Information request
Bogdan Cimpoesu
Let me start by first saying that I love the idea behind Smewt. The title determination mechanism is magical :)
However, how Smewt is at the moment it doesn't fit my needs :). I have a huge movie repository (call me obsessed if you wish :) ) about 3000-4000 movies in all formats (HD, standard dvd rips, etc) plus about 75 TV shows. While Smewt is working properly for most of the movies and TV shows there is a small percentage that are not detected correctly. A small percentage from such a large number of movies is a big number (as you can imagine). I have an idea how to furthermore improve the detection mechanism and I would like to give it a try. I am no real programmer and not very experimented but I like trying stuff in python especially when I do something for my (and others if they will) use as opposed to doing it as a job.
This is why I am emailing you to get more details about your development environment. What do you use? (don't tell me python and ubuntu or something like that :) I need more in-depth details )
Thanks in advance,
Bogdan
Nicolas Wack
Thanks for your interest! The auto-magical title finding is one of
Smewt's main goal, so I'm glad it worked for you. ;-)
At the moments, the movies for which no title could be found go into
the "Unknown" movie and you can see the files directly there. That is
of course not optimal. There are 2 solutions:
- let the user edit the metadata for the file so it shows up properly
- fix smewt so that it makes fewer errors
I am of course also more interested in seeing the 2nd one, I'd rather
have the computer do the work for me.
There are 2 steps in the title guessing:
- first we try to extract as much information as possible from only
the filename. Ideally we get a good candidate for the movie title
- then we look online (TMDb for movies) to see if there is a movie
with the title candidate
The first part is done by a part of Smewt which has been split into a
separate library and is called GuessIt. It would be interesting to see
whether your files that fail do so because GuessIt improperly guessed
a movie title, or because the movie could not actually be found in
TMDb. The repository is at https://gitorious.org/smewt/guessit .
As for my programming environment, I'm mostly on debian and use emacs
as a text editor. If you're not programming everyday, emacs has
probably a learning curve too steep, so if you're on linux, I'd
recommend kate or gedit depending on your desktop environment. Not too
sure about windows or mac... There is an IDE that seems to be quite
good, which is called PyCharm, but I never tried it so I can't say for
sure. It's not free either.
I also use IPython a lot. It is a must if you're going to do python in
the interpreter (always handy to try stuff). It's much better as it
has auto-completion, history, colors, etc...
Don't hesitate if you have further questions!
Nico.