def __call__(self, *args, **kwargs):
"""Check that all arguments are present
and load all tracks that are given as paths
instead of track objects. Also checks for
direct calls with generators."""
# Initialization #
generator_call = False # Special switch for direct generator calls
found_args = {} # Will contain a set of parameters extracted
found_tracks = {} # Will contain a set of track parameters extracted
found_generators = {} # Will contain a set of FeatureStream
extra_args = {} # Will contain a set of parameters computed
all_tracks = [] # Will contain all single tracks sent
tracks_to_close = [] # Will contain single tracks to close
virtual_tracks = [] # Will contain the results tracks
rest_of_fields = [] # Will contain variable output fields when required
### Parse arguments ###
for p in self.input_args:
if p['key'] in kwargs: value = kwargs[p['key']]
elif len(args) >= p['position']: value = args[p['position']-1]
elif 'default' in p: value = p['default']
elif p.get('optional'): continue
else: raise Exception("The argument '%s' is missing for the manipulation '%s'." \
% (p['key'], self.short_name))
# Cast it if it's not the right type #
if not isinstance(value, p['type']): value = p['type'](value)
# Add it to the dict #
found_args[p['key']] = value
### Parse tracks ###
for t in self.input_tracks:
if t['key'] in kwargs: value = kwargs[t['key']]
elif len(args) >= t['position']: value = args[t['position']-1]
elif 'default' in t: value = t['default']
elif t.get('optional'): continue
else: raise Exception("The argument '%s' is missing for the manipulation '%s'." \
% (t['key'], self.short_name))
# Check is track collection #
if t.get('kind') == 'many':
if not is_list(value):
message = "The track collection '%s' for the manipulation '%s' is not a list: %s"
raise Exception(message % (t['key'], self.short_name, value))
# Don't modify the input list #
if t.get('kind') == 'many': value = value[:]
# Check for generator case #
if is_gen(value): generator_call = True
if t.get('kind') == 'many' and is_gen(value[0]): generator_call = True
if generator_call:
found_tracks[t['key']] = value
continue
# Check is path #
if isinstance(value, basestring):
value = track.load(value, readonly=True)
tracks_to_close.append(value)
if t.get('kind') == 'many':
for i,_ in enumerate(value):
if isinstance(value[i], basestring):
value[i] = track.load(value[i], readonly=True)
tracks_to_close.append(value[i])
# Add to the list of all tracks #
if t.get('kind') == 'many': all_tracks += [x for x in value]
else: all_tracks += [value]
# Track collection must be combined #
if t.get('kind') == 'many':
value = TrackCollection(value, self.fields_collapse, self.chroms_collapse)
# Variable fields case (track collection must collapse fields) #
if t['fields'][-1] == '...':
first_fields = t['fields'][:-1]
rest_of_fields = [f for f in value.fields if f not in first_fields]
value.fields = first_fields + rest_of_fields
# Specific fields case #
else: value.fields = t['fields']
# What about track SimpleTrack case #
pass #TODO
# Add it to the dict #
found_tracks[t['key']] = value
# Check for generator case #
if generator_call: return self.from_generator(found_tracks, found_args, args, kwargs)
# Collapse chromosomes #
if not self.chroms_collapse: chromosomes = all_tracks[0].chromosomes
else: chromosomes = collapse(self.chroms_collapse, [t.chromosomes for t in all_tracks])
# Multiple output tracks disabled #
t = self.output_tracks[0]
# Make a new virtual track #
vtrack = VirtualTrack()
# Output chromosome metadata #
for chrom in chromosomes:
vtrack.chrmeta[chrom] = {'length': max([i.chrmeta[chrom]['length'] for i in all_tracks])}
# Output attributes #
if t.get('datatype'): vtrack.datatype = t['datatype']
# Output name #
vtrack.name = self.long_name + ' on ' + andify_strings([i.name for i in all_tracks])
### Iterate on chromosomes ###
for chrom in chromosomes:
# Get special input arguments #
for p in self.input_meta:
if p['kind'] == 'chrom_len':
extra_args[p['key']] = vtrack.chrmeta[chrom]['length']
# Call read on tracks #
for k,input_track in found_tracks.items():
if is_list(input_track): found_generators[k] = [i.read(chrom) for i in input_track]
else: found_generators[k] = input_track.read(chrom)
# What about track collapse and recursion #
pass #TODO
# Final argument list #
final_args = {}
for d in (found_args, found_generators, extra_args): final_args.update(d)
# Call generate #
data = self.generate(**final_args)
# Variable fields case #
if t['fields'][-1] == '...': fields = t['fields'][:-1] + rest_of_fields
else: fields = t['fields']
# Make a FeatureStream #
stream = FeatureStream(data, fields)
# Add it to the virtual track #
vtrack.write(chrom, stream)
# Close tracks later #
vtrack.tracks_to_close = tracks_to_close
# Add it #
virtual_tracks.append(vtrack)
# Return one virutal track or list of virtual tracks #
return len(virtual_tracks) == 1 and virtual_tracks[0] or virtual_tracks
def fields(self):
return collapse(self.fields_collapse, [t.fields for t in self.tracks])
def chromosomes(self):
if not hasattr(self.tracks[0], "chromosomes"): return []
chroms = collapse(self.chroms_collapse, [t.chromosomes for t in self.tracks])
chroms.sort(key=natural_sort)
return chroms
def __call__(self, *args, **kwargs):
"""Check that all arguments are present
and load all tracks that are given as paths
instead of track objects. Also checks for
direct calls with generators."""
# Initialization #
generator_call = False # Special switch for direct generator calls
found_args = {} # Will contain a set of parameters extracted
found_tracks = {} # Will contain a set of track parameters extracted
found_generators = {} # Will contain a set of FeatureStream
extra_args = {} # Will contain a set of parameters computed
all_tracks = [] # Will contain all single tracks sent
tracks_to_close = [] # Will contain single tracks to close
virtual_tracks = [] # Will contain the results tracks
rest_of_fields = [
] # Will contain variable output fields when required
### Parse arguments ###
for p in self.input_args:
if p['key'] in kwargs: value = kwargs[p['key']]
elif len(args) >= p['position']: value = args[p['position'] - 1]
elif 'default' in p: value = p['default']
elif p.get('optional'): continue
else: raise Exception("The argument '%s' is missing for the manipulation '%s'." \
% (p['key'], self.short_name))
# Cast it if it's not the right type #
if not isinstance(value, p['type']): value = p['type'](value)
# Add it to the dict #
found_args[p['key']] = value
### Parse tracks ###
for t in self.input_tracks:
if t['key'] in kwargs: value = kwargs[t['key']]
elif len(args) >= t['position']: value = args[t['position'] - 1]
elif 'default' in t: value = t['default']
elif t.get('optional'): continue
else: raise Exception("The argument '%s' is missing for the manipulation '%s'." \
% (t['key'], self.short_name))
# Check is track collection #
if t.get('kind') == 'many':
if not is_list(value):
message = "The track collection '%s' for the manipulation '%s' is not a list: %s"
raise Exception(message %
(t['key'], self.short_name, value))
# Don't modify the input list #
if t.get('kind') == 'many': value = value[:]
# Check for generator case #
if is_gen(value): generator_call = True
if t.get('kind') == 'many' and is_gen(value[0]):
generator_call = True
if generator_call:
found_tracks[t['key']] = value
continue
# Check is path #
if isinstance(value, basestring):
value = track.load(value, readonly=True)
tracks_to_close.append(value)
if t.get('kind') == 'many':
for i, _ in enumerate(value):
if isinstance(value[i], basestring):
value[i] = track.load(value[i], readonly=True)
tracks_to_close.append(value[i])
# Add to the list of all tracks #
if t.get('kind') == 'many': all_tracks += [x for x in value]
else: all_tracks += [value]
# Track collection must be combined #
if t.get('kind') == 'many':
value = TrackCollection(value, self.fields_collapse,
self.chroms_collapse)
# Variable fields case (track collection must collapse fields) #
if t['fields'][-1] == '...':
first_fields = t['fields'][:-1]
rest_of_fields = [
f for f in value.fields if f not in first_fields
]
value.fields = first_fields + rest_of_fields
# Specific fields case #
else:
value.fields = t['fields']
# What about track SimpleTrack case #
pass #TODO
# Add it to the dict #
found_tracks[t['key']] = value
# Check for generator case #
if generator_call:
return self.from_generator(found_tracks, found_args, args, kwargs)
# Collapse chromosomes #
if not self.chroms_collapse: chromosomes = all_tracks[0].chromosomes
else:
chromosomes = collapse(self.chroms_collapse,
[t.chromosomes for t in all_tracks])
# Multiple output tracks disabled #
t = self.output_tracks[0]
# Make a new virtual track #
vtrack = VirtualTrack()
# Output chromosome metadata #
for chrom in chromosomes:
vtrack.chrmeta[chrom] = {
'length': max([i.chrmeta[chrom]['length'] for i in all_tracks])
}
# Output attributes #
if t.get('datatype'): vtrack.datatype = t['datatype']
# Output name #
vtrack.name = self.long_name + ' on ' + andify_strings(
[i.name for i in all_tracks])
### Iterate on chromosomes ###
for chrom in chromosomes:
# Get special input arguments #
for p in self.input_meta:
if p['kind'] == 'chrom_len':
extra_args[p['key']] = vtrack.chrmeta[chrom]['length']
# Call read on tracks #
for k, input_track in found_tracks.items():
if is_list(input_track):
found_generators[k] = [i.read(chrom) for i in input_track]
else:
found_generators[k] = input_track.read(chrom)
# What about track collapse and recursion #
pass #TODO
# Final argument list #
final_args = {}
for d in (found_args, found_generators, extra_args):
final_args.update(d)
# Call generate #
data = self.generate(**final_args)
# Variable fields case #
if t['fields'][-1] == '...':
fields = t['fields'][:-1] + rest_of_fields
else:
fields = t['fields']
# Make a FeatureStream #
stream = FeatureStream(data, fields)
# Add it to the virtual track #
vtrack.write(chrom, stream)
# Close tracks later #
vtrack.tracks_to_close = tracks_to_close
# Add it #
virtual_tracks.append(vtrack)
# Return one virutal track or list of virtual tracks #
return len(virtual_tracks) == 1 and virtual_tracks[0] or virtual_tracks
def fields(self):
return collapse(self.fields_collapse, [t.fields for t in self.tracks])
def chromosomes(self):
if not hasattr(self.tracks[0], "chromosomes"): return []
chroms = collapse(self.chroms_collapse,
[t.chromosomes for t in self.tracks])
chroms.sort(key=natural_sort)
return chroms