1 | <?php |
||
2 | /* |
||
3 | * This file is part of dispositif/wikibot application (@github) |
||
4 | * 2019-2023 © Philippe M./Irønie <[email protected]> |
||
5 | * For the full copyright and MIT license information, view the license file. |
||
6 | */ |
||
7 | |||
8 | declare(strict_types=1); |
||
9 | |||
10 | namespace App\Domain\Utils; |
||
11 | |||
12 | use App\Domain\Models\Wiki\AbstractWikiTemplate; |
||
13 | use App\Domain\WikiTemplateFactory; |
||
14 | use Exception; |
||
15 | use LogicException; |
||
16 | use Throwable; |
||
17 | |||
18 | /** |
||
19 | * todo legacy. |
||
20 | * Class TemplateParser. |
||
21 | */ |
||
22 | abstract class TemplateParser extends WikiTextUtil |
||
23 | { |
||
24 | /** |
||
25 | * todo : simplify array if only one occurrence ? |
||
26 | * todo refac extract/logic. |
||
27 | * |
||
28 | * |
||
29 | * @return array |
||
30 | * @throws Exception |
||
31 | */ |
||
32 | public static function parseAllTemplateByName(string $tplName, string $text): array |
||
33 | { |
||
34 | 2 | $result = []; |
|
35 | // Extract wikiText from that template |
||
36 | $arrayTplText = self::findAllTemplatesByName($tplName, $text); |
||
37 | 2 | ||
38 | if ($arrayTplText === [] || empty($arrayTplText[0])) { |
||
39 | 2 | return []; |
|
40 | } |
||
41 | |||
42 | $result[$tplName] = []; |
||
43 | 2 | $inc = -1; |
|
44 | 2 | foreach ($arrayTplText as $tplText) { |
|
45 | 2 | ++$inc; |
|
46 | 2 | // store the raw text of the template |
|
47 | $result[$tplName][$inc] = ['raw' => $tplText]; |
||
48 | 2 | ||
49 | // create an object of the template |
||
50 | try { |
||
51 | $tplObject = WikiTemplateFactory::create($tplName); |
||
52 | } catch (Throwable $e) { |
||
53 | unset($e); |
||
54 | continue; |
||
55 | 2 | } |
|
56 | |||
57 | if (!is_object($tplObject) || !is_subclass_of($tplObject, AbstractWikiTemplate::class)) { |
||
58 | continue; |
||
59 | } |
||
60 | |||
61 | 2 | $data = self::parseDataFromTemplate($tplName, $tplText); |
|
62 | $tplObject->hydrate($data); |
||
63 | $tplObject->detectUserSeparator($tplText); |
||
64 | |||
65 | 2 | $result[$tplName][$inc] += ['model' => $tplObject]; |
|
66 | 2 | } |
|
67 | 2 | ||
68 | return (array)$result; |
||
69 | 2 | } |
|
70 | |||
71 | /** |
||
72 | 2 | * Find all the recurrences of a wiki's template in a text. |
|
73 | * Compatible with inclusion of sub-templates. |
||
74 | * Example : |
||
75 | * {{Infobox |pays={{pays|France}} }} |
||
76 | * retourne array {{modèle|...}}. |
||
77 | * |
||
78 | * @return array [ 0=>{{bla|...}}, 1=>{{bla|...}} ] |
||
79 | */ |
||
80 | public static function findAllTemplatesByName(string $templateName, string $text): array |
||
81 | { |
||
82 | // TODO check {{fr}} |
||
83 | $res = preg_match_all( |
||
84 | "#{{[ \n]*".preg_quote(trim($templateName), '#')."[ \t \n\r]*\|[^{}]*(?:{{[^{}]+}}[^{}]*)*}}#i", |
||
85 | $text, |
||
86 | $matches |
||
87 | 2 | ); |
|
88 | |||
89 | if (false === $res) { |
||
90 | 2 | return []; |
|
91 | 2 | } |
|
92 | 2 | ||
93 | 2 | return $matches[0]; |
|
94 | //OK : preg_match_all("#\{\{".preg_quote(trim($nommodele), '#')."[ \t \n\r]*\|([^\{\}]*(\{\{[^\{\}]+\}\}[^\{\}]*)*)\}\}#i", $text, $matches); |
||
95 | } |
||
96 | 2 | ||
97 | /** |
||
98 | * todo refactor + check if @notused |
||
99 | * Parsing of any wiki template from text and templateName |
||
100 | 2 | * Using the first {{template}} definition found in text |
|
101 | * todo legacy. |
||
102 | * |
||
103 | * |
||
104 | * @return array |
||
105 | */ |
||
106 | public static function parseDataFromTemplate(string $tplName, string $text): array |
||
107 | { |
||
108 | $text = str_replace("\n", '', $text); // todo WTF ? |
||
109 | |||
110 | // check {{template}} in text |
||
111 | $tplFounded = self::findFirstTemplateInText($tplName, $text); |
||
112 | |||
113 | // $matches[0] : {{template|...}} |
||
114 | if (empty($tplFounded)) { |
||
115 | 27 | throw new LogicException("Template $tplName not found in text"); |
|
116 | } |
||
117 | 27 | // $matches[1] : url=blabla|titre=Popo |
|
118 | if (false === $tplFounded[1]) { |
||
119 | throw new LogicException("No parameters found in $tplName"); |
||
120 | 27 | } |
|
121 | // sub-template pipe | encoding |
||
122 | $tplFounded[1] = self::encodeTemplatePipes($tplFounded[1]); |
||
123 | 27 | ||
124 | // x flag => "\ " for space |
||
125 | $res = preg_match_all( |
||
126 | "/ |
||
127 | 27 | ( |
|
128 | [^|=]*=? # parameter name (or nothing) |
||
129 | ( |
||
130 | [^|{}\[\]<>]* # reject <i>,<ref> |
||
131 | 27 | (?:\[[^\[\]]+])? # [url text] or [text] |
|
132 | (?:<!--[^<>]+-->)? # commentary <!-- --> |
||
133 | (?:{{[^}{]+}})? # {{template}} but KO with {{tmp|...}} |
||
134 | 27 | # test : {{bla@PIPE@bla}} |
|
135 | 27 | (?:\[\[[^]]+]])? # [[fu|bar]] |
|
136 | [^|{}\[\]]* # accept <i>,<ref> |
||
137 | )* |
||
138 | )\|? |
||
139 | /x", |
||
140 | $tplFounded[1], |
||
141 | $wikiParams |
||
142 | ); |
||
143 | |||
144 | if (false === $res || 0 === $res || empty($wikiParams[1])) { |
||
145 | throw new LogicException("Parameters from template '$tplName' can't be parsed"); |
||
146 | } |
||
147 | |||
148 | return self::explodeParameterValue($wikiParams[1]); |
||
149 | 27 | } |
|
150 | 27 | ||
151 | /** |
||
152 | * For multiple occurrences see findAllTemplatesByName(). |
||
153 | 27 | * |
|
154 | * |
||
155 | * @return array|null |
||
156 | */ |
||
157 | 27 | private static function findFirstTemplateInText(string $templateName, string $text): ?array |
|
158 | { |
||
159 | // BUG marche pas avec : |
||
160 | // $text = '{{Ouvrage|auteur1 = Clément|titre = Les Borgia {{nobr|Alexandre {{VI}}}}}}'; // to debug |
||
161 | // $templateName = 'ouvrage'; // to debug |
||
162 | |||
163 | // $text = str_replace("\n", '', $text); // ??? todo regex multiline or encode char |
||
164 | |||
165 | // todo: replace <!-- --> by encode char and memorize in var |
||
166 | |||
167 | // hack : replace solitary { and } by encoded string CURLYBRACKET |
||
168 | 27 | $text = preg_replace('#([^{]){([^{])#', '${1}CURLYBRACKETO$2', $text); |
|
169 | $text = preg_replace('#([^}])}([^}])#', '${1}CURLYBRACKETC$2', $text); |
||
170 | |||
171 | // TODO: implement better regex :( |
||
172 | if (preg_match( |
||
173 | '~{{ ?'.preg_quote($templateName, '~')."[ \t \n\r]*\|([^{}]*(?:{{[^{}]+}}[^{}]*)*)}}~i", |
||
174 | $text, |
||
175 | $matches |
||
176 | ) > 0 |
||
177 | ) { |
||
178 | array_walk( |
||
179 | 27 | $matches, |
|
180 | 27 | function (&$value) { |
|
181 | $value = str_replace(['CURLYBRACKETO', 'CURLYBRACKETC'], ['{', '}'], $value); |
||
182 | } |
||
183 | 27 | ); |
|
184 | 27 | ||
185 | 27 | return $matches; |
|
186 | 27 | } |
|
187 | 27 | ||
188 | return null; |
||
189 | 27 | } |
|
190 | 27 | ||
191 | /** |
||
192 | 27 | * replace sub-templates pipes | by @PIPE@ in text. |
|
193 | 27 | */ |
|
194 | protected static function encodeTemplatePipes(string $text): string |
||
195 | { |
||
196 | 27 | if (preg_match_all('#{{(?:[^{}]+)}}#m', $text, $subTmpl) > 0) { |
|
197 | foreach ($subTmpl[0] as $sub) { |
||
198 | $subSanit = str_replace('|', '@PIPE@', (string) $sub); |
||
199 | $text = str_replace($sub, $subSanit, $text); |
||
200 | } |
||
201 | } |
||
202 | |||
203 | return $text; |
||
204 | } |
||
205 | |||
206 | /** |
||
207 | * From ['fr', 'url=blabla', 'titre=popo'] |
||
208 | * To [ '1'=> 'fr', url' => 'blabla', 'titre' => 'popo' ]. |
||
209 | 27 | * |
|
210 | * @param array $wikiLines ['url=blabla', 'titre=popo'] |
||
211 | 27 | * |
|
212 | 1 | * @return array |
|
213 | 1 | */ |
|
214 | 1 | protected static function explodeParameterValue(array $wikiLines): array |
|
215 | { |
||
216 | $data = []; |
||
217 | $keyNum = 1; |
||
218 | 27 | foreach ($wikiLines as $line) { |
|
219 | if (empty($line)) { |
||
220 | continue; |
||
221 | } |
||
222 | $line = str_replace( |
||
223 | ["\t", "\n", "\r", ' '], |
||
224 | ['', '', '', ' '], |
||
225 | (string) $line |
||
226 | ); // perte cosmétique : est-ce bien ? + espace insécable remplacé par espace sécable |
||
227 | |||
228 | // $line : fu = bar (OK : fu=bar=coco) |
||
229 | 27 | $pos = strpos($line, '='); |
|
230 | $param = null; |
||
231 | 27 | if (is_int($pos) && $pos >= 0) { |
|
232 | 27 | $param = mb_strtolower(substr($line, 0, $pos), 'UTF-8'); |
|
233 | 27 | $value = substr($line, $pos + 1); |
|
234 | 27 | } |
|
235 | 27 | // No param name => take $keyNum as param name |
|
236 | if (false === $pos) { |
||
237 | 27 | $param = (string)$keyNum; |
|
238 | 27 | $value = $line; |
|
239 | 27 | ++$keyNum; |
|
240 | 27 | } |
|
241 | |||
242 | if (empty($param) || !isset($value)) { |
||
243 | throw new LogicException('param/value variable not defined'); |
||
244 | 27 | } |
|
245 | 27 | ||
246 | 26 | // TODO : accept empty value ? |
|
247 | 26 | if (trim($value) === '') { |
|
0 ignored issues
–
show
Comprehensibility
Best Practice
introduced
by
![]() |
|||
248 | continue; |
||
249 | } |
||
250 | 27 | // reverse the sub-template pipe encoding |
|
251 | 2 | $value = str_replace('@PIPE@', '|', $value); |
|
252 | 2 | $data[trim($param)] = trim($value); |
|
253 | 2 | } |
|
254 | |||
255 | return $data; |
||
256 | 27 | } |
|
257 | |||
258 | /** |
||
259 | * Find text style of template : only pipe, space-pipe-space, pipe-space, return-pipe, etc. |
||
260 | */ |
||
261 | 27 | public static function findUserStyleSeparator(string $tplText): string |
|
262 | 5 | { |
|
263 | // Fixed : {{fu\n | bar}} |
||
264 | if (preg_match('#{{[^}|]+\n +\|( ?)[^}]+}}#i', $tplText, $matches) > 0) { |
||
265 | 27 | return "\n |".$matches[1]; |
|
266 | 27 | } |
|
267 | // {{fu | bar}} (duplicate : because [^}|\n]+ allows final space...) |
||
268 | if (preg_match('#{{[^}|\n]+([ \n]\|[ \n]?)[^}]+}}#i', $tplText, $matches) > 0) { |
||
269 | 27 | return $matches[1]; |
|
270 | } |
||
271 | // others : {{fu|bar}} ; {{fu\n|bar}} ; {{fu |bar}} ... |
||
272 | if (preg_match('#{{[^}|\n]+([ \n]?\|[ \n]?)[^}]+}}#i', $tplText, $matches) > 0) { |
||
273 | return $matches[1]; |
||
274 | } |
||
275 | |||
276 | return ' |'; |
||
277 | } |
||
278 | |||
279 | 31 | /** |
|
280 | * Detect if "param = bla". |
||
281 | */ |
||
282 | 31 | public static function isMultispacedTemplate(string $tplText): bool |
|
283 | 2 | { |
|
284 | // detect 4 spaces chars |
||
285 | return (bool) preg_match('#{{[^}]+ {4}[^}]+}}#i', $tplText); |
||
286 | 29 | } |
|
287 | 6 | ||
288 | /** |
||
289 | * https://fr.wikipedia.org/wiki/Mod%C3%A8le:P. |
||
290 | 24 | * Examples: |
|
291 | 24 | * 'bla {{p.|125-133}} bla' => ['{{p.|125-133}}', '125-133'] |
|
292 | * 'bla {{p.}}10, 20, 35-36 bla' => ['{{p.}}10, 20, 35-36', '10, 20, 35-36'] |
||
293 | */ |
||
294 | public static function extractPageTemplateContent(string $text): ?array |
||
295 | { |
||
296 | if (preg_match('#\{\{p\.(?:\|([0-9,\-—\/ ]+))?\}\}([0-9,\-—\/ ]+)?#i', $text, $matches) !== false) { |
||
297 | if (!empty($matches[1]) && trim($matches[1]) !== '') { // {{p.|125}} |
||
298 | return [trim($matches[0]), trim($matches[1])]; |
||
299 | } |
||
300 | if (!empty($matches[2]) && trim($matches[2]) !== '') { // {{p.}}125 |
||
301 | return [trim($matches[0]), trim($matches[2])]; |
||
302 | } |
||
303 | } |
||
304 | 25 | ||
305 | return null; |
||
306 | } |
||
307 | } |
||
308 |