def load_meta(self, group, reformat=True):
""" Load the meta data as a Table
Parameters
----------
group : str
"""
import json
self.meta = spdbu.hdf_decode(self.hdf[group + '/meta'].value,
itype='Table')
# Attributes
self.meta_attr = {}
for key in self.hdf[group + '/meta'].attrs.keys():
if 'SSA' in key:
self.meta_attr[key] = json.loads(
spdbu.hdf_decode(self.hdf[group + '/meta'].attrs[key]))
else:
self.meta_attr[key] = spdbu.hdf_decode(
self.hdf[group + '/meta'].attrs[key])
# Reformat
if reformat:
try:
self.meta['RA_GROUP'].format = '8.4f'
except KeyError: # Backwards compatible, will deprecate
self.meta['RA'].format = '8.4f'
self.meta['DEC'].format = '8.4f'
self.meta['zem'].format = '6.3f'
else:
self.meta['DEC_GROUP'].format = '8.4f'
self.meta['zem_GROUP'].format = '6.3f'
self.meta['WV_MIN'].format = '6.1f'
self.meta['WV_MAX'].format = '6.1f'
# Add group
self.meta.meta['group'] = group
def __init__(self, skip_test=True, db_file=None, verbose=False, **kwargs):
"""
"""
if db_file is None:
try:
db_file = self.grab_dbfile(**kwargs)
except:
raise IOError(
"DB not found. Please either check the corresponding environmental "
"variable or directly provide the db_file")
# Init
self.verbose = verbose
self.open_db(db_file)
# Catalog
self.qcat = QueryCatalog(self.hdf, verbose=self.verbose, **kwargs)
self.cat = self.qcat.cat # For convenience
self.qcat.verbose = verbose
self.groups = self.qcat.groups
self.group_dict = self.qcat.group_dict
self.idkey = self.qcat.idkey
# Groups
self._gdict = {}
# Name, Creation date
self.name = spdbu.hdf_decode(self.qcat.cat_attr['NAME'])
print("Database is {:s}".format(self.name))
print("Created on {:s}".format(
spdbu.hdf_decode(self.qcat.cat_attr['CREATION_DATE'])))
if self.qcat.version is not None:
print("Version: {:s}".format(self.qcat.version))
# Return
return
def load_cat(self, hdf, idkey=None):
""" Open the DB catalog file
Parameters
----------
db_file : str
Name of DB file
idkey : str, optional
Key for ID indices
Returns
-------
"""
import json
# Catalog and attributes
self.cat = Table(hdf['catalog'].value)
self.cat_attr = {}
for key in hdf['catalog'].attrs.keys():
self.cat_attr[key] = spdbu.hdf_decode(hdf['catalog'].attrs[key])
# Set ID key
self.idkey = idkey
if idkey is None:
for key in self.cat.keys():
if 'ID' in key:
if self.idkey is not None:
raise ValueError("Two keys with ID in them. You must specify idkey directly.")
self.idkey = key
# Group dict
self.group_dict = json.loads(spdbu.hdf_decode(hdf['catalog'].attrs['GROUP_DICT']))
self.groups = list(self.group_dict.keys())
try:
self.version = self.cat_attr['VERSION']
except KeyError:
self.version = None
if self.verbose:
print("Available groups: {}".format(self.groups))
def main(pargs):
""" Run
Parameters
----------
pargs
Returns
-------
"""
import h5py
import warnings
from specdb import defs
from specdb import utils as sdbu
dbinfo = defs.dbase_info()
# Open file
hdf = h5py.File(pargs.db_file, 'r')
# Deal with Python 3
catalog = {}
for key in hdf['catalog'].attrs:
catalog[key] = hdf['catalog'].attrs[key]
catalog = sdbu.hdf_decode(catalog)
# Name
try:
dbname = catalog['NAME']
except:
warnings.warn('DB file has no name. Must be really old.')
dbname = None
else:
print("specdb DB file is from the {:s} database".format(str(dbname)))
# Check Creation Date
try:
cdate = catalog['CREATION_DATE']
except:
warnings.warn('DB file has no Creation Date. Must be really old..')
return
else:
version = catalog['VERSION']
print("specdb DB file version={} was created on {}".format(str(version),str(cdate)))
if dbname is not None:
try:
print("Latest version for specdb DB type={} is version={}".format(
dbname, str(dbinfo[dbname]['latest_version'])))
except KeyError:
print("No version version history for {:s}".format(str(dbname)))
else:
# Check
print("Latest creation date for this DB version was {}".format(
dbinfo[dbname][version]['newest_date']))
print("Oldest valid DB file for this DB version was {}".format(
dbinfo[dbname][version]['oldest_ok_date']))
# Compare?
# Sources and spectra
print("-------------------------------------------------")
print("There are {:d} unique sources in the source catalog".format(len(hdf['catalog'].value)))
# List datasets
nspec = 0
for key in hdf.keys():
if key == 'catalog':
continue
print("Dataset: {:s}".format(key))
# Spectra
try:
nspec += hdf[key]['spec'].size
except:
pass
print("There are a total of {:d} spectra in the database".format(nspec))