def _initialize(self):
# If we're already initialized, then do nothing.
if self._basedir == get_basedir(): return
# Make sure the corpus is installed.
basedir = get_basedir()
if not os.path.isdir(os.path.join(basedir, self._original_rootdir)):
raise IOError('%s is not installed' % self._name)
self._basedir = basedir
self._rootdir = os.path.join(basedir, self._original_rootdir)
# Read in the data file.
datapath = os.path.join(self._rootdir, self._data_file)
data = open(datapath).read()
# Extract the license
self._license = self.LICENSE_RE.search(data).group(1)
# Extract the description
self._description = self.DESCRIPTION_RE.search(data).group(1)
# Remove line number markings and other comments
data = re.sub(r'<--\s+.*?-->', '', data)
#p\.\s+\d+\s+-->', '', data)
#data = re.sub(r'<--\s+p\.\s+\d+\s+-->', '', data)
# Divide the thesaurus into items.
items = re.split('\n #', data)
self._itemlist = []
self._items = {}
for item in items[1:]:
(key, contents) = item.split('--', 1)
key = ' '.join(key.split()) # Normalize the key.
self._itemlist.append(key)
self._items[key] = contents.strip()
self._itemlist = tuple(self._itemlist)
def _initialize(self):
"Make sure that we're initialized."
# If we're already initialized, then do nothing.
if self._initialized: return
# Make sure the corpus is installed.
basedir = get_basedir()
if not os.path.isabs(self._original_rootdir):
if not os.path.isdir(os.path.join(basedir, self._original_rootdir)):
raise IOError('%s is not installed' % self._name)
self._basedir = basedir
self._rootdir = os.path.join(basedir, self._original_rootdir)
else:
if not os.path.isdir(self._original_rootdir):
raise IOError('%s is not installed' % self._name)
self._basedir = '' # empty
self._rootdir = self._original_rootdir
# # Check the directory for 'merged', and change it to
# # 'combined' if appropriate.
# if 'merged' in self._groups:
# if os.path.isdir(os.path.join(self._rootdir, 'combined')):
# self._group_directory['merged'] = 'combined'
# Get the list of items in each group.
self._group_items = {}
for group in self._groups:
self._find_items(group)
if not self._group_items.has_key('combined'):
self._virtual_merged = 1
self._find_virtual_merged_items()
# Get the overall list of items
self._items = []
for items in self._group_items.values():
self._items += items
# Read metadata from files
if self._description is None and self._description_file is not None:
path = os.path.join(self._rootdir, self._description_file)
self._description = open(path).read()
if self._license is None and self._license_file is not None:
path = os.path.join(self._rootdir, self._license_file)
self._license = open(path).read()
if self._copyright is None and self._copyright_file is not None:
path = os.path.join(self._rootdir, self._copyright_file)
self._copyright = open(path).read()
self._initialized = True
