clarify meaning of `sha1name`

[coolaj86]

I'm confused as to what sha1name means.

It isn't the sha1sum of the file, correct?

Is it the sha1sum of t he name of the file?

That doesn't seem very safe. If I upload a file called me.jpeg and someone else uploads a file called me.jpeg, what would prevent collision?

Or let's say I have a backup service and a client is uploading two files of the same name in one request? Will the sha1sums be different?

I would prefer a UUID or a sha1sum which includes some random element.

Or is each sha1name unique and I'm misunderstanding the documentation?

[rootslab]

Hi coolaj86, sha1name is the filename SHA1 hash !

two different user never causes a filename collisions! The directories are different!

If you upload, for example, two files with the same name, through the same POST, the second file are renamed wth a different sha1name (with a random seed).

If you upload a file with DIfferent POSTs:

• if no session identifier was found, the directory are different for every POST, there is no file collision .
• if session identifier was found (an authenticated user), the directory brings session id as name, so the upload directory remains the same, across multiple user POSTS, and the files are overwritten!

I have written about it in the 'File Uploads' section in the Readme!

[rootslab]

ops, yes sha1name is unique!

[rootslab]

I have added comments to Readme for sha1name . ;)

[lonevan]

'I would prefer a UUID or a sha1sum which includes some random element.'

Agreed..

'getSessionID' function cannot fit for all siturations. For example, some users upload files much more than others, so I don't want to create an unique ID Directory but some random dirs.

But the filename is created by SHA1 hash. When then upload files with the same name, collisions occurs..

[lonevan]

To give user a chance to use UUID or random string as filename, maybe formaline could add a config method?