ",@add),"'\n"; |
1039
|
|
|
|
|
|
|
#update updatedPosi array |
1040
|
0
|
|
|
|
|
0
|
for (my $i=$delPosi[0]+$numDel;$i<=$numt;$i++){ |
1041
|
0
|
|
|
|
|
0
|
$updatedPosi[$i]+=$numAdd-$numDel; |
1042
|
0
|
|
|
|
|
0
|
$reversePosi{"$updatedPosi[$i]"}=$i; |
1043
|
|
|
|
|
|
|
} |
1044
|
|
|
|
|
|
|
} |
1045
|
|
|
|
|
|
|
} |
1046
|
0
|
0
|
|
|
|
0
|
if ($verbose>0){print "--------------------------------------------------------------\n";} |
|
0
|
|
|
|
|
0
|
|
1047
|
|
|
|
|
|
|
} |
1048
|
|
|
|
|
|
|
}else{ |
1049
|
|
|
|
|
|
|
# sentence pair is in corpus: don't modify anything |
1050
|
|
|
|
|
|
|
} |
1051
|
0
|
|
|
|
|
0
|
$al->output($FH,$newFormat,$newFH,$newLocation,$internalSentPairNum); |
1052
|
0
|
|
|
|
|
0
|
$count++; |
1053
|
|
|
|
|
|
|
} |
1054
|
0
|
|
|
|
|
0
|
$inputSentPairNum++; |
1055
|
0
|
|
|
|
|
0
|
$internalSentPairNum++; |
1056
|
|
|
|
|
|
|
} #while |
1057
|
0
|
|
|
|
|
0
|
print "$count sentence pairs parsed in alignment set\n"; |
1058
|
0
|
|
|
|
|
0
|
closeFiles($newFH,$newFormat); |
1059
|
0
|
|
|
|
|
0
|
closeFiles($FH,$alSet->{format}); |
1060
|
0
|
|
|
|
|
0
|
$alSet->updateObject($newFormat,$newLocation,$internalSentPairNum-1); |
1061
|
|
|
|
|
|
|
} |
1062
|
|
|
|
|
|
|
|
1063
|
|
|
|
|
|
|
###################################################################### |
1064
|
|
|
|
|
|
|
### PRIVATE SUBS |
1065
|
|
|
|
|
|
|
###################################################################### |
1066
|
|
|
|
|
|
|
|
1067
|
|
|
|
|
|
|
sub readLocation{ |
1068
|
2
|
|
|
2
|
0
|
4
|
my $location = shift; |
1069
|
|
|
|
|
|
|
|
1070
|
2
|
100
|
|
|
|
6
|
if (!ref($location)){ #if it is a path, put it in a location hash |
1071
|
1
|
|
|
|
|
4
|
$location = {"sourceToTarget"=>$location} |
1072
|
|
|
|
|
|
|
} |
1073
|
2
|
|
|
|
|
5
|
return $location; |
1074
|
|
|
|
|
|
|
} |
1075
|
|
|
|
|
|
|
|
1076
|
|
|
|
|
|
|
sub setRange { |
1077
|
2
|
|
|
2
|
0
|
4
|
my ($alSet,$range) = @_; |
1078
|
|
|
|
|
|
|
|
1079
|
2
|
|
|
|
|
7
|
my @limits = split /-/, $range; |
1080
|
2
|
|
|
|
|
3
|
my $numLimits = scalar(@limits); |
1081
|
2
|
50
|
33
|
|
|
16
|
if ($numLimits == 0 || $numLimits >2){ |
|
|
100
|
|
|
|
|
|
1082
|
0
|
|
|
|
|
0
|
die "Invalid Range:$range\n"; |
1083
|
|
|
|
|
|
|
}elsif ($numLimits == 1){ |
1084
|
1
|
|
|
|
|
3
|
$limits[1]=""; |
1085
|
|
|
|
|
|
|
} |
1086
|
2
|
|
|
|
|
8
|
$limits[0] =~ s/^\s+|\s+$//g; |
1087
|
2
|
|
|
|
|
6
|
$limits[1] =~ s/^\s+|\s+$//g; |
1088
|
2
|
50
|
33
|
|
|
20
|
if ($limits[0] !~ /\d+/ || $limits[0] == 0){ |
1089
|
0
|
|
|
|
|
0
|
$alSet->{firstSentPair}="1"; |
1090
|
|
|
|
|
|
|
}else{ |
1091
|
2
|
|
|
|
|
6
|
$alSet->{firstSentPair}=$limits[0]; |
1092
|
|
|
|
|
|
|
} |
1093
|
2
|
100
|
66
|
|
|
11
|
if ($limits[1] !~ /\d+/ || $limits[1] == 0){ |
1094
|
1
|
|
|
|
|
3
|
$alSet->{lastSentPair}="eof"; |
1095
|
|
|
|
|
|
|
}else{ |
1096
|
1
|
|
|
|
|
5
|
$alSet->{lastSentPair}=$limits[1]; |
1097
|
|
|
|
|
|
|
} |
1098
|
|
|
|
|
|
|
} |
1099
|
|
|
|
|
|
|
|
1100
|
|
|
|
|
|
|
#for future ease we save detailed infos contained in the source sample path |
1101
|
|
|
|
|
|
|
#input: sourceToTarget dir (not optional), targetToSource dir (if exists), source path (optional) and target path (if necessary) |
1102
|
|
|
|
|
|
|
#output: target (if not specified in input), sampleNum (sample number) |
1103
|
|
|
|
|
|
|
sub completeBlinkerLocation{ |
1104
|
1
|
|
|
1
|
0
|
2
|
my $refToLocation = shift; |
1105
|
1
|
|
|
|
|
4
|
my ($sourceLang,$targetLang); |
1106
|
|
|
|
|
|
|
|
1107
|
1
|
50
|
|
|
|
5
|
if ($refToLocation->{source}){ |
1108
|
0
|
|
|
|
|
0
|
my ($sourceDir,$sourceFileName)=split /\/([^\/]+)$/,$refToLocation->{source}; |
1109
|
0
|
0
|
|
|
|
0
|
if ($sourceFileName =~ /^(EN|FR)\.sample.\d+$/){ |
1110
|
|
|
|
|
|
|
#extract the sample number and target file: |
1111
|
0
|
|
|
|
|
0
|
my ($sourceLang,$nothing,$sampleNum) = split /\./,$sourceFileName; |
1112
|
0
|
|
|
|
|
0
|
$refToLocation->{sampleNum} = $sampleNum; |
1113
|
|
|
|
|
|
|
} |
1114
|
|
|
|
|
|
|
} |
1115
|
1
|
50
|
|
|
|
4
|
if (!$refToLocation->{sampleNum}){ |
1116
|
1
|
|
|
|
|
4
|
$refToLocation->{sampleNum} = 1; |
1117
|
|
|
|
|
|
|
} |
1118
|
|
|
|
|
|
|
} |
1119
|
|
|
|
|
|
|
|
1120
|
|
|
|
|
|
|
# open (for read or write) the files contained in a "location" hash (ex. at the {location} key of the alignment set hash) |
1121
|
|
|
|
|
|
|
# if opens for write needs old location hash to check you won't delete the old format files |
1122
|
|
|
|
|
|
|
# returns a ref to a hash containing the filehandle variables (hash with same keys as "location" except for Blinker format) |
1123
|
|
|
|
|
|
|
sub openLocation { |
1124
|
2
|
|
|
2
|
0
|
4
|
my ($location,$format,$openMode,$oldLocation) = @_; #oldLocation: optional parameter |
1125
|
2
|
|
|
|
|
4
|
my %FH; |
1126
|
|
|
|
|
|
|
|
1127
|
2
|
50
|
|
|
|
6
|
if ($openMode eq ">"){ |
1128
|
0
|
0
|
|
|
|
0
|
if ($format eq "BLINKER"){ |
1129
|
0
|
|
|
|
|
0
|
completeBlinkerLocation($location); |
1130
|
|
|
|
|
|
|
} |
1131
|
|
|
|
|
|
|
# check that your new files are different to prevent from deleting the old ones |
1132
|
0
|
|
|
|
|
0
|
my %oldFiles = reverse %$oldLocation; |
1133
|
0
|
|
|
|
|
0
|
my ($key,$newFile); |
1134
|
0
|
|
|
|
|
0
|
while (($key, $newFile)=each %$location){ |
1135
|
0
|
0
|
0
|
|
|
0
|
if ($oldFiles{$newFile} && $key ne "sampleNum"){ |
1136
|
0
|
|
|
|
|
0
|
die "Convert function: you are opening for write one of the old format file: $newFile\n"; |
1137
|
|
|
|
|
|
|
} |
1138
|
|
|
|
|
|
|
} |
1139
|
|
|
|
|
|
|
#end of check |
1140
|
|
|
|
|
|
|
|
1141
|
|
|
|
|
|
|
# create directory structure where to create the file/directory if it doesn't exist, create it |
1142
|
0
|
|
|
|
|
0
|
my $type; |
1143
|
0
|
0
|
|
|
|
0
|
if ($format eq "BLINKER"){$type = "dir"} |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
1144
|
|
|
|
|
|
|
else {$type = "file"} |
1145
|
0
|
|
|
|
|
0
|
createDirStructure($location->{sourceToTarget},$type); |
1146
|
0
|
0
|
|
|
|
0
|
if ($location->{targetToSource}){ |
1147
|
0
|
|
|
|
|
0
|
createDirStructure($location->{targetToSource},$type); |
1148
|
|
|
|
|
|
|
} |
1149
|
0
|
0
|
|
|
|
0
|
if ($location->{source}){ |
1150
|
0
|
|
|
|
|
0
|
createDirStructure($location->{source},"file"); |
1151
|
|
|
|
|
|
|
} |
1152
|
0
|
0
|
|
|
|
0
|
if ($location->{target}){ |
1153
|
0
|
|
|
|
|
0
|
createDirStructure($location->{target},"file"); |
1154
|
|
|
|
|
|
|
} |
1155
|
|
|
|
|
|
|
#end create directory structure |
1156
|
|
|
|
|
|
|
} |
1157
|
|
|
|
|
|
|
|
1158
|
2
|
50
|
66
|
|
|
24
|
if ($format eq "GIZA"){ |
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1159
|
0
|
0
|
|
|
|
0
|
$FH{sourceToTarget} = IO::File->new($openMode.$location->{sourceToTarget}) or die "GIZA file (".$location->{sourceToTarget}.") opening error:$!"; |
1160
|
0
|
0
|
|
|
|
0
|
if ($location->{targetToSource}){ |
1161
|
0
|
0
|
|
|
|
0
|
$FH{targetToSource} = IO::File->new($openMode.$location->{targetToSource}) or die "GIZA file (".$location->{targetToSource}.") opening error:$!"; |
1162
|
|
|
|
|
|
|
} |
1163
|
|
|
|
|
|
|
} elsif ($format eq "NAACL" || $format eq "TALP"){ |
1164
|
1
|
50
|
|
|
|
4
|
if ($location->{source}){ |
1165
|
1
|
50
|
|
|
|
13
|
$FH{source} = IO::File->new($openMode.$location->{source}) or die "Source file (".$location->{source}.") opening error:$!"; |
1166
|
|
|
|
|
|
|
} |
1167
|
1
|
50
|
|
|
|
109
|
if ($location->{target}){ |
1168
|
1
|
50
|
|
|
|
6
|
$FH{target} = IO::File->new($openMode.$location->{target}) or die "Target file (".$location->{target}.") opening error:$!"; |
1169
|
|
|
|
|
|
|
} |
1170
|
1
|
50
|
|
|
|
52
|
$FH{sourceToTarget} = IO::File->new($openMode.$location->{sourceToTarget}) or die "Alignment file (".$location->{sourceToTarget}.") opening error:$!"; |
1171
|
1
|
50
|
|
|
|
56
|
if ($location->{targetToSource}){ |
1172
|
0
|
0
|
|
|
|
0
|
$FH{targetToSource} = IO::File->new($openMode.$location->{targetToSource}) or die "Alignment file (".$location->{targetToSource}.") opening error:$!"; |
1173
|
|
|
|
|
|
|
} |
1174
|
|
|
|
|
|
|
} elsif ($format eq "BLINKER"){ |
1175
|
1
|
50
|
|
|
|
3
|
if ($location->{source}){ |
1176
|
0
|
0
|
|
|
|
0
|
$FH{source} = IO::File->new($openMode.$location->{source}) or die "BLINKER source file (".$location->{source}.") opening error:$!"; |
1177
|
|
|
|
|
|
|
} |
1178
|
1
|
50
|
|
|
|
6
|
if ($location->{target}){ |
1179
|
0
|
0
|
|
|
|
0
|
$FH{target} = IO::File->new($openMode.$location->{target}) or die "BLINKER source file (".$location->{target}.") opening error:$!"; |
1180
|
|
|
|
|
|
|
} |
1181
|
|
|
|
|
|
|
} |
1182
|
2
|
|
|
|
|
10
|
return (\%FH); |
1183
|
|
|
|
|
|
|
} |
1184
|
|
|
|
|
|
|
|
1185
|
|
|
|
|
|
|
|
1186
|
|
|
|
|
|
|
# if you want to create a file of path "directory_structure/file", makes "directory_structure" if necessary. |
1187
|
|
|
|
|
|
|
# if you want to create a directory of path "directory_structure", makes it if it doesn't exist |
1188
|
|
|
|
|
|
|
# type is "dir" (if you want to create a directory) or "file" (a file) |
1189
|
|
|
|
|
|
|
sub createDirStructure { |
1190
|
0
|
|
|
0
|
0
|
0
|
my ($path,$type)=@_; |
1191
|
|
|
|
|
|
|
|
1192
|
0
|
0
|
|
|
|
0
|
if ($type eq "dir"){ |
|
|
0
|
|
|
|
|
|
1193
|
0
|
0
|
0
|
|
|
0
|
unless(-e $path && -d _){ |
1194
|
0
|
|
|
|
|
0
|
system('mkdir -p '.$path); |
1195
|
|
|
|
|
|
|
} |
1196
|
|
|
|
|
|
|
}elsif ($type eq "file"){ |
1197
|
0
|
|
|
|
|
0
|
$path =~ s/\/$//; |
1198
|
0
|
0
|
|
|
|
0
|
if ($path =~ /\//){ |
1199
|
0
|
|
|
|
|
0
|
my ($dir,$file)=split /\/[^\/]+$/,$path; |
1200
|
0
|
0
|
|
|
|
0
|
unless (-e $dir){ |
1201
|
0
|
|
|
|
|
0
|
system('mkdir -p '.$dir); |
1202
|
|
|
|
|
|
|
} |
1203
|
|
|
|
|
|
|
} |
1204
|
|
|
|
|
|
|
} |
1205
|
|
|
|
|
|
|
} |
1206
|
|
|
|
|
|
|
|
1207
|
|
|
|
|
|
|
# open files of an alignment set for read and go to first sentence pair |
1208
|
|
|
|
|
|
|
sub openFiles { |
1209
|
2
|
|
|
2
|
0
|
3
|
my $alSet = shift; |
1210
|
2
|
|
|
|
|
4
|
my %FH = %{openLocation($alSet->{location},$alSet->{format},"<")}; |
|
2
|
|
|
|
|
9
|
|
1211
|
2
|
|
|
|
|
5
|
my $fhPos; |
1212
|
|
|
|
|
|
|
my $lineNb; |
1213
|
|
|
|
|
|
|
|
1214
|
|
|
|
|
|
|
# go to first Sentence pair: |
1215
|
2
|
50
|
|
|
|
18
|
if ($alSet->{format} eq "TALP"){ |
|
|
50
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1216
|
0
|
|
|
|
|
0
|
for ($lineNb=$alSet->{firstSentPair}-1;$lineNb>0;$lineNb--){ #go to first Sentence pair |
1217
|
0
|
0
|
|
|
|
0
|
if ($FH{source}){ |
1218
|
0
|
|
|
|
|
0
|
$FH{source}->getline(); |
1219
|
|
|
|
|
|
|
} |
1220
|
0
|
0
|
|
|
|
0
|
if ($FH{target}){ |
1221
|
0
|
|
|
|
|
0
|
$FH{target}->getline(); |
1222
|
|
|
|
|
|
|
} |
1223
|
0
|
0
|
|
|
|
0
|
if ($FH{sourceToTarget}){ |
1224
|
0
|
|
|
|
|
0
|
$FH{sourceToTarget}->getline(); |
1225
|
|
|
|
|
|
|
} |
1226
|
0
|
0
|
|
|
|
0
|
if ($FH{targetToSource}){ |
1227
|
0
|
|
|
|
|
0
|
$FH{targetToSource}->getline(); |
1228
|
|
|
|
|
|
|
} |
1229
|
|
|
|
|
|
|
} |
1230
|
|
|
|
|
|
|
}elsif ($alSet->{format} eq "GIZA"){ |
1231
|
0
|
|
|
|
|
0
|
for ($lineNb=$alSet->{firstSentPair}-1;$lineNb>0;$lineNb--){ #go to first Sentence pair |
1232
|
0
|
|
|
|
|
0
|
$FH{sourceToTarget}->getline(); |
1233
|
0
|
|
|
|
|
0
|
$FH{sourceToTarget}->getline(); |
1234
|
0
|
|
|
|
|
0
|
$FH{sourceToTarget}->getline(); |
1235
|
0
|
0
|
|
|
|
0
|
if ($FH{targetToSource}){ |
1236
|
0
|
|
|
|
|
0
|
$FH{targetToSource}->getline(); |
1237
|
0
|
|
|
|
|
0
|
$FH{targetToSource}->getline(); |
1238
|
0
|
|
|
|
|
0
|
$FH{targetToSource}->getline(); |
1239
|
|
|
|
|
|
|
} |
1240
|
|
|
|
|
|
|
} |
1241
|
|
|
|
|
|
|
} elsif ($alSet->{format} eq "NAACL"){ |
1242
|
1
|
|
|
|
|
9
|
for ($lineNb=$alSet->{firstSentPair}-1;$lineNb>0;$lineNb--){ |
1243
|
0
|
0
|
|
|
|
0
|
if ($FH{source}){ |
1244
|
0
|
|
|
|
|
0
|
$FH{source}->getline(); |
1245
|
|
|
|
|
|
|
} |
1246
|
0
|
0
|
|
|
|
0
|
if ($FH{target}){ |
1247
|
0
|
|
|
|
|
0
|
$FH{target}->getline(); |
1248
|
|
|
|
|
|
|
} |
1249
|
0
|
|
|
|
|
0
|
$fhPos = $FH{sourceToTarget}->getpos; |
1250
|
0
|
|
0
|
|
|
0
|
while ($FH{sourceToTarget}->getline() !~ m/^0*$alSet->{firstSentPair} .*/ && !$FH{sourceToTarget}->eof()) { |
1251
|
0
|
|
|
|
|
0
|
$fhPos = $FH{sourceToTarget}->getpos; |
1252
|
|
|
|
|
|
|
} |
1253
|
0
|
0
|
|
|
|
0
|
if ($FH{sourceToTarget}->eof()){ |
1254
|
0
|
|
|
|
|
0
|
die "First sentence pair of range (number ".$alSet->{firstSentPair}.") not found in ".$alSet->{location}{sourceToTarget}; |
1255
|
|
|
|
|
|
|
} |
1256
|
0
|
|
|
|
|
0
|
$FH{sourceToTarget}->setpos($fhPos); #if we changed sentences, we read the first line of next sentence=>go back one line |
1257
|
0
|
0
|
|
|
|
0
|
if ($FH{targetToSource}){ |
1258
|
0
|
|
|
|
|
0
|
$fhPos = $FH{targetToSource}->getpos; |
1259
|
0
|
|
0
|
|
|
0
|
while ($FH{targetToSource}->getline() !~ m/^0*$alSet->{firstSentPair} .*/ && !$FH{targetToSource}->eof()) { |
1260
|
0
|
|
|
|
|
0
|
$fhPos = $FH{targetToSource}->getpos; |
1261
|
|
|
|
|
|
|
} |
1262
|
0
|
0
|
|
|
|
0
|
if ($FH{targetToSource}->eof()){ |
1263
|
0
|
|
|
|
|
0
|
die "First sentence pair of range (number ".$alSet->{firstSentPair}.") not found in ".$alSet->{location}{targetToSource}; |
1264
|
|
|
|
|
|
|
} |
1265
|
0
|
|
|
|
|
0
|
$FH{targetToSource}->setpos($fhPos); |
1266
|
|
|
|
|
|
|
} |
1267
|
|
|
|
|
|
|
} |
1268
|
|
|
|
|
|
|
} elsif ($alSet->{format} eq "BLINKER"){ |
1269
|
1
|
50
|
|
|
|
4
|
if ($FH{source}){ |
1270
|
0
|
|
|
|
|
0
|
for ($lineNb=$alSet->{firstSentPair}-1;$lineNb>0;$lineNb--){ |
1271
|
0
|
|
|
|
|
0
|
$FH{source}->getline(); |
1272
|
|
|
|
|
|
|
} |
1273
|
|
|
|
|
|
|
} |
1274
|
1
|
50
|
|
|
|
3
|
if ($FH{target}){ |
1275
|
0
|
|
|
|
|
0
|
for ($lineNb=$alSet->{firstSentPair}-1;$lineNb>0;$lineNb--){ |
1276
|
0
|
|
|
|
|
0
|
$FH{target}->getline(); |
1277
|
|
|
|
|
|
|
} |
1278
|
|
|
|
|
|
|
} |
1279
|
|
|
|
|
|
|
} |
1280
|
2
|
|
|
|
|
5
|
return (\%FH); |
1281
|
|
|
|
|
|
|
} |
1282
|
|
|
|
|
|
|
|
1283
|
|
|
|
|
|
|
# close the files contained in the hash at the {location} key of the alignment set hash |
1284
|
|
|
|
|
|
|
sub closeFiles { |
1285
|
2
|
|
|
2
|
0
|
5
|
my ($FH,$format) = @_; |
1286
|
|
|
|
|
|
|
|
1287
|
2
|
50
|
66
|
|
|
20
|
if ($format eq "GIZA"){ |
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1288
|
0
|
|
|
|
|
0
|
$FH->{sourceToTarget}->close(); |
1289
|
0
|
0
|
|
|
|
0
|
if ($$FH{targetToSource}){ |
1290
|
0
|
|
|
|
|
0
|
$FH->{targetToSource}->close(); |
1291
|
|
|
|
|
|
|
} |
1292
|
|
|
|
|
|
|
} elsif ($format eq "NAACL" || $format eq "TALP"){ |
1293
|
1
|
50
|
|
|
|
6
|
if ($FH->{source}){ |
1294
|
1
|
|
|
|
|
13
|
$FH->{source}->close(); |
1295
|
|
|
|
|
|
|
} |
1296
|
1
|
50
|
|
|
|
24
|
if ($FH->{target}){ |
1297
|
1
|
|
|
|
|
5
|
$FH->{target}->close(); |
1298
|
|
|
|
|
|
|
} |
1299
|
1
|
50
|
|
|
|
20
|
if ($FH->{sourceToTarget}){ |
1300
|
1
|
|
|
|
|
4
|
$FH->{sourceToTarget}->close(); |
1301
|
|
|
|
|
|
|
} |
1302
|
1
|
50
|
|
|
|
17
|
if ($FH->{targetToSource}){ |
1303
|
0
|
|
|
|
|
0
|
$FH->{targetToSource}->close(); |
1304
|
|
|
|
|
|
|
} |
1305
|
|
|
|
|
|
|
} elsif ($format eq "BLINKER"){ |
1306
|
1
|
50
|
|
|
|
3
|
if ($FH->{source}){ |
1307
|
0
|
|
|
|
|
0
|
$FH->{source}->close(); |
1308
|
|
|
|
|
|
|
} |
1309
|
1
|
50
|
|
|
|
3
|
if ($FH->{target}){ |
1310
|
0
|
|
|
|
|
0
|
$FH->{target}->close(); |
1311
|
|
|
|
|
|
|
} |
1312
|
|
|
|
|
|
|
} |
1313
|
|
|
|
|
|
|
} |
1314
|
|
|
|
|
|
|
|
1315
|
|
|
|
|
|
|
# convert a chunk of alignment set file to an array of references to simple (1 sentence) alignment objects |
1316
|
|
|
|
|
|
|
# returns 0 if the file is at eof |
1317
|
|
|
|
|
|
|
sub loadChunk { |
1318
|
22
|
|
|
22
|
0
|
38
|
my ($alSet,$alFH,$sentPairNum,$alignMode) = @_; |
1319
|
22
|
|
|
|
|
25
|
my ($sourceString,$targetString,$alString,$reverseAlString); |
1320
|
22
|
|
|
|
|
27
|
my $st_alignments=[]; |
1321
|
22
|
|
|
|
|
32
|
my $ts_alignments=[]; |
1322
|
22
|
|
|
|
|
28
|
my $al; |
1323
|
|
|
|
|
|
|
my $theEnd; |
1324
|
|
|
|
|
|
|
|
1325
|
22
|
50
|
33
|
|
|
207
|
if (!defined($alignMode) || $alignMode =~ /^as.?is$/i){ |
|
|
50
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1326
|
0
|
|
|
|
|
0
|
$alignMode = "as-is"; |
1327
|
|
|
|
|
|
|
}elsif ($alignMode =~ /^null.?align$/i){ |
1328
|
0
|
|
|
|
|
0
|
$alignMode = "null-align"; |
1329
|
|
|
|
|
|
|
}elsif ($alignMode =~ /^no.?null.?align$/i){ |
1330
|
22
|
|
|
|
|
35
|
$alignMode = "no-null-align"; |
1331
|
|
|
|
|
|
|
}else{ |
1332
|
0
|
|
|
|
|
0
|
die 'Incorrect alignment mode. Correct modes are "as-is","null-align" or "no-null-align".'."\n"; |
1333
|
|
|
|
|
|
|
} |
1334
|
22
|
50
|
|
|
|
97
|
if ($alSet->{format} eq "TALP"){ |
|
|
50
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
1335
|
0
|
0
|
|
|
|
0
|
if ($alSet->{lastSentPair} eq "eof"){ |
1336
|
0
|
|
|
|
|
0
|
$theEnd = $$alFH{sourceToTarget}->eof(); |
1337
|
|
|
|
|
|
|
}else{ |
1338
|
0
|
|
0
|
|
|
0
|
$theEnd = ($$alFH{sourceToTarget}->eof() || $sentPairNum > $alSet->{lastSentPair}); |
1339
|
|
|
|
|
|
|
} |
1340
|
0
|
0
|
|
|
|
0
|
if ($theEnd){ |
1341
|
0
|
|
|
|
|
0
|
return 0; |
1342
|
|
|
|
|
|
|
}else{ |
1343
|
0
|
0
|
|
|
|
0
|
if ($alFH->{source}){ |
1344
|
0
|
|
|
|
|
0
|
$sourceString = $alFH->{source}->getline(); |
1345
|
|
|
|
|
|
|
} |
1346
|
0
|
0
|
|
|
|
0
|
if ($alFH->{target}){ |
1347
|
0
|
|
|
|
|
0
|
$targetString = $alFH->{target}->getline(); |
1348
|
|
|
|
|
|
|
} |
1349
|
0
|
0
|
|
|
|
0
|
if ($alFH->{sourceToTarget}){ |
1350
|
0
|
|
|
|
|
0
|
$alString = $alFH->{sourceToTarget}->getline(); |
1351
|
|
|
|
|
|
|
} |
1352
|
0
|
0
|
|
|
|
0
|
if ($alFH->{targetToSource}){ |
1353
|
0
|
|
|
|
|
0
|
$reverseAlString = $alFH->{targetToSource}->getline(); |
1354
|
|
|
|
|
|
|
} |
1355
|
0
|
|
|
|
|
0
|
$al = Lingua::Alignment->new; |
1356
|
0
|
|
|
|
|
0
|
$al->loadFromTalp($alString,$reverseAlString,$sourceString,$targetString); |
1357
|
|
|
|
|
|
|
} |
1358
|
|
|
|
|
|
|
}elsif ($alSet->{format} eq "GIZA"){ |
1359
|
0
|
0
|
|
|
|
0
|
if ($alSet->{lastSentPair} eq "eof"){ |
1360
|
0
|
|
|
|
|
0
|
$theEnd = $$alFH{sourceToTarget}->eof(); |
1361
|
|
|
|
|
|
|
}else{ |
1362
|
0
|
|
0
|
|
|
0
|
$theEnd = ($$alFH{sourceToTarget}->eof() || $sentPairNum > $alSet->{lastSentPair}); |
1363
|
|
|
|
|
|
|
} |
1364
|
0
|
0
|
|
|
|
0
|
if ($theEnd){ |
1365
|
0
|
|
|
|
|
0
|
return 0; |
1366
|
|
|
|
|
|
|
}else{ |
1367
|
0
|
|
|
|
|
0
|
$$alFH{sourceToTarget}->getline(); |
1368
|
0
|
|
|
|
|
0
|
$targetString = $$alFH{sourceToTarget}->getline(); |
1369
|
0
|
|
|
|
|
0
|
$alString = $$alFH{sourceToTarget}->getline(); |
1370
|
0
|
0
|
|
|
|
0
|
if ($$alFH{targetToSource}){ |
1371
|
0
|
|
|
|
|
0
|
$$alFH{targetToSource}->getline(); |
1372
|
0
|
|
|
|
|
0
|
$$alFH{targetToSource}->getline(); |
1373
|
0
|
|
|
|
|
0
|
$reverseAlString = $$alFH{targetToSource}->getline(); |
1374
|
|
|
|
|
|
|
} |
1375
|
0
|
|
|
|
|
0
|
$al = Lingua::Alignment->new; |
1376
|
0
|
|
|
|
|
0
|
$al->loadFromGiza($alString,$targetString,$reverseAlString); |
1377
|
|
|
|
|
|
|
} |
1378
|
|
|
|
|
|
|
} elsif ($alSet->{format} eq "NAACL"){ |
1379
|
11
|
|
|
|
|
11
|
my $fhPos; |
1380
|
11
|
50
|
|
|
|
21
|
if ($alSet->{lastSentPair} eq "eof"){ |
1381
|
0
|
|
|
|
|
0
|
$theEnd = $$alFH{sourceToTarget}->eof(); |
1382
|
|
|
|
|
|
|
}else{ |
1383
|
11
|
|
66
|
|
|
39
|
$theEnd = ($$alFH{sourceToTarget}->eof() || $sentPairNum > $alSet->{lastSentPair}); |
1384
|
|
|
|
|
|
|
} |
1385
|
11
|
100
|
|
|
|
174
|
if ($theEnd){ |
1386
|
1
|
|
|
|
|
7
|
return 0; |
1387
|
|
|
|
|
|
|
}else{ |
1388
|
10
|
50
|
|
|
|
25
|
if ($$alFH{source}){ |
1389
|
10
|
|
|
|
|
281
|
$sourceString = $$alFH{source}->getline(); |
1390
|
|
|
|
|
|
|
#strip tags and memorize snum: |
1391
|
10
|
|
|
|
|
332
|
$sourceString =~ s/(.*)<\/s>/$1/; |
1392
|
|
|
|
|
|
|
} |
1393
|
10
|
50
|
|
|
|
29
|
if ($$alFH{target}){ |
1394
|
10
|
|
|
|
|
218
|
$targetString = $$alFH{target}->getline(); |
1395
|
|
|
|
|
|
|
#strip tags and memorize snum: |
1396
|
10
|
|
|
|
|
308
|
$targetString =~ s/(.*)<\/s>/$2/; |
1397
|
|
|
|
|
|
|
} |
1398
|
10
|
|
|
|
|
43
|
$fhPos = $$alFH{sourceToTarget}->getpos; |
1399
|
10
|
|
|
|
|
211
|
$alString = $$alFH{sourceToTarget}->getline(); |
1400
|
10
|
|
|
|
|
264
|
my ($num,$theRest)=split " ",$alString,2; |
1401
|
10
|
50
|
|
|
|
27
|
if ($num==$sentPairNum){ #skip if there is no link for this sentence pair |
1402
|
10
|
|
|
|
|
30
|
$fhPos = $$alFH{sourceToTarget}->getpos; |
1403
|
10
|
|
|
|
|
20
|
push @$st_alignments,$theRest; |
1404
|
10
|
|
|
|
|
211
|
while ($$alFH{sourceToTarget}->getline() =~ m/^$sentPairNum (.*)$/) { |
1405
|
58
|
|
|
|
|
1919
|
push @$st_alignments,$1; |
1406
|
58
|
|
|
|
|
1408
|
$fhPos = $$alFH{sourceToTarget}->getpos; |
1407
|
|
|
|
|
|
|
} |
1408
|
|
|
|
|
|
|
} |
1409
|
10
|
|
|
|
|
384
|
$$alFH{sourceToTarget}->setpos($fhPos); #if we changed sentences, we read the first line of next sentence=>go back one line |
1410
|
10
|
50
|
|
|
|
32
|
if ($$alFH{targetToSource}){ |
1411
|
0
|
|
|
|
|
0
|
$fhPos = $$alFH{targetToSource}->getpos; |
1412
|
0
|
|
|
|
|
0
|
$alString = $$alFH{targetToSource}->getline(); |
1413
|
0
|
|
|
|
|
0
|
my ($num,$theRest)=split " ",$alString,2; |
1414
|
0
|
0
|
|
|
|
0
|
if ($num==$sentPairNum){ #skip if there is no link for this sentence pair |
1415
|
0
|
|
|
|
|
0
|
$fhPos = $$alFH{targetToSource}->getpos; |
1416
|
0
|
|
|
|
|
0
|
push @$ts_alignments,$theRest; |
1417
|
0
|
|
|
|
|
0
|
while ($$alFH{targetToSource}->getline() =~ m/^$sentPairNum (.*)$/) { |
1418
|
0
|
|
|
|
|
0
|
push @$ts_alignments,$1; |
1419
|
0
|
|
|
|
|
0
|
$fhPos = $$alFH{targetToSource}->getpos; |
1420
|
|
|
|
|
|
|
} |
1421
|
|
|
|
|
|
|
} |
1422
|
0
|
|
|
|
|
0
|
$$alFH{targetToSource}->setpos($fhPos); #if we changed sentences, we read the first line of next sentence=>go back one line |
1423
|
|
|
|
|
|
|
} |
1424
|
10
|
|
|
|
|
36
|
$al = Lingua::Alignment->new; |
1425
|
10
|
|
|
|
|
30
|
$al->loadFromBlinker($st_alignments,$ts_alignments,$sourceString,$targetString); |
1426
|
|
|
|
|
|
|
} |
1427
|
|
|
|
|
|
|
} elsif ($alSet->{format} eq "BLINKER"){ |
1428
|
11
|
50
|
|
|
|
21
|
if ($alSet->{lastSentPair} eq "eof"){ |
1429
|
11
|
|
|
|
|
485
|
$theEnd = !(-e $alSet->{location}->{sourceToTarget}."/samp".$alSet->{location}->{sampleNum}.".SentPair".($sentPairNum-1)); |
1430
|
|
|
|
|
|
|
}else{ |
1431
|
0
|
|
0
|
|
|
0
|
$theEnd = !(-e $alSet->{location}->{sourceToTarget}."/samp".$alSet->{location}->{sampleNum}.".SentPair".($sentPairNum-1)) || $sentPairNum > $alSet->{lastSentPair}; |
1432
|
|
|
|
|
|
|
} |
1433
|
11
|
100
|
|
|
|
22
|
if ($theEnd){ |
1434
|
1
|
|
|
|
|
4
|
return 0; |
1435
|
|
|
|
|
|
|
}else{ |
1436
|
10
|
50
|
|
|
|
24
|
if ($alFH->{source}){ |
1437
|
0
|
|
|
|
|
0
|
$sourceString = $alFH->{source}->getline(); |
1438
|
|
|
|
|
|
|
} |
1439
|
10
|
50
|
|
|
|
21
|
if ($alFH->{target}){ |
1440
|
0
|
|
|
|
|
0
|
$targetString = $alFH->{target}->getline(); |
1441
|
|
|
|
|
|
|
} |
1442
|
10
|
|
|
|
|
346
|
open(AL,"< ".$alSet->{location}->{sourceToTarget}."/samp".$alSet->{location}->{sampleNum}.".SentPair".($sentPairNum-1)); |
1443
|
10
|
|
|
|
|
241
|
@$st_alignments = ; |
1444
|
10
|
|
|
|
|
103
|
close(AL); |
1445
|
10
|
50
|
|
|
|
31
|
if ($alSet->{location}->{targetToSource}){ |
1446
|
0
|
|
|
|
|
0
|
open(AL,"< ".$alSet->{location}->{targetToSource}."/samp".$alSet->{location}->{sampleNum}.".SentPair".($sentPairNum-1)); |
1447
|
0
|
|
|
|
|
0
|
@$ts_alignments = ; |
1448
|
0
|
|
|
|
|
0
|
close(AL); |
1449
|
|
|
|
|
|
|
} |
1450
|
10
|
|
|
|
|
44
|
$al = Lingua::Alignment->new; |
1451
|
10
|
|
|
|
|
35
|
$al->loadFromBlinker($st_alignments,$ts_alignments,$sourceString,$targetString); |
1452
|
|
|
|
|
|
|
} |
1453
|
|
|
|
|
|
|
} |
1454
|
20
|
50
|
|
|
|
68
|
if ($alignMode eq "null-align"){ |
|
|
50
|
|
|
|
|
|
1455
|
0
|
|
|
|
|
0
|
$al->forceNullAlign(); |
1456
|
|
|
|
|
|
|
}elsif ($alignMode eq "no-null-align"){ |
1457
|
20
|
|
|
|
|
55
|
$al->forceNoNullAlign(); |
1458
|
|
|
|
|
|
|
} |
1459
|
20
|
|
|
|
|
110
|
return [$al]; |
1460
|
|
|
|
|
|
|
} |
1461
|
|
|
|
|
|
|
|
1462
|
|
|
|
|
|
|
|
1463
|
|
|
|
|
|
|
sub updateObject { |
1464
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newFormat,$newLocation,$lastSentPairNum)=@_; |
1465
|
0
|
|
|
|
|
0
|
$alSet->{location}->{sourceToTarget}=$newLocation->{sourceToTarget}; |
1466
|
0
|
|
|
|
|
0
|
$alSet->{location}->{targetToSource}=$newLocation->{targetToSource}; |
1467
|
0
|
0
|
|
|
|
0
|
if ($newLocation->{source}){ |
1468
|
0
|
|
|
|
|
0
|
$alSet->{location}->{source}=$newLocation->{source}; |
1469
|
|
|
|
|
|
|
}else{ |
1470
|
0
|
0
|
0
|
|
|
0
|
if ($alSet->{firstSentPair} != 1 || $alSet->{format} ne $newFormat){ |
1471
|
|
|
|
|
|
|
# in this case the numeration of the converted alignment file and that of the (not converted) source file will not correspond |
1472
|
0
|
|
|
|
|
0
|
delete($alSet->{location}->{source}); |
1473
|
|
|
|
|
|
|
# warn "After converting into ",$newLocation->{sourceToTarget},", the numeration of the source words file", |
1474
|
|
|
|
|
|
|
# " didn't correspond any more to that of the alignment file. So the 'source' entry has been removed from the location hash."; |
1475
|
|
|
|
|
|
|
} |
1476
|
|
|
|
|
|
|
} |
1477
|
0
|
0
|
|
|
|
0
|
if ($newLocation->{target}){ |
1478
|
0
|
|
|
|
|
0
|
$alSet->{location}->{target}=$newLocation->{target}; |
1479
|
|
|
|
|
|
|
}else{ |
1480
|
0
|
0
|
0
|
|
|
0
|
if ($alSet->{firstSentPair} != 1 || $alSet->{format} ne $newFormat){ |
1481
|
|
|
|
|
|
|
# in this case the numeration of the converted alignment file and that of the (not converted) source file will not correspond |
1482
|
0
|
|
|
|
|
0
|
delete($alSet->{location}->{target}); |
1483
|
|
|
|
|
|
|
# warn "After converting into ",$newLocation->{sourceToTarget},", the numeration of the target words file ", |
1484
|
|
|
|
|
|
|
# "didn't correspond any more to that of the alignment file. So the 'target' entry has been removed from the location hash."; |
1485
|
|
|
|
|
|
|
} |
1486
|
|
|
|
|
|
|
} |
1487
|
0
|
|
|
|
|
0
|
$alSet->{format}=$newFormat; |
1488
|
0
|
0
|
|
|
|
0
|
if ($newFormat eq "BLINKER"){ |
|
|
0
|
|
|
|
|
|
1489
|
0
|
|
|
|
|
0
|
$alSet->{location}->{sampleNum}=$newLocation->{sampleNum}; |
1490
|
|
|
|
|
|
|
}elsif(exists($alSet->{location}->{sampleNum})){ |
1491
|
0
|
|
|
|
|
0
|
delete($alSet->{location}->{sampleNum}); |
1492
|
|
|
|
|
|
|
} |
1493
|
0
|
|
|
|
|
0
|
$alSet->{firstSentPair}=1; |
1494
|
0
|
|
|
|
|
0
|
$alSet->{lastSentPair}=$lastSentPairNum; |
1495
|
|
|
|
|
|
|
} |
1496
|
|
|
|
|
|
|
|
1497
|
|
|
|
|
|
|
# returns the alignment set, with a unique new file set that has the required location,format and range values. |
1498
|
|
|
|
|
|
|
# TO DO: conversion to Giza++ format |
1499
|
|
|
|
|
|
|
sub convert { |
1500
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat,$alignMode,$AlignmentSub)=@_; |
1501
|
0
|
0
|
|
|
|
0
|
if (!defined($newFormat)){$newFormat="TALP"} |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
1502
|
|
|
|
|
|
|
else {$newFormat = uc $newFormat} |
1503
|
0
|
|
|
|
|
0
|
$newLocation = readLocation($newLocation); |
1504
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
1505
|
0
|
|
|
|
|
0
|
my $newFH = openLocation($newLocation,$newFormat,">",$alSet->{location}); |
1506
|
0
|
|
|
|
|
0
|
my ($i,$al,$alSetChunk,$line,$lines); |
1507
|
0
|
|
|
|
|
0
|
my $inputSentPairNum=$alSet->{firstSentPair}; |
1508
|
0
|
|
|
|
|
0
|
my $internalSentPairNum = 1; |
1509
|
0
|
|
|
|
|
0
|
while ($alSetChunk = $alSet->loadChunk($FH,$inputSentPairNum,$alignMode)){ # returns 0 if eof or last sentence pair |
1510
|
|
|
|
|
|
|
# print $inputSentPairNum."\n"; |
1511
|
0
|
|
|
|
|
0
|
for ($i=0;$i<@$alSetChunk;$i++){ |
1512
|
0
|
|
|
|
|
0
|
$al = $$alSetChunk[$i]; |
1513
|
0
|
0
|
|
|
|
0
|
if (defined($AlignmentSub)){ |
1514
|
|
|
|
|
|
|
#look if $AlignmentSub is a ref to an Array or a subroutine |
1515
|
0
|
0
|
|
|
|
0
|
if (ref($AlignmentSub) eq "ARRAY"){ |
1516
|
0
|
|
|
|
|
0
|
my ($sub,@params) = @$AlignmentSub; |
1517
|
0
|
|
|
|
|
0
|
$al->$sub(@params); |
1518
|
|
|
|
|
|
|
}else{ |
1519
|
0
|
|
|
|
|
0
|
$al->$AlignmentSub(); |
1520
|
|
|
|
|
|
|
} |
1521
|
|
|
|
|
|
|
} |
1522
|
0
|
|
|
|
|
0
|
$al->output($FH,$newFormat,$newFH,$newLocation,$internalSentPairNum); |
1523
|
|
|
|
|
|
|
} #for |
1524
|
0
|
|
|
|
|
0
|
$inputSentPairNum++; |
1525
|
0
|
|
|
|
|
0
|
$internalSentPairNum++; |
1526
|
|
|
|
|
|
|
} #while |
1527
|
0
|
|
|
|
|
0
|
closeFiles($newFH,$newFormat); |
1528
|
0
|
|
|
|
|
0
|
closeFiles($FH,$alSet->{format}); |
1529
|
0
|
|
|
|
|
0
|
$alSet->updateObject($newFormat,$newLocation,$internalSentPairNum-1); |
1530
|
|
|
|
|
|
|
} |
1531
|
|
|
|
|
|
|
|
1532
|
|
|
|
|
|
|
|
1533
|
|
|
|
|
|
|
# returns the alignment set, with only manyToMany links (source and target words file don't change) |
1534
|
|
|
|
|
|
|
# at the moment, only works for sourceToTarget alignment (sourceAl) |
1535
|
|
|
|
|
|
|
sub printManyToMany { |
1536
|
0
|
|
|
0
|
0
|
0
|
my ($alSet,$newLocation,$newFormat)=@_; |
1537
|
0
|
0
|
|
|
|
0
|
if (!defined($newFormat)){$newFormat="TALP"} |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
1538
|
|
|
|
|
|
|
else {$newFormat = uc $newFormat} |
1539
|
0
|
|
|
|
|
0
|
$newLocation = readLocation($newLocation); |
1540
|
0
|
|
|
|
|
0
|
my $dumper=new Dumpvalue; |
1541
|
0
|
|
|
|
|
0
|
my $FH = $alSet->openFiles(); |
1542
|
0
|
|
|
|
|
0
|
my $newFH = openLocation($newLocation,$newFormat,">",$alSet->{location}); |
1543
|
0
|
|
|
|
|
0
|
my ($i,$al,$alSetChunk,$line,$lines); |
1544
|
0
|
|
|
|
|
0
|
my $inputSentPairNum=$alSet->{firstSentPair}; |
1545
|
0
|
|
|
|
|
0
|
my $internalSentPairNum = 1; |
1546
|
|
|
|
|
|
|
|
1547
|
0
|
|
|
|
|
0
|
while ($alSetChunk = $alSet->loadChunk($FH,$inputSentPairNum,"no-null-align")){ # returns 0 if eof or last sentence pair |
1548
|
0
|
|
|
|
|
0
|
for ($i=0;$i<@$alSetChunk;$i++){ |
1549
|
0
|
|
|
|
|
0
|
$al = $$alSetChunk[$i]; |
1550
|
0
|
|
|
|
|
0
|
my $clone = $al->clone; |
1551
|
|
|
|
|
|
|
# look for manyToMany links |
1552
|
0
|
|
|
|
|
0
|
my $clusters = $al->getAlClusters; |
1553
|
0
|
|
|
|
|
0
|
$al->{sourceAl}=[]; |
1554
|
|
|
|
|
|
|
# print $dumper->dumpValue($clusters); |
1555
|
0
|
|
|
|
|
0
|
for (my $c=0;$c<@$clusters;$c++){ |
1556
|
0
|
0
|
0
|
|
|
0
|
if ( @{$clusters->[$c]{source}}>1 || @{$clusters->[$c]{target}}>1 ){ |
|
0
|
|
|
|
|
0
|
|
|
0
|
|
|
|
|
0
|
|
1557
|
|
|
|
|
|
|
# this is a many to many alignment |
1558
|
0
|
|
|
|
|
0
|
foreach my $j (@{$clusters->[$c]{source}}){ |
|
0
|
|
|
|
|
0
|
|
1559
|
0
|
|
|
|
|
0
|
foreach my $k (@{$clusters->[$c]{target}}){ |
|
0
|
|
|
|
|
0
|
|
1560
|
0
|
0
|
|
|
|
0
|
if ($clone->isIn("sourceAl",$j,$k)){ |
1561
|
0
|
|
|
|
|
0
|
push @{$al->{sourceAl}[$j]},$k; |
|
0
|
|
|
|
|
0
|
|
1562
|
|
|
|
|
|
|
} |
1563
|
|
|
|
|
|
|
} |
1564
|
|
|
|
|
|
|
} |
1565
|
|
|
|
|
|
|
} |
1566
|
|
|
|
|
|
|
} |
1567
|
0
|
|
|
|
|
0
|
$al->output($FH,$newFormat,$newFH,$newLocation,$internalSentPairNum); |
1568
|
|
|
|
|
|
|
} #for |
1569
|
0
|
|
|
|
|
0
|
$inputSentPairNum++; |
1570
|
0
|
|
|
|
|
0
|
$internalSentPairNum++; |
1571
|
|
|
|
|
|
|
} #while |
1572
|
0
|
|
|
|
|
0
|
closeFiles($newFH,$newFormat); |
1573
|
0
|
|
|
|
|
0
|
closeFiles($FH,$alSet->{format}); |
1574
|
0
|
|
|
|
|
0
|
$alSet->updateObject($newFormat,$newLocation,$internalSentPairNum-1); |
1575
|
|
|
|
|
|
|
} |
1576
|
|
|
|
|
|
|
|
1577
|
|
|
|
|
|
|
|
1578
|
|
|
|
|
|
|
# identifies a link as sure or possible |
1579
|
|
|
|
|
|
|
# input: a Naacl-file line containing the link and refs to sure and possible hashes |
1580
|
|
|
|
|
|
|
# action: add to the relevant hash a key corresponding to this link |
1581
|
|
|
|
|
|
|
sub identifySurePossible{ |
1582
|
149
|
|
|
149
|
0
|
241
|
my ($line,$sure,$possible)=@_; |
1583
|
149
|
|
|
|
|
126
|
my @components; |
1584
|
|
|
|
|
|
|
my $alignment; |
1585
|
|
|
|
|
|
|
|
1586
|
|
|
|
|
|
|
# print "line:$line\n"; |
1587
|
|
|
|
|
|
|
|
1588
|
|
|
|
|
|
|
#code adapted from Rada Mihalcea's wa_eval_align.pl, rada@cs.unt.edu |
1589
|
|
|
|
|
|
|
# get all line components: format should be |
1590
|
|
|
|
|
|
|
# sentence_no position_L1 position_L2 [S|P] [confidence] |
1591
|
149
|
|
|
|
|
376
|
@components = split /\s+/, $line; |
1592
|
|
|
|
|
|
|
|
1593
|
149
|
50
|
|
|
|
411
|
if(scalar(@components) < 3) { |
1594
|
0
|
|
|
|
|
0
|
print STDERR "Incorrect format in answer file\n"; |
1595
|
0
|
|
|
|
|
0
|
exit; |
1596
|
|
|
|
|
|
|
} |
1597
|
|
|
|
|
|
|
|
1598
|
149
|
|
|
|
|
262
|
$alignment = $components[0]." ".$components[1]." ".$components[2]; |
1599
|
|
|
|
|
|
|
|
1600
|
|
|
|
|
|
|
# identify the S[ure] alignments |
1601
|
149
|
0
|
0
|
|
|
279
|
if( scalar (@components) == 3 || (scalar (@components) == 4 && ($components[3] =~ /^[\d\.]+$/ || $components[3] eq 'S')) || |
|
|
|
0
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
1602
|
|
|
|
|
|
|
(scalar (@components) == 5 && ($components[3] eq 'S' || $components[4] eq 'S'))) { |
1603
|
149
|
|
|
|
|
312
|
$sure->{$alignment} = 1; |
1604
|
|
|
|
|
|
|
} |
1605
|
|
|
|
|
|
|
|
1606
|
|
|
|
|
|
|
# identify the P[robable] alignments |
1607
|
149
|
0
|
33
|
|
|
812
|
if( (scalar (@components) == 4 && $components[3] eq 'P') || (scalar (@components) == 5 && |
|
|
|
0
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
|
33
|
|
|
|
|
1608
|
|
|
|
|
|
|
($components[3] eq 'P' || $components[4] eq 'P'))) { |
1609
|
0
|
|
|
|
|
|
$possible->{$alignment} = 1; |
1610
|
|
|
|
|
|
|
} |
1611
|
|
|
|
|
|
|
} |
1612
|
|
|
|
|
|
|
|
1613
|
|
|
|
|
|
|
1; |
1614
|
|
|
|
|
|
|
__END__ |