Resolves #2.

mjordan · Dec 28, 2018 · f0d6464 · f0d6464
1 parent 3f9f2be
commit f0d6464
Show file tree

Hide file tree

Showing 7 changed files with 51 additions and 10 deletions.
diff --git a/README.md b/README.md
@@ -1,6 +1,6 @@
 # Islandora Bagger
 
-Tool to generate [Bags](https://en.wikipedia.org/wiki/BagIt) for objects using Islandora's REST interface. Specific content is added to the Bag's `data` directory and `bag-info.txt` file using plugins. Bags are compliant with version 0.96 of the BagIt specification.
+Command-line tool to generate [Bags](https://en.wikipedia.org/wiki/BagIt) for objects using Islandora's REST interface. Specific content is added to the Bag's `data` directory and `bag-info.txt` file using plugins. Bags are compliant with version 0.96 of the BagIt specification.
 
 This utility is for Islandora 8.x-1.x (CLAW). For creating Bags for Islandora 7.x, use [Islandora Fetch Bags](https://github.com/mjordan/islandora_fetch_bags).
 
@@ -79,6 +79,11 @@ fedora_base_url: 'http://localhost:8080/fcrepo/rest/'
 # from the "Islandora Media Use" vocabulary. Use an emply list (e.g., [])
 # to include all media.
 drupal_media_tags: ['/taxonomy/term/15']
+
+# Used by the 'AddMedia' plugin. Indicates whether the Bag should contain a file
+# named 'media_use_summary.tsv' that lists all the media files plus the taxonomy
+# name corresponding to the 'drupal_media_tags' list. Default is false.
+include_media_use_list: true.
 ```
 
 The command to generate a Bag takes two required parameters, `--settings` and `--node`. Assuming the above configuration file is named `sample_config.yml`, and the Drupal node ID you want to generate a Bag from is 112, the command would look like this:
@@ -120,7 +125,7 @@ Apart from the static tags mentioned in the previous section, all file content a
 
 #### Included plugins
 
-The following plugins are bunlded with Islandora Bagger:
+The following plugins are bundled with Islandora Bagger:
 
 * AddBasicTags: Adds the `Internal-Sender-Identifier` bag-info.txt tag using the Drupal URL for the node as its value, and the `Bagging-Date` tag using the current date as its value.
 * AddNodeJson: Adds the Drupal JSON representation of the node, specifically, the response to a request to `/node/1234?_format=json`.

diff --git a/sample_config.yml b/sample_config.yml
@@ -55,3 +55,8 @@ fedora_base_url: 'http://localhost:8080/fcrepo/rest/'
 # from the "Islandora Media Use" vocabulary. Use an emply list (e.g., [])
 # to include all media.
 drupal_media_tags: ['/taxonomy/term/15']
+
+# Used by the 'AddMedia' plugin. Indicates whether the Bag should contain a file
+# named 'media_use_summary.tsv' that lists all the media files plus the taxonomy
+# name corresponding to the 'drupal_media_tags' list. Default is false.
+include_media_use_list: true.
diff --git a/src/Command/CreateBagCommand.php b/src/Command/CreateBagCommand.php
@@ -43,7 +43,6 @@ protected function execute(InputInterface $input, OutputInterface $output)
         $nid = $input->getOption('node');
         $settings_path = $input->getOption('settings');
         $this->settings = Yaml::parseFile($settings_path);
-        $this->settings['drupal_base_url'] .= '/node/';
 
         // Set some configuration defaults.
         $this->settings['http_timeout'] = (!isset($this->settings['http_timeout'])) ?
@@ -63,7 +62,7 @@ protected function execute(InputInterface $input, OutputInterface $output)
         $client = new \GuzzleHttp\Client();
 
         // Get the node's UUID from Drupal.
-        $drupal_url = $this->settings['drupal_base_url'] . $nid . '?_format=json';
+        $drupal_url = $this->settings['drupal_base_url'] . '/node/' . $nid . '?_format=json';
         $response = $client->get($drupal_url);
         $response_body = (string) $response->getBody();
         $node_json = $response_body;
@@ -113,12 +112,12 @@ protected function execute(InputInterface $input, OutputInterface $output)
             $bag_name = $bag_name . '.' . $package;
         }
 
-        $io->success("Bag created for " . $this->settings['drupal_base_url'] . $nid . " at " . $bag_dir);
+        $io->success("Bag created for " . $this->settings['drupal_base_url'] . '/node/' . $nid . " at " . $bag_dir);
         if ($this->settings['log_bag_creation']) {
             $this->logger->info(
                 "Bag created.",
                 array(
-                    'node URL' => $this->settings['drupal_base_url'] . $nid,
+                    'node URL' => $this->settings['drupal_base_url'] . '/node/' . $nid,
                     'node UUID' => $uuid,
                     'Bag location' => $this->settings['output_dir'],
                     'Bag name' => $bag_name

diff --git a/src/Plugin/AddMedia.php b/src/Plugin/AddMedia.php
@@ -30,9 +30,12 @@ public function __construct($settings, $logger)
      */
     public function execute($bag, $bag_temp_dir, $nid, $node_json)
     {
+        $this->settings['include_media_use_list'] = (!isset($this->settings['include_media_use_list'])) ?
+            false: $this->settings['include_media_use_list'];
+
         // Get the media associated with this node using the Islandora-supplied Manage Media View.
         $media_client = new \GuzzleHttp\Client();
-        $media_url = $this->settings['drupal_base_url'] . $nid . '/media';
+        $media_url = $this->settings['drupal_base_url'] . '/node/' . $nid . '/media';
         $media_response = $media_client->request('GET', $media_url, [
             'http_errors' => false,
             'auth' => $this->settings['drupal_media_auth'],
@@ -45,6 +48,9 @@ public function execute($bag, $bag_temp_dir, $nid, $node_json)
 
         // Loop through all the media and pick the ones that are tagged with terms in $this->settings['drupal_media_tags'].
         // If that list is empty, add all media to the Bag.
+        if ($this->settings['include_media_use_list']) {
+            $file_use_list = '';
+        }
         foreach ($media_list as $media) {
             if (count($media['field_media_use'])) {
                 foreach ($media['field_media_use'] as $term) {
@@ -56,6 +62,13 @@ public function execute($bag, $bag_temp_dir, $nid, $node_json)
                             $file_url = $media['field_media_file'][0]['url'];
                         }
                         $filename = $this->getFilenameFromUrl($file_url);
+
+                        if ($this->settings['include_media_use_list']) {
+                            $term_info = $this->fetchTermInfo($term['url']);
+                            $term_external_uri = $term_info['field_external_uri'][0]['uri'];
+                            $file_use_list .= $filename . "\t" . $term_external_uri . PHP_EOL;
+                        }
+
                         $temp_file_path = $bag_temp_dir . DIRECTORY_SEPARATOR . $filename;
                         // Fetch file and save it to $bag_temp_dir with its original filename.
                         // @todo: Determine what to do if the file already exists.
@@ -74,6 +87,11 @@ public function execute($bag, $bag_temp_dir, $nid, $node_json)
                 }
             }
         }
+        if ($this->settings['include_media_use_list']) {
+            $temp_file_path = $bag_temp_dir . DIRECTORY_SEPARATOR . 'media_use_summary.tsv';
+            file_put_contents($temp_file_path, $file_use_list);
+            $bag->addFile($temp_file_path, 'media_use_summary.tsv');
+        }
         return $bag;
     }
 
@@ -83,4 +101,18 @@ protected function getFilenameFromUrl($url)
         $filename = pathinfo($path, PATHINFO_BASENAME);
         return $filename;
     }
+
+    protected function fetchTermInfo($term)
+    {
+        $client = new \GuzzleHttp\Client();
+        $url = $this->settings['drupal_base_url'] . $term;
+        $response = $client->request('GET', $url, [
+            'http_errors' => false,
+            'auth' => $this->settings['drupal_media_auth'],
+            'query' => ['_format' => 'json']
+        ]);
+        $body = (string) $response->getBody();
+        $tag_info = json_decode($body, true);
+        return $tag_info;
+    }
 }
diff --git a/src/Plugin/AddMediaJson.php b/src/Plugin/AddMediaJson.php
@@ -31,7 +31,7 @@ public function __construct($settings, $logger)
     public function execute($bag, $bag_temp_dir, $nid, $node_json)
     {
         $client = new \GuzzleHttp\Client();
-        $media_url = $this->settings['drupal_base_url'] . $nid . '/media';
+        $media_url = $this->settings['drupal_base_url'] . '/node/' . $nid . '/media';
         $response = $client->request('GET', $media_url, [
             'http_errors' => false,
             'auth' => $this->settings['drupal_media_auth'],

diff --git a/src/Plugin/AddMediaJsonld.php b/src/Plugin/AddMediaJsonld.php
@@ -31,7 +31,7 @@ public function __construct($settings, $logger)
     public function execute($bag, $bag_temp_dir, $nid, $node_json)
     {
         $client = new \GuzzleHttp\Client();
-        $media_url = $this->settings['drupal_base_url'] . $nid . '/media';
+        $media_url = $this->settings['drupal_base_url'] . '/node/' . $nid . '/media';
         $response = $client->request('GET', $media_url, [
             'http_errors' => false,
             'auth' => $this->settings['drupal_media_auth'],

diff --git a/src/Plugin/AddNodeJsonld.php b/src/Plugin/AddNodeJsonld.php
@@ -31,7 +31,7 @@ public function __construct($settings, $logger)
     public function execute($bag, $bag_temp_dir, $nid, $node_json)
     {
         $client = new \GuzzleHttp\Client();
-        $url = $this->settings['drupal_base_url'] . $nid;
+        $url = $this->settings['drupal_base_url'] . '/node/' . $nid;
         $response = $client->request('GET', $url, [
             'http_errors' => false,
             'query' => ['_format' => 'jsonld']