[mythtv-commits] Ticket #8374: update mythvideo's scan.php to support hash tags

MythTV mythtv at cvs.mythtv.org
Fri Apr 23 02:33:32 UTC 2010


#8374: update mythvideo's scan.php to support hash tags
-----------------------------------------------+----------------------------
 Reporter:  Dave Miller <mythtv@…>             |       Owner:  kormoc    
     Type:  patch                              |      Status:  new       
 Priority:  minor                              |   Milestone:  unknown   
Component:  Plugin - MythWeb                   |     Version:  Trunk Head
 Severity:  medium                             |     Mlocked:  0         
-----------------------------------------------+----------------------------
 The attached patches update scan.php in mythweb's video module to support
 handling the hashes that are now kept on video files.  They also fix a
 minor potential security issue where the path being scanned isn't properly
 shell-escaped before being passed to the find command.  Even in the non-
 security case, the lack of shell-escape could have undesired consequences
 if your storage path happens to contain any spaces or other funny
 characters.

 From what I can tell, php appears to not support unsigned 64-bit integers
 at all, which are required for proper calculation of the hash.  So I had
 to write a python script to do the hash calculation (it borrows the hash
 routine from JAMU.py and just makes it executable standalone from the
 command line) and dump it to standard out, then we shell out to it.  This
 will probably need some massaging, as I don't know where the proper place
 is to put such a script, and right now scan.php with this patch just
 depends on it being in your default PATH (which I'm assuming probably
 isn't the best way to do it).

 I'm more than happy to update the patches if there are suggestions of a
 better way to do it.

 As patched, the new scanning process is:
 1) Search for new files first.  If it exists in the DB already, continue
 like before.
 2) If there isn't a file with this filename in the DB already, then we
 hash the file, and look to see if there's a file with the same hash
 already (this step is skipped if the file is too small to hash).  If
 there's a matching hash, then we update the filename on that entry in the
 database, on the assumption that the file's been moved.
 3) If we still haven't matched anything at this point, it's a new file and
 it's added to the database, hash included.
 4) After the above is done, so we can catch any files that were moved,
 then we go back and scan for files in the database that are no longer on
 the filesystem and remove them from the database.

-- 
Ticket URL: <http://svn.mythtv.org/trac/ticket/8374>
MythTV <http://www.mythtv.org/>
MythTV


More information about the mythtv-commits mailing list