/**
* Generate and display the CAPTCHA image.
* @param $captcha object Captcha
*/
function generateImage(&$captcha)
{
$width = $this->getWidth();
$height = $this->getHeight();
$length = String::strlen($captcha->getValue());
$value = $captcha->getValue();
$image = imagecreatetruecolor($width, $height);
$fg = imagecolorallocate($image, rand(128, 255), rand(128, 255), rand(128, 255));
$bg = imagecolorallocate($image, rand(0, 64), rand(0, 64), rand(0, 64));
imagefill($image, $width / 2, $height / 2, $bg);
$xStart = rand($width / 12, $width / 3);
$xEnd = rand($width * 2 / 3, $width * 11 / 12);
for ($i = 0; $i < $length; $i++) {
imagefttext($image, rand(20, 34), rand(-15, 15), $xStart + ($xEnd - $xStart) * $i / $length + rand(-5, 5), rand(40, 60), $fg, Config::getVar('captcha', 'font_location'), String::substr($value, $i, 1));
}
// Add some noise to the image.
for ($i = 0; $i < 20; $i++) {
$color = imagecolorallocate($image, rand(0, 255), rand(0, 255), rand(0, 255));
for ($j = 0; $j < 20; $j++) {
imagesetpixel($image, rand(0, $this->getWidth()), rand(0, $this->getHeight()), $color);
}
}
header('Content-type: ' . $this->getMimeType());
imagepng($image);
imagedestroy($image);
}
/**
* Split a string into a clean array of keywords
* @param $text string
* @param $allowWildcards boolean
* @return array of keywords
*/
static function filterKeywords($text, $allowWildcards = false)
{
$minLength = Config::getVar('search', 'min_word_length');
$stopwords = self::_loadStopwords();
// Join multiple lines into a single string
if (is_array($text)) {
$text = join("\n", $text);
}
$cleanText = Core::cleanVar($text);
// Remove punctuation
$cleanText = String::regexp_replace('/[!"\\#\\$%\'\\(\\)\\.\\?@\\[\\]\\^`\\{\\}~]/', '', $cleanText);
$cleanText = String::regexp_replace('/[\\+,:;&\\/<=>\\|\\\\]/', ' ', $cleanText);
$cleanText = String::regexp_replace('/[\\*]/', $allowWildcards ? '%' : ' ', $cleanText);
$cleanText = String::strtolower($cleanText);
// Split into words
$words = String::regexp_split('/\\s+/', $cleanText);
// FIXME Do not perform further filtering for some fields, e.g., author names?
// Remove stopwords
$keywords = array();
foreach ($words as $k) {
if (!isset($stopwords[$k]) && String::strlen($k) >= $minLength && !is_numeric($k)) {
$keywords[] = String::substr($k, 0, SEARCH_KEYWORD_MAX_LENGTH);
}
}
return $keywords;
}
/**
* Generate a filename for a library file.
* @param $type int LIBRARY_FILE_TYPE_...
* @param $originalFileName string
* @return string
*/
function generateFileName($type, $originalFileName)
{
$libraryFileDao =& DAORegistry::getDAO('LibraryFileDAO');
$suffix = $this->getFileSuffixFromType($type);
$ext = $this->getExtension($originalFileName);
$truncated = $this->truncateFileName($originalFileName, 127 - String::strlen($suffix) - 1);
$baseName = String::substr($truncated, 0, String::strpos($originalFileName, $ext) - 1);
// Try a simple syntax first
$fileName = $baseName . '-' . $suffix . '.' . $ext;
if (!$libraryFileDao->filenameExists($this->pressId, $fileName)) {
return $fileName;
}
for ($i = 1;; $i++) {
$fullSuffix = $suffix . '-' . $i;
//truncate more if necessary
$truncated = $this->truncateFileName($originalFileName, 127 - String::strlen($fullSuffix) - 1);
// get the base name and append the suffix
$baseName = String::substr($truncated, 0, String::strpos($originalFileName, $ext) - 1);
//try the following
$fileName = $baseName . '-' . $fullSuffix . '.' . $ext;
if (!$libraryFileDao->filenameExists($this->pressId, $fileName)) {
return $fileName;
}
}
}
/**
* @see FormValidator::isValid()
* Value is valid if it is empty and optional or meets the specified length requirements.
* @return boolean
*/
function isValid()
{
if ($this->isEmptyAndOptional()) {
return true;
} else {
$length = String::strlen($this->getFieldValue());
switch ($this->_comparator) {
case '==':
return $length == $this->_length;
case '!=':
return $length != $this->_length;
case '<':
return $length < $this->_length;
case '>':
return $length > $this->_length;
case '<=':
return $length <= $this->_length;
case '>=':
return $length >= $this->_length;
}
return false;
}
}
/**
* Check if field value is valid.
* Value is valid if it is empty and optional or meets the specified length requirements.
* @return boolean
*/
function isValid()
{
if ($this->isEmptyAndOptional()) {
return true;
} else {
$length = String::strlen(trim($this->form->getData($this->field)));
switch ($this->comparator) {
case '==':
return $length == $this->length;
case '!=':
return $length != $this->length;
case '<':
return $length < $this->length;
case '>':
return $length > $this->length;
case '<=':
return $length <= $this->length;
case '>=':
return $length >= $this->length;
}
return false;
}
}
/**
* Truncate a filename to fit in the specified length.
*/
function truncateFileName($fileName, $length = 127)
{
if (String::strlen($fileName) <= $length) {
return $fileName;
}
$ext = $this->getExtension($fileName);
$truncated = String::substr($fileName, 0, $length - 1 - String::strlen($ext)) . '.' . $ext;
return String::substr($truncated, 0, $length);
}
/**
* 格式化字段
*
* @access private
* @param mixed $typeInfo
* @param mixed $value
* @return mixed
*/
private function formatField($typeInfo, $value)
{
preg_match("/(\\w+)(\\((\\d+)\\))?/", $typeInfo, $matches);
if (isset($matches[1])) {
$type = $matches[1];
}
if (isset($matches[3])) {
$len = $matches[3];
}
$_type = "string";
$type = strtolower($type);
switch ($type) {
case 'bit':
case 'bigbit':
case 'bool':
case 'boolean':
case 'decimal':
case 'decimal':
case 'dec':
case 'double':
case 'float':
case 'int':
case 'bigint':
case 'mediumint':
case 'smallint':
case 'tinyint':
case 'real':
if (!is_numeric($value)) {
$value = 0;
}
if ($value == '' || $value == null || empty($value)) {
$value = 0;
}
$_type = 'numeric';
break;
}
if (isset($len)) {
if (!is_array($value) && String::strlen($value) > $len) {
$value = String::msubstr($value, 0, $len, 'utf-8', '');
}
}
if (is_array($value)) {
$value = serialize($value);
}
if ($_type == 'string') {
$value = '\'' . $value . '\'';
}
return $value;
}
/**
* Calculate the differences between two strings and
* produce an array with three types of entries: added
* substrings, deleted substrings and unchanged substrings.
*
* The calculation is optimized to identify the common
* largest substring.
*
* The return value is an array of the following format:
*
* array(
* array( diff-type => substring ),
* array(...)
* )
*
* whereby diff-type can be one of:
* -1 = deletion
* 0 = common substring
* 1 = addition
*
* @param $originalString string
* @param $editedString string
* @return array
*/
static function diff($originalString, $editedString)
{
// Split strings into character arrays (multi-byte compatible).
foreach (array('originalStringCharacters' => $originalString, 'editedStringCharacters' => $editedString) as $characterArrayName => $string) {
${$characterArrayName} = array();
String::regexp_match_all('/./', $string, ${$characterArrayName});
if (isset(${$characterArrayName}[0])) {
${$characterArrayName} = ${$characterArrayName}[0];
}
}
// Determine the length of the strings.
$originalStringLength = count($originalStringCharacters);
$editedStringLength = count($editedStringCharacters);
// Is there anything to compare?
if ($originalStringLength == 0 && $editedStringLength == 0) {
return array();
}
// Is the original string empty?
if ($originalStringLength == 0) {
// Return the edited string as addition.
return array(array(1 => $editedString));
}
// Is the edited string empty?
if ($editedStringLength == 0) {
// Return the original string as deletion.
return array(array(-1 => $originalString));
}
// Initialize the local indices:
// 1) Create a character index for the edited string.
$characterIndex = array();
for ($characterPosition = 0; $characterPosition < $editedStringLength; $characterPosition++) {
$characterIndex[$editedStringCharacters[$characterPosition]][] = $characterPosition;
}
// 2) Initialize the substring and the length index.
$substringIndex = $lengthIndex = array();
// Iterate over the original string to identify
// the largest common string.
for ($originalPosition = 0; $originalPosition < $originalStringLength; $originalPosition++) {
// Find all occurrences of the original character
// in the target string.
$comparedCharacter = $originalStringCharacters[$originalPosition];
// Do we have a commonality between the original string
// and the edited string?
if (isset($characterIndex[$comparedCharacter])) {
// Loop over all commonalities.
foreach ($characterIndex[$comparedCharacter] as $editedPosition) {
// Calculate the current and the preceding position
// ids for indexation.
$currentPosition = $originalPosition . '-' . $editedPosition;
$previousPosition = $originalPosition - 1 . '-' . ($editedPosition - 1);
// Does the occurrence in the target string continue
// an existing common substring or does it start
// a new one?
if (isset($substringIndex[$previousPosition])) {
// This is a continuation of an existing common
// substring...
$newSubstring = $substringIndex[$previousPosition] . $comparedCharacter;
$newSubstringLength = String::strlen($newSubstring);
// Move the substring in the substring index.
$substringIndex[$currentPosition] = $newSubstring;
unset($substringIndex[$previousPosition]);
// Move the substring in the length index.
$lengthIndex[$newSubstringLength][$currentPosition] = $newSubstring;
unset($lengthIndex[$newSubstringLength - 1][$previousPosition]);
} else {
// Start a new common substring...
// Add the substring to the substring index.
$substringIndex[$currentPosition] = $comparedCharacter;
// Add the substring to the length index.
$lengthIndex[1][$currentPosition] = $comparedCharacter;
}
}
}
}
// If we have no commonalities at all then mark the original
// string as deleted and the edited string as added and
// return.
if (empty($lengthIndex)) {
return array(array(-1 => $originalString), array(1 => $editedString));
}
// Pop the largest common substrings from the length index.
end($lengthIndex);
$largestSubstringLength = key($lengthIndex);
// Take the first common substring if we have more than
// one substring with the same length.
// FIXME: Find a better heuristic for this decision.
reset($lengthIndex[$largestSubstringLength]);
$largestSubstringPosition = key($lengthIndex[$largestSubstringLength]);
list($largestSubstringEndOriginal, $largestSubstringEndEdited) = explode('-', $largestSubstringPosition);
$largestSubstring = $lengthIndex[$largestSubstringLength][$largestSubstringPosition];
// Add the largest common substring to the result set
$diffResult = array(array(0 => $largestSubstring));
// Prepend the diff of the substrings before the common substring
// to the result diff (by recursion).
$precedingSubstringOriginal = String::substr($originalString, 0, $largestSubstringEndOriginal - $largestSubstringLength + 1);
$precedingSubstringEdited = String::substr($editedString, 0, $largestSubstringEndEdited - $largestSubstringLength + 1);
$diffResult = array_merge(String::diff($precedingSubstringOriginal, $precedingSubstringEdited), $diffResult);
// Append the diff of the substrings after thr common substring
// to the result diff (by recursion).
$succeedingSubstringOriginal = String::substr($originalString, $largestSubstringEndOriginal + 1);
$succeedingSubstringEdited = String::substr($editedString, $largestSubstringEndEdited + 1);
$diffResult = array_merge($diffResult, String::diff($succeedingSubstringOriginal, $succeedingSubstringEdited));
// Return the array representing the diff.
return $diffResult;
}
/**
* Fills the given citation object with
* meta-data retrieved from PubMed.
* @param $pmid string
* @param $citationDescription MetadataDescription
* @return MetadataDescription
*/
function &_lookup($pmid, &$citationDescription)
{
$nullVar = null;
// Use eFetch to get XML metadata for the given PMID
$lookupParams = array('db' => 'pubmed', 'mode' => 'xml', 'tool' => 'pkp-wal', 'id' => $pmid);
if (!is_null($this->getEmail())) {
$lookupParams['email'] = $this->getEmail();
}
// Call the eFetch URL and get an XML result
if (is_null($resultDOM = $this->callWebService(PUBMED_WEBSERVICE_EFETCH, $lookupParams))) {
return $nullVar;
}
$metadata = array('pub-id[@pub-id-type="pmid"]' => $pmid, 'article-title' => $resultDOM->getElementsByTagName("ArticleTitle")->item(0)->textContent, 'source' => $resultDOM->getElementsByTagName("MedlineTA")->item(0)->textContent);
if ($resultDOM->getElementsByTagName("Volume")->length > 0) {
$metadata['volume'] = $resultDOM->getElementsByTagName("Volume")->item(0)->textContent;
}
if ($resultDOM->getElementsByTagName("Issue")->length > 0) {
$metadata['issue'] = $resultDOM->getElementsByTagName("Issue")->item(0)->textContent;
}
// get list of author full names
$nlmNameSchema = new NlmNameSchema();
foreach ($resultDOM->getElementsByTagName("Author") as $authorNode) {
if (!isset($metadata['person-group[@person-group-type="author"]'])) {
$metadata['person-group[@person-group-type="author"]'] = array();
}
// Instantiate an NLM name description
$authorDescription = new MetadataDescription($nlmNameSchema, ASSOC_TYPE_AUTHOR);
// Surname
$authorDescription->addStatement('surname', $authorNode->getElementsByTagName("LastName")->item(0)->textContent);
// Given names
$givenNamesString = '';
if ($authorNode->getElementsByTagName("FirstName")->length > 0) {
$givenNamesString = $authorNode->getElementsByTagName("FirstName")->item(0)->textContent;
} elseif ($authorNode->getElementsByTagName("ForeName")->length > 0) {
$givenNamesString = $authorNode->getElementsByTagName("ForeName")->item(0)->textContent;
}
if (!empty($givenNamesString)) {
foreach (explode(' ', $givenNamesString) as $givenName) {
$authorDescription->addStatement('given-names', String::trimPunctuation($givenName));
}
}
// Suffix
if ($authorNode->getElementsByTagName("Suffix")->length > 0) {
$authorDescription->addStatement('suffix', $authorNode->getElementsByTagName("Suffix")->item(0)->textContent);
}
// Include collective names
/*if ($resultDOM->getElementsByTagName("CollectiveName")->length > 0 && $authorNode->getElementsByTagName("CollectiveName")->item(0)->textContent != '') {
// FIXME: This corresponds to an NLM-citation <collab> tag and should be part of the Metadata implementation
}*/
$metadata['person-group[@person-group-type="author"]'][] =& $authorDescription;
unset($authorDescription);
}
// Extract pagination
if (String::regexp_match_get("/^[:p\\.\\s]*(?P<fpage>[Ee]?\\d+)(-(?P<lpage>\\d+))?/", $resultDOM->getElementsByTagName("MedlinePgn")->item(0)->textContent, $pages)) {
$fPage = (int) $pages['fpage'];
$metadata['fpage'] = $fPage;
if (!empty($pages['lpage'])) {
$lPage = (int) $pages['lpage'];
// Deal with shortcuts like '382-7'
if ($lPage < $fPage) {
$lPage = (int) (String::substr($pages['fpage'], 0, -String::strlen($pages['lpage'])) . $pages['lpage']);
}
$metadata['lpage'] = $lPage;
}
}
// Get publication date
// TODO: The publication date could be in multiple places
if ($resultDOM->getElementsByTagName("ArticleDate")->length > 0) {
$publicationDate = $resultDOM->getElementsByTagName("ArticleDate")->item(0)->getElementsByTagName("Year")->item(0)->textContent . '-' . $resultDOM->getElementsByTagName("ArticleDate")->item(0)->getElementsByTagName("Month")->item(0)->textContent . '-' . $resultDOM->getElementsByTagName("ArticleDate")->item(0)->getElementsByTagName("Day")->item(0)->textContent;
$metadata['date'] = $publicationDate;
}
// Get publication type
if ($resultDOM->getElementsByTagName("PublicationType")->length > 0) {
foreach ($resultDOM->getElementsByTagName("PublicationType") as $publicationType) {
// The vast majority of items on PubMed are articles so catch these...
if (String::strpos(String::strtolower($publicationType->textContent), 'article') !== false) {
$metadata['[@publication-type]'] = NLM_PUBLICATION_TYPE_JOURNAL;
break;
}
}
}
// Get DOI if it exists
foreach ($resultDOM->getElementsByTagName("ArticleId") as $idNode) {
if ($idNode->getAttribute('IdType') == 'doi') {
$metadata['pub-id[@pub-id-type="doi"]'] = $idNode->textContent;
}
}
// Use eLink utility to find fulltext links
$lookupParams = array('dbfrom' => 'pubmed', 'cmd' => 'llinks', 'tool' => 'pkp-wal', 'id' => $pmid);
if (!is_null($resultDOM = $this->callWebService(PUBMED_WEBSERVICE_ELINK, $lookupParams))) {
// Get a list of possible links
foreach ($resultDOM->getElementsByTagName("ObjUrl") as $linkOut) {
$attributes = '';
foreach ($linkOut->getElementsByTagName("Attribute") as $attribute) {
$attributes .= String::strtolower($attribute->textContent) . ' / ';
}
// Only add links to open access resources
if (String::strpos($attributes, "subscription") === false && String::strpos($attributes, "membership") === false && String::strpos($attributes, "fee") === false && $attributes != "") {
$links[] = $linkOut->getElementsByTagName("Url")->item(0)->textContent;
}
}
// Take the first link if we have any left (presumably pubmed returns them in preferential order)
if (isset($links[0])) {
$metadata['uri'] = $links[0];
}
}
return $this->addMetadataArrayToNlmCitationDescription($metadata, $citationDescription);
}
/**
* Retrieve auto-suggestions from the faceting service.
* @param $url string
* @param $searchRequest SolrSearchRequest
* @param $userInput string
* @param $fieldName string
* @return array The generated suggestions.
*/
function _getFacetingAutosuggestions($url, $searchRequest, $userInput, $fieldName)
{
// Remove special characters from the user input.
$searchTerms = strtr($userInput, '"()+-|&!', ' ');
// Cut off the last search term.
$searchTerms = explode(' ', $searchTerms);
$facetPrefix = array_pop($searchTerms);
if (empty($facetPrefix)) {
return array();
}
// Use the remaining search query to pre-filter
// facet results. This may be an invalid query
// but edismax will deal gracefully with syntax
// errors.
$userInput = String::substr($userInput, 0, -String::strlen($facetPrefix));
switch ($fieldName) {
case 'query':
// The 'query' filter goes agains all fields.
$articleSearch = new ArticleSearch();
$solrFields = array_values($articleSearch->getIndexFieldMap());
break;
case 'indexTerms':
// The 'index terms' filter goes against keyword index fields.
$solrFields = array('discipline', 'subject', 'type', 'coverage');
break;
default:
// All other filters can be used directly.
$solrFields = array($fieldName);
}
$solrFieldString = implode('|', $solrFields);
$searchRequest->addQueryFieldPhrase($solrFieldString, $userInput);
// Construct the main query.
$params = $this->_getSearchQueryParameters($searchRequest);
if (!isset($params['q'])) {
// Use a catch-all query in case we have no limiting
// search.
$params['q'] = '*:*';
}
if ($fieldName == 'query') {
$params['facet.field'] = 'default_spell';
} else {
$params['facet.field'] = $fieldName . '_spell';
}
$facetPrefixLc = String::strtolower($facetPrefix);
$params['facet.prefix'] = $facetPrefixLc;
// Make the request.
$response = $this->_makeRequest($url, $params);
if (!is_a($response, 'DOMXPath')) {
return array();
}
// Extract term suggestions.
$nodeList = $response->query('//lst[@name="facet_fields"]/lst/int/@name');
if ($nodeList->length == 0) {
return array();
}
$termSuggestions = array();
foreach ($nodeList as $childNode) {
$termSuggestions[] = $childNode->value;
}
// Add the term suggestion to the remaining user input.
$suggestions = array();
foreach ($termSuggestions as $termSuggestion) {
// Restore case if possible.
if (strpos($termSuggestion, $facetPrefixLc) === 0) {
$termSuggestion = $facetPrefix . String::substr($termSuggestion, String::strlen($facetPrefix));
}
$suggestions[] = $userInput . $termSuggestion;
}
return $suggestions;
}
/**
* Parse an XML file using the specified handler.
* If no handler has been specified, XMLParserDOMHandler is used by default, returning a tree structure representing the document.
* @param $file string full path to the XML file
* @return object actual return type depends on the handler
*/
function &parse($file)
{
$parser =& $this->createParser();
if (!isset($this->handler)) {
// Use default handler for parsing
$handler = new XMLParserDOMHandler();
$this->setHandler($handler);
}
xml_set_object($parser, $this->handler);
xml_set_element_handler($parser, "startElement", "endElement");
xml_set_character_data_handler($parser, "characterData");
import('lib.pkp.classes.file.FileWrapper');
$wrapper =& FileWrapper::wrapper($file);
// Handle responses of various types
while (true) {
$newWrapper = $wrapper->open();
if (is_object($newWrapper)) {
// Follow a redirect
unset($wrapper);
$wrapper =& $newWrapper;
unset($newWrapper);
} elseif (!$newWrapper) {
// Could not open resource -- error
$returner = false;
return $returner;
} else {
// OK, we've found the end result
break;
}
}
if (!$wrapper) {
$result = false;
return $result;
}
while (!$wrapper->eof() && ($data = $wrapper->read()) !== false) {
// if the string contains non-UTF8 characters, convert it to UTF-8 for parsing
if (Config::getVar('i18n', 'charset_normalization') == 'On' && !String::utf8_compliant($data)) {
$utf8_last = String::substr($data, String::strlen($data) - 1);
// if the string ends in a "bad" UTF-8 character, maybe it's truncated
while (!$wrapper->eof() && String::utf8_bad_find($utf8_last) === 0) {
// read another chunk of data
$data .= $wrapper->read();
$utf8_last = String::substr($data, String::strlen($data) - 1);
}
$data = String::utf8_normalize($data);
// strip any invalid UTF-8 sequences
$data = String::utf8_bad_strip($data);
// convert named entities to numeric entities
$data = strtr($data, String::getHTMLEntities());
}
// strip any invalid ASCII control characters
$data = String::utf8_strip_ascii_ctrl($data);
if (!xml_parse($parser, $data, $wrapper->eof())) {
$this->addError(xml_error_string(xml_get_error_code($parser)));
}
}
$wrapper->close();
$result =& $this->handler->getResult();
$this->destroyParser($parser);
if (isset($handler)) {
$handler->destroy();
unset($handler);
}
return $result;
}
function mb_substr_replace($string, $replacement, $start, $length = null)
{
if (extension_loaded('mbstring') === true) {
$string_length = String::strlen($string);
if ($start < 0) {
$start = max(0, $string_length + $start);
} else {
if ($start > $string_length) {
$start = $string_length;
}
}
if ($length < 0) {
$length = max(0, $string_length - $start + $length);
} else {
if (is_null($length) === true || $length > $string_length) {
$length = $string_length;
}
}
if ($start + $length > $string_length) {
$length = $string_length - $start;
}
return String::substr($string, 0, $start) . $replacement . String::substr($string, $start + $length, $string_length - $start - $length);
}
}
/**
* @function abntDateFormatWithDay Format date taking in consideration ABNT month abbreviations
* @param $string string
* @return string
*/
function abntDateFormatWithDay($string)
{
if (is_numeric($string)) {
// it is a numeric string, we handle it as timestamp
$timestamp = (int) $string;
} else {
$timestamp = strtotime($string);
}
$format = "%d %B %Y";
if (String::strlen(strftime("%B", $timestamp)) > 4) {
$format = "%d %b. %Y";
}
return String::strtolower(strftime($format, $timestamp));
}
<?php
class String
{
private $_string;
public function __construct($string)
{
$this->_string = $string;
}
public function __call($method, $arguments)
{
$this->_string = call_user_func($method, $this->_string);
return $this;
}
public function getValue()
{
return $this->_string;
}
}
$test = new String(' test, test2 ');
$test->trim();
var_dump($test->getValue());
$test->strlen();
var_dump($test->getValue());
/**
* Override the built-in smarty truncate modifier to support mbstring
* text properly, if possible.
*/
function smartyTruncate($string, $length = 80, $etc = '...', $break_words = false, $middle = false)
{
// Re-implement Smarty version, with multibyte-capable calls.
if ($length == 0) {
return '';
}
if (String::strlen($string) > $length) {
$length -= min($length, String::strlen($etc));
if (!$break_words && !$middle) {
$string = String::regexp_replace('/\\s+?(\\S+)?$/', '', substr($string, 0, $length + 1));
}
if (!$middle) {
return String::substr($string, 0, $length) . $etc;
} else {
return String::substr($string, 0, $length / 2) . $etc . String::substr($string, -$length / 2);
}
} else {
return $string;
}
}
/**
* Construct an array of search strings from a citation
* description and an array of search templates.
* The templates may contain the placeholders
* %aulast%: the first author's surname
* %au%: the first author full name
* %title%: the article-title (if it exists),
* otherwise the source
* %date%: the publication year
* %isbn%: ISBN
* @param $searchTemplates an array of templates
* @param $citationDescription MetadataDescription
* @return array
*/
function constructSearchStrings(&$searchTemplates, &$citationDescription)
{
// Convert first authors' name description to a string
import('lib.pkp.plugins.metadata.nlm30.filter.Nlm30NameSchemaPersonStringFilter');
$personStringFilter = new Nlm30NameSchemaPersonStringFilter();
// Retrieve the authors
$firstAuthorSurname = $firstAuthor = '';
$authors = $citationDescription->getStatement('person-group[@person-group-type="author"]');
if (is_array($authors) && count($authors)) {
$firstAuthorSurname = (string) $authors[0]->getStatement('surname');
$firstAuthor = $personStringFilter->execute($authors[0]);
}
// Retrieve the editors
$firstEditorSurname = $firstEditor = '';
$editors = $citationDescription->getStatement('person-group[@person-group-type="editor"]');
if (is_array($editors) && count($editors)) {
$firstEditorSurname = (string) $editors[0]->getStatement('surname');
$firstEditor = $personStringFilter->execute($editors[0]);
}
// Retrieve (default language) title
$title = (string) ($citationDescription->hasStatement('article-title') ? $citationDescription->getStatement('article-title') : $citationDescription->getStatement('source'));
// Extract the year from the publication date
$year = (string) $citationDescription->getStatement('date');
$year = String::strlen($year) > 4 ? String::substr($year, 0, 4) : $year;
// Retrieve ISBN
$isbn = (string) $citationDescription->getStatement('isbn');
// Replace the placeholders in the templates
$searchStrings = array();
foreach ($searchTemplates as $searchTemplate) {
// Try editors and authors separately
$searchStrings[] = str_replace(array('%aulast%', '%au%', '%title%', '%date%', '%isbn%'), array($firstAuthorSurname, $firstAuthor, $title, $year, $isbn), $searchTemplate);
$searchStrings[] = str_replace(array('%aulast%', '%au%', '%title%', '%date%', '%isbn%'), array($firstEditorSurname, $firstEditor, $title, $year, $isbn), $searchTemplate);
}
// Remove empty or duplicate searches
$searchStrings = array_map(array('String', 'trimPunctuation'), $searchStrings);
$searchStrings = array_unique($searchStrings);
$searchStrings = arrayClean($searchStrings);
return $searchStrings;
}
/**
* Checks whether the given string is an ISBN.
* @param $isbn
* @return boolean
*/
function isValidIsbn($isbn)
{
return is_string($isbn) && is_numeric($isbn) && String::strlen($isbn) == 13;
}
/**
* Fills the given citation object with
* meta-data retrieved from PubMed.
* @param $pmid string
* @return MetadataDescription
*/
function &_lookup($pmid)
{
$nullVar = null;
// Use eFetch to get XML metadata for the given PMID
$lookupParams = array('db' => 'pubmed', 'mode' => 'xml', 'tool' => 'pkp-wal', 'id' => $pmid);
if (!is_null($this->getEmail())) {
$lookupParams['email'] = $this->getEmail();
}
// Call the eFetch URL and get an XML result
if (is_null($resultDOM = $this->callWebService(PUBMED_WEBSERVICE_EFETCH, $lookupParams))) {
return $nullVar;
}
$articleTitleNodes =& $resultDOM->getElementsByTagName("ArticleTitle");
$articleTitleFirstNode =& $articleTitleNodes->item(0);
$medlineTaNodes =& $resultDOM->getElementsByTagName("MedlineTA");
$medlineTaFirstNode =& $medlineTaNodes->item(0);
$metadata = array('pub-id[@pub-id-type="pmid"]' => $pmid, 'article-title' => $articleTitleFirstNode->textContent, 'source' => $medlineTaFirstNode->textContent);
$volumeNodes =& $resultDOM->getElementsByTagName("Volume");
$issueNodes =& $resultDOM->getElementsByTagName("Issue");
if ($volumeNodes->length > 0) {
$volumeFirstNode =& $volumeNodes->item(0);
}
$metadata['volume'] = $volumeFirstNode->textContent;
if ($issueNodes->length > 0) {
$issueFirstNode =& $issueNodes->item(0);
}
$metadata['issue'] = $issueFirstNode->textContent;
// Get list of author full names
foreach ($resultDOM->getElementsByTagName("Author") as $authorNode) {
if (!isset($metadata['person-group[@person-group-type="author"]'])) {
$metadata['person-group[@person-group-type="author"]'] = array();
}
// Instantiate an NLM name description
$authorDescription = new MetadataDescription('lib.pkp.plugins.metadata.nlm30.schema.Nlm30NameSchema', ASSOC_TYPE_AUTHOR);
// Surname
$lastNameNodes =& $authorNode->getElementsByTagName("LastName");
$lastNameFirstNode =& $lastNameNodes->item(0);
$authorDescription->addStatement('surname', $lastNameFirstNode->textContent);
// Given names
$givenNamesString = '';
$firstNameNodes =& $authorNode->getElementsByTagName("FirstName");
if ($firstNameNodes->length > 0) {
$firstNameFirstNode =& $firstNameNodes->item(0);
$givenNamesString = $firstNameFirstNode->textContent;
} else {
$foreNameNodes =& $authorNode->getElementsByTagName("ForeName");
if ($foreNameNodes->length > 0) {
$foreNameFirstNode =& $foreNameNodes->item(0);
$givenNamesString = $foreNameFirstNode->textContent;
}
}
if (!empty($givenNamesString)) {
foreach (explode(' ', $givenNamesString) as $givenName) {
$authorDescription->addStatement('given-names', String::trimPunctuation($givenName));
}
}
// Suffix
$suffixNodes =& $authorNode->getElementsByTagName("Suffix");
if ($suffixNodes->length > 0) {
$suffixFirstNode =& $suffixNodes->item(0);
$authorDescription->addStatement('suffix', $suffixFirstNode->textContent);
}
// Include collective names
// FIXME: This corresponds to an NLM-citation <collab> tag and should be part of the Metadata implementation
/*if ($resultDOM->getElementsByTagName("CollectiveName")->length > 0 && $authorNode->getElementsByTagName("CollectiveName")->item(0)->textContent != '') {
}*/
$metadata['person-group[@person-group-type="author"]'][] =& $authorDescription;
unset($authorDescription);
}
// Extract pagination
$medlinePgnNodes =& $resultDOM->getElementsByTagName("MedlinePgn");
$medlinePgnFirstNode =& $medlinePgnNodes->item(0);
if (String::regexp_match_get("/^[:p\\.\\s]*(?P<fpage>[Ee]?\\d+)(-(?P<lpage>\\d+))?/", $medlinePgnFirstNode->textContent, $pages)) {
$fPage = (int) $pages['fpage'];
$metadata['fpage'] = $fPage;
if (!empty($pages['lpage'])) {
$lPage = (int) $pages['lpage'];
// Deal with shortcuts like '382-7'
if ($lPage < $fPage) {
$lPage = (int) (String::substr($pages['fpage'], 0, -String::strlen($pages['lpage'])) . $pages['lpage']);
}
$metadata['lpage'] = $lPage;
}
}
// Get publication date (can be in several places in PubMed).
$dateNode = null;
$articleDateNodes =& $resultDOM->getElementsByTagName("ArticleDate");
if ($articleDateNodes->length > 0) {
$dateNode =& $articleDateNodes->item(0);
} else {
$pubDateNodes =& $resultDOM->getElementsByTagName("PubDate");
if ($pubDateNodes->length > 0) {
$dateNode =& $pubDateNodes->item(0);
}
}
// Retrieve the data parts and assemble date.
if (!is_null($dateNode)) {
$publicationDate = '';
$requiresNormalization = false;
foreach (array('Year' => 4, 'Month' => 2, 'Day' => 2) as $dateElement => $padding) {
$dateElementNodes =& $dateNode->getElementsByTagName($dateElement);
if ($dateElementNodes->length > 0) {
if (!empty($publicationDate)) {
$publicationDate .= '-';
}
$dateElementFirstNode =& $dateElementNodes->item(0);
$datePart = str_pad($dateElementFirstNode->textContent, $padding, '0', STR_PAD_LEFT);
if (!is_numeric($datePart)) {
$requiresNormalization = true;
}
$publicationDate .= $datePart;
} else {
break;
}
}
// Normalize the date to NLM standard if necessary.
if ($requiresNormalization) {
$dateFilter = new DateStringNormalizerFilter();
$publicationDate = $dateFilter->execute($publicationDate);
}
if (!empty($publicationDate)) {
$metadata['date'] = $publicationDate;
}
}
// Get publication type
$publicationTypeNodes =& $resultDOM->getElementsByTagName("PublicationType");
if ($publicationTypeNodes->length > 0) {
foreach ($publicationTypeNodes as $publicationType) {
// The vast majority of items on PubMed are articles so catch these...
if (String::strpos(String::strtolower($publicationType->textContent), 'article') !== false) {
$metadata['[@publication-type]'] = NLM30_PUBLICATION_TYPE_JOURNAL;
break;
}
}
}
// Get DOI if it exists
$articleIdNodes =& $resultDOM->getElementsByTagName("ArticleId");
foreach ($articleIdNodes as $idNode) {
if ($idNode->getAttribute('IdType') == 'doi') {
$metadata['pub-id[@pub-id-type="doi"]'] = $idNode->textContent;
}
}
// Use eLink utility to find fulltext links
$lookupParams = array('dbfrom' => 'pubmed', 'cmd' => 'llinks', 'tool' => 'pkp-wal', 'id' => $pmid);
if (!is_null($resultDOM = $this->callWebService(PUBMED_WEBSERVICE_ELINK, $lookupParams))) {
// Get a list of possible links
foreach ($resultDOM->getElementsByTagName("ObjUrl") as $linkOut) {
$attributes = '';
foreach ($linkOut->getElementsByTagName("Attribute") as $attribute) {
$attributes .= String::strtolower($attribute->textContent) . ' / ';
}
// Only add links to open access resources
if (String::strpos($attributes, "subscription") === false && String::strpos($attributes, "membership") === false && String::strpos($attributes, "fee") === false && $attributes != "") {
$urlNodes =& $linkOut->getElementsByTagName("Url");
$urlFirstNode =& $urlNodes->item(0);
$links[] = $urlFirstNode->textContent;
}
}
// Take the first link if we have any left (presumably pubmed returns them in preferential order)
if (isset($links[0])) {
$metadata['uri'] = $links[0];
}
}
return $this->getNlm30CitationDescriptionFromMetadataArray($metadata);
}
function parseDate($fieldName, $value, $attributes = null)
{
switch ($fieldName) {
case '005':
// YYYYMMDDHHMMSS.0 Date and time of latest transaction
if (String::strlen($value) < 14) {
return null;
}
$year = String::substr($value, 0, 4);
$month = String::substr($value, 4, 2);
$day = String::substr($value, 6, 2);
$hour = String::substr($value, 8, 2);
$minute = String::substr($value, 10, 2);
$second = String::substr($value, 12);
// Make sure the values fetched are all numeric
foreach (array('year', 'month', 'day', 'hour', 'minute', 'second') as $var) {
if (!is_numeric(${$var})) {
return null;
}
}
return mktime($hour, $minute, $second, $month, $day, $year);
case '008':
// YYMMDD[junk] Date entered on file
$date = String::substr($value, 0, 6);
$date = strtotime($date);
if ($date !== -1 && $date !== false) {
return $date;
}
break;
case '260':
if (isset($attributes['label']) && $attributes['label'] == 'c') {
$date = strtotime($value);
if ($date !== -1 && $date !== false) {
return $date;
}
}
break;
}
return null;
}
/**
* Get the localized, truncated description of the book for review.
* @return string
*/
function getLocalizedDescriptionShort()
{
$end = '';
if (String::strlen($this->getLocalizedData('description'))) {
$end = ' ...';
}
return String::substr($this->getLocalizedData('description'), 0, 250) . $end;
}
/**
* Transform a single NLM name description to a person string.
* NB: We use the style: surname suffix, initials (first-name) prefix
* which is relatively easy to parse back.
* @param $personDescription MetadataDescription
* @return string
*/
function _flattenPersonDescription(&$personDescription)
{
$nameVars['%surname%'] = (string) $personDescription->getStatement('surname');
$givenNames = $personDescription->getStatement('given-names');
$nameVars['%firstname%'] = $nameVars['%initials%'] = '';
if (is_array($givenNames) && count($givenNames)) {
if (String::strlen($givenNames[0]) > 1) {
$nameVars['%firstname%'] = array_shift($givenNames);
}
foreach ($givenNames as $givenName) {
$nameVars['%initials%'] .= String::substr($givenName, 0, 1) . '.';
}
}
if (!empty($nameVars['%initials%'])) {
$nameVars['%initials%'] = ' ' . $nameVars['%initials%'];
}
$nameVars['%prefix%'] = (string) $personDescription->getStatement('prefix');
if (!empty($nameVars['%prefix%'])) {
$nameVars['%prefix%'] = ' ' . $nameVars['%prefix%'];
}
$nameVars['%suffix%'] = (string) $personDescription->getStatement('suffix');
if (!empty($nameVars['%suffix%'])) {
$nameVars['%suffix%'] = ' ' . $nameVars['%suffix%'];
}
// Fill placeholders in person template.
$personString = str_replace(array_keys($nameVars), array_values($nameVars), $this->getTemplate());
// Remove empty brackets and trailing/leading whitespace
$personString = trim(str_replace('()', '', $personString));
return $personString;
}
/**
* Converts a string with a single person
* to an NLM name description.
*
* TODO: add initials from all given names to initials
* element
*
* @param $personString string
* @param $title boolean true to parse for title
* @param $degrees boolean true to parse for degrees
* @return MetadataDescription an NLM name description or null
* if the string could not be converted
*/
function &_parsePersonString($personString, $title, $degrees)
{
// Expressions to parse person strings, ported from CiteULike person
// plugin, see http://svn.citeulike.org/svn/plugins/person.tcl
static $personRegex = array('title' => '(?:His (?:Excellency|Honou?r)\\s+|Her (?:Excellency|Honou?r)\\s+|The Right Honou?rable\\s+|The Honou?rable\\s+|Right Honou?rable\\s+|The Rt\\.? Hon\\.?\\s+|The Hon\\.?\\s+|Rt\\.? Hon\\.?\\s+|Mr\\.?\\s+|Ms\\.?\\s+|M\\/s\\.?\\s+|Mrs\\.?\\s+|Miss\\.?\\s+|Dr\\.?\\s+|Sir\\s+|Dame\\s+|Prof\\.?\\s+|Professor\\s+|Doctor\\s+|Mister\\s+|Mme\\.?\\s+|Mast(?:\\.|er)?\\s+|Lord\\s+|Lady\\s+|Madam(?:e)?\\s+|Priv\\.-Doz\\.\\s+)+', 'degrees' => '(,\\s+(?:[A-Z\\.]+))+', 'initials' => '(?:(?:[A-Z]\\.){1,3}[A-Z]\\.?)|(?:(?:[A-Z]\\.\\s){1,3}[A-Z]\\.?)|(?:[A-Z]{1,4})|(?:(?:[A-Z]\\.-?){1,4})|(?:(?:[A-Z]\\.-?){1,3}[A-Z]\\.?)|(?:(?:[A-Z]-){1,3}[A-Z])|(?:(?:[A-Z]\\s){1,3}[A-Z]\\.?)|(?:(?:[A-Z]-){1,3}[A-Z]\\.?)', 'prefix' => 'Dell(?:[a|e])?(?:\\s|$)|Dalle(?:\\s|$)|D[a|e]ll\'(?:\\s|$)|Dela(?:\\s|$)|Del(?:\\s|$)|[Dd]e(?:\\s|$)(?:La(?:\\s|$)|Los(?:\\s|$))?|[Dd]e(?:\\s|$)|[Dd][a|i|u](?:\\s|$)|L[a|e|o](?:\\s|$)|[D|L|O]\'|St\\.?(?:\\s|$)|San(?:\\s|$)|[Dd]en(?:\\s|$)|[Vv]on(?:\\s|$)(?:[Dd]er(?:\\s|$))?|(?:[Ll][ea](?:\\s|$))?[Vv]an(?:\\s|$)(?:[Dd]e(?:n|r)?(?:\\s|$))?', 'givenName' => '(?:[^ \\t\\n\\r\\f\\v,.;()]{2,}|[^ \\t\\n\\r\\f\\v,.;()]{2,}\\-[^ \\t\\n\\r\\f\\v,.;()]{2,})');
// The expressions for given name, suffix and surname are the same
$personRegex['surname'] = $personRegex['suffix'] = $personRegex['givenName'];
$personRegex['double-surname'] = "(?:" . $personRegex['surname'] . "\\s)*" . $personRegex['surname'];
// Shortcut for prefixed surname
$personRegexPrefixedSurname = "(?P<prefix>(?:" . $personRegex['prefix'] . ")?)(?P<surname>" . $personRegex['surname'] . ")";
$personRegexPrefixedDoubleSurname = "(?P<prefix>(?:" . $personRegex['prefix'] . ")?)(?P<surname>" . $personRegex['double-surname'] . ")";
// Instantiate the target person description
$personDescription = new MetadataDescription('lib.pkp.plugins.metadata.nlm30.schema.Nlm30NameSchema', $this->_assocType);
// Clean the person string
$personString = trim($personString);
// 1. Extract title and degree from the person string and use this as suffix
$suffixString = '';
$results = array();
if ($title && String::regexp_match_get('/^(' . $personRegex['title'] . ')/i', $personString, $results)) {
$suffixString = trim($results[1], ',:; ');
$personString = String::regexp_replace('/^(' . $personRegex['title'] . ')/i', '', $personString);
}
if ($degrees && String::regexp_match_get('/(' . $personRegex['degrees'] . ')$/i', $personString, $results)) {
$degreesArray = explode(',', trim($results[1], ','));
foreach ($degreesArray as $key => $degree) {
$degreesArray[$key] = String::trimPunctuation($degree);
}
$suffixString .= ' - ' . implode('; ', $degreesArray);
$personString = String::regexp_replace('/(' . $personRegex['degrees'] . ')$/i', '', $personString);
}
if (!empty($suffixString)) {
$personDescription->addStatement('suffix', $suffixString);
}
// Space initials when followed by a given name or last name.
$personString = String::regexp_replace('/([A-Z])\\.([A-Z][a-z])/', '\\1. \\2', $personString);
// 2. Extract names and initials from the person string
// The parser expressions are ordered by specificity. The most specific expressions
// come first. Only if these specific expressions don't work will we turn to less
// specific ones. This avoids parsing errors. It also explains why we don't use the
// ?-quantifier for optional elements like initials or middle name where they could
// be misinterpreted.
$personExpressions = array('/^' . $personRegexPrefixedSurname . '$/i', '/^(?P<initials>' . $personRegex['initials'] . ')\\s' . $personRegexPrefixedSurname . '$/', '/^' . $personRegexPrefixedSurname . ',?\\s(?P<initials>' . $personRegex['initials'] . ')$/', '/^' . $personRegexPrefixedDoubleSurname . ',\\s(?P<givenName>' . $personRegex['givenName'] . ')\\s(?P<initials>' . $personRegex['initials'] . ')$/', '/^(?P<givenName>' . $personRegex['givenName'] . ')\\s(?P<initials>' . $personRegex['initials'] . ')\\s' . $personRegexPrefixedSurname . '$/', '/^' . $personRegexPrefixedDoubleSurname . ',\\s(?P<givenName>(?:' . $personRegex['givenName'] . '\\s)+)(?P<initials>' . $personRegex['initials'] . ')$/', '/^(?P<givenName>(?:' . $personRegex['givenName'] . '\\s)+)(?P<initials>' . $personRegex['initials'] . ')\\s' . $personRegexPrefixedSurname . '$/', '/^' . $personRegexPrefixedDoubleSurname . ',(?P<givenName>(?:\\s' . $personRegex['givenName'] . ')+)$/', '/^(?P<givenName>(?:' . $personRegex['givenName'] . '\\s)+)' . $personRegexPrefixedSurname . '$/', '/^\\s*(?P<surname>' . $personRegex['surname'] . ')(?P<suffix>(?:\\s+' . $personRegex['suffix'] . ')?)\\s*,\\s*(?P<initials>(?:' . $personRegex['initials'] . ')?)\\s*\\((?P<givenName>(?:\\s*' . $personRegex['givenName'] . ')+)\\s*\\)\\s*(?P<prefix>(?:' . $personRegex['prefix'] . ')?)$/', '/^(?P<givenName>' . $personRegex['givenName'] . ')\\.(?P<surname>' . $personRegex['double-surname'] . ')$/', '/^(?P<surname>.*)$/');
$results = array();
foreach ($personExpressions as $expressionId => $personExpression) {
if ($nameFound = String::regexp_match_get($personExpression, $personString, $results)) {
// Given names
if (!empty($results['givenName'])) {
// Split given names
$givenNames = explode(' ', trim($results['givenName']));
foreach ($givenNames as $givenName) {
$personDescription->addStatement('given-names', $givenName);
unset($givenName);
}
}
// Initials (will also be saved as given names)
if (!empty($results['initials'])) {
$results['initials'] = str_replace(array('.', '-', ' '), array('', '', ''), $results['initials']);
for ($initialNum = 0; $initialNum < String::strlen($results['initials']); $initialNum++) {
$initial = $results['initials'][$initialNum];
$personDescription->addStatement('given-names', $initial);
unset($initial);
}
}
// Surname
if (!empty($results['surname'])) {
// Correct all-upper surname
if (strtoupper($results['surname']) == $results['surname']) {
$results['surname'] = ucwords(strtolower($results['surname']));
}
$personDescription->addStatement('surname', $results['surname']);
}
// Prefix/Suffix
foreach (array('prefix', 'suffix') as $propertyName) {
if (!empty($results[$propertyName])) {
$results[$propertyName] = trim($results[$propertyName]);
$personDescription->addStatement($propertyName, $results[$propertyName]);
}
}
break;
}
}
return $personDescription;
}
/**
* Derive a confidence score calculated as the similarity of the
* original raw citation and the citation text generated from the
* citation description.
* @param $metadataDescription MetadataDescription
* @return integer filter confidence score
*/
function _filterConfidenceScore(&$metadataDescription)
{
// Retrieve the original plain text citation.
$originalCitation = $this->getOriginalRawCitation();
// Generate the formatted citation output from the description.
$citationOutputFilter =& $this->getCitationOutputFilter();
$generatedCitation = $citationOutputFilter->execute($metadataDescription);
// Strip formatting and the Google Scholar tag so that we get a plain
// text string that is comparable with the raw citation.
$generatedCitation = trim(str_replace(GOOGLE_SCHOLAR_TAG, '', strip_tags($generatedCitation)));
// Compare the original to the generated citation.
$citationDiff = String::diff($originalCitation, $generatedCitation);
// Calculate similarity as the number of deleted characters in relation to the
// number of characters in the original citation. This intentionally excludes
// additions as these can represent useful data like a DOI or an external link.
$deletedCharacters = 0;
foreach ($citationDiff as $diffPart) {
// Identify deletions.
if (key($diffPart) == -1) {
$deletedCharacters += String::strlen(current($diffPart));
}
}
$originalCharacters = String::strlen($originalCitation);
$partOfCommonCharacters = ($originalCharacters - $deletedCharacters) / $originalCharacters;
$filterConfidenceScore = (int) round(min($partOfCommonCharacters * 100, 100));
return $filterConfidenceScore;
}
/**
* Determine whether or not the lengths of the two supplied values are
* "similar".
* @param $reference string
* @param $value string
* @return boolean True if the lengths match very roughly.
*/
function checkLengths($reference, $value)
{
$referenceLength = String::strlen($reference);
$length = String::strlen($value);
$lengthDifference = abs($referenceLength - $length);
if ($referenceLength == 0) {
return $length == 0;
}
if ($lengthDifference / $referenceLength > 1 && $lengthDifference > 10) {
return false;
}
return true;
}
function truncate($value, $length = 80, $ellipsis = '...')
{
if (String::strlen($value) > $length) {
$value = String::substr($value, 0, $length - String::strlen($ellipsis));
return $value . $ellipsis;
}
return $value;
}
/**
* Helper function: Reinsert tags from the tag array into their original position in the string
* @author Matt Crider
* @param string
* @param array
* @param boolean Set to true to reinsert tags starting at the back of the string
* @return string
*/
function _reinsertTags($string, &$tags, $reverse = false)
{
if (empty($tags)) {
return $string;
}
for ($i = 0; $i < count($tags); $i++) {
$length = String::strlen($string);
if ($tags[$i][1] < String::strlen($string)) {
if ($reverse) {
if ($tags[$i][1] == 0) {
// Cannot use -0 as the start index (its same as +0)
$string = String::substr_replace($string, $tags[$i][0], $length, 0);
} else {
$string = String::substr_replace($string, $tags[$i][0], -$tags[$i][1], 0);
}
} else {
$string = String::substr_replace($string, $tags[$i][0], $tags[$i][1], 0);
}
}
}
return $string;
}
private function parseCondition()
{
$page = intval(Req::args("p"));
$page_size = 36;
$sort = Filter::int(Req::args("sort"));
$sort = $sort == null ? 0 : $sort;
$cid = Filter::int(Req::args("cid"));
$cid = $cid == null ? 0 : $cid;
$brand = Filter::int(Req::args("brand"));
$price = Req::args("price");
//下面已进行拆分过滤
$keyword = urldecode(Req::args('keyword'));
$keyword = Filter::text($keyword);
$keyword = Filter::commonChar($keyword);
//初始化数据
$attrs = $specs = $spec_attr = $category_child = $spec_attr_selected = $selected = $has_category = $category = $current_category = array();
$where = $spec_attr_where = $url = "";
$condition_num = 0;
$model = $this->model;
//基本条件的建立
//关于搜索的处理
$action = strtolower(Req::args("act"));
if ($action == 'search') {
// xuzhongyi
$seo_title = "分类检索";
$seo_keywords = "全部分类";
$seo_description = "所有分类商品";
//取得商品的子分类
$category_ids = "";
$categ = Category::getInstance();
//set:$cid = 5;
if ($cid == 0) {
$category_child = $categ->getCategoryChild(0, 1);
} else {
$current_category = $this->model->table("goods_category as gc")->fields("gc.*,gt.name as gname,gt.attr,gt.spec,gc.seo_title,gc.seo_keywords,gc.seo_description")->join("left join goods_type as gt on gc.type_id = gt.id")->where("gc.id = {$cid}")->find();
if ($current_category) {
$path = trim($current_category['path'], ',');
$rows = $this->model->table("goods_category")->where("path like '{$current_category['path']}%'")->order("field(`id`,{$path})")->findAll();
$category = $this->model->table("goods_category")->where("id in ({$path})")->order("field(`id`,{$path})")->findAll();
foreach ($rows as $row) {
$category_ids .= $row['id'] . ',';
}
$category_ids = trim($category_ids, ",");
$category_child = $categ->getCategoryChild($path, 1);
$attrs = unserialize($current_category['attr']);
$specs = unserialize($current_category['spec']);
$attrs = is_array($attrs) ? $attrs : array();
$specs = is_array($specs) ? $specs : array();
}
}
$seo_category = $model->table('goods_category')->where("id={$cid}")->find();
if ($seo_category) {
if ($seo_category['seo_title'] != '') {
$seo_title = $seo_category['seo_title'];
} else {
$seo_title = $seo_category['name'];
}
if ($seo_category['seo_keywords'] != '') {
$seo_keywords = $seo_category['name'] . ',' . $seo_category['seo_keywords'];
} else {
$seo_keywords = $seo_category['name'];
}
if ($seo_category['seo_description'] != '') {
$seo_description = $seo_category['seo_description'];
} else {
$seo_description = $seo_category['name'];
}
}
if ($category_ids != "") {
$where = "go.category_id in ({$category_ids})";
} else {
$where = "1=1";
}
//关于类型的处理
////提取商品下的类型
$seo_title = $seo_keywords = $keyword;
$where = "name like '%{$keyword}%'";
$rows = $model->table("goods")->fields("category_id,count(id) as num")->where($where)->group("category_id")->findAll();
$category_ids = "";
$category_count = array();
foreach ($rows as $row) {
$category_ids .= $row['category_id'] . ',';
$category_count[$row['category_id']] = $row['num'];
}
$category_ids = trim($category_ids, ",");
$has_category = array();
$seo_description = '';
if ($category_ids) {
//搜索到内容且真正的点击搜索时进行统计
if ($this->getModule()->checkToken()) {
$keyword = urldecode(Req::args('keyword'));
$keyword = Filter::sql($keyword);
$keyword = trim($keyword);
$len = String::strlen($keyword);
if ($len >= 2 && $len <= 8) {
$model = new Model("tags");
$obj = $model->where("name='{$keyword}'")->find();
if ($obj) {
$model->data(array('num' => "`num`+1"))->where("id=" . $obj['id'])->update();
} else {
$model->data(array('name' => $keyword))->insert();
}
}
}
$rows = $model->table("goods_category")->where("id in ({$category_ids})")->findAll();
foreach ($rows as $row) {
$path = trim($row['path'], ',');
$paths = explode(',', $path);
$root = 0;
if (is_array($paths)) {
$root = $paths[0];
}
$row['num'] = $category_count[$row['id']];
$has_category[$root][] = $row;
$seo_description .= $row['name'] . ',';
}
}
if ($cid != 0) {
$where = "category_id={$cid} and name like '%{$keyword}%'";
$category = $model->table("goods_category as gc ")->join("left join goods_type as gt on gc.type_id = gt.id")->where("gc.id={$cid}")->find();
if ($category) {
$attrs = unserialize($category['attr']);
$specs = unserialize($category['spec']);
if ($category['seo_title'] != '') {
$seo_title = $category['seo_title'];
} else {
$seo_title = $category['name'];
}
if ($category['seo_keywords'] != '') {
$seo_keywords = $category['seo_keywords'];
}
if ($category['seo_description'] != '') {
$seo_description = $category['seo_description'];
}
}
}
//关于分类检索的处理
} else {
if ($action == 'category') {
$seo_title = "分类检索";
$seo_keywords = "全部分类";
$seo_description = "所有分类商品";
//取得商品的子分类
$category_ids = "";
$categ = Category::getInstance();
if ($cid == 0) {
$category_child = $categ->getCategoryChild(0, 1);
} else {
$current_category = $this->model->table("goods_category as gc")->fields("gc.*,gt.name as gname,gt.attr,gt.spec,gc.seo_title,gc.seo_keywords,gc.seo_description")->join("left join goods_type as gt on gc.type_id = gt.id")->where("gc.id = {$cid}")->find();
if ($current_category) {
$path = trim($current_category['path'], ',');
$rows = $this->model->table("goods_category")->where("path like '{$current_category['path']}%'")->order("field(`id`,{$path})")->findAll();
$category = $this->model->table("goods_category")->where("id in ({$path})")->order("field(`id`,{$path})")->findAll();
foreach ($rows as $row) {
$category_ids .= $row['id'] . ',';
}
$category_ids = trim($category_ids, ",");
$category_child = $categ->getCategoryChild($path, 1);
$attrs = unserialize($current_category['attr']);
$specs = unserialize($current_category['spec']);
$attrs = is_array($attrs) ? $attrs : array();
$specs = is_array($specs) ? $specs : array();
}
}
$seo_category = $model->table('goods_category')->where("id={$cid}")->find();
if ($seo_category) {
if ($seo_category['seo_title'] != '') {
$seo_title = $seo_category['seo_title'];
} else {
$seo_title = $seo_category['name'];
}
if ($seo_category['seo_keywords'] != '') {
$seo_keywords = $seo_category['name'] . ',' . $seo_category['seo_keywords'];
} else {
$seo_keywords = $seo_category['name'];
}
if ($seo_category['seo_description'] != '') {
$seo_description = $seo_category['seo_description'];
} else {
$seo_description = $seo_category['name'];
}
}
if ($category_ids != "") {
$where = "go.category_id in ({$category_ids})";
} else {
$where = "1=1";
}
}
}
//品牌筛选
$rows = $model->table("goods as go")->fields("brand_id,count(id) as num")->where($where)->group("brand_id")->findAll();
$brand_ids = '';
$brand_num = $has_brand = array();
foreach ($rows as $row) {
$brand_ids .= $row['brand_id'] . ',';
$brand_num[$row['brand_id']] = $row['num'];
}
$brand_ids = trim($brand_ids, ',');
//价格区间
$prices = $model->table("goods as go")->fields("max(sell_price) as max,min(sell_price) as min,avg(sell_price) as avg")->where($where)->find();
$price_range = Common::priceRange($prices);
if ($brand_ids) {
$has_brand = $model->table("brand")->where("id in ({$brand_ids})")->findAll();
}
//var_dump($price_range);exit();
if (!empty($price_range)) {
$has_price = array_flip($price_range);
} else {
$has_price = array();
}
if ($price && isset($has_price[$price])) {
$prices = explode('-', $price);
if (count($prices) == 2) {
$where .= " and sell_price>=" . Filter::int($prices[0]) . " and sell_price <=" . Filter::int($prices[1]);
} else {
$where .= " and sell_price>=" . Filter::int($prices[0]);
}
$url .= "/price/{$price}";
}
if ($brand && isset($brand_num[$brand])) {
$url .= "/brand/{$brand}";
$where .= " and brand_id = {$brand} ";
}
//规格与属性的处理
if ($attrs != null) {
foreach ($attrs as $attr) {
if ($attr['show_type'] == 1) {
$spec_attr[$attr['id']] = $attr;
}
}
}
if ($specs != null) {
foreach ($specs as $spec) {
$spec['values'] = unserialize($spec['value']);
unset($spec['value'], $spec['spec']);
$spec_attr[$spec['id']] = $spec;
}
}
foreach ($selected as $key => $value) {
if (isset($spec_attr[$key])) {
$spec_attr_selected[$key] = $spec_attr[$key];
foreach ($spec_attr_selected[$key]['values'] as $k => $v) {
if ($value == $v['id']) {
$spec_attr_selected[$key]['values'] = $v;
break;
}
}
}
}
//规格处属性的筛选
$args = Req::args();
unset($args['con'], $args['act'], $args['p'], $args['sort'], $args['brand'], $args['price']);
foreach ($args as $key => $value) {
if (is_numeric($key) && is_numeric($value)) {
if (isset($spec_attr[$key])) {
$spec_attr_where .= "or (`key`={$key} and `value` = {$value}) ";
$condition_num++;
$url .= '/' . $key . '/' . $value;
}
}
$selected[$key] = $value;
}
$selected['price'] = $price;
$selected['brand'] = $brand;
$spec_attr_where = trim($spec_attr_where, "or");
$where .= ' and go.is_online =0';
if ($condition_num > 0) {
$where .= " and go.id in (select goods_id from tiny_spec_attr where {$spec_attr_where} group by goods_id having count(goods_id) >= {$condition_num})";
}
//排序的处理
switch ($sort) {
case '1':
$goods_model = $model->table("goods as go")->join("left join tiny_order_goods as og on go.id = og.goods_id")->fields("go.*,sum(og.goods_nums) as sell_num")->order("sell_num desc")->group("go.id");
break;
case '2':
$goods_model = $model->table("goods as go")->join("left join tiny_review as re on go.id = re.goods_id")->fields("go.*,count(re.goods_id) as renum")->group("go.id")->order("renum desc");
break;
case '3':
$goods_model = $model->table("goods as go")->order("sell_price desc");
break;
case '4':
$goods_model = $model->table("goods as go")->order("sell_price");
break;
case '5':
$goods_model = $model->table("goods as go")->order("id desc");
break;
default:
$goods_model = $model->table("goods as go")->order("sort desc");
break;
}
//var_dump($where);exit;
//提取商品
$goods = $goods_model->where($where)->findPage($page, $page_size);
//品牌处理
preg_match_all('!(<(a|span)[^>]+>(上一页|下一页)</\\2>)!', $goods['html'], $matches);
$topPageBar = "";
if (count($matches[0]) > 0) {
$topPageBar = implode("", $matches[0]);
}
$this->assign("topPageBar", $topPageBar);
//赋值处理
$this->assign('seo_title', $seo_title);
$this->assign('seo_keywords', $seo_keywords);
$this->assign('seo_description', '对应的商品共有' . $goods['page']['total'] . '件商品,包括以下分类：' . $seo_description);
$this->assign("keyword", $keyword);
$this->assign("sort", $sort);
$this->assign("has_brand", $has_brand);
$this->assign("brand_num", $brand_num);
$this->assign("current_category", $current_category);
$this->assign("goods", $goods);
$this->assign("selected", $selected);
$this->assign("spec_attr", $spec_attr);
$this->assign("spec_attr_selected", $spec_attr_selected);
$this->assign("category_child", $category_child);
$this->assign("price_range", $price_range);
$this->assign("category_nav", $category);
$this->assign("has_category", $has_category);
if ($action == 'search') {
$this->assign("url", "/index/search/keyword/" . $keyword . "/cid/{$cid}/sort/{$sort}" . $url);
} else {
$this->assign("url", "/index/category/cid/" . $cid . "/sort/{$sort}" . $url);
}
$this->redirect();
}
