Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
76.67% |
230 / 300 |
|
40.00% |
4 / 10 |
CRAP | |
0.00% |
0 / 1 |
| Connector | |
76.67% |
230 / 300 |
|
40.00% |
4 / 10 |
173.64 | |
0.00% |
0 / 1 |
| __construct | |
90.00% |
9 / 10 |
|
0.00% |
0 / 1 |
5.03 | |||
| query | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
2 | |||
| performSearch | |
51.47% |
35 / 68 |
|
0.00% |
0 / 1 |
89.83 | |||
| call | |
86.36% |
19 / 22 |
|
0.00% |
0 / 1 |
8.16 | |||
| process | |
90.57% |
96 / 106 |
|
0.00% |
0 / 1 |
27.61 | |||
| getRecord | |
100.00% |
13 / 13 |
|
100.00% |
1 / 1 |
3 | |||
| getRecords | |
0.00% |
0 / 15 |
|
0.00% |
0 / 1 |
12 | |||
| getInstitutionCode | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| processHighlighting | |
80.49% |
33 / 41 |
|
0.00% |
0 / 1 |
9.60 | |||
| processDescription | |
100.00% |
7 / 7 |
|
100.00% |
1 / 1 |
2 | |||
| 1 | <?php |
| 2 | |
| 3 | /** |
| 4 | * Primo Central connector. |
| 5 | * |
| 6 | * PHP version 8 |
| 7 | * |
| 8 | * Copyright (C) Villanova University 2010. |
| 9 | * |
| 10 | * This program is free software; you can redistribute it and/or modify |
| 11 | * it under the terms of the GNU General Public License version 2, |
| 12 | * as published by the Free Software Foundation. |
| 13 | * |
| 14 | * This program is distributed in the hope that it will be useful, |
| 15 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 16 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 17 | * GNU General Public License for more details. |
| 18 | * |
| 19 | * You should have received a copy of the GNU General Public License |
| 20 | * along with this program; if not, write to the Free Software |
| 21 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 22 | * |
| 23 | * @category VuFind |
| 24 | * @package Search |
| 25 | * @author Spencer Lamm <slamm1@swarthmore.edu> |
| 26 | * @author Anna Headley <aheadle1@swarthmore.edu> |
| 27 | * @author Chelsea Lobdell <clobdel1@swarthmore.edu> |
| 28 | * @author Demian Katz <demian.katz@villanova.edu> |
| 29 | * @author Ere Maijala <ere.maijala@helsinki.fi> |
| 30 | * @author Oliver Goldschmidt <o.goldschmidt@tuhh.de> |
| 31 | * @license http://opensource.org/licenses/gpl-2.0.php GNU General Public License |
| 32 | * @link https://vufind.org |
| 33 | */ |
| 34 | |
| 35 | namespace VuFindSearch\Backend\Primo; |
| 36 | |
| 37 | use Laminas\Http\Client as HttpClient; |
| 38 | |
| 39 | use function array_key_exists; |
| 40 | use function count; |
| 41 | use function in_array; |
| 42 | use function is_array; |
| 43 | use function strlen; |
| 44 | |
| 45 | /** |
| 46 | * Primo Central connector. |
| 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 | * @deprecated Use RestConnector instead |
| 60 | */ |
| 61 | class Connector implements ConnectorInterface, \Laminas\Log\LoggerAwareInterface |
| 62 | { |
| 63 | use \VuFind\Log\LoggerAwareTrait; |
| 64 | use \VuFindSearch\Backend\Feature\ConnectorCacheTrait; |
| 65 | |
| 66 | /** |
| 67 | * HTTP client used for API transactions |
| 68 | * |
| 69 | * @var HttpClient |
| 70 | */ |
| 71 | public $client; |
| 72 | |
| 73 | /** |
| 74 | * Institution code |
| 75 | * |
| 76 | * @var string |
| 77 | */ |
| 78 | protected $inst; |
| 79 | |
| 80 | /** |
| 81 | * Base URL for API |
| 82 | * |
| 83 | * @var string |
| 84 | */ |
| 85 | protected $host; |
| 86 | |
| 87 | /** |
| 88 | * Response for an empty search |
| 89 | * |
| 90 | * @var array |
| 91 | */ |
| 92 | protected static $emptyQueryResponse = [ |
| 93 | 'recordCount' => 0, |
| 94 | 'documents' => [], |
| 95 | 'facets' => [], |
| 96 | 'error' => 'empty_search_disallowed', |
| 97 | ]; |
| 98 | |
| 99 | /** |
| 100 | * Regular expression to match highlighted terms |
| 101 | * |
| 102 | * @var string |
| 103 | */ |
| 104 | protected $highlightRegEx = '{<span[^>]*>([^<]*?)</span>}si'; |
| 105 | |
| 106 | /** |
| 107 | * Constructor |
| 108 | * |
| 109 | * Sets up the Primo API Client |
| 110 | * |
| 111 | * @param string $url Primo API URL (either a host name and port or a full |
| 112 | * path to the brief search including a trailing question mark) |
| 113 | * @param string $inst Institution code |
| 114 | * @param HttpClient $client HTTP client |
| 115 | */ |
| 116 | public function __construct($url, $inst, $client) |
| 117 | { |
| 118 | $parts = parse_url($url); |
| 119 | if (empty($parts['path']) || $parts['path'] == '/') { |
| 120 | $parts['path'] = '/PrimoWebServices/xservice/search/brief'; |
| 121 | } |
| 122 | $this->host = $parts['scheme'] . '://' . $parts['host'] |
| 123 | . (!empty($parts['port']) ? ':' . $parts['port'] : '') |
| 124 | . $parts['path'] . '?'; |
| 125 | if (!empty($parts['query'])) { |
| 126 | $this->host .= $parts['query'] . '&'; |
| 127 | } |
| 128 | |
| 129 | $this->inst = $inst; |
| 130 | $this->client = $client; |
| 131 | } |
| 132 | |
| 133 | /** |
| 134 | * Execute a search. Adds all the querystring parameters into |
| 135 | * $this->client and returns the parsed response |
| 136 | * |
| 137 | * @param string $institution Institution |
| 138 | * @param array $terms Associative array: |
| 139 | * index string: primo index to search (default "any") |
| 140 | * lookfor string: actual search terms |
| 141 | * @param array $params Associative array of optional arguments: |
| 142 | * phrase bool: true if it's a quoted phrase (default false) |
| 143 | * onCampus bool: (default true) |
| 144 | * didyoumean bool: (default false) |
| 145 | * filterList array: (field, value) pairs to filter results (def null) |
| 146 | * pageNumber string: index of first record (default 1) |
| 147 | * limit string: number of records to return (default 20) |
| 148 | * sort string: value to be used by for sorting (default null) |
| 149 | * highlight bool: whether to highlight search term matches in records |
| 150 | * highlightStart string: Prefix for a highlighted term |
| 151 | * highlightEnd string: Suffix for a Highlighted term |
| 152 | * Anything in $params not listed here will be ignored. |
| 153 | * |
| 154 | * Note: some input parameters accepted by Primo are not implemented here: |
| 155 | * - dym (did you mean) |
| 156 | * - more (get more) |
| 157 | * - lang (specify input language so engine can do lang. recognition) |
| 158 | * - displayField (has to do with highlighting somehow) |
| 159 | * |
| 160 | * @throws \Exception |
| 161 | * @return array An array of query results |
| 162 | * |
| 163 | * @link http://www.exlibrisgroup.org/display/PrimoOI/Brief+Search |
| 164 | */ |
| 165 | public function query($institution, $terms, $params = null) |
| 166 | { |
| 167 | // defaults for params |
| 168 | $args = [ |
| 169 | 'phrase' => false, |
| 170 | 'onCampus' => true, |
| 171 | 'didYouMean' => false, |
| 172 | 'filterList' => null, |
| 173 | 'pcAvailability' => false, |
| 174 | 'pageNumber' => 1, |
| 175 | 'limit' => 20, |
| 176 | 'sort' => null, |
| 177 | 'highlight' => false, |
| 178 | 'highlightStart' => '', |
| 179 | 'highlightEnd' => '', |
| 180 | ]; |
| 181 | if (isset($params)) { |
| 182 | $args = array_merge($args, $params); |
| 183 | } |
| 184 | |
| 185 | $result = $this->performSearch($institution, $terms, $args); |
| 186 | return $result; |
| 187 | } |
| 188 | |
| 189 | /** |
| 190 | * Support method for query() -- perform inner search logic |
| 191 | * |
| 192 | * @param string $institution Institution |
| 193 | * @param array $terms Associative array: |
| 194 | * index string: primo index to search (default "any") |
| 195 | * lookfor string: actual search terms |
| 196 | * @param array $args Associative array of optional arguments (see query |
| 197 | * method for more information) |
| 198 | * |
| 199 | * @throws \Exception |
| 200 | * @return array An array of query results |
| 201 | */ |
| 202 | protected function performSearch($institution, $terms, $args) |
| 203 | { |
| 204 | // we have to build a querystring because I think adding them |
| 205 | // incrementally is implemented as a dictionary, but we are allowed |
| 206 | // multiple querystring parameters with the same key. |
| 207 | $qs = []; |
| 208 | |
| 209 | // QUERYSTRING: query (search terms) |
| 210 | // re: phrase searches, turns out we can just pass whatever we got |
| 211 | // to primo and they will interpret it correctly. |
| 212 | // leaving this flag in b/c it's not hurting anything, but we |
| 213 | // don't currently have a situation where we need to use "exact" |
| 214 | $precision = 'contains'; |
| 215 | if ($args['phrase']) { |
| 216 | $precision = 'exact'; |
| 217 | } |
| 218 | // determine which primo index to search |
| 219 | |
| 220 | //default index is any and initialize lookfor to an empty string |
| 221 | $lookin = 'any'; |
| 222 | $lookfor = ''; |
| 223 | |
| 224 | if (is_array($terms)) { |
| 225 | foreach ($terms as $thisTerm) { |
| 226 | //set the index to search |
| 227 | switch ($thisTerm['index']) { |
| 228 | case 'AllFields': |
| 229 | $lookin = 'any'; |
| 230 | break; |
| 231 | case 'Title': |
| 232 | $lookin = 'title'; |
| 233 | break; |
| 234 | case 'Author': |
| 235 | $lookin = 'creator'; |
| 236 | break; |
| 237 | case 'Subject': |
| 238 | $lookin = 'sub'; |
| 239 | break; |
| 240 | case 'Abstract': |
| 241 | $lookin = 'desc'; |
| 242 | break; |
| 243 | case 'ISSN': |
| 244 | $lookin = 'issn'; |
| 245 | break; |
| 246 | } |
| 247 | |
| 248 | //set the lookfor terms to search |
| 249 | $lookfor = str_replace(',', ' ', $thisTerm['lookfor']); |
| 250 | |
| 251 | //set precision |
| 252 | if (array_key_exists('op', $thisTerm) && !empty($thisTerm['op'])) { |
| 253 | $precision = $thisTerm['op']; |
| 254 | } |
| 255 | |
| 256 | $qs[] = "query=$lookin,$precision," . urlencode($lookfor); |
| 257 | } |
| 258 | } |
| 259 | |
| 260 | // continue only if lookfor is not an empty string |
| 261 | if (strlen($lookfor) > 0) { |
| 262 | // It's a giant nested thing! This is because we really have to |
| 263 | // have a query to send to primo or it hates us |
| 264 | |
| 265 | // QUERYSTRING: institution |
| 266 | $qs[] = "institution=$institution"; |
| 267 | |
| 268 | // QUERYSTRING: onCampus |
| 269 | if ($args['onCampus']) { |
| 270 | $qs[] = 'onCampus=true'; |
| 271 | } else { |
| 272 | $qs[] = 'onCampus=false'; |
| 273 | } |
| 274 | |
| 275 | // QUERYSTRING: didYouMean |
| 276 | if ($args['didYouMean']) { |
| 277 | $qs[] = 'dym=true'; |
| 278 | } else { |
| 279 | $qs[] = 'dym=false'; |
| 280 | } |
| 281 | |
| 282 | // QUERYSTRING: query (filter list) |
| 283 | // Date-related TODO: |
| 284 | // - provide additional support / processing for [x to y] limits? |
| 285 | // - sys/Summon.php messes with publication date to enable date |
| 286 | // range facet control in the interface. look for injectPubDate |
| 287 | if (!empty($args['filterList'])) { |
| 288 | foreach ($args['filterList'] as $current) { |
| 289 | $facet = $current['field']; |
| 290 | $facetOp = $current['facetOp']; |
| 291 | $values = $current['values']; |
| 292 | $values = array_map( |
| 293 | function ($value) { |
| 294 | return urlencode(str_replace(',', ' ', $value)); |
| 295 | }, |
| 296 | $values |
| 297 | ); |
| 298 | if ('OR' === $facetOp) { |
| 299 | $qs[] = "query_inc=facet_$facet,exact," . |
| 300 | implode(',', $values); |
| 301 | } elseif ('NOT' === $facetOp) { |
| 302 | $qs[] = "query_exc=facet_$facet,exact," . |
| 303 | implode(',', $values); |
| 304 | } else { |
| 305 | foreach ($values as $value) { |
| 306 | $qs[] = "query_inc=facet_$facet,exact,$value"; |
| 307 | } |
| 308 | } |
| 309 | } |
| 310 | } |
| 311 | |
| 312 | // QUERYSTRING: pcAvailability |
| 313 | // by default, PrimoCentral only returns matches, |
| 314 | // which are available via Holdingsfile |
| 315 | // pcAvailability = false |
| 316 | // By setting this value to true, also matches, which |
| 317 | // are NOT available via Holdingsfile are returned |
| 318 | // (yes, right, set this to true - thats ExLibris Logic) |
| 319 | if ($args['pcAvailability']) { |
| 320 | $qs[] = 'pcAvailability=true'; |
| 321 | } |
| 322 | |
| 323 | // QUERYSTRING: indx (start record) |
| 324 | $recordStart = ($args['pageNumber'] - 1) * $args['limit'] + 1; |
| 325 | $qs[] = "indx=$recordStart"; |
| 326 | |
| 327 | // TODO: put bulksize in conf file? set a reasonable cap... |
| 328 | // or is it better to grab each set of 20 through this api module? |
| 329 | // Look at how vufind/Summon does this... |
| 330 | // QUERYSTRING: bulkSize (limit, # of records to return) |
| 331 | $qs[] = 'bulkSize=' . $args['limit']; |
| 332 | |
| 333 | // QUERYSTRING: sort |
| 334 | // Looks like the possible values are "popularity" or "scdate" |
| 335 | // omit the field for default sorting |
| 336 | if (isset($args['sort']) && ($args['sort'] != 'relevance')) { |
| 337 | $qs[] = 'sortField=' . $args['sort']; |
| 338 | } |
| 339 | |
| 340 | // Highlighting |
| 341 | $qs[] = 'highlight=' . (empty($args['highlight']) ? 'false' : 'true'); |
| 342 | |
| 343 | // QUERYSTRING: loc |
| 344 | // all primocentral queries need this |
| 345 | $qs[] = 'loc=adaptor,primo_central_multiple_fe'; |
| 346 | |
| 347 | // Send Request |
| 348 | $result = $this->call(implode('&', $qs), $args); |
| 349 | } else { |
| 350 | return self::$emptyQueryResponse; |
| 351 | } |
| 352 | |
| 353 | return $result; |
| 354 | } |
| 355 | |
| 356 | /** |
| 357 | * Small wrapper for sendRequest, process to simplify error handling. |
| 358 | * |
| 359 | * @param string $qs Query string |
| 360 | * @param array $params Request parameters |
| 361 | * @param string $method HTTP method |
| 362 | * @param bool $cacheable Whether the request is cacheable |
| 363 | * |
| 364 | * @return object The parsed primo data |
| 365 | * @throws \Exception |
| 366 | */ |
| 367 | protected function call($qs, $params = [], $method = 'GET', $cacheable = true) |
| 368 | { |
| 369 | $this->debug("{$method}: {$this->host}{$qs}"); |
| 370 | $this->client->resetParameters(); |
| 371 | $baseUrl = null; |
| 372 | if ($method == 'GET') { |
| 373 | $baseUrl = $this->host . $qs; |
| 374 | } elseif ($method == 'POST') { |
| 375 | throw new \Exception('POST not supported'); |
| 376 | } |
| 377 | |
| 378 | $this->client->setUri($baseUrl); |
| 379 | $this->client->setMethod($method); |
| 380 | // Check cache: |
| 381 | $resultBody = null; |
| 382 | $cacheKey = null; |
| 383 | if ($cacheable && $this->cache) { |
| 384 | $cacheKey = $this->getCacheKey($this->client); |
| 385 | $resultBody = $this->getCachedData($cacheKey); |
| 386 | } |
| 387 | if (null === $resultBody) { |
| 388 | // Send request: |
| 389 | $result = $this->client->send(); |
| 390 | $resultBody = $result->getBody(); |
| 391 | if (!$result->isSuccess()) { |
| 392 | throw new \Exception($resultBody); |
| 393 | } |
| 394 | if ($cacheKey) { |
| 395 | $this->putCachedData($cacheKey, $resultBody); |
| 396 | } |
| 397 | } |
| 398 | return $this->process($resultBody, $params); |
| 399 | } |
| 400 | |
| 401 | /** |
| 402 | * Translate Primo's XML into array of arrays. |
| 403 | * |
| 404 | * @param string $data The raw xml from Primo |
| 405 | * @param array $params Request parameters |
| 406 | * |
| 407 | * @return array The processed response from Primo |
| 408 | */ |
| 409 | protected function process($data, $params = []) |
| 410 | { |
| 411 | // make sure data exists |
| 412 | if (strlen($data) == 0) { |
| 413 | throw new \Exception('Primo did not return any data'); |
| 414 | } |
| 415 | |
| 416 | // Load API content as XML objects |
| 417 | $sxe = new \SimpleXmlElement($data); |
| 418 | |
| 419 | if ($sxe === false) { |
| 420 | throw new \Exception('Error while parsing the document'); |
| 421 | } |
| 422 | |
| 423 | // some useful data about these results |
| 424 | $totalhitsarray = $sxe->xpath('//@TOTALHITS'); |
| 425 | |
| 426 | // if totalhits is missing but we have a message, this is an error |
| 427 | // situation. |
| 428 | if (!isset($totalhitsarray[0])) { |
| 429 | $messages = $sxe->xpath('//@MESSAGE'); |
| 430 | $message = isset($messages[0]) |
| 431 | ? (string)$messages[0] : 'TOTALHITS attribute missing.'; |
| 432 | throw new \Exception($message); |
| 433 | } else { |
| 434 | $totalhits = (int)$totalhitsarray[0]; |
| 435 | } |
| 436 | // TODO: would these be useful? |
| 437 | //$firsthit = $sxe->xpath('//@FIRSTHIT'); |
| 438 | //$lasthit = $sxe->xpath('//@LASTHIT'); |
| 439 | |
| 440 | // Register the 'sear' namespace at the top level to avoid problems: |
| 441 | $sxe->registerXPathNamespace( |
| 442 | 'sear', |
| 443 | 'http://www.exlibrisgroup.com/xsd/jaguar/search' |
| 444 | ); |
| 445 | |
| 446 | // Get the available namespaces. The Primo API uses multiple namespaces. |
| 447 | // Will be used to navigate the DOM for elements that have namespaces |
| 448 | $namespaces = $sxe->getNameSpaces(true); |
| 449 | |
| 450 | // Get results set data and add to $items array |
| 451 | // This foreach grabs all the child elements of sear:DOC, |
| 452 | // except those with namespaces |
| 453 | $items = []; |
| 454 | |
| 455 | $docset = $sxe->xpath('//sear:DOC'); |
| 456 | if (empty($docset) && isset($sxe->JAGROOT->RESULT->DOCSET->DOC)) { |
| 457 | $docset = $sxe->JAGROOT->RESULT->DOCSET->DOC; |
| 458 | } |
| 459 | |
| 460 | foreach ($docset as $doc) { |
| 461 | $item = []; |
| 462 | // Due to a bug in the primo API, the first result has |
| 463 | // a namespace (prim:) while the rest of the results do not. |
| 464 | // Those child elements do not get added to $doc. |
| 465 | // If the bib parent element (PrimoNMBib) is missing for a $doc, |
| 466 | // that means it has the prim namespace prefix. |
| 467 | // So first set the right prefix |
| 468 | $prefix = $doc; |
| 469 | if ($doc->PrimoNMBib != 'true' && isset($namespaces['prim'])) { |
| 470 | // Use the namespace prefix to get those missing child |
| 471 | // elements out of $doc. |
| 472 | $prefix = $doc->children($namespaces['prim']); |
| 473 | } |
| 474 | // Now, navigate the DOM and set values to the array |
| 475 | // cast to (string) to get the element's value not an XML object |
| 476 | $item['recordid'] |
| 477 | = substr((string)$prefix->PrimoNMBib->record->control->recordid, 3); |
| 478 | $item['title'] |
| 479 | = (string)$prefix->PrimoNMBib->record->display->title; |
| 480 | $item['format'] = [(string)$prefix->PrimoNMBib->record->display->type]; |
| 481 | // creators |
| 482 | $creator |
| 483 | = trim((string)$prefix->PrimoNMBib->record->display->creator); |
| 484 | if (strlen($creator) > 0) { |
| 485 | $item['creator'] = array_map('trim', explode(';', $creator)); |
| 486 | } |
| 487 | // subjects |
| 488 | $subject |
| 489 | = trim((string)$prefix->PrimoNMBib->record->display->subject); |
| 490 | if (strlen($subject) > 0) { |
| 491 | $item['subjects'] = explode(';', $subject); |
| 492 | } |
| 493 | $item['ispartof'] |
| 494 | = (string)$prefix->PrimoNMBib->record->display->ispartof; |
| 495 | // description is sort of complicated and will be processed after |
| 496 | // highlighting tags are handled. |
| 497 | $description = isset($prefix->PrimoNMBib->record->display->description) |
| 498 | ? (string)$prefix->PrimoNMBib->record->display->description |
| 499 | : (string)$prefix->PrimoNMBib->record->search->description; |
| 500 | $item['description'] = $description; |
| 501 | // and the rest! |
| 502 | $item['language'] |
| 503 | = (string)$prefix->PrimoNMBib->record->display->language; |
| 504 | $item['source'] |
| 505 | = implode('; ', (array)$prefix->PrimoNMBib->record->display->source); |
| 506 | $item['identifier'] |
| 507 | = (string)$prefix->PrimoNMBib->record->display->identifier; |
| 508 | $item['fulltext'] |
| 509 | = (string)$prefix->PrimoNMBib->record->delivery->fulltext; |
| 510 | |
| 511 | $item['issn'] = []; |
| 512 | foreach ($prefix->PrimoNMBib->record->search->issn as $issn) { |
| 513 | $item['issn'][] = (string)$issn; |
| 514 | } |
| 515 | |
| 516 | //Are these two needed? |
| 517 | //$item['publisher'] = |
| 518 | // (string)$prefix->PrimoNMBib->record->display->publisher; |
| 519 | //$item['peerreviewed'] = |
| 520 | // (string)$prefix->PrimoNMBib->record->display->lds50; |
| 521 | |
| 522 | // Get the URL, which has a separate namespace |
| 523 | $sear = $doc->children($namespaces['sear']); |
| 524 | $item['url'] = !empty($sear->LINKS->openurl) |
| 525 | ? (string)$sear->LINKS->openurl |
| 526 | : (string)$sear->GETIT->attributes()->GetIt2; |
| 527 | |
| 528 | // Container data |
| 529 | $addata = $prefix->PrimoNMBib->record->addata; |
| 530 | $item['container_title'] = (string)$addata->jtitle; |
| 531 | $item['container_volume'] = (string)$addata->volume; |
| 532 | $item['container_issue'] = (string)$addata->issue; |
| 533 | $item['container_start_page'] = (string)$addata->spage; |
| 534 | $item['container_end_page'] = (string)$addata->epage; |
| 535 | foreach ($addata->eissn as $eissn) { |
| 536 | if (!in_array((string)$eissn, $item['issn'])) { |
| 537 | $item['issn'][] = (string)$eissn; |
| 538 | } |
| 539 | } |
| 540 | foreach ($addata->issn as $issn) { |
| 541 | if (!in_array((string)$issn, $item['issn'])) { |
| 542 | $item['issn'][] = (string)$issn; |
| 543 | } |
| 544 | } |
| 545 | foreach ($addata->doi as $doi) { |
| 546 | $item['doi_str_mv'][] = (string)$doi; |
| 547 | } |
| 548 | |
| 549 | $processCitations = function ($data): array { |
| 550 | $result = []; |
| 551 | foreach ($data as $item) { |
| 552 | $result[] = 'cdi_' . (string)$item; |
| 553 | } |
| 554 | return $result; |
| 555 | }; |
| 556 | |
| 557 | // These require the cdi_ prefix in search, so add it right away: |
| 558 | $item['cites'] = $processCitations($prefix->PrimoNMBib->record->display->cites ?? []); |
| 559 | $item['cited_by'] = $processCitations($prefix->PrimoNMBib->record->display->citedby ?? []); |
| 560 | |
| 561 | // Remove dash-less ISSNs if there are corresponding dashed ones |
| 562 | // (We could convert dash-less ISSNs to dashed ones, but try to stay |
| 563 | // true to the metadata) |
| 564 | $callback = function ($issn) use ($item) { |
| 565 | return strlen($issn) != 8 |
| 566 | || !in_array( |
| 567 | substr($issn, 0, 4) . '-' . substr($issn, 4), |
| 568 | $item['issn'] |
| 569 | ); |
| 570 | }; |
| 571 | $item['issn'] = array_values(array_filter($item['issn'], $callback)); |
| 572 | |
| 573 | // Always process highlighting data as it seems Primo sometimes returns |
| 574 | // it (e.g. for CDI search) even if highlight parameter is set to false. |
| 575 | $this->processHighlighting($item, $params); |
| 576 | |
| 577 | // Fix description now that highlighting is done: |
| 578 | $item['description'] = $this->processDescription($item['description']); |
| 579 | |
| 580 | $item['fullrecord'] = $prefix->PrimoNMBib->record->asXml(); |
| 581 | $items[] = $item; |
| 582 | } |
| 583 | |
| 584 | // Set up variables with needed attribute names |
| 585 | // Makes matching attributes and getting their values easier |
| 586 | $att = 'NAME'; |
| 587 | $key = 'KEY'; |
| 588 | $value = 'VALUE'; |
| 589 | |
| 590 | // Get facet data and add to multidimensional $facets array |
| 591 | // Start by getting XML for each FACET element, |
| 592 | // which has the name of the facet as an attribute. |
| 593 | // We only get the first level of elements |
| 594 | // because child elements have a namespace prefix |
| 595 | $facets = []; |
| 596 | |
| 597 | $facetSet = $sxe->xpath('//sear:FACET'); |
| 598 | if (empty($facetSet)) { |
| 599 | if (!empty($sxe->JAGROOT->RESULT->FACETLIST)) { |
| 600 | $facetSet = $sxe->JAGROOT->RESULT->FACETLIST |
| 601 | ->children($namespaces['sear']); |
| 602 | } |
| 603 | } |
| 604 | |
| 605 | foreach ($facetSet as $facetlist) { |
| 606 | // Set first level of array with the facet name |
| 607 | $facet_name = (string)$facetlist->attributes()->$att; |
| 608 | |
| 609 | // Use the namespace prefix to get second level child elements |
| 610 | // (the facet values) out of $facetlist. |
| 611 | $sear_facets = $facetlist->children($namespaces['sear']); |
| 612 | foreach ($sear_facets as $facetvalues) { |
| 613 | // Second level of the array is facet values and their counts |
| 614 | $facet_key = (string)$facetvalues->attributes()->$key; |
| 615 | $facets[$facet_name][$facet_key] |
| 616 | = (string)$facetvalues->attributes()->$value; |
| 617 | } |
| 618 | } |
| 619 | |
| 620 | $didYouMean = []; |
| 621 | $suggestions = $sxe->xpath('//sear:QUERYTRANSFORMS'); |
| 622 | foreach ($suggestions as $suggestion) { |
| 623 | $didYouMean[] = (string)$suggestion->attributes()->QUERY; |
| 624 | } |
| 625 | |
| 626 | return [ |
| 627 | 'recordCount' => $totalhits, |
| 628 | 'documents' => $items, |
| 629 | 'facets' => $facets, |
| 630 | 'didYouMean' => $didYouMean, |
| 631 | ]; |
| 632 | } |
| 633 | |
| 634 | /** |
| 635 | * Retrieves a document specified by the ID. |
| 636 | * |
| 637 | * @param string $recordId The document to retrieve from the Primo API |
| 638 | * @param ?string $inst_code Institution code (optional) |
| 639 | * @param bool $onCampus Whether the user is on campus |
| 640 | * |
| 641 | * @throws \Exception |
| 642 | * @return array An array of query results |
| 643 | */ |
| 644 | public function getRecord(string $recordId, $inst_code = null, $onCampus = false) |
| 645 | { |
| 646 | if ('' === $recordId) { |
| 647 | return self::$emptyQueryResponse; |
| 648 | } |
| 649 | // Query String Parameters |
| 650 | $qs = []; |
| 651 | // There is currently (at 2015-12-17) a problem with Primo fetching |
| 652 | // records that have colons in the id (e.g. |
| 653 | // doaj_xmloai:doaj.org/article:94935655971c4917aab4fcaeafeb67b9). |
| 654 | // According to Ex Libris support we must use contains search without |
| 655 | // quotes for the time being. |
| 656 | // Escaping the - character causes problems getting records like |
| 657 | // wj10.1111/j.1475-679X.2011.00421.x |
| 658 | $qs[] = 'query=rid,contains,' |
| 659 | . urlencode(addcslashes($recordId, '":()')); |
| 660 | $qs[] = "institution=$inst_code"; |
| 661 | $qs[] = 'onCampus=' . ($onCampus ? 'true' : 'false'); |
| 662 | $qs[] = 'indx=1'; |
| 663 | $qs[] = 'bulkSize=1'; |
| 664 | $qs[] = 'loc=adaptor,primo_central_multiple_fe'; |
| 665 | // pcAvailability=true is needed for records, which |
| 666 | // are NOT in the PrimoCentral Holdingsfile. |
| 667 | // It won't hurt to have this parameter always set to true. |
| 668 | // But it'd hurt to have it not set in case you want to get |
| 669 | // a record, which is not in the Holdingsfile. |
| 670 | $qs[] = 'pcAvailability=true'; |
| 671 | |
| 672 | // Send Request |
| 673 | $result = $this->call(implode('&', $qs)); |
| 674 | |
| 675 | return $result; |
| 676 | } |
| 677 | |
| 678 | /** |
| 679 | * Retrieves multiple documents specified by the ID. |
| 680 | * |
| 681 | * @param array $recordIds The documents to retrieve from the Primo API |
| 682 | * @param ?string $inst_code Institution code (optional) |
| 683 | * @param bool $onCampus Whether the user is on campus |
| 684 | * |
| 685 | * @throws \Exception |
| 686 | * @return array An array of query results |
| 687 | */ |
| 688 | public function getRecords($recordIds, $inst_code = null, $onCampus = false) |
| 689 | { |
| 690 | // Callback function for formatting IDs: |
| 691 | $formatIds = function ($id) { |
| 692 | return addcslashes($id, '":()'); |
| 693 | }; |
| 694 | |
| 695 | // Query String Parameters |
| 696 | if ($recordIds) { |
| 697 | $qs = []; |
| 698 | $recordIds = array_map($formatIds, $recordIds); |
| 699 | $qs[] = 'query=rid,contains,' . urlencode(implode(' OR ', $recordIds)); |
| 700 | $qs[] = "institution=$inst_code"; |
| 701 | $qs[] = 'onCampus=' . ($onCampus ? 'true' : 'false'); |
| 702 | $qs[] = 'indx=1'; |
| 703 | $qs[] = 'bulkSize=' . count($recordIds); |
| 704 | $qs[] = 'loc=adaptor,primo_central_multiple_fe'; |
| 705 | // pcAvailability=true is needed for records, which |
| 706 | // are NOT in the PrimoCentral Holdingsfile. |
| 707 | // It won't hurt to have this parameter always set to true. |
| 708 | // But it'd hurt to have it not set in case you want to get |
| 709 | // a record, which is not in the Holdingsfile. |
| 710 | $qs[] = 'pcAvailability=true'; |
| 711 | |
| 712 | // Send Request |
| 713 | $result = $this->call(implode('&', $qs)); |
| 714 | } else { |
| 715 | return self::$emptyQueryResponse; |
| 716 | } |
| 717 | |
| 718 | return $result; |
| 719 | } |
| 720 | |
| 721 | /** |
| 722 | * Get the institution code based on user IP. If user is coming from |
| 723 | * off campus return |
| 724 | * |
| 725 | * @return string |
| 726 | */ |
| 727 | public function getInstitutionCode() |
| 728 | { |
| 729 | return $this->inst; |
| 730 | } |
| 731 | |
| 732 | /** |
| 733 | * Process highlighting tags of the record fields |
| 734 | * |
| 735 | * @param array $record Record data |
| 736 | * @param array $params Request params |
| 737 | * |
| 738 | * @return void |
| 739 | */ |
| 740 | protected function processHighlighting(&$record, $params) |
| 741 | { |
| 742 | $highlight = !empty($params['highlight']); |
| 743 | $startTag = $params['highlightStart'] ?? ''; |
| 744 | $endTag = $params['highlightEnd'] ?? ''; |
| 745 | |
| 746 | $highlightFields = [ |
| 747 | 'title' => 'title', |
| 748 | 'creator' => 'author', |
| 749 | 'description' => 'description', |
| 750 | ]; |
| 751 | |
| 752 | $hilightDetails = []; |
| 753 | foreach ($record as $field => $fieldData) { |
| 754 | $values = (array)$fieldData; |
| 755 | |
| 756 | // Collect highlighting details: |
| 757 | if (isset($highlightFields[$field])) { |
| 758 | $highlightedValues = []; |
| 759 | foreach ($values as $value) { |
| 760 | $count = 0; |
| 761 | $value = preg_replace( |
| 762 | $this->highlightRegEx, |
| 763 | "$startTag$1$endTag", |
| 764 | $value, |
| 765 | -1, |
| 766 | $count |
| 767 | ); |
| 768 | if ($count) { |
| 769 | // Account for double tags. Yes, it's possible. |
| 770 | $value = preg_replace( |
| 771 | $this->highlightRegEx, |
| 772 | '$1', |
| 773 | $value |
| 774 | ); |
| 775 | $highlightedValues[] = $value; |
| 776 | } |
| 777 | } |
| 778 | if ($highlightedValues) { |
| 779 | $hilightDetails[$highlightFields[$field]] = $highlightedValues; |
| 780 | } |
| 781 | } |
| 782 | |
| 783 | // Strip highlighting tags from all fields: |
| 784 | foreach ($values as &$value) { |
| 785 | $value = preg_replace( |
| 786 | $this->highlightRegEx, |
| 787 | '$1', |
| 788 | $value |
| 789 | ); |
| 790 | // Account for double tags. Yes, it's possible. |
| 791 | $value = preg_replace( |
| 792 | $this->highlightRegEx, |
| 793 | '$1', |
| 794 | $value |
| 795 | ); |
| 796 | } |
| 797 | // Unset reference: |
| 798 | unset($value); |
| 799 | $record[$field] = is_array($fieldData) ? $values : $values[0]; |
| 800 | |
| 801 | if ($highlight) { |
| 802 | $record['highlightDetails'] = $hilightDetails; |
| 803 | } |
| 804 | } |
| 805 | } |
| 806 | |
| 807 | /** |
| 808 | * Fix the description field by removing tags etc. |
| 809 | * |
| 810 | * @param string $description Description |
| 811 | * |
| 812 | * @return string |
| 813 | */ |
| 814 | protected function processDescription($description) |
| 815 | { |
| 816 | // Sometimes the entire article is in the description, so just take a chunk |
| 817 | // from the beginning. |
| 818 | $description = trim(mb_substr($description, 0, 2500, 'UTF-8')); |
| 819 | // These may contain all kinds of metadata, and just stripping |
| 820 | // tags mushes it all together confusingly. |
| 821 | $description = str_replace('<P>', '<p>', $description); |
| 822 | $paragraphs = explode('<p>', $description); |
| 823 | foreach ($paragraphs as &$value) { |
| 824 | // Strip tags, trim so array_filter can get rid of |
| 825 | // entries that would just have spaces |
| 826 | $value = trim(strip_tags($value)); |
| 827 | } |
| 828 | $paragraphs = array_filter($paragraphs); |
| 829 | // Now join paragraphs using line breaks |
| 830 | return implode('<br>', $paragraphs); |
| 831 | } |
| 832 | } |