1
|
|
|
<?php |
2
|
|
|
/** |
3
|
|
|
* Created by PhpStorm. |
4
|
|
|
* @author hashashiyyin [email protected] / [email protected] |
5
|
|
|
* Date: 02/08/24 |
6
|
|
|
* Time: 11:45 |
7
|
|
|
* |
8
|
|
|
*/ |
9
|
|
|
|
10
|
|
|
namespace Matecat\XliffParser\XliffReplacer; |
11
|
|
|
|
12
|
|
|
use Matecat\XliffParser\Utils\Strings; |
13
|
|
|
|
14
|
|
|
class Xliff12 extends AbstractXliffReplacer { |
15
|
|
|
|
16
|
|
|
/** |
17
|
|
|
* @var array |
18
|
|
|
*/ |
19
|
|
|
protected array $nodesToBuffer = [ |
20
|
|
|
'source', |
21
|
|
|
'seg-source', |
22
|
|
|
'note', |
23
|
|
|
'context', |
24
|
|
|
'context-group' |
25
|
|
|
]; |
26
|
|
|
|
27
|
|
|
/** |
28
|
|
|
* @inheritDoc |
29
|
|
|
*/ |
30
|
|
|
protected function tagOpen( $parser, $name, $attr ) { |
31
|
|
|
|
32
|
|
|
$this->handleOpenUnit( $name, $attr ); |
33
|
|
|
|
34
|
|
|
// check if we are entering into a <target> |
35
|
|
|
if ( 'target' === $name ) { |
36
|
|
|
|
37
|
|
|
if ( $this->currentTransUnitIsTranslatable === 'no' ) { |
38
|
|
|
$this->inTarget = false; |
39
|
|
|
} else { |
40
|
|
|
$this->inTarget = true; |
41
|
|
|
} |
42
|
|
|
} |
43
|
|
|
|
44
|
|
|
// open buffer |
45
|
|
|
if ( in_array( $name, $this->nodesToBuffer ) ) { |
46
|
|
|
$this->bufferIsActive = true; |
47
|
|
|
} |
48
|
|
|
|
49
|
|
|
// check if we are inside a <target>, obviously this happen only if there are targets inside the trans-unit |
50
|
|
|
// <target> must be stripped to be replaced, so this check avoids <target> reconstruction |
51
|
|
|
if ( !$this->inTarget ) { |
52
|
|
|
|
53
|
|
|
$tag = ''; |
54
|
|
|
|
55
|
|
|
// construct tag |
56
|
|
|
$tag .= "<$name "; |
57
|
|
|
|
58
|
|
|
foreach ( $attr as $k => $v ) { |
59
|
|
|
|
60
|
|
|
//if tag name is file, we must replace the target-language attribute |
61
|
|
|
if ( $name === 'file' && $k === 'target-language' && !empty( $this->targetLang ) ) { |
62
|
|
|
//replace Target language with job language provided from constructor |
63
|
|
|
$tag .= "$k=\"$this->targetLang\" "; |
64
|
|
|
} else { |
65
|
|
|
$tag .= "$k=\"$v\" "; |
66
|
|
|
} |
67
|
|
|
|
68
|
|
|
} |
69
|
|
|
|
70
|
|
|
$seg = $this->getCurrentSegment(); |
71
|
|
|
|
72
|
|
|
if ( $name === $this->tuTagName && !empty( $seg ) and isset( $seg[ 'sid' ] ) ) { |
|
|
|
|
73
|
|
|
|
74
|
|
|
if ( strpos( $tag, 'help-id' ) === false ) { |
75
|
|
|
if ( !empty( $seg[ 'sid' ] ) ) { |
76
|
|
|
$tag .= "help-id=\"{$seg[ 'sid' ]}\" "; |
77
|
|
|
} |
78
|
|
|
} |
79
|
|
|
|
80
|
|
|
} |
81
|
|
|
|
82
|
|
|
// Add MateCat specific namespace. |
83
|
|
|
// Add trgLang |
84
|
|
|
if ( $name === 'xliff' ) { |
85
|
|
|
if ( !array_key_exists( 'xmlns:mtc', $attr ) ) { |
86
|
|
|
$tag .= ' xmlns:mtc="https://www.matecat.com" '; |
87
|
|
|
} |
88
|
|
|
$tag = preg_replace( '/trgLang="(.*?)"/', 'trgLang="' . $this->targetLang . '"', $tag ); |
89
|
|
|
} |
90
|
|
|
|
91
|
|
|
$this->checkForSelfClosedTagAndFlush( $parser, $tag ); |
92
|
|
|
|
93
|
|
|
} |
94
|
|
|
|
95
|
|
|
} |
96
|
|
|
|
97
|
|
|
|
98
|
|
|
/** |
99
|
|
|
* @inheritDoc |
100
|
|
|
*/ |
101
|
|
|
protected function tagClose( $parser, $name ) { |
102
|
|
|
$tag = ''; |
103
|
|
|
|
104
|
|
|
/** |
105
|
|
|
* if is a tag within <target> or |
106
|
|
|
* if it is an empty tag, do not add closing tag because we have already closed it in |
107
|
|
|
* |
108
|
|
|
* self::tagOpen method |
109
|
|
|
*/ |
110
|
|
|
if ( !$this->isEmpty ) { |
111
|
|
|
|
112
|
|
|
if ( !$this->inTarget ) { |
113
|
|
|
$tag = "</$name>"; |
114
|
|
|
} |
115
|
|
|
|
116
|
|
|
if ( 'target' == $name ) { |
117
|
|
|
|
118
|
|
|
if ( isset( $this->transUnits[ $this->currentTransUnitId ] ) ) { |
119
|
|
|
|
120
|
|
|
// get translation of current segment, by indirect indexing: id -> positional index -> segment |
121
|
|
|
// actually there may be more than one segment to that ID if there are two mrk of the same source segment |
122
|
|
|
$tag = $this->rebuildTarget(); |
123
|
|
|
|
124
|
|
|
} |
125
|
|
|
|
126
|
|
|
$this->targetWasWritten = true; |
127
|
|
|
// signal we are leaving a target |
128
|
|
|
$this->inTarget = false; |
129
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag, true ); |
130
|
|
|
|
131
|
|
|
} elseif ( in_array( $name, $this->nodesToBuffer ) ) { // we are closing a critical CDATA section |
132
|
|
|
|
133
|
|
|
$this->bufferIsActive = false; |
134
|
|
|
$tag = $this->CDATABuffer . "</$name>"; |
135
|
|
|
$this->CDATABuffer = ""; |
136
|
|
|
|
137
|
|
|
//flush to the pointer |
138
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag ); |
139
|
|
|
|
140
|
|
|
} elseif ( $name === $this->tuTagName ) { |
141
|
|
|
|
142
|
|
|
$tag = ""; |
143
|
|
|
|
144
|
|
|
// handling </trans-unit> closure |
145
|
|
|
if ( !$this->targetWasWritten ) { |
146
|
|
|
|
147
|
|
|
if ( isset( $this->transUnits[ $this->currentTransUnitId ] ) ) { |
148
|
|
|
$tag = $this->rebuildTarget(); |
149
|
|
|
} else { |
150
|
|
|
$tag = $this->createTargetTag( "", "" ); |
151
|
|
|
} |
152
|
|
|
|
153
|
|
|
} |
154
|
|
|
|
155
|
|
|
$tag .= "</$this->tuTagName>"; |
156
|
|
|
$this->targetWasWritten = false; |
157
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag ); |
158
|
|
|
|
159
|
|
|
} elseif ( $this->bufferIsActive ) { // this is a tag ( <g | <mrk ) inside a seg or seg-source tag |
160
|
|
|
$this->CDATABuffer .= "</$name>"; |
161
|
|
|
// Do NOT Flush |
162
|
|
|
} else { //generic tag closure do Nothing |
163
|
|
|
// flush to pointer |
164
|
|
|
$this->postProcAndFlush( $this->outputFP, $tag ); |
165
|
|
|
} |
166
|
|
|
|
167
|
|
|
} else { |
168
|
|
|
//ok, nothing to be done; reset flag for next coming tag |
169
|
|
|
$this->isEmpty = false; |
170
|
|
|
} |
171
|
|
|
|
172
|
|
|
// check if we are leaving a <trans-unit> (xliff v1.*) or <unit> (xliff v2.*) |
173
|
|
|
if ( $this->tuTagName === $name ) { |
174
|
|
|
$this->currentTransUnitIsTranslatable = null; |
175
|
|
|
$this->inTU = false; |
176
|
|
|
$this->hasWrittenCounts = false; |
177
|
|
|
|
178
|
|
|
$this->resetCounts(); |
179
|
|
|
} |
180
|
|
|
} |
181
|
|
|
|
182
|
|
|
/** |
183
|
|
|
* prepare segment tagging for xliff insertion |
184
|
|
|
* |
185
|
|
|
* @param array $seg |
186
|
|
|
* @param string $transUnitTranslation |
187
|
|
|
* |
188
|
|
|
* @return string |
189
|
|
|
*/ |
190
|
|
|
protected function prepareTranslation( array $seg, string $transUnitTranslation = "" ): string { |
191
|
|
|
|
192
|
|
|
$segment = Strings::removeDangerousChars( $seg [ 'segment' ] ); |
193
|
|
|
$translation = Strings::removeDangerousChars( $seg [ 'translation' ] ); |
194
|
|
|
|
195
|
|
|
if ( $seg [ 'translation' ] == '' ) { |
196
|
|
|
$translation = $segment; |
197
|
|
|
} else { |
198
|
|
|
if ( $this->callback instanceof XliffReplacerCallbackInterface ) { |
199
|
|
|
$error = ( !empty( $seg[ 'error' ] ) ) ? $seg[ 'error' ] : null; |
200
|
|
|
if ( $this->callback->thereAreErrors( $seg[ 'sid' ], $segment, $translation, [], $error ) ) { |
201
|
|
|
$translation = '|||UNTRANSLATED_CONTENT_START|||' . $segment . '|||UNTRANSLATED_CONTENT_END|||'; |
202
|
|
|
} |
203
|
|
|
} |
204
|
|
|
} |
205
|
|
|
|
206
|
|
|
$transUnitTranslation .= $seg[ 'prev_tags' ] . $this->rebuildMarks( $seg, $translation ) . ltrim( $seg[ 'succ_tags' ] ); |
207
|
|
|
|
208
|
|
|
return $transUnitTranslation; |
209
|
|
|
} |
210
|
|
|
|
211
|
|
|
protected function rebuildMarks( array $seg, string $translation ): string { |
212
|
|
|
|
213
|
|
|
if ( $seg[ 'mrk_id' ] !== null && $seg[ 'mrk_id' ] != '' ) { |
214
|
|
|
if ( $this->targetLang === 'ja-JP' ) { |
215
|
|
|
$seg[ 'mrk_succ_tags' ] = ltrim( $seg[ 'mrk_succ_tags' ] ); |
216
|
|
|
} |
217
|
|
|
|
218
|
|
|
$translation = "<mrk mid=\"" . $seg[ 'mrk_id' ] . "\" mtype=\"seg\">" . $seg[ 'mrk_prev_tags' ] . $translation . $seg[ 'mrk_succ_tags' ] . "</mrk>"; |
219
|
|
|
} |
220
|
|
|
|
221
|
|
|
return $translation; |
222
|
|
|
|
223
|
|
|
} |
224
|
|
|
|
225
|
|
|
/** |
226
|
|
|
* This function creates a <target> |
227
|
|
|
* |
228
|
|
|
* @param string $translation |
229
|
|
|
* @param string $stateProp |
230
|
|
|
* |
231
|
|
|
* @return string |
232
|
|
|
*/ |
233
|
|
|
private function createTargetTag( string $translation, string $stateProp ): string { |
234
|
|
|
$targetLang = ' xml:lang="' . $this->targetLang . '"'; |
235
|
|
|
$tag = "<target $targetLang $stateProp>$translation</target>"; |
236
|
|
|
$tag .= "\n<count-group name=\"$this->currentTransUnitId\"><count count-type=\"x-matecat-raw\">" . $this->counts[ 'raw_word_count' ] . "</count><count count-type=\"x-matecat-weighted\">" . $this->counts[ 'eq_word_count' ] . '</count></count-group>'; |
237
|
|
|
|
238
|
|
|
return $tag; |
239
|
|
|
|
240
|
|
|
} |
241
|
|
|
|
242
|
|
|
protected function rebuildTarget(): string { |
243
|
|
|
|
244
|
|
|
// init translation and state |
245
|
|
|
$translation = ''; |
246
|
|
|
$lastMrkState = null; |
247
|
|
|
$stateProp = null; |
248
|
|
|
|
249
|
|
|
// we must reset the lastMrkId found because this is a new segment. |
250
|
|
|
$lastMrkId = -1; |
251
|
|
|
|
252
|
|
|
foreach ( $this->lastTransUnit as $pos => $seg ) { |
253
|
|
|
|
254
|
|
|
/* |
255
|
|
|
* This routine works to respect the positional orders of markers. |
256
|
|
|
* In every cycle we check if the mrk of the segment is below or equal the last one. |
257
|
|
|
* When this is true, means that the mrk id belongs to the next segment with the same internal_id |
258
|
|
|
* so we MUST stop to apply markers and translations |
259
|
|
|
* and stop to add eq_word_count |
260
|
|
|
* |
261
|
|
|
* Begin: |
262
|
|
|
* pre-assign zero to the new mrk if this is the first one ( in this segment ) |
263
|
|
|
* If it is null leave it NULL |
264
|
|
|
*/ |
265
|
|
|
if ( (int)$seg[ "mrk_id" ] < 0 && $seg[ "mrk_id" ] !== null ) { |
266
|
|
|
$seg[ "mrk_id" ] = 0; |
267
|
|
|
} |
268
|
|
|
|
269
|
|
|
/* |
270
|
|
|
* WARNING: |
271
|
|
|
* For those seg-source that doesn't have a mrk ( having a mrk id === null ) |
272
|
|
|
* ( null <= -1 ) === true |
273
|
|
|
* so, cast to int |
274
|
|
|
*/ |
275
|
|
|
if ( (int)$seg[ "mrk_id" ] <= $lastMrkId ) { |
276
|
|
|
break; |
277
|
|
|
} |
278
|
|
|
|
279
|
|
|
// update counts |
280
|
|
|
if ( !empty( $seg ) ) { |
281
|
|
|
$this->updateSegmentCounts( $seg ); |
282
|
|
|
} |
283
|
|
|
|
284
|
|
|
// delete translations so the prepareSegment |
285
|
|
|
// will put source content in target tag |
286
|
|
|
if ( $this->sourceInTarget ) { |
287
|
|
|
$seg[ 'translation' ] = ''; |
288
|
|
|
$this->resetCounts(); |
289
|
|
|
} |
290
|
|
|
|
291
|
|
|
// append $translation |
292
|
|
|
$translation = $this->prepareTranslation( $seg, $translation ); |
293
|
|
|
|
294
|
|
|
$lastMrkId = $seg[ "mrk_id" ]; |
295
|
|
|
|
296
|
|
|
[ $stateProp, $lastMrkState ] = StatusToStateAttribute::getState( $seg[ 'status' ], $this->xliffVersion, $stateProp, $lastMrkState ); |
297
|
|
|
|
298
|
|
|
} |
299
|
|
|
|
300
|
|
|
//append translation |
301
|
|
|
return $this->createTargetTag( $translation, $stateProp ); |
|
|
|
|
302
|
|
|
|
303
|
|
|
} |
304
|
|
|
|
305
|
|
|
} |
PHP has two types of connecting operators (logical operators, and boolean operators):
and
&&
or
||
The difference between these is the order in which they are executed. In most cases, you would want to use a boolean operator like
&&
, or||
.Let’s take a look at a few examples:
Logical Operators are used for Control-Flow
One case where you explicitly want to use logical operators is for control-flow such as this:
Since
die
introduces problems of its own, f.e. it makes our code hardly testable, and prevents any kind of more sophisticated error handling; you probably do not want to use this in real-world code. Unfortunately, logical operators cannot be combined withthrow
at this point:These limitations lead to logical operators rarely being of use in current PHP code.