def __init__(self, data):
from corpus import Corpus
# handle a folder containing corpora
if type(data) == str or type(data) == str:
import os
from os.path import join, isfile, isdir
if not os.path.isdir(data):
raise ValueError('Corpora(str) needs to point to a directory.')
data = sorted([join(data, d) for d in os.listdir(data) \
if isdir(join(data, d))])
# otherwise, make a list of Corpus objects
for index, i in enumerate(data):
if type(i) == str:
data[index] = Corpus(i)
# now turn it into a Datalist
Datalist.__init__(self, data)
def __init__(self, data):
from corpus import Corpus
# handle a folder containing corpora
if type(data) == str or type(data) == str:
import os
from os.path import join, isfile, isdir
if not os.path.isdir(data):
raise ValueError('Corpora(str) needs to point to a directory.')
data = sorted([join(data, d) for d in os.listdir(data) \
if isdir(join(data, d))])
# otherwise, make a list of Corpus objects
for index, i in enumerate(data):
if type(i) == str:
data[index] = Corpus(i)
# now turn it into a Datalist
Datalist.__init__(self, data)
def __getitem__(self, key):
from process import makesafe
if isinstance(key, slice):
#Get the start, stop, and step from the slice
return Datalist(
[self[ii] for ii in range(*key.indices(len(self.subcorpora)))])
elif type(key) == int:
return self.subcorpora.__getitem__(makesafe(self.subcorpora[key]))
else:
try:
return self.subcorpora.__getattribute__(key)
except:
from process import is_number
if is_number(key):
return self.__getattribute__('c' + key)
def subcorpora(self):
"""A list-like object containing a corpus' subcorpora."""
import re, os, operator
from os.path import join, isdir
if self.data.__class__ == Datalist or type(self.data) == list:
return self.data
if self.level == 'c':
variable_safe_r = re.compile('[\W0-9_]+', re.UNICODE)
sbs = Datalist(sorted([Subcorpus(join(self.path, d), self.datatype) \
for d in os.listdir(self.path) \
if isdir(join(self.path, d))], \
key=operator.attrgetter('name')))
for subcorpus in sbs:
variable_safe = re.sub(variable_safe_r, '', \
subcorpus.name.lower().split(',')[0])
setattr(self, variable_safe, subcorpus)
return sbs
def files(self):
"""A list-like object containing the files in a folder
>>> corpus.subcorpora[0].files
"""
import re, os, operator
from os.path import join, isdir
if self.level == 's':
#variable_safe_r = re.compile('[\W0-9_]+', re.UNICODE)
fs = sorted([File(f, self.path, self.datatype) for f in os.listdir(self.path) \
if not f.startswith('.')], key=operator.attrgetter('name'))
fs = Datalist(fs)
#for f in fs:
# variable_safe = re.sub(variable_safe_r, '', f.name.lower().split('.')[0])
# setattr(self, variable_safe, f)
return fs