relax SVN archive

Modified how the Exec_info data container is handled.

The class was shifted from prompt.interpreter to prompt.base_class. In addition if not container
is passed into the user function classes, then one is created with the intro flag set to a default
of True.

The relax base object is no longer passed into the class namespace of the user functions.

This was only being used for access to the intro flag (a few other cases exist, but these should be
fixed). Now a new container object with this flag is being passed in. This significantly
simplifies the code paths and removes the need for the test suite to create 'fake' relax instances.

Eliminated the dummy mode.

This is only used by the fetch_docstrings script and will soon not be necessary.

Removed some unused and commented out code.

Redesign of the setup of the names spaces for the user functions.

All the user functions and other objects located in the run() method namespace have been shifted
into the interpreter self._locals dictionary. These are then placed into the run() method namespace
by looping over the keys and using the 'locals()[key] = object' syntax. There may be better ways to
do this.

This has no visible effect on the prompt or script based interfaces. It will change how this is
seen by other modules though, and will allow the namespace to be more easily accessed.

Merged revisions 10296-10302 via svnmerge from
svn+ssh://bugman@.../svn/relax/1.3

........
r10296 | bugman | 2010-01-25 11:13:06 +0100 (Mon, 25 Jan 2010) | 14 lines

Spun out the automatic model-free protocol code from the full_analysis.py script.

This is discussed in the post https://mail.gna.org/public/relax-devel/2010-01/msg00075.html.

The aim is to have a single code base for the automatic protocols for all UI designs (prompt,
script, GUI, web interface, etc.). The API interface will be via class initialisations. The
automatic model-free protocol has been placed into the dAuvergne_protocol class.

Future aims of this code is to have the UI fill a single data pipe with all the data required for
the analysis, and then to pass in this pipe into the protocol. Then the auto-analysis will copy
this info into any other data pipes it requires. This will allow for better GUI interaction, data
display, etc.
........
r10297 | bugman | 2010-01-25 11:33:22 +0100 (Mon, 25 Jan 2010) | 3 lines

Removed all the now unused imports and a useless comment.
........
r10298 | bugman | 2010-01-25 11:34:54 +0100 (Mon, 25 Jan 2010) | 3 lines

Updated the copyright notice.
........
r10299 | bugman | 2010-01-25 11:37:25 +0100 (Mon, 25 Jan 2010) | 3 lines

Fixed the import of the dAuvergne_protocol class.
........
r10300 | bugman | 2010-01-25 12:13:59 +0100 (Mon, 25 Jan 2010) | 3 lines

Modifications to the phthalic_acid_noes.py system test script to allow for standalone operation.
........
r10301 | bugman | 2010-01-25 12:40:54 +0100 (Mon, 25 Jan 2010) | 5 lines

Fixes for the N-state model for when only NOEs are present.

This fixes a bug introduced at r10243.
........
r10302 | bugman | 2010-01-25 12:54:34 +0100 (Mon, 25 Jan 2010) | 5 lines

Bug fix for the bug fix of r10301!

The bug fix actually introduced a bug, and this commit fixes that!
........

Bug fix for the bug fix of r10301!

The bug fix actually introduced a bug, and this commit fixes that!

Fixes for the N-state model for when only NOEs are present.

This fixes a bug introduced at r10243.

Modifications to the phthalic_acid_noes.py system test script to allow for standalone operation.

Fixed the import of the dAuvergne_protocol class.