Enhance image_import by allowing submitter to provide captions (plus a general-case solution)

I've been doing a LOT of big gallery uploads and imports ... on behalf of a client who's trying to supply captions as fast as they can, but ... I also need to force naming conventions on the images, and renaming images, whilst negotiating lists of captions to match up is a terror.

I'm now using a flat-file syntax to drop the relevant info into my directories as we move them around. I'm choosing to use a version popularized years ago by ACDSee image browser, as I've since seen it used by other image software as well.

Now, when importing my galleries, all the fields are filled out for me!

In keeping with the very important goal of keeping the image package as simple as possible, all I've done is introduce a single hook callback that allows a third-party module to provide metadata to the image_import form.
It looks like this:

function image_import_form() {
...
..
   $files = file_scan_directory($dirpath, '.*');
 
+  // Allow other modules to supply metadata about the images being imported
+  // hook_mediadescribe() returns a metadata array to add extra information, possibly including a description or title.
+  $metadata = module_invoke_all('describemedia', $dirpath);
...
.. 
     foreach ($files as $file) {
       $info = image_get_info($file->filename);
+      $meta = $metadata[$file->filename];
...
..
         $form['files']['body'][] = array(
           '#type' => 'textfield',
+          '#default_value' => $meta['description'],
           '#size' => 20,
         );
       }

Then I worked my logic into a small module that exists to read the standard description file metadata and feed it back to image_import. The idea then came to me that if I made the hook useful enough, more modules could use it, both as consumers and suppliers, so I now have hook_describemedia() !

I'm thinking of releasing this import/export module (yes, it exports image galleries also). At the moment it's only really an add-on to image_import.module, but I like the way the code is going.

I don't mind rolling it for D6 also, but I'd like to request this hook be made available in image_import.module. What are the chances? Adding hooks is good mojo :)

Releasing a module with a patch that needs to be applied to another one is not :(

My module does the following:

mediadescriber.module README

/**
 * @file mediadescriber.module
 *
 * Accesses flat-file descriptions alongside images (or other managed files).
 * 
 * In the first case, this is designed to enhance image_import functionality, to
 * enable rapid uploading of image galleries prepared offline, and to enable
 * those galleries to be migrated between Drupal sites easily.
 * 
 * This module creates and updates 'descript.ion' files as used by some old-
 * school file metadata managers - specifically ACDSee.
 * 
 * It slso declares a hook for image_import to use to extract that info when
 * scanning directories.
 * 
 * This module provides no interface or options, it just enhances contextual
 * metadata behind-the-scenes when importing or modifying image resources.
 * It does NOT re-scan image directories for modifications made after import, it
 * just assists filling out the input fields the first time around.
 *
 * Currently it takes ACDSee-type format of 
 * 
 * {(optionally quoted) filename}{whitespace}{caption}{newline}
 * 
 * It should also handle an equivalent simple CSV file, 
 * replacing [whitespace] with [,]
 * 
 * WHY READ - to assist adding captions to images when importing them in bulk
 * from other places ... like a desktop image management tool.
 * 
 * WHY WRITE - so that files/image directories can be accessed using other
 * means, or exported/archived to disk with at least some of the extra metadata
 * retained.
 * 
 * WHY? - Because flat-file storage of information about files is still one of
 * the most robust and portable ways of keeping information about them available
 * when archiving, zipping, emailing, or transferring them. Databases are more
 * powerful for real classifications, but you are always locked into using the
 * database to get that information out again. Flat-file information from 10
 * years ago is still available if you dig up the old disks. Not so with any
 * CMS or software package.
 *
 * This type of functionality (the hook_describemedia()) can be re-implemented
 * in other modules to read, for example:
 *  
 *  =  metadata.xml as found in some fileshare apps, 
 *  =  playlist.m3u MP3 Playlist information 
 *  =  Embedded EXIF data or ID3 tags 
 *  =  RSS media or podcast formats 
 *  =  xmlsitemap information about files or URLs
 */

hook_describemedia

/**
 * A mechanism for modules to extract and return information (metadata or
 * content) about given files, eg media images. 
 * 
 * Possibly even URLs one day.
 * ###########################################
 * 
 * $metadata = hook_describemedia('files/images');
 * OR
 * $metadata = module_invoke_all('mediadescribe', 'files/images');
 *  
 * $metadata = array(
 *   'files/images/cat.jpg' => array(
 *     'title'       => 'LOLCats #42',
 *     'description' => 'I can haz cheezburger',   
 *   )
 *  
 *   'files/images/hello.jpg' => array( 
 *     'title'       => 'The Goatse',     
 *     'description' => 'You've been goatse\d'!',
 *    ) 
 *  );
 * 
 * ###########################################
 * 
 * @see mediadescriber_describemedia()
 * 
 * @param $dirpath a folder to scan, recursively.
 * - may also be a single file getting information requested about it
 * 
 * @return $metadata an array, keyed by filename, that may contain attributes,
 * such as title, caption, or other relevant values.
 * 
 */
function hook_describemedia($dir) {