diff --git a/apachesolr_attachments.admin.inc b/apachesolr_attachments.admin.inc
index 7876473..062fe6d 100644
--- a/apachesolr_attachments.admin.inc
+++ b/apachesolr_attachments.admin.inc
@@ -105,13 +105,31 @@ function apachesolr_attachments_entity_bundle_settings() {
$indexed_bundles = apachesolr_get_index_bundles($env_id, $entity_type);
foreach ($indexed_bundles as $bundle) {
- if ($bundle != 'file') {
- $form['apachesolr_attachments_entity_bundle_indexing_' . $bundle] = array(
- '#type' => 'select',
- /* NOTE: If http://drupal.org/node/969180 ever gets committed we could use that function instead
- and remove apachesolr_attachments_entity_bundle_label from the bottom of this file.
- */
- '#title' => apachesolr_attachments_entity_bundle_label($entity_type, $bundle),
+ $form['apachesolr_attachments_entity_bundle_indexing_' . $bundle] = array(
+ '#type' => 'select',
+ );
+ // NOTE: If http://drupal.org/node/969180 ever gets committed we could
+ // use that function instead and remove
+ // apachesolr_attachments_entity_bundle_label() from the bottom of this
+ // file.
+ $bundle_label = apachesolr_attachments_entity_bundle_label($entity_type, $bundle);
+ // Unattached file entities need different options in the user
+ // interface and (for backwards compatibility reasons) a different
+ // default choice than files attached to other entities.
+ if ($entity_type == 'file') {
+ $form['apachesolr_attachments_entity_bundle_indexing_' . $bundle] += array(
+ '#title' => t('@bundle (unattached)', array('@bundle' => $bundle_label)),
+ '#default_value' => variable_get('apachesolr_attachments_entity_bundle_indexing_' . $bundle, 'none'),
+ '#options' => array(
+ 'seperate' => t('Index as separate entities'),
+ 'none' => t("Don't index"),
+ ),
+ '#description' => t('Determine how these files should be indexed for search when they are not attached to any other entity.'),
+ );
+ }
+ else {
+ $form['apachesolr_attachments_entity_bundle_indexing_' . $bundle] += array(
+ '#title' => check_plain($bundle_label),
'#default_value' => variable_get('apachesolr_attachments_entity_bundle_indexing_' . $bundle, 'seperate'),
'#options' => array(
'seperate' => t('Attachments as separate entities'),
diff --git a/apachesolr_attachments.index.inc b/apachesolr_attachments.index.inc
index ac11c41..ab1c08a 100644
--- a/apachesolr_attachments.index.inc
+++ b/apachesolr_attachments.index.inc
@@ -185,11 +185,11 @@ function apachesolr_attachments_extract_using_solr($filepath) {
* (optional) The number of references to add to the object. Defaults to 1.
*
*/
-function apachesolr_attachments_add_file_usage(stdClass $stub_file, $parent_entity_type, $parent_entity_id) {
+function apachesolr_attachments_add_file_usage(stdClass $stub_file, $parent_entity_type = NULL, $parent_entity_id = NULL) {
// Only add this file type if the parent entity type can be indexed.
// Example : node is mostly indexed but media module is not. So
// exclude all media entities from being added
- $entity_info = entity_get_info($parent_entity_type);
+ $entity_info = isset($parent_entity_type) ? entity_get_info($parent_entity_type) : entity_get_info('file');
if (!empty($entity_info['apachesolr']['indexable'])) {
// We do have to load the file, because there is no way to get the
// bundle type, and media adds many bundles, so fixing this here
@@ -210,8 +210,8 @@ function apachesolr_attachments_add_file_usage(stdClass $stub_file, $parent_enti
->key(array(
'entity_type' => 'file',
'entity_id' => $file->fid,
- 'parent_entity_type' => $parent_entity_type,
- 'parent_entity_id' => $parent_entity_id,
+ 'parent_entity_type' => isset($parent_entity_type) ? $parent_entity_type : '',
+ 'parent_entity_id' => isset($parent_entity_id) ? $parent_entity_id : 0,
))
->fields(array(
'bundle' => $file->type,
@@ -245,10 +245,6 @@ function apachesolr_attachments_delete_file_usage(stdClass $file, $parent_entity
function apachesolr_attachments_clean_index_table() {
$indexer_table = apachesolr_get_indexer_table('file');
- // Clean all entries where parent_entity_id is empty
- db_delete($indexer_table)
- ->condition('parent_entity_id', 0)
- ->execute();
// Clean all entries from entity types that should not be indexed
foreach (entity_get_info() as $entity_type => $entity_info) {
if (empty($entity_info['apachesolr']['indexable'])) {
diff --git a/apachesolr_attachments.module b/apachesolr_attachments.module
index 4a95d37..4e0a129 100644
--- a/apachesolr_attachments.module
+++ b/apachesolr_attachments.module
@@ -93,40 +93,108 @@ function apachesolr_attachments_solr_document(ApacheSolrDocument $document, $fil
// Text is saved in the index table. Will be used by the node indexing if
// available.
- $text = apachesolr_attachments_get_attachment_text($file);
- // If we don't have extracted text we should stop our process here
-
- if (empty($text)) {
- return $documents;
+ $text = '';
+ if (apachesolr_attachments_allowed_mime($file->filemime)) {
+ $text = (string) apachesolr_attachments_get_attachment_text($file);
}
// Get the list of parents that we should index from the indexing table
- $parents = db_select($table, 'aie')
+ $file_records = db_select($table, 'aie')
->fields('aie')
->condition('entity_type', 'file')
->condition('entity_id', $file->fid)
+ // Prioritize records with a parent entity over those that don't have one,
+ // since we will only index the latter separately if the file isn't being
+ // indexed as an attachment somewhere else.
+ ->orderBy('parent_entity_id', 'DESC')
->execute();
- foreach ($parents as $parent) {
- // load the parent entity and reset cache
- $parent_entities = entity_load($parent->parent_entity_type, array($parent->parent_entity_id), NULL, TRUE);
- $parent_entity = reset($parent_entities);
+ // Start by assuming that we'll index the file as its own entity, but as soon
+ // as we find a record of it attached somewhere else we'll change this.
+ $index_standalone_file = TRUE;
+ foreach ($file_records as $file_record) {
+ // If this is an attachment, check if it should be indexed.
+ if (!empty($file_record->parent_entity_id)) {
+ // Check if there is extracted text to index.
+ if (empty($text)) {
+ continue;
+ }
- // Skip invalid entities
- if (empty($parent_entity)) {
- continue;
- }
+ // Check if the parent can be loaded.
+ $parent_entities = entity_load($file_record->parent_entity_type, array($file_record->parent_entity_id), NULL, TRUE);
+ $parent_entity = reset($parent_entities);
+ if (empty($parent_entity)) {
+ continue;
+ }
- // Retrieve the parent entity id and bundle
- list($parent_entity_id, $parent_entity_vid, $parent_entity_bundle) = entity_extract_ids($parent->parent_entity_type, $parent_entity);
- $parent_entity_type = $parent->parent_entity_type;
+ // Retrieve the parent entity id and bundle.
+ list($parent_entity_id, $parent_entity_vid, $parent_entity_bundle) = entity_extract_ids($file_record->parent_entity_type, $parent_entity);
+ $parent_entity_type = $file_record->parent_entity_type;
+
+ // Check if the parent entity is flagged for indexing attachments as
+ // separate documents.
+ $indexing_method = variable_get('apachesolr_attachments_entity_bundle_indexing_' . $parent_entity_bundle, 'seperate');
+
+ // If it's not flagged for indexing at all, skip it.
+ if ($indexing_method == 'none') {
+ continue;
+ }
- // proceed with building this document only if the parent entity is not flagged for
- // indexing attachments with parent entity or not indexing attachements
- if (variable_get('apachesolr_attachments_entity_bundle_indexing_' . $parent_entity_bundle, 'seperate') == 'seperate') {
- // Get a clone of the bare minimum document
- $filedocument = clone $document;
+ // If we got this far, it's a legitimate attachment. It will either be
+ // indexed somewhere else (along with its parent), or it shouldn't be
+ // indexed at all due to restrictions on the parent (see below). In
+ // either case, we need to prevent the file from being indexed as a
+ // standalone entity also.
+ $index_standalone_file = FALSE;
+
+ // Only continue if the parent is flagged for indexing as a separate
+ // document. (Instead it might be indexed as part of the parent entity
+ // itself, which happens in apachesolr_attachments_node_solr_document().
+ if ($indexing_method != 'seperate') {
+ continue;
+ }
+ // Check if the parent entity is excluded from search indexing.
+ if (apachesolr_attachments_is_parent_excluded($file->fid, 'file', $parent_entity_id, $parent_entity_type, $env_id)) {
+ continue;
+ }
+ // Mark this for indexing, with a link to its parent.
+ $text_to_index = $text;
+ $has_parent = TRUE;
+ }
+ // If this is an unattached file record, check if it should be indexed.
+ elseif ($index_standalone_file) {
+ // Check if we are indexing unattached files of this type.
+ if (variable_get('apachesolr_attachments_entity_bundle_indexing_' . $file_record->bundle, 'none') != 'seperate') {
+ continue;
+ }
+ // The file entity itself may have rendered content (e.g. due to attached
+ // fields) if the File Entity module is installed. Since the file is
+ // being presented as a standalone entity in search results (with no link
+ // to any parent with fields of its own) we should index its fields just
+ // like any other entity.
+ // @todo Should these fields be indexed in the other cases also?
+ $rendered_file = '';
+ if (function_exists('file_view')) {
+ $build = file_view($file, 'search_index');
+ // Remove unwanted HTML.
+ unset($build['#theme']);
+ $content = drupal_render($build);
+ $rendered_file = apachesolr_clean_text($content);
+ }
+ // Mark this for indexing, without a link to any parent entity.
+ $text_to_index = $rendered_file . $text;
+ $has_parent = FALSE;
+ }
+ // Unattached file record, but we aren't going to index it.
+ else {
+ continue;
+ }
+
+ // Get a clone of the bare minimum document
+ $filedocument = clone $document;
+ // If the file record has a parent entity, we need to index data from that.
+ if ($has_parent) {
//Get the callback array to add stuff to the document
$callbacks = apachesolr_entity_get_callback($parent_entity_type, 'document callback');
$build_documents = array();
@@ -144,22 +212,6 @@ function apachesolr_attachments_solr_document(ApacheSolrDocument $document, $fil
// Build our separate document and overwrite basic information
$filedocument->id = apachesolr_document_id($file->fid . '-' . $parent_entity_id, $entity_type);
- $filedocument->url = file_create_url($file->uri);
- $path = file_stream_wrapper_get_instance_by_uri($file->uri)->getExternalUrl();
- // A path is not a requirement of an entity
- if (!empty($path)) {
- $filedocument->path = $path;
- }
-
- // Add extra info to our document
- $filedocument->label = apachesolr_clean_text($file->filename);
- $filedocument->content = apachesolr_clean_text($file->filename) . ' ' . $text;
-
- $filedocument->ds_created = apachesolr_date_iso($file->timestamp);
- $filedocument->ds_changed = $filedocument->ds_created;
-
- $filedocument->created = apachesolr_date_iso($file->timestamp);
- $filedocument->changed = $filedocument->created;
// Add Parent information fields. See http://drupal.org/node/1515822 for explanation
$parent_entity_info = entity_get_info($parent_entity_type);
@@ -182,13 +234,32 @@ function apachesolr_attachments_solr_document(ApacheSolrDocument $document, $fil
$filedocument->zm_parent_entity = drupal_json_encode($small_parent_entity);
$filedocument->sm_parent_entity_bundle = $parent_entity_type . "-" . $parent_entity_bundle;
$filedocument->sm_parent_entity_type = $parent_entity_type;
+ }
+
+ // Now add the data from the file itself that we will index.
+ $filedocument->url = file_create_url($file->uri);
+ $path = file_stream_wrapper_get_instance_by_uri($file->uri)->getExternalUrl();
+ // A path is not a requirement of an entity
+ if (!empty($path)) {
+ $filedocument->path = $path;
+ }
+
+ // Add extra info to our document
+ $filedocument->label = apachesolr_clean_text($file->filename);
+ $filedocument->content = apachesolr_clean_text($file->filename) . ' ' . $text_to_index;
+
+ $filedocument->ds_created = apachesolr_date_iso($file->timestamp);
+ $filedocument->ds_changed = $filedocument->ds_created;
+
+ $filedocument->created = apachesolr_date_iso($file->timestamp);
+ $filedocument->changed = $filedocument->created;
+
+ // Add Apachesolr Attachments specific fields.
+ $filedocument->ss_filemime = $file->filemime;
+ $filedocument->ss_filesize = $file->filesize;
- // Add Apachesolr Attachments specific fields.
- $filedocument->ss_filemime = $file->filemime;
- $filedocument->ss_filesize = $file->filesize;
$documents[] = $filedocument;
- }
}
return $documents;
@@ -266,13 +337,10 @@ function apachesolr_attachments_solr_reindex() {
->condition('entity_type', 'file')
->isNull('body')
->execute();
- $files = _apachesolr_attachments_get_all_files();
- // If we do not have files, return success
- if (empty($files)) {
- return TRUE;
- }
- // Loop over all the files and add them to our indexing table
+ // Loop over all the attachments and add them to our indexing table along
+ // with a record of their parents.
+ $files = _apachesolr_attachments_get_all_attachments();
foreach ($files as $parent_entity_type => $parent_entities) {
foreach ($parent_entities as $parent_entity_info) {
// Fake our file class
@@ -286,6 +354,28 @@ function apachesolr_attachments_solr_reindex() {
apachesolr_attachments_add_file_usage($file, $parent_entity_type, $parent_entity_id);
}
}
+
+ // Loop over all indexable files and add them to our indexing table as
+ // standalone entities.
+ if ($indexable_bundles = apachesolr_get_index_bundles($env_id, 'file')) {
+ $select = db_select('file_managed', 'f');
+ $select->fields('f', array('fid'));
+ // Limit by bundle. We need to handle the case of the File Entity module
+ // which alters the {file_managed} table to add a "type" field there.
+ // Otherwise, there is only one bundle, so we don't need to limit
+ // anything. See also apachesolr_attachments_add_file_usage().
+ if (db_field_exists('file_managed', 'type')) {
+ $select->condition('type', $indexable_bundles, 'IN');
+ }
+ $fids = $select->execute()->fetchCol();
+ foreach ($fids as $fid) {
+ // Fake our file class.
+ $file = new stdClass();
+ $file->fid = $fid;
+ apachesolr_attachments_add_file_usage($file);
+ }
+ }
+
}
catch (Exception $e) {
$transaction->rollback();
@@ -301,7 +391,7 @@ function apachesolr_attachments_solr_reindex() {
* regardless of the widget
* @return type
*/
-function _apachesolr_attachments_get_all_files() {
+function _apachesolr_attachments_get_all_attachments() {
$results = array();
$fields = field_info_field_by_ids();
foreach ($fields as $field_id => $field_info) {
@@ -341,16 +431,7 @@ function apachesolr_attachments_status_callback($entity_id, $entity_type) {
// load the entity and reset cache
$entities = entity_load($entity_type, array($entity_id), NULL, TRUE);
$entity = reset($entities);
- // Check if the mimetype is allowed
- if (apachesolr_attachments_allowed_mime($entity->filemime) == FALSE) {
- // Set status to 0 and remove from the index
- return FALSE;
- }
- // Check if the file is a real file
- if (apachesolr_attachments_is_file($entity) == FALSE) {
- // Set status to 0 and remove from the index
- return FALSE;
- }
+
// Check if the entity status is active
if ($entity->status != 1) {
// Set status to 0 and remove from the index
@@ -360,37 +441,6 @@ function apachesolr_attachments_status_callback($entity_id, $entity_type) {
return TRUE;
}
-/**
- * Implenents hook_apachesolr_ENTITY_TYPE_exclude().
- *
- * This is invoked for each file entity that is being inspected to be added to the
- * index. if any module returns TRUE, the entity is skipped for indexing.
- *
- * @param integer $entity_id
- * @param integer $row
- * A complete set of data from the indexing table.
- * @param string $env_id
- * @return boolean
- */
-function apachesolr_attachments_apachesolr_file_exclude($entity_id, $row, $env_id) {
- module_load_include('inc', 'apachesolr_attachments', 'apachesolr_attachments.index');
- // Make sure we have a boolean value.
- // Anything different from 1 becomes zero
- if (!$entity_id || !$row->parent_entity_id) {
- // Exclude
- return TRUE;
- }
- // Check if the parent entity is excluded
- $parent_entity_id = $row->parent_entity_id;
- $parent_entity_type = $row->parent_entity_type;
- $exclude = apachesolr_attachments_is_parent_excluded($entity_id, 'file', $parent_entity_id, $parent_entity_type, $env_id);
- if ($exclude) {
- // Exclude
- return TRUE;
- }
- // Do not exclude
- return FALSE;
-}
function apachesolr_attachments_is_file($entity) {
if (!empty($entity->uri)) {
@@ -696,20 +746,22 @@ function theme_apachesolr_search_snippets__file($vars) {
// Retrieve our parent entities. They have been saved as
// a small serialized entity
- foreach ($doc->zm_parent_entity as $parent_entity_encoded) {
- $parent_decoded = (object) drupal_json_decode($parent_entity_encoded);
+ if (!empty($doc->zm_parent_entity)) {
+ foreach ($doc->zm_parent_entity as $parent_entity_encoded) {
+ $parent_decoded = (object) drupal_json_decode($parent_entity_encoded);
- // Extract entity id from parent. List fetches the first item from the array.
- list($id) = entity_extract_ids($parent_decoded->entity_type, $parent_decoded);
+ // Extract entity id from parent. List fetches the first item from the array.
+ list($id) = entity_extract_ids($parent_decoded->entity_type, $parent_decoded);
- // Load parent entity
- $load = entity_load($parent_decoded->entity_type, array($id));
- $parent_entity = array_shift($load);
+ // Load parent entity
+ $load = entity_load($parent_decoded->entity_type, array($id));
+ $parent_entity = array_shift($load);
- $parent_entity_uri = entity_uri($parent_decoded->entity_type, $parent_entity);
- $parent_entity_uri['options']['absolute'] = TRUE;
- $parent_label = entity_label($parent_decoded->entity_type, $parent_entity);
- $parent_entity_links[] = l($parent_label, $parent_entity_uri['path'], $parent_entity_uri['options']);
+ $parent_entity_uri = entity_uri($parent_decoded->entity_type, $parent_entity);
+ $parent_entity_uri['options']['absolute'] = TRUE;
+ $parent_label = entity_label($parent_decoded->entity_type, $parent_entity);
+ $parent_entity_links[] = l($parent_label, $parent_entity_uri['path'], $parent_entity_uri['options']);
+ }
}
if (module_exists('file')) {
@@ -719,7 +771,15 @@ function theme_apachesolr_search_snippets__file($vars) {
$file_type = t('@filemime', array('@filemime' => $doc->ss_filemime));
}
- return implode(' ... ', $snippets) . '' . $file_type . ' attached to:' . implode(', ', $parent_entity_links) . '';
+ $output = implode(' ... ', $snippets);
+ $output .= '' . $file_type;
+ if (!empty($parent_entity_links)) {
+ $output .= ' attached to:' . implode(', ', $parent_entity_links);
+ }
+ $output .= '';
+
+ return $output;
+
}
/**