Skip to content

Latest commit

 

History

History
155 lines (104 loc) · 4.3 KB

README.rst

File metadata and controls

155 lines (104 loc) · 4.3 KB
Raw

PyPDS

This python package is suitable for working with Planetary Data System (PDS) data products in your own projects.

Several command line tools are also available for viewing labels and image contents as well as image conversion.

Overview

PyPDS is a python interface to Planetary Data System (PDS) data products. For more information please visit the PyPDS Wiki and the Sphinx documentation included in the source.

Documentation

The latest documentation is available on Read the Docs.

Installation

  1. Install PIL

  2. Install PyPDS

    • From the Python Package Index

      pip install PyPDS

    • From the GitHub repository

      git clone git://github.com/RyanBalfanz/PyPDS.git

If you want to contribute, consider creating your own fork and submitting a pull request.

Examples

First, you may want to run tests.py to verify that things seem to be working. That files was based on the interacitve examples below. There are more examples in the documentation.

Working with labels

Before we may do anything, let's import some modules.

>>> import pprint
>>> from pds.core.common import open_pds
>>> from pds.core.parser import Parser

If we are to get all the labels from a PDS file, we'll need a Parser object.

>>> parser = Parser()

Now we use use it's parse method on a file. The file should first be opened with open_pds.

>>> labels = parser.parse(open_pds("../test_data/pds.img"))

The return object is just a dictionary. Let's take a look.

>>> pprint.pprint(labels.keys()) # Inspect the top-level labels.
['IMAGE_HISTOGRAM',
 'IMAGE_MAP_PROJECTION_CATALOG',
 'FILE_RECORDS',
 'IMAGE_ID',
 'TARGET_NAME',
 'SOURCE_IMAGE_ID',
 'IMAGE',
 'LABEL_RECORDS',
 'SPACECRAFT_NAME',
 'NOTE',
 'RECORD_TYPE',
 'CCSD3ZF0000100000001NJPL3IF0PDS200000001',
 '^IMAGE',
 'INSTRUMENT_NAME',
 '^IMAGE_HISTOGRAM',
 'DATA_SET_ID',
 'RECORD_BYTES']

Some of the labels contain other labels, which are also dictionary objects.

>>> pprint.pprint(labels["IMAGE"]) # Inspect the image labels.
{'CHECKSUM': '12054227',
 'LINES': '320',
 'LINE_SAMPLES': '306',
 'SAMPLE_BITS': '8',
 'SAMPLE_BIT_MASK': '2#11111111#',
 'SAMPLE_TYPE': 'UNSIGNED_INTEGER'}

Working with the image

PyPDS also takes care of the details about creating image objects. Behind the scenes all images are instances of PIL's Image class.

To get an image from a PDS file, create an ImageExtractor object and use its extract method. Don't forget to first open the PDS file.

>>> from pds.core.common import open_pds
>>> from pds.imageextractor import ImageExtractor
>>> ie = ImageExtractor()
>>> img, labels = ie.extract(open_pds("../test_data/pds.img"))

The extract method first parses the file, then creates the image. Since it goes to the trouble of doing so anyway, it provides the label as a freebie, along with the image.

Here, img is an instance of PIL's Image class. Do whatever you want to it.

>>> print (img.mode, img.size)
('L', (306, 320))
>>> img.show() # Open the image in the default viewer.
>>> img.save("pds.img.jpeg") # Write the image to disk in JPEG format.

Verify that the returned image has the proper dimensions.

>>> imageSize = map(int, \
... (labels["IMAGE"]["LINE_SAMPLES"], \
... labels["IMAGE"]["LINES"])) # Save the image dimensions with integers.
>>> tuple(imageSize) == img.size # The built-in map returns a list, but Image.size is a tuple.
True

By the way, an Image has a show method which should happily open the image in your default viewer.

Command Line Tools

Sometimes you might not want to interact with PDS files programmatically. PyPDS also comes with several tools which are handy for working with PDS files at the command line.

Each has several options. For detailed information use --help.

pds-convert.py
Convert images to the specified format.
pds-image.py
Like pds-convert.py but dump to standard output.
pds-labels.py
Dump the labels to standard output.
pds-view.py
View an image contained in a PDS file in the default viewer.