*** Changes in ParselTongue 1.0.5

* New functionality

The so called "random parameters" associated with each visibility can
now be changed:

>>> for vis in uvdata:
...   vis.weight *= 2
...   vis.uvw = [-x for x in vis.uvw]
...   vis.update()

Changing the raw visibilities themselves is still not possible.

Also, you no lomger need to guess what keywords are attached to a data
set or table:

>>> image = AIPSImage('4C39.25', 'ICL001', 1, 1)
>>> image.keywords
{'ZENANGLE': 20.997356414794922, 'PARANGLE': -49.165603637695312}

It's also possible to change keywords on data sets now:

>>> from Wizardry.AIPSData import AIPSImage as WizAIPSImage
>>> image = WizAIPSImage('4C39.25', 'ICL001', 1, 1)
>>> image.keywords['ZENANGLE'] = 21

Don't forget to update the keywords to make the changes permamnent:

>>> image.keywords.update()

It's now easier to use Wizardry functionality on "normal" datasets,
since you can convert one into the other:

>>> from Wizardry.AIPSData import AIPSImage as WizAIPSImage
>>> image = AIPSImage('4C39.25', 'ICL001', 1, 1)
>>> wiz_image = WizAIPSImage(image)
>>> image = AIPSImage(wiz_image)

There's also a new AIPSCat class that allows you to access the AIPS
catalog:

>>> print AIPSCat()
Catalog on disk  1
 Cat Mapname      Class   Seq  Pt     Last access
   1 4C39.25     .ICL001.    1 MA 19-May-2006 09:50:43
Catalog on disk  2
 Cat Mapname      Class   Seq  Pt     Last access

It can be used to zap multiple files in one go:

>>> AIPSCat(2).zap(klass='SUBIM')

will zap all files with class 'SUBIM' on AIPS disk 2.

ParselTongue can now deal with interactive AIPS tasks.

* Bug fixes

Better limits on disk, box and channel adverbs.

Better support for "experimental" adverbs.

Default values for array adverbs should stay constant now.


*** Changes in ParselTongue 1.0.4

* New functionality

Modifying headers of images and UV data should work now:

>>> from Wizardry.AIPSData import AIPSUVData as WizAIPSUVData
>>> uvdata = WizAIPSUVData('N05C2', 'UVDATA', 1, 1)
>>> uvdata.header.date_map = '2006-04-05'

Don't forget to update the header to make the changes permanent:

>>> uvdata.header.update()

It is now possible to construct instances of the Wizardry classes from
non-Wizardry instances and vice versa:

>>> from Wizardry.AIPSData import AIPSUVData as WizAIPSUVData
>>> data = AIPSUVData('N05C2', 'UVDATA', 1, 1)
>>> wizdata = WizAIPSUVData(data)
>>> data2 = AIPSUVData(wizdata)

* Bug fixes

Opening the highest version of an extension table of an image should
works now.


*** Changes in ParselTongue 1.0.3

* New functionality

Attaching CC tables to images should work now.  By default, a table
for point source models is created:

>>> image.attach_table('CC', 0)

For gaussian models (used for multi-scale CLEAN) you have to do

>>> image.attach_table('CC', 0, no_parms=4)

However, it is not clear whether AIPS reads tables created with
no_parms=4 correctly.

AIPSData.zap() has been changed such that you can now forcefully zap
data that's still marked as busy:

>>> uvdata.zap(force=True)

ParselTongue will now display its version number when starting an
interactive session or when invoking ParselTongue with the --version
option.

* Bug fixes

ParselTongue will no longer zap tables from data that's still marked
as busy (unless you specify force=True of course).


*** Changes in ParselTongue 1.0.2

* New functionality

CLRSTAT-like functionality is now implemented.  If an AIPS task falls
over and leaves a data set in a READ or WRITE state, you can use:

>>> uvdata = AIPSUVData('MULTI', 'UVDATA', 1, 1)
>>> uvdata.clrstat()

to clear all status flags.

* Bug fixes

Setting adverbs that are multi-dimensional arrays (like DOWIDTH for
SAD) should work correctly now.

Pathnames specified for INFILE2, OFMFILE, BOXFILE and OBOXFILE can now
be longer than 48 characters, just like INFILE, OUTFILE and OUTPRINT.

*** Changes in ParselTongue 1.0.1

* Bug fixes

ParselTongue will now deal gracefully with missing readline support in
Python.  The AIPSUVData.antennas, AIPSUVData.polarizations and
AIPSUVData.sources attributes no longer incorrectly cache their
results.

Obit no longer leaks huge amounts of memory, can now access FG tables
and now produces CL tables that no longer confuse AIPS.

* New functionality

Dictonary-like access to table rows is now possible, and

>>> uvdata = AIPSUVData('MULTI', 'UVDATA', 1, 1)
>>> print len(uvdata)

prints the number of visibilities in the data set.

*** Changes in ParselTongue 1.0

* Renaming data sets

It is now possible to rename images and UV data:

>>> uvdata = AIPSUVData('MULTI', 'UVDATA', 1, 1)
>>> uvdata.rename('N05C2', 'UVDATA', 1, 1)

* History access

You can now access the history table of a data set:

>>> print uvdata.history[6]

will print the 7th record in the history.

* Changing keywords

It is now possible to change non-standard keywords:

>>> uvdata.keywords['OLDFRQ'] = 1.64249000e9

will set the keyword 'OLDFRQ' to 1.6 GHz.  This also works for tables,
but make sure you close the table explicitly when you're done with it,
otherwise you'll lose the change:

>>> antable.keywords['ARRNAM'] = 'EVN'
>>> antable.close()


*** Changes in ParselTongue 0.99

* Support for the AIPS TV

It is now possible to run AIPS tasks that use the AIPS TV, provided
the AIPS TV is running on the local machine and it is set up to use
internet sockets for communication with AIPS tasks.  A TVServer script
to start the AIPS TV server is provided.  Some limited support to
control the TV from ParselTongue is also included:

>>> tv = AIPSTV()
>>> tv.start()

will start the AIPS TV from ParselTongue

>>> tv.clear()

will clear the AIPS TV ("TVINIT").

>>> tv.kill()

will close down the AIPS TV in a somewhat gracefull way.

* Support for more AIPS disks

ParselTongue now supports up to 35 AIPS disks.


*** Changes in ParselTongue 0.97

* Interface changes

AIPSUVData.header, AIPSImage.header, AIPSUVData.tables and
AIPSImage.tables are now properties of the data set.  The header
property has its own attributes.  The names of these attributes match
the names used by AIPS' GETHEAD and PUTHEAD verbs.  So instead of:

>>> print AIPSImage.header()['OBSDAT']
>>> print AIPSUVData.tables()

you now should use:

>>> print AIPSImage.header.date_obs
>>> print AIPSUVData.tables


AIPSTable.header should no longer be used since a much better syntax
now exists to deal with tables.  So instead of:

    sutable = uvdata.table('AIPS SU', 0)
    suheader = sutable.header()
    nrow = suheader['nrow']
    for rowno in xrange(1, nrow + 1):
        row = sutable.getrow(rowno)
        print row['SOURCE'][0].strip()

you now should use:

    sutable = uvdata.table('SU', 0)
    for row in sutable:
        print row.source.strip()


This should make the ParselTongue interface much more consistent.