Drupal modules for browsing and managing Fedora-based digital repositories.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

309 lines
9.8 KiB

<?php
// $Id$
/**
* @file
* Base utilities used by the Islandora fedora module.
*/
if (!function_exists('str_getcsv')) {
/**
* Functions that emulate php5.3 functionality for backwards compatiablity
* @param type $input
* @param type $delimiter
* @param type $enclosure
* @param type $escape
* @param type $eol
* @return type
*/
function str_getcsv($input, $delimiter=',', $enclosure='"', $escape=NULL, $eol=NULL) {
$temp = fopen("php://memory", "rw");
fwrite($temp, $input);
fseek($temp, 0);
$r = fgetcsv($temp, 4096, $delimiter, $enclosure);
fclose($temp);
return $r;
}
}
/**
*
* Utility method to get data from a url location using curl_exec and return
*
* This method takes identical parameters to do_curl_ext and passes them
* directly to do_curl_ext. The output from do_curl_ext is processed into
* either a TRUE, FALSE, or NULL value.
*
* This method exists for historical reasons as existing code makes use of
* do_curl directly and expects a simple return value and not an array.
*
* @param $url
* URL to be accessed by the function
* @param $return_to_variable
* Indicates whether the resource accessed by the curl call (HTML page,
* XML page, etc.) is returned directly from the function to the calling
* program or if it is sent directly to output.
* @param $number_of_post_vars
* Number of variable sot be posted
* @param $post
* Whether the curl_exec is done as a "get" or a "post"
*
* @return
* TRUE, FALSE, NULL, or the data returned from accessing the URL
*/
function do_curl($url, $return_to_variable = 1, $number_of_post_vars = 0, $post = NULL) {
$return_array = do_curl_ext($url, $return_to_variable, $number_of_post_vars, $post);
return ($return_array != NULL) ? $return_array[0] : NULL;
}
/**
*
* Utility method to get data from a url location using curl_exec
*
* This method takes a URL and three associated parameters and initializes the
* structure required to make a call to curl_exec. As a part of the
* initialization the permissions associated with the (global) user are added
* to the call. This ensures access to any potentially user restricted URLs.
*
* Various defaults are used for the parameters required by curl_exec including
* the user agent and timeout. These are hard-coded.
*
* @param $url
* URL to be accessed by the function
* @param $return_to_variable
* Indicates whether the resource accessed by the curl call (HTML page,
* XML page, etc.) is returned directly from the function to the calling
* program or if it is sent directly to output.
* @param $number_of_post_vars
* Number of variable sot be posted
* @param $post
* Whether the curl_exec is done as a "get" or a "post"
*
* @return
* an array that consists of three value or NULL if curl is not suported:
* - element 0:
* The value returned from the curl_exec function call.
* This is either a TRUE or FALSE value or the actual data returned from
* accessing the URL.
* - element 1:
* The error code reslting from attempting to access the URL with curl_exec
* - element 2:
* A string representing a textual representation of the error code that
* resulted from attempting to access the URL with curl_exec
*/
function do_curl_ext($url, $return_to_variable = 1, $number_of_post_vars = 0, $post = NULL) {
global $user;
// Check if we are inside Drupal and there is a valid user.
// If the user is not valid for a Fedora call curl will throw an exception.
if ((!isset($user)) || $user->uid == 0) {
$fedora_user = 'anonymous';
$fedora_pass = 'anonymous';
}
else {
$fedora_user = $user->name;
$fedora_pass = $user->pass;
}
if (function_exists("curl_init")) {
$ch = curl_init();
$user_agent = "Mozilla/4.0 pp(compatible; MSIE 5.01; Windows NT 5.0)";
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, FALSE);
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, FALSE);
curl_setopt($ch, CURLOPT_FAILONERROR, TRUE); // Fail on errors
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1); // allow redirects
curl_setopt($ch, CURLOPT_TIMEOUT, 90); // times out after 90s
curl_setopt($ch, CURLOPT_USERAGENT, $user_agent);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, $return_to_variable); // return into a variable
curl_setopt($ch, CURLOPT_URL, $url);
curl_setopt($ch, CURLOPT_USERPWD, "$fedora_user:$fedora_pass");
//curl_setopt($ch, CURLOPT_HTTPAUTH, CURLAUTH_ANY);
if ($number_of_post_vars > 0 && $post) {
curl_setopt($ch, CURLOPT_POST, $number_of_post_vars);
curl_setopt($ch, CURLOPT_POSTFIELDS, "$post");
}
$ret_val = curl_exec($ch);
$error_code = curl_errno($ch);
$error_string = curl_error($ch);
return array ($ret_val, $error_code, $error_string);
}
else {
if (function_exists(drupal_set_message)) {
drupal_set_message(t('No curl support.'), 'error');
}
return NULL;
}
}
/**
* Fedora available
* @return type
*/
function fedora_available() {
$response = do_curl_ext(variable_get('fedora_base_url', 'http://localhost:8080/fedora') . '/describe');
return ($response != NULL) ? ($response[1] == 0) : FALSE;
}
/**
* Resource index search available
* @return type
*/
function risearch_available() {
$response = do_curl_ext(variable_get('fedora_repository_url', 'http://localhost:8080/fedora/risearch'));
return ($response != NULL) ? ($response[1] == 0): FALSE;
}
/**
* Returns a UTF-8-encoded transcripiton of the string given in $in_str.
* @param string $in_str
* @return string A UTF-8 encoded string.
*/
function fix_encoding($in_str) {
$cur_encoding = mb_detect_encoding($in_str);
if ($cur_encoding == "UTF-8" && mb_check_encoding($in_str, "UTF-8")) {
return $in_str;
}
else {
return utf8_encode($in_str);
}
}
/**
* valid pid ??
* @param type $pid
* @return boolean
*/
function valid_pid($pid) {
$valid = FALSE;
if (strlen(trim($pid)) <= 64 && preg_match('/^([A-Za-z0-9]|-|\.)+:(([A-Za-z0-9])|-|\.|~|_|(%[0-9A-F]{2}))+$/', trim($pid))) {
$valid = TRUE;
}
return $valid;
}
/**
* Valid Dsid ??
* @param type $dsid
* @return boolean
*/
function valid_dsid($dsid) {
$valid = FALSE;
if (strlen(trim($dsid)) <= 64 && preg_match('/^[a-zA-Z0-9\_\-\.]+$/', trim($dsid))) {
$valid = TRUE;
}
return $valid;
}
/**
* fixDsid ??
* @param type $dsid
* @return string
*/
function fix_dsid($dsid) {
$new_dsid = trim($dsid);
$find = '/[^a-zA-Z0-9\.\_\-]/';
$replace = '';
$new_dsid = preg_replace($find, $replace, $new_dsid);
if (strlen($new_dsid) > 63)
$new_dsid = substr($new_dsid, -63);
if (preg_match('/^[^a-zA-Z]/', $dsid))
$new_dsid = 'x' . $new_dsid;
if (strlen($new_dsid) == 0)
$new_dsid = 'item' . rand(1, 100);
return $new_dsid;
}
/**
* Function: get_collections_as_option_array
*
* Description: Returns an associative array of all collection objects in Fedora instance
*
* @return array
*/
function get_collections_as_option_array() {
module_load_include('inc', 'fedora_repository', 'api/fedora_utils');
$query = 'select $object $title from <#ri>
where ($object <dc:title> $title
and $object <info:fedora/fedora-system:def/model#hasModel> <info:fedora/islandora:collectionCModel>
and $object <fedora-model:state> <info:fedora/fedora-system:def/model#Active>)
order by $title';
$url = variable_get('fedora_repository_url', 'http://localhost:8080/fedora/risearch');
$url .= "?type=tuples&flush=true&format=csv&limit=1000&lang=itql&stream=on&query=";
$content = do_curl($url . htmlentities(urlencode($query)));
$list = explode("\n", $content);
array_shift($list);
$list = preg_replace('/info:fedora\//', '', $list);
$options = array();
foreach ($list as $item) { //removes blanks
if ($item) {
$parts = explode(',', $item);
$options[$parts[0]] = $parts[1];
}
}
unset($options['islandora:ContentModelCollection']);
return $options;
}
/**
* Function: get_content_models_as_option_array
*
* Description: Returns an associative array of all collection objects in Fedora instance
*
* @return array
*/
function get_content_models_as_option_array() {
module_load_include('inc', 'fedora_repository', 'api/fedora_item');
module_load_include('inc', 'fedora_repository', 'api/fedora_utils');
$restricted = variable_get('fedora_namespace_restriction_enforced', true);
$allowed_string = variable_get('fedora_pids_allowed', 'default: demo: changeme: islandora:');
$namespaces = explode(':', $allowed_string);
foreach ($namespaces as $namespace) {
if ($namespace) {
$allowed[] = trim($namespace);
}
}
$query = 'select $object $title from <#ri>
where ($object <dc:title> $title
and ($object <fedora-model:hasModel> <info:fedora/fedora-system:ContentModel-3.0>
or $object <fedora-rels-ext:isMemberOfCollection> <info:fedora/islandora:ContentModelsCollection>)
and $object <fedora-model:state> <info:fedora/fedora-system:def/model#Active>)
order by $title';
$url = variable_get('fedora_repository_url', 'http://localhost:8080/fedora/risearch');
$url .= "?type=tuples&flush=true&format=csv&limit=1000&lang=itql&stream=on&query=";
$content = do_curl($url . htmlentities(urlencode($query)));
$list = explode("\n", $content);
array_shift($list);
$list = preg_replace('/info:fedora\//', '', $list);
foreach ($list as $item) { //removes blanks
if ($item) {
$parts = explode(',', $item);
$item = new fedora_item($parts[0]);
$nameparts = explode(':', $parts[0]);
if (!$restricted || in_array($nameparts[0], $allowed))
if ($item->get_datastream_dissemination('ISLANDORACM')) {
$options[$parts[0]] = $parts[1];
}
}
}
return $options;
}