|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace Matecat\XliffParser\XliffReplacer; |
|
4
|
|
|
|
|
5
|
|
|
use Matecat\XliffParser\Utils\Strings; |
|
6
|
|
|
|
|
7
|
|
|
class SdlXliffSAXTranslationReplacer extends XliffSAXTranslationReplacer { |
|
8
|
|
|
protected $markerPos = ""; |
|
9
|
|
|
|
|
10
|
|
|
/** |
|
11
|
|
|
* @inheritDoc |
|
12
|
|
|
*/ |
|
13
|
|
|
protected function tagOpen( $parser, $name, $attr ) { |
|
14
|
|
|
// check if we are entering into a <trans-unit> (xliff v1.*) or <unit> (xliff v2.*) |
|
15
|
|
|
if ( $this->tuTagName === $name ) { |
|
16
|
|
|
$this->inTU = true; |
|
17
|
|
|
|
|
18
|
|
|
// get id |
|
19
|
|
|
// trim to first 100 characters because this is the limit on Matecat's DB |
|
20
|
|
|
$this->currentTransUnitId = substr( $attr[ 'id' ], 0, 100 ); |
|
21
|
|
|
|
|
22
|
|
|
// current 'translate' attribute of the current trans-unit |
|
23
|
|
|
$this->currentTransUnitTranslate = isset( $attr[ 'translate' ] ) ? $attr[ 'translate' ] : 'yes'; |
|
24
|
|
|
} |
|
25
|
|
|
|
|
26
|
|
|
// check if we are entering into a <target> |
|
27
|
|
|
if ( 'target' == $name ) { |
|
28
|
|
|
if ( $this->currentTransUnitTranslate === 'no' ) { |
|
29
|
|
|
$this->inTarget = false; |
|
30
|
|
|
} else { |
|
31
|
|
|
$this->inTarget = true; |
|
32
|
|
|
} |
|
33
|
|
|
} |
|
34
|
|
|
|
|
35
|
|
|
// reset Marker positions |
|
36
|
|
|
if ( 'sdl:seg-defs' == $name ) { |
|
37
|
|
|
$this->markerPos = 0; |
|
38
|
|
|
} |
|
39
|
|
|
|
|
40
|
|
|
// check if we are inside a <target>, obviously this happen only if there are targets inside the trans-unit |
|
41
|
|
|
// <target> must be stripped to be replaced, so this check avoids <target> reconstruction |
|
42
|
|
|
if ( !$this->inTarget ) { |
|
43
|
|
|
|
|
44
|
|
|
// costruct tag |
|
45
|
|
|
$tag = "<$name "; |
|
46
|
|
|
|
|
47
|
|
|
// needed to avoid multiple conf writing inside the same tag |
|
48
|
|
|
// because the "conf" attribute could be not present in the tag, |
|
49
|
|
|
// so the check on it's name is not enough |
|
50
|
|
|
$_sdlStatus_confWritten = false; |
|
51
|
|
|
|
|
52
|
|
|
foreach ( $attr as $k => $v ) { |
|
53
|
|
|
|
|
54
|
|
|
// if tag name is file, we must replace the target-language attribute |
|
55
|
|
|
if ( $name == 'file' && $k == 'target-language' && !empty( $this->targetLang ) ) { |
|
56
|
|
|
//replace Target language with job language provided from constructor |
|
57
|
|
|
$tag .= "$k=\"$this->targetLang\" "; |
|
58
|
|
|
|
|
59
|
|
|
if ( null !== $this->logger ) { |
|
60
|
|
|
$this->logger->debug( $k . " => " . $this->targetLang ); |
|
61
|
|
|
} |
|
62
|
|
|
} elseif ( 'sdl:seg' == $name ) { |
|
63
|
|
|
|
|
64
|
|
|
// write the confidence level for this segment ( Translated, Draft, etc. ) |
|
65
|
|
|
if ( isset( $this->segments[ 'matecat|' . $this->currentTransUnitId ] ) && $_sdlStatus_confWritten === false ) { |
|
66
|
|
|
|
|
67
|
|
|
// append definition attribute |
|
68
|
|
|
$tag .= $this->prepareTargetStatuses( $this->lastTransUnit[ $this->markerPos ] ); |
|
69
|
|
|
|
|
70
|
|
|
//prepare for an eventual next cycle |
|
71
|
|
|
$this->markerPos++; |
|
72
|
|
|
$_sdlStatus_confWritten = true; |
|
73
|
|
|
} |
|
74
|
|
|
|
|
75
|
|
|
// Warning, this is NOT an elseif |
|
76
|
|
|
if ( $k != 'conf' ) { |
|
77
|
|
|
//put also the current attribute in it if it is not a "conf" attribute |
|
78
|
|
|
$tag .= "$k=\"$v\" "; |
|
79
|
|
|
} |
|
80
|
|
|
} else { |
|
81
|
|
|
//normal tag flux, put attributes in it |
|
82
|
|
|
$tag .= "$k=\"$v\" "; |
|
83
|
|
|
} |
|
84
|
|
|
} |
|
85
|
|
|
|
|
86
|
|
|
// this logic helps detecting empty tags |
|
87
|
|
|
// get current position of SAX pointer in all the stream of data is has read so far: |
|
88
|
|
|
// it points at the end of current tag |
|
89
|
|
|
$idx = xml_get_current_byte_index( $parser ); |
|
90
|
|
|
|
|
91
|
|
|
// check whether the bounds of current tag are entirely in current buffer || the end of the current tag |
|
92
|
|
|
// is outside current buffer (in the latter case, it's in next buffer to be read by the while loop); |
|
93
|
|
|
// this check is necessary because we may have truncated a tag in half with current read, |
|
94
|
|
|
// and the other half may be encountered in the next buffer it will be passed |
|
95
|
|
|
if ( isset( $this->currentBuffer[ $idx - $this->offset ] ) ) { |
|
96
|
|
|
// if this tag entire lenght fitted in the buffer, the last char must be the last |
|
97
|
|
|
// symbol before the '>'; if it's an empty tag, it is assumed that it's a '/' |
|
98
|
|
|
$tmp_offset = $idx - $this->offset; |
|
99
|
|
|
$lastChar = $this->currentBuffer[ $tmp_offset ]; |
|
100
|
|
|
} else { |
|
101
|
|
|
//if it's out, simple use the last character of the chunk |
|
102
|
|
|
$tmp_offset = $this->len - 1; |
|
103
|
|
|
$lastChar = $this->currentBuffer[ $tmp_offset ]; |
|
104
|
|
|
} |
|
105
|
|
|
|
|
106
|
|
|
// trim last space |
|
107
|
|
|
$tag = rtrim( $tag ); |
|
108
|
|
|
|
|
109
|
|
|
// detect empty tag |
|
110
|
|
|
$this->isEmpty = ( $lastChar == '/' || $name == 'x' ); |
|
111
|
|
|
if ( $this->isEmpty ) { |
|
112
|
|
|
$tag .= '/'; |
|
113
|
|
|
} |
|
114
|
|
|
|
|
115
|
|
|
// add tag ending |
|
116
|
|
|
$tag .= ">"; |
|
117
|
|
|
|
|
118
|
|
|
// set a a Buffer for the segSource Source tag |
|
119
|
|
|
if ( 'source' == $name |
|
120
|
|
|
|| 'seg-source' === $name |
|
121
|
|
|
|| $this->bufferIsActive |
|
122
|
|
|
|| 'value' === $name |
|
123
|
|
|
|| 'bpt' === $name |
|
124
|
|
|
|| 'ept' === $name |
|
125
|
|
|
|| 'ph' === $name |
|
126
|
|
|
|| 'st' === $name |
|
127
|
|
|
|| 'note' === $name |
|
128
|
|
|
|| 'context' === $name ) { // we are opening a critical CDATA section |
|
129
|
|
|
|
|
130
|
|
|
// WARNING BECAUSE SOURCE AND SEG-SOURCE TAGS CAN BE EMPTY IN SOME CASES!!!!! |
|
131
|
|
|
// so check for isEmpty also in conjunction with name |
|
132
|
|
|
if ( $this->isEmpty && ( 'source' == $name || 'seg-source' == $name ) ) { |
|
133
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag ); |
|
134
|
|
|
} else { |
|
135
|
|
|
//these are NOT source/seg-source/value empty tags, THERE IS A CONTENT, write it in buffer |
|
136
|
|
|
$this->bufferIsActive = true; |
|
137
|
|
|
$this->CDATABuffer .= $tag; |
|
138
|
|
|
} |
|
139
|
|
|
} else { |
|
140
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag ); |
|
141
|
|
|
} |
|
142
|
|
|
} |
|
143
|
|
|
} |
|
144
|
|
|
|
|
145
|
|
|
/** |
|
146
|
|
|
* prepare segment tagging for xliff insertion |
|
147
|
|
|
* |
|
148
|
|
|
* @param array $seg |
|
149
|
|
|
* @param string $transUnitTranslation |
|
150
|
|
|
* |
|
151
|
|
|
* @return string |
|
152
|
|
|
*/ |
|
153
|
|
|
protected function prepareTranslation( $seg, $transUnitTranslation = "" ) { |
|
154
|
|
|
$endTags = ""; |
|
155
|
|
|
|
|
156
|
|
|
$segment = Strings::removeDangerousChars( $seg [ 'segment' ] ); |
|
157
|
|
|
$translation = Strings::removeDangerousChars( $seg [ 'translation' ] ); |
|
158
|
|
|
$dataRefMap = ( isset( $seg[ 'data_ref_map' ] ) && $seg[ 'data_ref_map' ] !== null ) ? Strings::jsonToArray( $seg[ 'data_ref_map' ] ) : []; |
|
159
|
|
|
|
|
160
|
|
|
if ( is_null( $seg [ 'translation' ] ) || $seg [ 'translation' ] == '' ) { |
|
161
|
|
|
$translation = $segment; |
|
162
|
|
|
} else { |
|
163
|
|
|
if ( $this->callback ) { |
|
164
|
|
|
$error = (isset($seg['error'])) ? $seg['error'] : null; |
|
165
|
|
|
if ( $this->callback->thereAreErrors( $seg[ 'sid' ], $segment, $translation, $dataRefMap, $error ) ) { |
|
166
|
|
|
$translation = '|||UNTRANSLATED_CONTENT_START|||' . $segment . '|||UNTRANSLATED_CONTENT_END|||'; |
|
167
|
|
|
} |
|
168
|
|
|
} |
|
169
|
|
|
} |
|
170
|
|
|
|
|
171
|
|
|
// for Trados the trailing spaces after </mrk> are meaningful |
|
172
|
|
|
// so we trim the translation from Matecat DB and add them after </mrk> |
|
173
|
|
|
$trailingSpaces = ''; |
|
174
|
|
|
for ( $s = 0; $s < Strings::getTheNumberOfTrailingSpaces( $translation ); $s++ ) { |
|
175
|
|
|
$trailingSpaces .= ' '; |
|
176
|
|
|
} |
|
177
|
|
|
|
|
178
|
|
|
if ( $seg[ 'mrk_id' ] !== null && $seg[ 'mrk_id' ] != '' ) { |
|
179
|
|
|
if ( $this->targetLang === 'ja-JP' ) { |
|
180
|
|
|
$seg[ 'mrk_succ_tags' ] = ltrim( $seg[ 'mrk_succ_tags' ] ); |
|
181
|
|
|
} |
|
182
|
|
|
|
|
183
|
|
|
$translation = "<mrk mid=\"" . $seg[ 'mrk_id' ] . "\" mtype=\"seg\">" . $seg[ 'mrk_prev_tags' ] . rtrim( $translation ) . $seg[ 'mrk_succ_tags' ] . "</mrk>" . $trailingSpaces; |
|
184
|
|
|
} |
|
185
|
|
|
|
|
186
|
|
|
// we need to trim succ_tags here because we already added the trailing spaces after </mrk> |
|
187
|
|
|
$transUnitTranslation .= $seg[ 'prev_tags' ] . $translation . $endTags . ltrim( $seg[ 'succ_tags' ] ); |
|
188
|
|
|
|
|
189
|
|
|
return $transUnitTranslation; |
|
190
|
|
|
} |
|
191
|
|
|
|
|
192
|
|
|
/** |
|
193
|
|
|
* @param $segment |
|
194
|
|
|
* |
|
195
|
|
|
* @return string |
|
196
|
|
|
*/ |
|
197
|
|
|
protected function prepareTargetStatuses( $segment ) { |
|
198
|
|
|
$statusMap = [ |
|
199
|
|
|
'NEW' => '', |
|
200
|
|
|
'DRAFT' => 'Draft', |
|
201
|
|
|
'TRANSLATED' => 'Translated', |
|
202
|
|
|
'APPROVED' => 'ApprovedTranslation', |
|
203
|
|
|
'REJECTED' => 'RejectedTranslation', |
|
204
|
|
|
]; |
|
205
|
|
|
|
|
206
|
|
|
return "conf=\"{$statusMap[ $segment[ 'status' ] ]}\" "; |
|
207
|
|
|
} |
|
208
|
|
|
|
|
209
|
|
|
/** |
|
210
|
|
|
* @param $seg |
|
211
|
|
|
* @param $state_prop |
|
212
|
|
|
* @param $lastMrkState |
|
213
|
|
|
* |
|
214
|
|
|
* @return array |
|
215
|
|
|
*/ |
|
216
|
|
|
protected function setTransUnitState( $seg, $state_prop, $lastMrkState ) { |
|
217
|
|
|
return [ null, null ]; |
|
218
|
|
|
} |
|
219
|
|
|
|
|
220
|
|
|
/** |
|
221
|
|
|
* @param $raw_word_count |
|
222
|
|
|
* @param $eq_word_count |
|
223
|
|
|
* |
|
224
|
|
|
* @return string |
|
225
|
|
|
*/ |
|
226
|
|
|
protected function getWordCountGroup( $raw_word_count, $eq_word_count ) { |
|
227
|
|
|
return ''; |
|
228
|
|
|
} |
|
229
|
|
|
} |
|
230
|
|
|
|
matecat /
xliff-parser
