
Development
***********

Tablib is under active development, and contributors are welcome.

If you have a feature request, suggestion, or bug report, please open
a new issue on GitHub. To submit patches, please send a pull request
on GitHub.

If you'd like to contribute, there's plenty to do. Here's a short todo
list.

   * Hooks System - pre/post-append - pre/post-import - pre/post-
     export

   * Add Tablib.ext namespace

   * Width detection for XLS output

   * Documentation Improvements


Design Considerations
=====================

Tablib was developed with a few **PEP 20** idioms in mind.

1. Beautiful is better than ugly.

2. Explicit is better than implicit.

3. Simple is better than complex.

4. Complex is better than complicated.

5. Readability counts.

A few other things to keep in mind:

1. Keep your code DRY.

2. Strive to be as simple (to use) as possible.


Source Control
==============

Tablib source is controlled with Git, the lean, mean, distributed
source control machine.

The repository is publicly accessable.

   "git clone git://github.com/kennethreitz/tablib.git"

The project is hosted on **GitHub**.

   GitHub:
      http://github.com/kennethreitz/tablib


Git Branch Structure
--------------------

Feature / Hotfix / Release branches follow a Successful Git Branching
Model . Git-flow is a great tool for managing the repository. I highly
recommend it.

"develop"
   The "next release" branch. Likely unstable.

"master"
   Current production release (0.9.11) on PyPi.

"gh-pages"
   Current release of http://tablib.org.

Each release is tagged.

When submitting patches, please place your feature/change in its own
branch prior to opening a pull reqeust on GitHub.


Adding New Formats
==================

Tablib welcomes new format additions! Format suggestions include:

* Tab Separated Values

* MySQL Dump

* HTML Table


Coding by Convention
--------------------

Tablib features a micro-framework for adding format support. The
easiest way to understand it is to use it. So, let's define our own
format, named *xxx*.

1. Write a new format interface.

      "tablib.core" follows a simple pattern for automatically
      utilizing your format throughout Tablib. Function names are
      crucial.

      Example **tablib/formats/_xxx.py**:

         title = 'xxx'

         def export_set(dset):
             ....
             # returns string representation of given dataset

         def export_book(dbook):
             ....
             # returns string representation of given databook

         def import_set(dset, in_stream):
             ...
             # populates given Dataset with given datastream

         def import_book(dbook, in_stream):
             ...
             # returns Databook instance

         def detect(stream):
             ...
             # returns True if given stream is parsable as xxx

Excluding Support: If the format excludes support for an import/export
mechanism (*eg.* "csv" excludes "Databook" support), simply don't
define the respective functions. Appropriate errors will be raised.

1. Add your new format module to the "tablib.formats.avalable" tuple.

2. Add a mock property to the "Dataset" class with verbose
   reStructured Text docstring. This alleviates IDE confusion, and
   allows for pretty auto-generated Sphinx documentation.

3. Write respective *tests*.


Testing Tablib
==============

Testing is crucial to Tablib's stability. This stable project is used
in production by many companies and developers, so it is important to
be certain that every version released is fully operational. When
developing a new feature for Tablib, be sure to write proper tests for
it as well.

When developing a feature for Tablib, the easiest way to test your
changes for potential issues is to simply run the test suite directly.

   $ ./test_tablib.py

Jenkins CI, amongst other tools, supports Java's xUnit testing report
format. Nose allows us to generate our own xUnit reports.

Installing nose is simple.

   $ pip install nose

Once installed, we can generate our xUnit report with a single
command.

   $ nosetests test_tablib.py --with-xunit

This will generate a **nosetests.xml** file, which can then be
analyzed.


Continuous Integration
======================

Every commit made to the **develop** branch is automatically tested
and inspected upon receipt with Jenkins CI. If you have access to the
main repository and broke the build, you will receive an email
accordingly.

Anyone may view the build status and history at any time.

   http://ci.kennethreitz.com/

If you are trustworthy and plan to contribute to tablib on a regular
basis, please contact Kenneth Reitz to get an account on the Jenkins
Server.

Additional reports will also be included here in the future, including
**PEP 8** checks and stress reports for extremely large datasets.


Building the Docs
=================

Documentation is written in the powerful, flexible, and standard
Python documentation format, reStructured Text. Documentation builds
are powered by the powerful Pocoo project, Sphinx. The *API
Documentation* is mostly documented inline throughout the module.

The Docs live in "tablib/docs". In order to build them, you will first
need to install Sphinx.

   $ pip install sphinx

Then, to build an HTML version of the docs, simply run the following
from the **docs** directory:

   $ make html

Your "docs/_build/html" directory will then contain an HTML
representation of the documentation, ready for publication on most web
servers.

You can also generate the documentation in **ebpub**, **latex**,
**json**, *&c* similarly.

GitHub Pages: To push the documentation up to GitHub Pages, you will
first need to run sphinx-to-github against your "docs/_build/html"
directory.GitHub Pages are powered by an HTML generation system called
Jeckyl, which is configured to ignore files and folders that begin
with ""_"" (*ie.* **_static**).

   and sphinx-to-github.

Installing sphinx-to-github is simple.

   $ pip install sphinx-to-github

Running it against the docs is even simpler.

   $ sphinx-to-github _build/html

Move the resulting files to the **gh-pages** branch of your
repository, and push it up to GitHub.

======================================================================

Make sure to check out the *API Documentation*.
