Support for additional syntax

[rsgalloway]

Support for more complex syntax, e.g. x10 for every tenth frame.

[nebukadhezer]

https://gist.github.com/nebukadhezer/452ec93d09a4b970d4b38a6cdce325af
Hey Ryan, @rsgalloway
I made a frameRange object, which is subclassed from list
It supports
1001-1500x2, 1550-1600x3 etc pp
syntax.
It can be created from such a string and ( that is more what I am after) be created from a list of frames.
those can be then formatted back into such a string, though it is hard to create the exact same string as the human input might be faulty. egg 1001-1020x2 will turn into 1001-1019x2,1020 ...

I guess this can be done in a much prettier way, thats why I made it a gist.
Id love to see that in pyseq as we often use that syntax accross the pipeline now

Cheers
Johannes

[rsgalloway]

Hey, cool. Yeah, this has been on the list for a while now. Obviously this does come up in production so it's important to support. I'd be happy to integrate something like this when it's ready and if it doesn't impact performance.

[rsgalloway]

The extended syntax S[i:j:d] is already supported, where d is the distance between successive frames.

>>> s = pyseq.Sequence(["file.%03d.exr" % d for d in range(100)])
>>> s.length()
100
>>> s[0:50:10]
[<pyseq.Item "file.000.exr">, <pyseq.Item "file.010.exr">, <pyseq.Item "file.020.exr">,
<pyseq.Item "file.030.exr">, <pyseq.Item "file.040.exr">]
>>> s1 = pyseq.Sequence(s[0:100:10])
>>> s1.length()
10

So ideally I think this is just adding support for a new %x extended range directive, ie. something like

>>> s1.format("%h%x%t")
<pyseq.Sequence "file.0-90x10.exr">

And support in uncompress. Does this need to support any other notations? Depending how complicated this gets, I might target this for the 1.0.0 release.

[nebukadhezer]

in #47 this is somewhat there, but I think if the frames become an object of there own, we then should move all string formatting methods related to frames, into the FrameRange object and call those from the Sequence object.

To me the FrameRange object is/would be pretty useful in other parts of the pipeline, so to have that in a central module like pyseq would be more than welcome...