Index ideas wanted.
 

I am working with data (actor, director, screenshot, etc.) from around one thousand movies, and I need a good way to name the folders, each with a unique name. Preferably using a piece of data that is contained in the movie itself. One way is to simply use the title. But filenames are better when short and have no special characters. Maybe use the initials from each word in the title? How do you do It? Ideas please?

Re: Index ideas wanted.
 

Usually the way people get around the downsides of what you're talking about is simply to use a "slug" (IE: a piece of text), but with a unique ID number. I assume you're doing something programmatic, so I'd recommend using the first X characters (or first X words) of the title, and then a dash or underscore and an incremented counter. Boom, unique, but still pretty descriptive.

That said, I'm not sure why the file names need to be short, in this context.

When programming, I like my varable names (each is used many times) to be descriptive but short (so the code does not over-fill the monitor). Also, with one thousand folders, short names take less space on the monitor. And that is why the file names should be short.

You have a good idea. Does each movie have a unique copyright or regristration code? I have many types of movies, mostly old. Any more ideas?

Are you building a streaming site?

Yes, I am working toward that. For now, I am preparing content. Gather files and check copyright. Later, I will need to find or write code for login and user account management. Later, uploading everything will be a challenge. Maybe later, evaluate and see if there is a market for the product. Maybe later, advertise. It is a fun and educational project.

I hope this is not going off topic too much.

I copied and pasted the filenames into my code. It works, but is a little messy. Later, end user might have a problem trying to download file onto thumb drive (filesystem incompatibily with filename). This thread is solved for now, and I might implement your ideas later. Thank you.

I don't know much about the subject but I'm curious about the result

Re: Index ideas wanted.
 

I worked with different data but the same need for a unique ID that was not anything published elsewhere. I used a combination that included year, the first bunch of letters in the name and a sequence number, like if you were doing Gone With the Wind from 1939, it would be 1939gone001 for the first 1939 movie that started with gone. The next one that year would be 1939gone002 and so on. Expanding the number of letters or characters increases the ID's exponentially if you expect that many records. It's probably better to have too many characters, given what happened with social security numbers.