def new_f(args,blargs):
tups = tuple(map(tuple,f(args,blargs)))
if args.dry_run:
print pipestr(tups,stdout=args.out,append=args.mode=='a')
return 0
with utils.quickopen(args.out,args.mode) as out:
process = pipe_together(tups,stdout=out,shell=False,bufsize=-1)
return process.wait()
def parseHeaderedCSV(fname,header=None,intflds=[],fltflds=[],
txtflds=[],delim=None,**kwds):
'''Converts a csv file, with a header line, to a sequence of
nameholder objects, with keys corresponding to the fields in the
header line. This is a generator function (i.e. returns an iterator).
*intflds*, *fltflds*, and *txtflds*, if specified, are sequences of
strings which are required to appear as field names (i.e. as fields
of the header). A key's presence in *fltflds* signal that its value
should be converted to a float; similarly, the value for a key in
*intfields* will be converted to a long.
Additional keywords may be specified. If KEY=CONV is included among
the arguments, then the converter function CONV will be applied to
records in the field KEY; additionally, similarly to txtfields and
others, the function will fail if a user-specified key is not found
in the header line. User-specified conversions take precedence over
float conversions, which in turn take precedence over long
conversions.
*header*, if specified, should be a sequence of labels corresponding
to the columns of the file. If omitted, this function assumes the
first line of the file is a header line; otherwise, that line is taken
to be data.
'''
def check(header):
head = header
header = map(str.upper,header)
missingfields = [nm for nm in conversions if nm not in header]
if missingfields: raise Error(('Required fields ({}) not found in ' +
'file {!r}; detected fields {} instead.').format(
', '.join(missingfields), fname, ', '.join(head)))
return header
def yieldable(tup,lineno,f):
if len(tup) != len(header): raise Error(
'line %d in file %s does not match header:'%(lineno,f.name) +
delim.join(header) + '\n' + delim.join(tup))
try: r = nameholder(zip(header,tup),_conversions=conversions)
except ValueError: raise Error('Conversion error at line %d'%lineno +
', file %r' % f.name)
r.close() ; return r
conversions = dict((x,str) for x in txtflds)
conversions.update((x,int) for x in intflds)
conversions.update((x,float) for x in fltflds)
conversions.update(kwds) # user-defined types / conversion functions
if header is not None: header = check(header)
with utils.quickopen(fname,'r') as f:
try: s = next(f).rstrip()
except StopIteration: return # indicates file was empty
if delim is None:
for delim in ',\t':
s_s = s.split(delim)
if len(s_s) > 1: break
else: raise Error('First line of file %r improperly '%fname +
'improperly formatted: must be comma-separated or tab-delimited\n'+
'problem line: %r'%s)
else:
s_s = s.split(delim)
names = {'\t':'tabs', ',':'commas'}
if len(s_s)==1: raise Error('First line of file %r' %fname +
'improperly formatted: expected fields delimited by ' +
names.get(delim,delim) + '\nproblem line: %r'%s)
if header is None: header = check(s_s)
else: yield yieldable(s_s,1,f)
for line,rec in enumerate(f,2):
yield yieldable(rec.rstrip().split(delim),line,f)
