def uninstall(language):
'''
Uninstall the dictionary of the specified language.
'language': is by convention a string of the form 'll_CC' whereby ll is the
language code and CC the country code.
'''
file_path = hyphen.dict_info[language].filepath
os.remove(file_path)
# Delete all references in dict_info to the removed file
for d in hyphen.dict_info.keys():
if hyphen.dict_info[d].filepath == file_path: hyphen.dict_info.pop(d)
save_dict_info()
def install(language, directory = config.default_dict_path,
repos = config.default_repository, use_description = True):
'''
Download and install a dictionary file.
language: a string of the form 'll_CC'. Example: 'en_US' for English, USA
directory: the installation directory. Defaults to the
value given in config.py. After installation this is the package root of 'hyphen'
repos: the url of the dictionary repository. (Default: as declared in config.py;
after installation of PyHyphen this is LibreOffice's GIT repository .).
'''
# Download the dictionaries.xcu file from the LibreOffice repository if needed
if use_description:
# first try full language name; it won't work in all cases...
language_ext_name = language
descr_url = repos + language_ext_name + '/dictionaries.xcu'
try:
descr_file = urlopen(descr_url)
except URLError:
# OK. So try with the country code.
language_ext_name = language[:2]
descr_url = repos + language_ext_name + '/dictionaries.xcu'
try:
descr_file = urlopen(descr_url)
except URLError:
descr_file = None
# Parse the xml file if it is present, and extract the data.
if use_description and descr_file:
descr_tree = ElementTree(file = descr_file)
# Find the nodes containing meta data of hyphenation dictionaries
# Iterate over all nodes
for node in descr_tree.getiterator('node'):
# Check if node relates to a hyphenation dict.
# We assume this is the case if an attribute value
# contains the substring 'hyph'
node_values = [i[1] for i in node.items()]
iter_values = (i for i in node_values if ('hyph' in i.lower()))
for v in iter_values:
# Found one! So extract the data and construct the local record
for property in node.getchildren():
prop_values = [j[1] for j in property.items()]
for pv in prop_values:
if pv.lower() == 'locations':
# Its only child's text is a list of strings of the form %origin%<filename>
# For simplicity, we only use the first filename in the list.
raw_dict_fn = property.getchildren()[0].text.split()[0]
dict_fn = raw_dict_fn[9:] # strip the prefix '%origin%'
dict_url = ''.join((repos, language_ext_name, '/', dict_fn))
break # skip any other values of this property
elif pv.lower() == 'locales':
# Its only child's text is a list of locales. .
dict_locales = property.getchildren()[0].text.replace('-', '_').split()
break # skip any other values of this property
# Install the dictionary file
dict_str = urlopen(dict_url).read()
filepath = directory + '/' + dict_fn
with open(filepath, 'wb') as dict_file:
dict_file.write(dict_str)
# Save the metadata
# Generate a record for each locale, overwrite any existing ones
new_dict = DictInfo(dict_locales, filepath, url = dict_url)
for l in dict_locales:
hyphen.dict_info[l] = new_dict
# handle the case that there is no xml metadata
else:
# Download the dictionary guessing its URL
dict_fn = ''.join(('hyph_dict_', language, '.dic'))
dict_url = ''.join((repos, dict_fn))
dict_str = urlopen(dict_url).read()
filepath = directory + '/' + dict_fn
with open(filepath, 'w') as dict_file:
dict_file.write(dict_str)
# Store the metadata
new_dict = DictInfo([language], filepath) # the URL is thus set to None.
hyphen.dict_info[language] = new_dict
# Save the modified metadata
save_dict_info()
