Skip to content
nested media container framework for python multimedia apps
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Q: What do zip files and mpeg4 files have in common?
A: Both are containers. In the first case for more files, in the latter case for one or more video and or audio streams. A video stream again is a container -- it contains still images. A file in the zip file most likely also is a container for something else.

In general, most file formats can be considered containers for subparts which in turn also are containers. This is a hierarchical recursive structure much like a directory tree.

pyramiding generalises traversal and read access of such a tree of containers, no matter if it is a directory tree or some structure inside a file.

pyramiding consists of several python modules that can be seen as 'drivers' for a specific file format. Such a driver is called 'Archive'. Archives know how to iterate over the contents of a file they support, provide data streams (file-like objects) for each item they contain and meta data about those items. Data stream and meta data are combined by 'Resources'.

An Archive contains Resourcees, a Resource is the input data stream of an Archive instance. Archives read a Resource to enumerate its children.

Archives are used by Resources internally. The class Resource is the only visible class outside pyrmaiding.

pyramiding will use lazy evaluation and caching to free the developer from efficiency concerns.

from pyRamiding import Resource

	data = Resource(file("videos.tar"))
	name = data["info.ini"]["section1"]["name"].read()
	# smae as data.find("info.ini/section1/name").as_string()
	image_resource, location = data.find("icon.jpg/image"), data.find("icon.jpg/exif/geo-location")
	width, height = image_resource.details.width, image_resource.details.height 
	rgb_data =

	for version in data.find("versions"):
		if version["video"].details.width >= desired_width:

	print "Selected video file %(name)s, size: %(width)d x %(height)d pixels" % version["video"].details

	last_frame_data = version["video"][version["video"].details.frame_count - 1].read()
except ResourceNotFound, e:
	print e.path, "does not exist"

As a first milestone, the following Archives will be implemented (probably even in this order)

- ini
- tar
- json, js
- png
- jpg, jpeg
- mpg, mpeg, mp4
- zip
Something went wrong with that request. Please try again.