project directory does not exist, i.e. if a new project should be created, the
new -n switch is required. New syntax to run laydi is now
{{{
laydi [options] </path/to/project>
}}}
* Added -c option to generate intial configuration file and directories. This will create
the file ~/.laydi and the directories ~/laydi/datasets and ~/laydi/workflows.
* Removed debug print statements.
Usages is txt2ftsv [-d dimension] [-o output.ftsv] [-n dsname] [FILE1 [FILE2 ...]]
Input and output can be - for STDIN/STDOUT respectively. The script is primarily ment
as a hack to easily load selections. Data is only half sanitized.
still allowing workflows to be placed in several locations in the file system.
The installation procedure is now ./configure && make install configure will
run m4 on fluents/paths.py.m4, Makefile.m4 and doc/Makefile.m4 so that the
installation system and the program will know where some important directories
are located.
The paths.py.m4 and consequently also pahts.py files are just listings of
directories configured during install. I did this to separate these from all
other files so that as little as possible is touched by m4. It is still
necessary to do an install now to get the program to run in a clean checkout.
Workflows can now be placed anywhere in the system. This is done by setting the
workflowdir variable in the configuration file. All workflow directories,
separated by semicolons, are added to the python path.
The use of setup.py is now deprecated.
- Navigator
- Aplication
- Workflow
- Project
- Options
Corresponding changes have been added in lots of other files to account for this, but
the access to these objects should now be a lot easier.
einarr