| static public regexp_match_all ( $pattern, $subject, &$matches ) : integer | boolean | ||
| $pattern | string Regular expression | |
| $subject | string String to apply regular expression to | |
| $matches | array Reference to receive matches | |
| return | integer | boolean | Returns number of full matches of given subject, or FALSE if an error occurred. |
/**
* Converts a string with multiple persons
* to an array of NLM name descriptions.
*
* @param $personsString string
* @param $title boolean true to parse for title
* @param $degrees boolean true to parse for degrees
* @return array an array of NLM name descriptions or null
* if the string could not be converted plus optionally a
* single 'et-al' string.
*/
function &_parsePersonsString($personsString, $title, $degrees)
{
// Check for 'et al'.
$personsStringBeforeEtal = PKPString::strlen($personsString);
$personsString = PKPString::regexp_replace('/et ?al$/', '', $personsString);
$etAl = $personsStringBeforeEtal == PKPString::strlen($personsString) ? false : true;
// Remove punctuation.
$personsString = trim($personsString, ':;, ');
// Cut the authors string into pieces.
$personStrings = PKPString::iterativeExplode(array(':', ';'), $personsString);
// If we did not have success with simple patterns then try more complex
// patterns to tokenize multiple-person strings.
if (count($personStrings) == 1) {
// The first pattern must match the whole string, the second is used
// to extract names.
$complexPersonsPatterns = array(array('/^((([^ \\t\\n\\r\\f\\v,.&]{2,}\\s*)+,\\s*([A-Z]\\.\\s*)+),\\s*)+(\\&|\\.\\s\\.\\s\\.)\\s*([^ \\t\\n\\r\\f\\v,.&]{2,}\\s*,\\s*([A-Z]\\.\\s*)+)$/i', '/(?:[^ \\t\\n\\r\\f\\v,.&]{2,}\\s*)+,\\s*(?:[A-Z]\\.\\s*)+/i'), array('/^((([^ \\t\\n\\r\\f\\v,&]+\\s+)+[^ \\t\\n\\r\\f\\v,&]+\\s*)[,&]\\s*)+(([^ \\t\\n\\r\\f\\v,&]+\\s+)+[^ \\t\\n\\r\\f\\v,&]+)/i', '/(?:(?:[^ \\t\\n\\r\\f\\v,&.]+|[^ \\t\\n\\r\\f\\v,&]{2,})\\s+)+(?:[^ \\t\\n\\r\\f\\v,&.]+|[^ \\t\\n\\r\\f\\v,&]{2,})/i'));
$matched = false;
foreach ($complexPersonsPatterns as $complexPersonsPattern) {
// Break at the first pattern that matches.
if ($matched = PKPString::regexp_match($complexPersonsPattern[0], $personsString)) {
// Retrieve names.
$success = PKPString::regexp_match_all($complexPersonsPattern[1], $personsString, $personStrings);
assert($success && count($personStrings) == 1);
$personStrings = $personStrings[0];
break;
}
}
if (!$matched) {
// If nothing matches then try to parse as a single person.
$personStrings = array($personsString);
}
}
// Parse persons.
$persons = array();
foreach ($personStrings as $personString) {
$persons[] =& $this->_parsePersonString($personString, $title, $degrees);
}
// Add et-al string.
if ($etAl) {
$persons[] = PERSON_STRING_FILTER_ETAL;
}
return $persons;
}
/**
* @see Filter::process()
* @param $citationDescription MetadataDescription
* @return string a DOI or null
*/
function &process(&$citationDescription)
{
$nullVar = null;
// Get the search strings
$searchTemplates =& $this->_getSearchTemplates();
$searchStrings = $this->constructSearchStrings($searchTemplates, $citationDescription);
// Run the searches, in order, until we have a result
$searchParams = array('qt' => 'worldcat_org_all');
foreach ($searchStrings as $searchString) {
$searchParams['q'] = $searchString;
// Worldcat Web search; results are (mal-formed) XHTML
if (is_null($result = $this->callWebService(WORLDCAT_WEBSERVICE_SEARCH, $searchParams, XSL_TRANSFORMER_DOCTYPE_STRING))) {
return $nullVar;
}
// parse the OCLC numbers from search results
PKPString::regexp_match_all('/id="itemid_(\\d+)"/', $result, $matches);
if (!empty($matches[1])) {
break;
}
}
// If we don't have an OCLC number, then we cannot get any metadata
if (empty($matches[1])) {
return $nullVar;
}
// use xISBN because it's free
foreach ($matches[1] as $oclcId) {
$isbns = $this->_oclcToIsbns($oclcId);
if (is_array($isbns)) {
break;
}
}
if (is_null($isbns)) {
return $nullVar;
}
$apiKey = $this->getApiKey();
if (empty($apiKey)) {
// Use the first ISBN if we have multiple
$citationDescription =& $this->_lookupXIsbn($isbns[0]);
return $citationDescription;
} elseif (!empty($isbns[0])) {
// Worldcat lookup only works with an API key
if (is_null($citationDescription =& $this->_lookupWorldcat($matches[1][0]))) {
return $nullVar;
}
// Prefer ISBN from xISBN if possible
if (!empty($isbns[0])) {
$citationDescription->addStatement('ibsn', $isbns[0], null, true);
}
return $citationDescription;
}
// Nothing found
return $nullVar;
}
/**
* Given a locale string, get the list of parameter references of the
* form {$myParameterName}.
* @param $source string
* @return array
*/
static function getParameterNames($source)
{
$matches = null;
PKPString::regexp_match_all('/({\\$[^}]+})/', $source, $matches);
array_shift($matches);
// Knock the top element off the array
if (isset($matches[0])) {
return $matches[0];
}
return array();
}
/**
* Parse SQL content into individual SQL statements.
* @param $sql string
* @return array
*/
function &parseStatements(&$sql)
{
$statements = array();
$statementsTmp = explode($this->statementDelim, $sql);
$currentStatement = '';
$numSingleQuotes = $numEscapedSingleQuotes = 0;
// This method for parsing the SQL statements was adapted from one used in phpBB (http://www.phpbb.com/)
for ($i = 0, $count = count($statementsTmp); $i < $count; $i++) {
// Get total number of single quotes in string
$numSingleQuotes += PKPString::substr_count($statementsTmp[$i], "'");
// Get number of escaped single quotes
$numEscapedSingleQuotes += PKPString::regexp_match_all("/(?<!\\\\)(\\\\\\\\)*\\\\'/", $statementsTmp[$i], $matches);
$currentStatement .= $statementsTmp[$i];
if (($numSingleQuotes - $numEscapedSingleQuotes) % 2 == 0) {
// Even number of unescaped single quotes, so statement must be complete
if (trim($currentStatement) !== '') {
array_push($statements, trim($currentStatement));
}
$currentStatement = '';
$numSingleQuotes = $numEscapedSingleQuotes = 0;
} else {
// The statement is not complete, the delimiter must be inside the statement
$currentStatement .= $this->statementDelim;
}
}
return $statements;
}
