Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
0.00% |
0 / 257 |
|
0.00% |
0 / 11 |
CRAP | |
0.00% |
0 / 1 |
| RestConnector | |
0.00% |
0 / 257 |
|
0.00% |
0 / 11 |
4556 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
| query | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
6 | |||
| getRecord | |
0.00% |
0 / 8 |
|
0.00% |
0 / 1 |
6 | |||
| getInstitutionCode | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| performSearch | |
0.00% |
0 / 45 |
|
0.00% |
0 / 1 |
462 | |||
| call | |
0.00% |
0 / 37 |
|
0.00% |
0 / 1 |
90 | |||
| processResponse | |
0.00% |
0 / 93 |
|
0.00% |
0 / 1 |
342 | |||
| processHighlighting | |
0.00% |
0 / 29 |
|
0.00% |
0 / 1 |
30 | |||
| processDescription | |
0.00% |
0 / 7 |
|
0.00% |
0 / 1 |
6 | |||
| getJWT | |
0.00% |
0 / 14 |
|
0.00% |
0 / 1 |
30 | |||
| getUrl | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| 1 | <?php |
| 2 | |
| 3 | /** |
| 4 | * Primo Central connector (REST API). |
| 5 | * |
| 6 | * PHP version 8 |
| 7 | * |
| 8 | * Copyright (C) Villanova University 2010. |
| 9 | * Copyright (C) The National Library of Finland 2023. |
| 10 | * |
| 11 | * This program is free software; you can redistribute it and/or modify |
| 12 | * it under the terms of the GNU General Public License version 2, |
| 13 | * as published by the Free Software Foundation. |
| 14 | * |
| 15 | * This program is distributed in the hope that it will be useful, |
| 16 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 17 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 18 | * GNU General Public License for more details. |
| 19 | * |
| 20 | * You should have received a copy of the GNU General Public License |
| 21 | * along with this program; if not, write to the Free Software |
| 22 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 23 | * |
| 24 | * @category VuFind |
| 25 | * @package Search |
| 26 | * @author Spencer Lamm <slamm1@swarthmore.edu> |
| 27 | * @author Anna Headley <aheadle1@swarthmore.edu> |
| 28 | * @author Chelsea Lobdell <clobdel1@swarthmore.edu> |
| 29 | * @author Demian Katz <demian.katz@villanova.edu> |
| 30 | * @author Ere Maijala <ere.maijala@helsinki.fi> |
| 31 | * @author Oliver Goldschmidt <o.goldschmidt@tuhh.de> |
| 32 | * @license http://opensource.org/licenses/gpl-2.0.php GNU General Public License |
| 33 | * @link https://vufind.org |
| 34 | */ |
| 35 | |
| 36 | namespace VuFindSearch\Backend\Primo; |
| 37 | |
| 38 | use Laminas\Session\Container as SessionContainer; |
| 39 | |
| 40 | use function array_key_exists; |
| 41 | use function in_array; |
| 42 | use function is_array; |
| 43 | use function strlen; |
| 44 | |
| 45 | /** |
| 46 | * Primo Central connector (REST API). |
| 47 | * |
| 48 | * @category VuFind |
| 49 | * @package Search |
| 50 | * @author Spencer Lamm <slamm1@swarthmore.edu> |
| 51 | * @author Anna Headley <aheadle1@swarthmore.edu> |
| 52 | * @author Chelsea Lobdell <clobdel1@swarthmore.edu> |
| 53 | * @author Demian Katz <demian.katz@villanova.edu> |
| 54 | * @author Ere Maijala <ere.maijala@helsinki.fi> |
| 55 | * @author Oliver Goldschmidt <o.goldschmidt@tuhh.de> |
| 56 | * @license http://opensource.org/licenses/gpl-2.0.php GNU General Public License |
| 57 | * @link https://vufind.org |
| 58 | */ |
| 59 | class RestConnector implements ConnectorInterface, \Laminas\Log\LoggerAwareInterface |
| 60 | { |
| 61 | use \VuFind\Log\LoggerAwareTrait; |
| 62 | use \VuFindSearch\Backend\Feature\ConnectorCacheTrait; |
| 63 | |
| 64 | /** |
| 65 | * HTTP client factory |
| 66 | * |
| 67 | * @var callable |
| 68 | */ |
| 69 | protected $clientFactory; |
| 70 | |
| 71 | /** |
| 72 | * Primo JWT API URL |
| 73 | * |
| 74 | * @var string |
| 75 | */ |
| 76 | protected $jwtUrl; |
| 77 | |
| 78 | /** |
| 79 | * Primo REST API search URL |
| 80 | * |
| 81 | * @var string |
| 82 | */ |
| 83 | protected $searchUrl; |
| 84 | |
| 85 | /** |
| 86 | * Institution code |
| 87 | * |
| 88 | * @var string |
| 89 | */ |
| 90 | protected $inst; |
| 91 | |
| 92 | /** |
| 93 | * Session container |
| 94 | * |
| 95 | * @var SessionContainer |
| 96 | */ |
| 97 | protected $session; |
| 98 | |
| 99 | /** |
| 100 | * Response for an empty search |
| 101 | * |
| 102 | * @var array |
| 103 | */ |
| 104 | protected static $emptyQueryResponse = [ |
| 105 | 'recordCount' => 0, |
| 106 | 'documents' => [], |
| 107 | 'facets' => [], |
| 108 | 'error' => 'empty_search_disallowed', |
| 109 | ]; |
| 110 | |
| 111 | /** |
| 112 | * Mappings from VuFind index names to Primo |
| 113 | * |
| 114 | * @var array |
| 115 | */ |
| 116 | protected $indexMappings = [ |
| 117 | 'AllFields' => 'any', |
| 118 | 'Title' => 'title', |
| 119 | 'Author' => 'creator', |
| 120 | 'Subject' => 'sub', |
| 121 | 'Abstract' => 'desc', |
| 122 | 'ISSN' => 'issn', |
| 123 | ]; |
| 124 | |
| 125 | /** |
| 126 | * Legacy sort mappings |
| 127 | * |
| 128 | * @var array |
| 129 | */ |
| 130 | protected $sortMappings = [ |
| 131 | 'scdate' => 'date', |
| 132 | 'screator' => 'author', |
| 133 | 'stitle' => 'title', |
| 134 | ]; |
| 135 | |
| 136 | /** |
| 137 | * Constructor |
| 138 | * |
| 139 | * Sets up the Primo API Client |
| 140 | * |
| 141 | * @param string $jwtUrl Primo JWT API URL |
| 142 | * @param string $searchUrl Primo REST API search URL |
| 143 | * @param string $instCode Institution code (used as view ID, i.e. the |
| 144 | * vid parameter unless specified in the URL) |
| 145 | * @param callable $clientFactory HTTP client factory |
| 146 | * @param SessionContainer $session Session container |
| 147 | */ |
| 148 | public function __construct( |
| 149 | string $jwtUrl, |
| 150 | string $searchUrl, |
| 151 | string $instCode, |
| 152 | callable $clientFactory, |
| 153 | SessionContainer $session |
| 154 | ) { |
| 155 | $this->jwtUrl = $jwtUrl; |
| 156 | $this->searchUrl = $searchUrl; |
| 157 | $this->inst = $instCode; |
| 158 | |
| 159 | $this->clientFactory = $clientFactory; |
| 160 | $this->session = $session; |
| 161 | } |
| 162 | |
| 163 | /** |
| 164 | * Execute a search. Adds all the querystring parameters into |
| 165 | * $this->client and returns the parsed response |
| 166 | * |
| 167 | * @param string $institution Institution |
| 168 | * @param array $terms Associative array: |
| 169 | * index string: primo index to search (default "any") |
| 170 | * lookfor string: actual search terms |
| 171 | * @param array $params Associative array of optional arguments: |
| 172 | * phrase bool: true if it's a quoted phrase (default false) |
| 173 | * onCampus bool: (default true) |
| 174 | * didyoumean bool: (default false) |
| 175 | * filterList array: (field, value) pairs to filter results (def null) |
| 176 | * pageNumber string: index of first record (default 1) |
| 177 | * limit string: number of records to return (default 20) |
| 178 | * sort string: value to be used by for sorting (default null) |
| 179 | * highlight bool: whether to highlight search term matches in records |
| 180 | * highlightStart string: Prefix for a highlighted term |
| 181 | * highlightEnd string: Suffix for a Highlighted term |
| 182 | * Anything in $params not listed here will be ignored. |
| 183 | * |
| 184 | * Note: some input parameters accepted by Primo are not implemented here: |
| 185 | * - dym (did you mean) |
| 186 | * - more (get more) |
| 187 | * - lang (specify input language so engine can do lang. recognition) |
| 188 | * - displayField (has to do with highlighting somehow) |
| 189 | * |
| 190 | * @throws \Exception |
| 191 | * @return array An array of query results |
| 192 | * |
| 193 | * @link http://www.exlibrisgroup.org/display/PrimoOI/Brief+Search |
| 194 | */ |
| 195 | public function query($institution, $terms, $params = null) |
| 196 | { |
| 197 | // defaults for params |
| 198 | $args = [ |
| 199 | 'phrase' => false, |
| 200 | 'onCampus' => true, |
| 201 | 'didYouMean' => false, |
| 202 | 'filterList' => null, |
| 203 | 'pcAvailability' => false, |
| 204 | 'pageNumber' => 1, |
| 205 | 'limit' => 20, |
| 206 | 'sort' => null, |
| 207 | 'highlight' => false, |
| 208 | 'highlightStart' => '', |
| 209 | 'highlightEnd' => '', |
| 210 | ]; |
| 211 | if (isset($params)) { |
| 212 | $args = array_merge($args, $params); |
| 213 | } |
| 214 | // Ensure limit is at least 1 since Primo seems to be flaky with 0: |
| 215 | $args['limit'] = max(1, $args['limit']); |
| 216 | |
| 217 | return $this->performSearch($terms, $args); |
| 218 | } |
| 219 | |
| 220 | /** |
| 221 | * Retrieves a document specified by the ID. |
| 222 | * |
| 223 | * @param string $recordId The document to retrieve from the Primo API |
| 224 | * @param ?string $inst_code Institution code (optional) |
| 225 | * @param bool $onCampus Whether the user is on campus |
| 226 | * |
| 227 | * @throws \Exception |
| 228 | * @return array An array of query results |
| 229 | */ |
| 230 | public function getRecord(string $recordId, $inst_code = null, $onCampus = false) |
| 231 | { |
| 232 | if ('' === $recordId) { |
| 233 | return self::$emptyQueryResponse; |
| 234 | } |
| 235 | // Query String Parameters |
| 236 | $qs = []; |
| 237 | // It would be tempting to use 'exact' matching here, but it does not work |
| 238 | // with all record IDs, so need to use 'contains'. Contrary to the old |
| 239 | // brief search API, quotes are necessary here for all IDs to work. |
| 240 | $qs['q'] = 'rid,contains,"' . str_replace(';', ' ', $recordId) . '"'; |
| 241 | $qs['offset'] = '0'; |
| 242 | $qs['limit'] = '1'; |
| 243 | // pcAvailability=true is needed for records, which |
| 244 | // are NOT in the PrimoCentral Holdingsfile. |
| 245 | // It won't hurt to have this parameter always set to true. |
| 246 | // But it'd hurt to have it not set in case you want to get |
| 247 | // a record, which is not in the Holdingsfile. |
| 248 | $qs['pcAvailability'] = 'true'; |
| 249 | |
| 250 | return $this->processResponse($this->call(http_build_query($qs))); |
| 251 | } |
| 252 | |
| 253 | /** |
| 254 | * Get the institution code based on user IP. If user is coming from |
| 255 | * off campus return |
| 256 | * |
| 257 | * @return string |
| 258 | */ |
| 259 | public function getInstitutionCode() |
| 260 | { |
| 261 | return $this->inst; |
| 262 | } |
| 263 | |
| 264 | /** |
| 265 | * Support method for query() -- perform inner search logic |
| 266 | * |
| 267 | * @param array $terms Associative array: |
| 268 | * index string: primo index to search (default "any") |
| 269 | * lookfor string: actual search terms |
| 270 | * @param array $args Associative array of optional arguments (see query method for more information) |
| 271 | * |
| 272 | * @throws \Exception |
| 273 | * @return array An array of query results |
| 274 | */ |
| 275 | protected function performSearch($terms, $args) |
| 276 | { |
| 277 | // we have to build a querystring because I think adding them |
| 278 | // incrementally is implemented as a dictionary, but we are allowed |
| 279 | // multiple querystring parameters with the same key. |
| 280 | $qs = []; |
| 281 | |
| 282 | // QUERYSTRING: query (search terms) |
| 283 | // re: phrase searches, turns out we can just pass whatever we got |
| 284 | // to primo and they will interpret it correctly. |
| 285 | // leaving this flag in b/c it's not hurting anything, but we |
| 286 | // don't currently have a situation where we need to use "exact" |
| 287 | $precision = $args['phrase'] ? 'exact' : 'contains'; |
| 288 | |
| 289 | $primoQuery = []; |
| 290 | if (is_array($terms)) { |
| 291 | foreach ($terms as $thisTerm) { |
| 292 | $lookfor = str_replace(';', ' ', $thisTerm['lookfor']); |
| 293 | if (!$lookfor) { |
| 294 | continue; |
| 295 | } |
| 296 | // Set the index to search |
| 297 | $index = $this->indexMappings[$thisTerm['index']] ?? 'any'; |
| 298 | |
| 299 | // Set precision |
| 300 | if (array_key_exists('op', $thisTerm) && !empty($thisTerm['op'])) { |
| 301 | $precision = $thisTerm['op']; |
| 302 | } |
| 303 | |
| 304 | $primoQuery[] = "$index,$precision,$lookfor"; |
| 305 | } |
| 306 | } |
| 307 | |
| 308 | // Return if we don't have any query terms: |
| 309 | if (!$primoQuery && empty($args['filterList'])) { |
| 310 | return self::$emptyQueryResponse; |
| 311 | } |
| 312 | |
| 313 | if ($primoQuery) { |
| 314 | $qs['q'] = implode(';', $primoQuery); |
| 315 | } |
| 316 | |
| 317 | // QUERYSTRING: query (filter list) |
| 318 | // Date-related TODO: |
| 319 | // - provide additional support / processing for [x to y] limits? |
| 320 | if (!empty($args['filterList'])) { |
| 321 | $multiFacets = []; |
| 322 | $qInclude = []; |
| 323 | $qExclude = []; |
| 324 | foreach ($args['filterList'] as $current) { |
| 325 | $facet = $current['field']; |
| 326 | $facetOp = $current['facetOp']; |
| 327 | $values = $current['values']; |
| 328 | |
| 329 | foreach ($values as $value) { |
| 330 | if ('OR' === $facetOp) { |
| 331 | $multiFacets[] = "facet_$facet,include,$value"; |
| 332 | } elseif ('NOT' === $facetOp) { |
| 333 | $qExclude[] = "facet_$facet,exact,$value"; |
| 334 | } else { |
| 335 | $qInclude[] = "facet_$facet,exact,$value"; |
| 336 | } |
| 337 | } |
| 338 | } |
| 339 | if ($multiFacets) { |
| 340 | $qs['multiFacets'] = implode('|,|', $multiFacets); |
| 341 | } |
| 342 | if ($qInclude) { |
| 343 | $qs['qInclude'] = implode('|,|', $qInclude); |
| 344 | } |
| 345 | if ($qExclude) { |
| 346 | $qs['qExclude'] = implode('|,|', $qExclude); |
| 347 | } |
| 348 | } |
| 349 | |
| 350 | // QUERYSTRING: pcAvailability |
| 351 | // by default, Primo Central only returns matches, |
| 352 | // which are available via Holdingsfile |
| 353 | // pcAvailability = false |
| 354 | // By setting this value to true, also matches, which |
| 355 | // are NOT available via Holdingsfile are returned |
| 356 | // (yes, right, set this to true - that's ExLibris Logic) |
| 357 | if ($args['pcAvailability']) { |
| 358 | $qs['pcAvailability'] = 'true'; |
| 359 | } |
| 360 | |
| 361 | // QUERYSTRING: offset and limit |
| 362 | $recordStart = ($args['pageNumber'] - 1) * $args['limit']; |
| 363 | $qs['offset'] = $recordStart; |
| 364 | $qs['limit'] = $args['limit']; |
| 365 | |
| 366 | // QUERYSTRING: sort |
| 367 | // Possible values are rank (default), title, author or date. |
| 368 | $sort = $args['sort'] ?? null; |
| 369 | if ($sort && 'relevance' !== $sort) { |
| 370 | // Map legacy sort options: |
| 371 | $qs['sort'] = $this->sortMappings[$sort] ?? $sort; |
| 372 | } |
| 373 | |
| 374 | return $this->processResponse($this->call(http_build_query($qs)), $args); |
| 375 | } |
| 376 | |
| 377 | /** |
| 378 | * Small wrapper for sendRequest, process to simplify error handling. |
| 379 | * |
| 380 | * @param string $qs Query string |
| 381 | * |
| 382 | * @return string Result body |
| 383 | * @throws \Exception |
| 384 | */ |
| 385 | protected function call(string $qs): string |
| 386 | { |
| 387 | $url = $this->getUrl($this->searchUrl); |
| 388 | $url .= (str_contains($url, '?') ? '&' : '?') . $qs; |
| 389 | $this->debug("GET: $url"); |
| 390 | $client = ($this->clientFactory)($url); |
| 391 | $client->setMethod('GET'); |
| 392 | // Check cache: |
| 393 | $resultBody = null; |
| 394 | $cacheKey = null; |
| 395 | if ($this->cache) { |
| 396 | $cacheKey = $this->getCacheKey($client); |
| 397 | $resultBody = $this->getCachedData($cacheKey); |
| 398 | } |
| 399 | if (null === $resultBody) { |
| 400 | if ($jwt = $this->getJWT()) { |
| 401 | $client->setHeaders( |
| 402 | [ |
| 403 | 'Authorization' => [ |
| 404 | "Bearer $jwt", |
| 405 | ], |
| 406 | ] |
| 407 | ); |
| 408 | } |
| 409 | // Send request: |
| 410 | $result = $client->send(); |
| 411 | if ($jwt && $result->getStatusCode() === 403) { |
| 412 | // Reset JWT and try again: |
| 413 | $jwt = $this->getJWT(true); |
| 414 | $client->setHeaders( |
| 415 | [ |
| 416 | 'Authorization' => [ |
| 417 | "Bearer $jwt", |
| 418 | ], |
| 419 | ] |
| 420 | ); |
| 421 | $result = $client->send(); |
| 422 | } |
| 423 | $resultBody = $result->getBody(); |
| 424 | if (!$result->isSuccess()) { |
| 425 | $this->logError("Request $url failed with error code " . $result->getStatusCode() . ": $resultBody"); |
| 426 | throw new \Exception($resultBody); |
| 427 | } |
| 428 | if ($cacheKey) { |
| 429 | $this->putCachedData($cacheKey, $resultBody); |
| 430 | } |
| 431 | } |
| 432 | return $resultBody; |
| 433 | } |
| 434 | |
| 435 | /** |
| 436 | * Translate Primo's JSON into array of arrays. |
| 437 | * |
| 438 | * @param string $data The raw xml from Primo |
| 439 | * @param array $params Request parameters |
| 440 | * |
| 441 | * @return array The processed response from Primo |
| 442 | */ |
| 443 | protected function processResponse(string $data, array $params = []): array |
| 444 | { |
| 445 | // Make sure data exists |
| 446 | if ('' === $data) { |
| 447 | throw new \Exception('Primo did not return any data'); |
| 448 | } |
| 449 | |
| 450 | // Parse API response |
| 451 | $response = json_decode($data); |
| 452 | |
| 453 | if (false === $response) { |
| 454 | throw new \Exception('Error while parsing Primo response'); |
| 455 | } |
| 456 | |
| 457 | $totalhits = (int)$response->info->total; |
| 458 | $items = []; |
| 459 | foreach ($response->docs as $doc) { |
| 460 | $item = []; |
| 461 | $pnx = $doc->pnx; |
| 462 | $addata = $pnx->addata; |
| 463 | $control = $pnx->control; |
| 464 | $display = $pnx->display; |
| 465 | $search = $pnx->search; |
| 466 | $item['recordid'] = substr($control->recordid[0], 3); |
| 467 | $item['title'] = $display->title[0] ?? ''; |
| 468 | $item['format'] = $display->type ?? []; |
| 469 | // creators (use the search fields instead of display to get them as an array instead of a long string) |
| 470 | if ($search->creator ?? null) { |
| 471 | $item['creator'] = array_map('trim', $search->creator); |
| 472 | } |
| 473 | // subjects (use the search fields instead of display to get them as an array instead of a long string) |
| 474 | if ($search->subject ?? null) { |
| 475 | $item['subjects'] = $search->subject; |
| 476 | } |
| 477 | $item['ispartof'] = $display->ispartof[0] ?? ''; |
| 478 | $item['description'] = $display->description[0] |
| 479 | ?? $search->description[0] |
| 480 | ?? ''; |
| 481 | // and the rest! |
| 482 | $item['language'] = $display->language[0] ?? ''; |
| 483 | $item['source'] = implode('; ', $display->source ?? []); |
| 484 | $item['identifier'] = $display->identifier[0] ?? ''; |
| 485 | $item['fulltext'] = $pnx->delivery->fulltext[0] ?? ''; |
| 486 | $item['issn'] = $search->issn ?? []; |
| 487 | $item['publisher'] = $display->publisher ?? []; |
| 488 | $item['peer_reviewed'] = ($display->lds50[0] ?? '') === 'peer_reviewed'; |
| 489 | $openurl = $pnx->links->openurl[0] ?? ''; |
| 490 | $item['url'] = $openurl && !str_starts_with($openurl, '$') |
| 491 | ? $openurl |
| 492 | : ($pnx->GetIt2->link ?? ''); |
| 493 | |
| 494 | $processCitations = function (array $data): array { |
| 495 | return array_map( |
| 496 | function ($s) { |
| 497 | return "cdi_$s"; |
| 498 | }, |
| 499 | $data |
| 500 | ); |
| 501 | }; |
| 502 | |
| 503 | // These require the cdi_ prefix in search, so add it right away: |
| 504 | $item['cites'] = $processCitations($display->cites ?? []); |
| 505 | $item['cited_by'] = $processCitations($display->citedby ?? []); |
| 506 | |
| 507 | // Container data |
| 508 | $item['container_title'] = $addata->jtitle[0] ?? ''; |
| 509 | $item['container_volume'] = $addata->volume[0] ?? ''; |
| 510 | $item['container_issue'] = $addata->issue[0] ?? ''; |
| 511 | $item['container_start_page'] = $addata->spage[0] ?? ''; |
| 512 | $item['container_end_page'] = $addata->epage[0] ?? ''; |
| 513 | foreach ($addata->eissn ?? [] as $eissn) { |
| 514 | if (!in_array($eissn, $item['issn'])) { |
| 515 | $item['issn'][] = $eissn; |
| 516 | } |
| 517 | } |
| 518 | foreach ($addata->issn ?? [] as $issn) { |
| 519 | if (!in_array($issn, $item['issn'])) { |
| 520 | $item['issn'][] = $issn; |
| 521 | } |
| 522 | } |
| 523 | $item['doi_str_mv'] = $addata->doi ?? []; |
| 524 | |
| 525 | // Remove dash-less ISSNs if there are corresponding dashed ones |
| 526 | // (We could convert dash-less ISSNs to dashed ones, but try to stay |
| 527 | // true to the metadata) |
| 528 | $callback = function ($issn) use ($item) { |
| 529 | return strlen($issn) != 8 |
| 530 | || !in_array( |
| 531 | substr($issn, 0, 4) . '-' . substr($issn, 4), |
| 532 | $item['issn'] |
| 533 | ); |
| 534 | }; |
| 535 | $item['issn'] = array_values(array_filter($item['issn'], $callback)); |
| 536 | |
| 537 | $this->processHighlighting($item, $params, $response->highlights); |
| 538 | |
| 539 | // Fix description now that highlighting is done: |
| 540 | $item['description'] = $this->processDescription($item['description']); |
| 541 | |
| 542 | $item['fullrecord'] = json_decode(json_encode($pnx), true); |
| 543 | $items[] = $item; |
| 544 | } |
| 545 | |
| 546 | // Add active filters to the facet list (Primo doesn't return them): |
| 547 | $facets = []; |
| 548 | foreach ($params['filterList'] ?? [] as $current) { |
| 549 | if ('NOT' === $current['facetOp']) { |
| 550 | continue; |
| 551 | } |
| 552 | $field = $current['field']; |
| 553 | foreach ($current['values'] as $value) { |
| 554 | $facets[$field][$value] = null; |
| 555 | } |
| 556 | } |
| 557 | |
| 558 | // Process received facets |
| 559 | foreach ($response->facets as $facet) { |
| 560 | // Handle facet values as strings to ensure that numeric values stay |
| 561 | // intact (no array_combine etc.): |
| 562 | foreach ($facet->values as $value) { |
| 563 | $facets[$facet->name][(string)$value->value] = $value->count; |
| 564 | } |
| 565 | uasort( |
| 566 | $facets[$facet->name], |
| 567 | function ($a, $b) { |
| 568 | // Put the selected facets (with null as value) on the top: |
| 569 | return ($b ?? PHP_INT_MAX) <=> ($a ?? PHP_INT_MAX); |
| 570 | } |
| 571 | ); |
| 572 | } |
| 573 | |
| 574 | // Apparently there's no "did you mean" data in the response.. |
| 575 | |
| 576 | return [ |
| 577 | 'recordCount' => $totalhits, |
| 578 | 'documents' => $items, |
| 579 | 'facets' => $facets, |
| 580 | 'didYouMean' => [], |
| 581 | 'error' => $response->info->errorDetails->errorMessages[0] ?? [], |
| 582 | ]; |
| 583 | } |
| 584 | |
| 585 | /** |
| 586 | * Process highlighting tags of the record fields |
| 587 | * |
| 588 | * @param array $record Record data |
| 589 | * @param array $params Request params |
| 590 | * @param \StdClass $highlight Highlighting data |
| 591 | * |
| 592 | * @return void |
| 593 | */ |
| 594 | protected function processHighlighting(array &$record, array $params, \StdClass $highlight): void |
| 595 | { |
| 596 | if (empty($params['highlight'])) { |
| 597 | return; |
| 598 | } |
| 599 | |
| 600 | $startTag = $params['highlightStart'] ?? ''; |
| 601 | $endTag = $params['highlightEnd'] ?? ''; |
| 602 | |
| 603 | $highlightFields = [ |
| 604 | 'title' => 'title', |
| 605 | 'creator' => 'author', |
| 606 | 'description' => 'description', |
| 607 | ]; |
| 608 | |
| 609 | $hilightDetails = []; |
| 610 | |
| 611 | foreach ($highlightFields as $primoField => $field) { |
| 612 | if ( |
| 613 | ($highlightValues = $highlight->$primoField ?? null) |
| 614 | && !empty($record[$field]) |
| 615 | ) { |
| 616 | $match = implode( |
| 617 | '|', |
| 618 | array_map( |
| 619 | function ($s) { |
| 620 | return preg_quote($s, '/'); |
| 621 | }, |
| 622 | $highlightValues |
| 623 | ) |
| 624 | ); |
| 625 | $hilightDetails[$field] = array_map( |
| 626 | function ($s) use ($match, $startTag, $endTag) { |
| 627 | return preg_replace("/(\b|-|–)($match)(\b|-|–)/", "$1$startTag$2$endTag$3", $s); |
| 628 | }, |
| 629 | (array)$record[$field] |
| 630 | ); |
| 631 | } |
| 632 | } |
| 633 | |
| 634 | $record['highlightDetails'] = $hilightDetails; |
| 635 | } |
| 636 | |
| 637 | /** |
| 638 | * Fix the description field by removing tags etc. |
| 639 | * |
| 640 | * @param string $description Description |
| 641 | * |
| 642 | * @return string |
| 643 | */ |
| 644 | protected function processDescription($description) |
| 645 | { |
| 646 | // Sometimes the entire article is in the description, so just take a chunk |
| 647 | // from the beginning. |
| 648 | $description = trim(mb_substr($description, 0, 2500, 'UTF-8')); |
| 649 | // These may contain all kinds of metadata, and just stripping |
| 650 | // tags mushes it all together confusingly. |
| 651 | $description = str_replace('<P>', '<p>', $description); |
| 652 | $paragraphs = explode('<p>', $description); |
| 653 | foreach ($paragraphs as &$value) { |
| 654 | // Strip tags, trim so array_filter can get rid of |
| 655 | // entries that would just have spaces |
| 656 | $value = trim(strip_tags($value)); |
| 657 | } |
| 658 | $paragraphs = array_filter($paragraphs); |
| 659 | // Now join paragraphs using line breaks |
| 660 | return implode('<br>', $paragraphs); |
| 661 | } |
| 662 | |
| 663 | /** |
| 664 | * Get a JWT token for the session |
| 665 | * |
| 666 | * @param bool $renew Whether to renew the token |
| 667 | * |
| 668 | * @return string |
| 669 | */ |
| 670 | protected function getJWT(bool $renew = false): string |
| 671 | { |
| 672 | if (!$this->jwtUrl) { |
| 673 | return ''; |
| 674 | } |
| 675 | |
| 676 | if (!$renew && isset($this->session->jwt)) { |
| 677 | return $this->session->jwt; |
| 678 | } |
| 679 | $client = ($this->clientFactory)($this->getUrl($this->jwtUrl)); |
| 680 | $result = $client->setMethod('GET')->send(); |
| 681 | $resultBody = $result->getBody(); |
| 682 | if (!$result->isSuccess()) { |
| 683 | $this->logError( |
| 684 | "Request {$this->jwtUrl} failed with error code " . $result->getStatusCode() . ": $resultBody" |
| 685 | ); |
| 686 | throw new \Exception($resultBody); |
| 687 | } |
| 688 | $this->session->jwt = trim($resultBody, '"'); |
| 689 | return $this->session->jwt; |
| 690 | } |
| 691 | |
| 692 | /** |
| 693 | * Build a URL from a configured one |
| 694 | * |
| 695 | * @param string $url URL |
| 696 | * |
| 697 | * @return string |
| 698 | */ |
| 699 | protected function getUrl(string $url): string |
| 700 | { |
| 701 | return str_replace('{{INSTCODE}}', urlencode($this->inst), $url); |
| 702 | } |
| 703 | } |