gcis-py-client

gcis-py-client provides a python-based abstraction of the Global Change Information System Structured Data Server API.

Installation

Requirements

These should be available through the package manager of your choice (Macports, apt, yum, etc.):

• Python 2.7.x (though it might work with previous versions)
• Setuptools 2.0.x
• pip is highly recommended

To install from git with pip:

pip install git+http://github.com/USGCRP/gcis-py-client

To install from tarball:

tar -xzvf GcisPyClient-x.y.tar.gz
cd GcisPyClient-x.y

python setup.py test (to run the test suite)
python setup.py install (to install)

Setup

User Credentials

GcisClient will use the first credentials it finds while searching the following places:

1. Strings passed directly to the GcisClient constructor.
2. Environment variables "GCIS_USER" and "GCIS_KEY" (both are required)
3. Config file at $HOME/etc/Gcis.conf (A sample of this file is included below)
4. Config file at $HOME/.gcis-py-client/Gcis.conf
5. Interactively, with a username and key prompt

Sample Gcis.conf file:

- url      : http://data-stage.globalchange.gov
  userinfo : me@example.com:298015f752D99E789056EF826A7D97afc38a8bbd6e3e23b3
  key      : M2FiLtG2n2qTyJHIztvHm5zweTYjNkM5ZWEtYjNkMS00LTgS00LTg2N2QtYZDFhzQyNGUxCg==

- url      : http://data.globalchange.gov
  userinfo : username:pass
  key      : key

Usage

from gcis_clients import GcisClient

base_url = 'http://data.gcis-dev-front.joss.ucar.edu'
gcis = GcisClient(base_url, 'andrew.buddenberg@noaa.gov', 'your api key here')

Make sure our credentials work:

status_code, resp_text = gcis.test_login()
print status_code, resp_text
assert 'auth_required' not in resp_text

Let's pull a list of all figures in Chapter 2 of the nca3:

for partial_figure in gcis.get_figure_listing('nca3', chapter_id='our-changing-climate'):
    full_figure = gcis.get_figure('nca3', partial_figure.identifier, chapter_id='our-changing-climate')
    print full_figure

Let's work with the infamous temperature figure:

fig2_7 = gcis.get_figure('nca3', 'observed-us-temperature-change')

Warning: Images and Chapters are specifically excluded from JSON output. This is what gets sent to GCIS. So...

print fig2_7.as_json(indent=4)

...you might want to view the original input:

print fig2_7.original

Let's see about the images:

for image in fig2_7.images:
    print image

How about the whole Image?:

print gcis.get_image('69da6d93-4426-4061-a2a1-7b3d01f2dc1c').as_json(indent=4)

Want to find the lineage of a publication?

fig2_1 = gcis.get_figure('nca3', 'ten-indicators-of-a-warming-world')
print fig2_1.parents

Add a parent to a figure directly...

cmip3 = gcis.get_dataset('nca3-cmip3-r201205')
gcis.associate_figure_with_parent('nca3', 'ten-indicators-of-a-warming-world', Parent.from_obj(cmip3))

...or as part of an update

fig2_1.add_parent(Parent.from_obj(cmip3))
gcis.update_figure('nca3', 'our-changing-climate', fig2_1)

NOTE: GcisClient will disassociate and then reassociate Parents as part of an update. Adding a new Parent during an update will cause "Parent dissociation failed" messages to appear--these can be safely ignored.

ALSO: Updates do NOT, presently, remove parents you have not included on the Figure object.

Remove a parent from a figure

gcis.delete_figure_parent_assoc('nca3', 'ten-indicators-of-a-warming-world', Parent.from_obj(cmip3))

Building a source distribution

$ python setup.py sdist

This builds a .tar.gz file of this package.