def get_filepath_or_buffer(filepath_or_buffer, encoding=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
return maybe_read_encoded_stream(req, encoding)
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
conn = boto.connect_s3()
b = conn.get_bucket(parsed_url.netloc)
k = boto.s3.key.Key(b)
k.key = parsed_url.path
filepath_or_buffer = StringIO(k.get_contents_as_string())
return filepath_or_buffer, None
return filepath_or_buffer, None
def get_filepath_or_buffer(filepath_or_buffer, encoding=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
return maybe_read_encoded_stream(req,encoding)
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
conn = boto.connect_s3()
b = conn.get_bucket(parsed_url.netloc)
k = boto.s3.key.Key(b)
k.key = parsed_url.path
filepath_or_buffer = StringIO(k.get_contents_as_string())
return filepath_or_buffer, None
return filepath_or_buffer, None
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compression == 'infer':
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
compression = 'gzip'
else:
compression = None
# cat on the compression to the tuple returned by the function
to_return = list(maybe_read_encoded_stream(req, encoding, compression)) + \
[compression]
return tuple(to_return)
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
try:
conn = boto.connect_s3()
except boto.exception.NoAuthHandlerFound:
conn = boto.connect_s3(anon=True)
b = conn.get_bucket(parsed_url.netloc, validate=False)
if compat.PY2 and (compression == 'gzip' or
(compression == 'infer' and
filepath_or_buffer.endswith(".gz"))):
k = boto.s3.key.Key(b, parsed_url.path)
filepath_or_buffer = BytesIO(k.get_contents_as_string(
encoding=encoding))
else:
k = BotoFileLikeReader(b, parsed_url.path, encoding=encoding)
k.open('r') # Expose read errors immediately
filepath_or_buffer = k
return filepath_or_buffer, None, compression
# It is a pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
return _expand_user(filepath_or_buffer), None, compression
def urlopen(url):
sys.stdout.flush()
url = url.replace('&max-results=20', '&max-results=100')
if '&key' not in url:
url += key
print url
return _urlopen(url, timeout=60).read()
def get_recent_changes(self):
"""Returns three lists of the newest weekly files (added,mod,obsolete).
Reads the directories with changed entries from the PDB server and
returns a tuple of three URL's to the files of new, modified and
obsolete entries from the most recent list. The directory with the
largest numerical name is used.
Returns None if something goes wrong.
Contents of the data/status dir (20031013 would be used);
drwxrwxr-x 2 1002 sysadmin 512 Oct 6 18:28 20031006
drwxrwxr-x 2 1002 sysadmin 512 Oct 14 02:14 20031013
-rw-r--r-- 1 1002 sysadmin 1327 Mar 12 2001 README
"""
url = self.pdb_server + '/pub/pdb/data/status/'
with contextlib.closing(_urlopen(url)) as handle:
recent = filter(str.isdigit,
(x.split()[-1] for x in handle.readlines())
)[-1]
path = self.pdb_server + '/pub/pdb/data/status/%s/' % (recent)
# Retrieve the lists
added = self.get_status_list(path + 'added.pdb')
modified = self.get_status_list(path + 'modified.pdb')
obsolete = self.get_status_list(path + 'obsolete.pdb')
return [added, modified, obsolete]
def get_all_obsolete(self):
"""Returns a list of all obsolete entries ever in the PDB.
Returns a list of all obsolete pdb codes that have ever been
in the PDB.
Gets and parses the file from the PDB server in the format
(the first pdb_code column is the one used). The file looks
like this:
LIST OF OBSOLETE COORDINATE ENTRIES AND SUCCESSORS
OBSLTE 31-JUL-94 116L 216L
...
OBSLTE 29-JAN-96 1HFT 2HFT
OBSLTE 21-SEP-06 1HFV 2J5X
OBSLTE 21-NOV-03 1HG6
OBSLTE 18-JUL-84 1HHB 2HHB 3HHB
OBSLTE 08-NOV-96 1HID 2HID
OBSLTE 01-APR-97 1HIU 2HIU
OBSLTE 14-JAN-04 1HKE 1UUZ
...
"""
url = self.pdb_server + '/pub/pdb/data/status/obsolete.dat'
with contextlib.closing(_urlopen(url)) as handle:
# Extract pdb codes. Could use a list comprehension, but I want
# to include an assert to check for mis-reading the data.
obsolete = []
for line in handle:
if not line.startswith("OBSLTE "):
continue
pdb = line.split()[2]
assert len(pdb) == 4
obsolete.append(pdb)
return obsolete
def get_all_obsolete(self):
"""Returns a list of all obsolete entries ever in the PDB.
Returns a list of all obsolete pdb codes that have ever been
in the PDB.
Gets and parses the file from the PDB server in the format
(the first pdb_code column is the one used). The file looks
like this:
LIST OF OBSOLETE COORDINATE ENTRIES AND SUCCESSORS
OBSLTE 31-JUL-94 116L 216L
...
OBSLTE 29-JAN-96 1HFT 2HFT
OBSLTE 21-SEP-06 1HFV 2J5X
OBSLTE 21-NOV-03 1HG6
OBSLTE 18-JUL-84 1HHB 2HHB 3HHB
OBSLTE 08-NOV-96 1HID 2HID
OBSLTE 01-APR-97 1HIU 2HIU
OBSLTE 14-JAN-04 1HKE 1UUZ
...
"""
url = self.pdb_server + '/pub/pdb/data/status/obsolete.dat'
with contextlib.closing(_urlopen(url)) as handle:
# Extract pdb codes. Could use a list comprehension, but I want
# to include an assert to check for mis-reading the data.
obsolete = []
for line in handle:
if not line.startswith("OBSLTE "):
continue
pdb = line.split()[2]
assert len(pdb) == 4
obsolete.append(pdb)
return obsolete
def _update_usa(self):
"""
Update whitelist based on usa.gov
"""
print 'Getting agencies from usa.gov...'
url_base = 'https://www.usa.gov'
letters = _string.ascii_lowercase
agency_dic = {}
for letter in letters:
url = url_base + '/federal-agencies/' + letter
soup = _BeautifulSoup(_urlopen(url).read())
links_content = [l for l in soup.find_all('ul') if 'class' in l.attrs and 'one_column_bullet' in l['class']]
if len(links_content) == 1:
links_list = links_content[0].find_all('a')
for agency_html in links_list:
name_short = self._preprocess_name(agency_html.string)
agency_dic[name_short] = {'html': agency_html,
'url': url_base + agency_html['href'],
'name_full': agency_html.string,
'source': 'usa.gov'}
print agency_html.string
elif len(links_content) == 0:
pass
else:
raise ValueError('Too many list elements found! Please modify the HTML parser.')
self.agency_dictionary.update(agency_dic)
def get_recent_changes(self):
"""Returns three lists of the newest weekly files (added,mod,obsolete).
Reads the directories with changed entries from the PDB server and
returns a tuple of three URL's to the files of new, modified and
obsolete entries from the most recent list. The directory with the
largest numerical name is used.
Returns None if something goes wrong.
Contents of the data/status dir (20031013 would be used);
drwxrwxr-x 2 1002 sysadmin 512 Oct 6 18:28 20031006
drwxrwxr-x 2 1002 sysadmin 512 Oct 14 02:14 20031013
-rw-r--r-- 1 1002 sysadmin 1327 Mar 12 2001 README
"""
url = self.pdb_server + '/pub/pdb/data/status/'
with contextlib.closing(_urlopen(url)) as handle:
recent = filter(str.isdigit,
(x.split()[-1] for x in handle.readlines()))[-1]
path = self.pdb_server + '/pub/pdb/data/status/%s/' % (recent)
# Retrieve the lists
added = self.get_status_list(path + 'added.pdb')
modified = self.get_status_list(path + 'modified.pdb')
obsolete = self.get_status_list(path + 'obsolete.pdb')
return [added, modified, obsolete]
def urlopen(url):
sys.stdout.flush()
url = url.replace('&max-results=20', '&max-results=100')
if '&key' not in url:
url += key
print url
return _urlopen(url, timeout=60).read()
def urlopen(url, data=None, lang='en'):
request = Request(url, data, {
"Accept-Language": "%s,en-us;q=0.7,en;q=0.3"%lang.lower(),
"User-Agent": UA,
})
logging.debug("urlopen: %s", url)
time.sleep(URLOPEN_DELAY)
return _urlopen(request)
def get_all_entries(self):
"""Retrieves a big file containing all the
PDB entries and some annotation to them.
Returns a list of PDB codes in the index file.
"""
print "retrieving index file. Takes about 5 MB."
url = _urlopen(self.pdb_server + "/pub/pdb/derived_data/index/entries.idx")
return [line[:4] for line in url.readlines()[2:] if len(line) > 4]
def fetch_film_info_from_criticker(film_data):
url = 'http://www.criticker.com/?f=' + film_data['criticker_id']
title_page = None
try:
page = unicode(_urlopen(url, None, 5).read(), 'iso-8859-1')
soup = BeautifulSoup(page)
title_page = soup.find("div", attrs={"id":"fi_info_filmname"})
except URLError, e:
logger.error("URL Error: " + str(e.reason) + ": " + url)
def get_all_entries(self):
"""Retrieves a big file containing all the
PDB entries and some annotation to them.
Returns a list of PDB codes in the index file.
"""
print "retrieving index file. Takes about 5 MB."
url = _urlopen(self.pdb_server +
'/pub/pdb/derived_data/index/entries.idx')
return [line[:4] for line in url.readlines()[2:] if len(line) > 4]
def get_filepath_or_buffer(filepath_or_buffer,
encoding=None,
compression=None,
mode=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
mode : str, optional
Returns
-------
tuple of ({a filepath_ or buffer or S3File instance},
encoding, str,
compression, str,
should_close, bool)
"""
filepath_or_buffer = _stringify_path(filepath_or_buffer)
if _is_url(filepath_or_buffer):
req = _urlopen(filepath_or_buffer)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
req.close()
return reader, encoding, compression, True
if is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression,
mode=mode)
if is_gcs_url(filepath_or_buffer):
from pandas.io import gcs
return gcs.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression,
mode=mode)
if isinstance(filepath_or_buffer,
(compat.string_types, compat.binary_type, mmap.mmap)):
return _expand_user(filepath_or_buffer), None, compression, False
if not is_file_like(filepath_or_buffer):
msg = "Invalid file path or buffer object type: {_type}"
raise ValueError(msg.format(_type=type(filepath_or_buffer)))
return filepath_or_buffer, None, compression, False
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compression == 'infer':
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
compression = 'gzip'
else:
compression = None
# cat on the compression to the tuple returned by the function
to_return = list(maybe_read_encoded_stream(req, encoding, compression)) + \
[compression]
return tuple(to_return)
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
try:
conn = boto.connect_s3()
except boto.exception.NoAuthHandlerFound:
conn = boto.connect_s3(anon=True)
b = conn.get_bucket(parsed_url.netloc, validate=False)
if compat.PY2 and (compression == 'gzip' or
(compression == 'infer' and
filepath_or_buffer.endswith(".gz"))):
k = boto.s3.key.Key(b, parsed_url.path)
filepath_or_buffer = BytesIO(k.get_contents_as_string(
encoding=encoding))
else:
k = BotoFileLikeReader(b, parsed_url.path, encoding=encoding)
k.open('r') # Expose read errors immediately
filepath_or_buffer = k
return filepath_or_buffer, None, compression
return _expand_user(filepath_or_buffer), None, compression
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None, mode=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
mode : str, optional
Returns
-------
tuple of ({a filepath_ or buffer or S3File instance},
encoding, str,
compression, str,
should_close, bool)
"""
filepath_or_buffer = _stringify_path(filepath_or_buffer)
if _is_url(filepath_or_buffer):
req = _urlopen(filepath_or_buffer)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
req.close()
return reader, encoding, compression, True
if is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression,
mode=mode)
if is_gcs_url(filepath_or_buffer):
from pandas.io import gcs
return gcs.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression,
mode=mode)
if isinstance(filepath_or_buffer, (compat.string_types,
compat.binary_type,
mmap.mmap)):
return _expand_user(filepath_or_buffer), None, compression, False
if not is_file_like(filepath_or_buffer):
msg = "Invalid file path or buffer object type: {_type}"
raise ValueError(msg.format(_type=type(filepath_or_buffer)))
return filepath_or_buffer, None, compression, False
def load(filename):
if filename[:4] == 'http':
resp = _urlopen(filename)
dict = _json.loads(resp.read(), object_pairs_hook=_parse_json)
else:
filename = os.path.expanduser(filename)
f = open(filename, 'r')
dict = _json.load(f, object_pairs_hook=_parse_json)
f.close()
return dict
def get_all_entries(self):
"""Retrieves a big file containing all the
PDB entries and some annotation to them.
Returns a list of PDB codes in the index file.
"""
print("retrieving index file. Takes about 5 MB.")
url = self.pdb_server + '/pub/pdb/derived_data/index/entries.idx'
with contextlib.closing(_urlopen(url)) as handle:
all_entries = [line[:4] for line in handle.readlines()[2:]
if len(line) > 4]
return all_entries
def get_seqres_file(self, savefile="pdb_seqres.txt"):
"""Retrieves a (big) file containing all the sequences of PDB entries
and writes it to a file.
"""
print "retrieving sequence file. Takes about 15 MB."
handle = _urlopen(self.pdb_server + "/pub/pdb/derived_data/pdb_seqres.txt")
lines = handle.readlines()
outfile = open(savefile, "w")
outfile.writelines(lines)
outfile.close()
handle.close()
def _update_wikipedia(self):
# do a little bit of name preprocessing here too
from requests import ConnectionError
from wikipedia import PageError
print 'Getting data from Wikipedia...'
page_current = _wikipedia.page('List_of_federal_agencies_in_the_United_States')
html = page_current.html()
subset = html[_re.search('<h2>.*?Legislative Branch', html).start():_re.search('<h2>.*?See also', html).start()]
soup = _BeautifulSoup(subset)
links = soup.find_all(lambda x: x.name == 'a' and x.has_attr('href') and '/wiki/' in x['href'] and
'File:' not in x['href'])
agency_dic = {self._preprocess_name(link['title']): {'html': link,
'url': 'https://en.wikipedia.org' + link['href'],
'name_full': link['title'],
'source': 'wikipedia'}
for link in links}
category_pages = ['https://en.wikipedia.org/w/api.php?action=query&list=categorymembers&' +
'cmtitle=Category:Defunct_agencies_of_the_United_States_government&cmlimit=500&format=json',
'https://en.wikipedia.org/w/api.php?action=query&list=categorymembers&' +
'cmtitle=Category:Committees_of_the_United_States_Senate&cmlimit=500&format=json',
'https://en.wikipedia.org/w/api.php?action=query&list=categorymembers&' +
'cmtitle=Category:Joint_committees_of_the_United_States_Congress' +
'&cmlimit=500&format=json',
'https://en.wikipedia.org/w/api.php?action=query&list=categorymembers&' +
'cmtitle=Category:Committees_of_the_United_States_House_of_Representatives' +
'&cmlimit=500&format=json',
]
for category_page in category_pages:
content_defunct = _json.loads(_urlopen(category_page).read())
for result in content_defunct['query']['categorymembers']:
if result['ns'] == 0:
url_defunct = 'https://en.wikipedia.org/wiki/' + _re.sub(' ', '_', result['title'])
print(result['title'])
try:
page_defunct = _wikipedia.page(result['title'])
name_short = self._preprocess_name(result['title'])
agency_dic[name_short] = {'html': page_defunct.html(),
'url': url_defunct,
'name_full': result['title'],
'source': 'wikipedia'}
except (ConnectionError, PageError):
print('Failed to get agency HTML!')
self.agency_dictionary.update(agency_dic)
def urlopen(url, data=None, *args, **kwargs):
if not isinstance(url, Request):
url = Request(url, data)
data = None
if 'basic_auth' in kwargs:
if kwargs['basic_auth']:
a = base64.b64encode(':'.join(kwargs['basic_auth']))
url.add_header('Authorization', 'Basic '+a)
del(kwargs['basic_auth'])
if 'authorization' in kwargs:
if kwargs['authorization']:
url.add_header('Authorization', kwargs['authorization'])
del(kwargs['authorization'])
if sys.version_info[0] == 2:
url.add_header('Host', url.get_origin_req_host())
return _urlopen(url, data, *args, **kwargs)
else:
url.add_header('Host', url.origin_req_host)
kwargs['cadefaults'] = True
return _urlopen(url, data, *args, **kwargs)
def get_xml(xml_link):
try:
xml_data = _urlopen(xml_link).read()
if 'xml' in xml_data[0:100]:
return xml_data
else:
return None
except HTTPError:
return None
def get_seqres_file(self,savefile='pdb_seqres.txt'):
"""Retrieves a (big) file containing all the sequences of PDB entries
and writes it to a file.
"""
print "retrieving sequence file. Takes about 15 MB."
handle = _urlopen(self.pdb_server +
'/pub/pdb/derived_data/pdb_seqres.txt')
lines = handle.readlines()
outfile = open(savefile, 'w')
outfile.writelines(lines)
outfile.close()
handle.close()
def get_all_entries(self):
"""Retrieves a big file containing all the
PDB entries and some annotation to them.
Returns a list of PDB codes in the index file.
"""
print("retrieving index file. Takes about 5 MB.")
url = self.pdb_server + '/pub/pdb/derived_data/index/entries.idx'
with contextlib.closing(_urlopen(url)) as handle:
all_entries = [
line[:4] for line in handle.readlines()[2:] if len(line) > 4
]
return all_entries
def get_filepath_or_buffer(filepath_or_buffer,
encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compression == 'infer':
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
compression = 'gzip'
# cat on the compression to the tuple returned by the function
to_return = list(maybe_read_encoded_stream(req, encoding, compression)) + \
[compression]
return tuple(to_return)
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
try:
conn = boto.connect_s3()
except boto.exception.NoAuthHandlerFound:
conn = boto.connect_s3(anon=True)
b = conn.get_bucket(parsed_url.netloc, validate=False)
k = boto.s3.key.Key(b)
k.key = parsed_url.path
filepath_or_buffer = BytesIO(
k.get_contents_as_string(encoding=encoding))
return filepath_or_buffer, None, compression
return _expand_user(filepath_or_buffer), None, compression
def get_status_list(self, url):
"""Retrieves a list of pdb codes in the weekly pdb status file
from the given URL. Used by get_recent_files.
Typical contents of the list files parsed by this method is now
very simply one PDB name per line.
"""
with contextlib.closing(_urlopen(url)) as handle:
answer = []
for line in handle:
pdb = line.strip()
assert len(pdb) == 4
answer.append(pdb)
return answer
def get_status_list(self, url):
"""Retrieves a list of pdb codes in the weekly pdb status file
from the given URL. Used by get_recent_files.
Typical contents of the list files parsed by this method is now
very simply one PDB name per line.
"""
with contextlib.closing(_urlopen(url)) as handle:
answer = []
for line in handle:
pdb = line.strip()
assert len(pdb) == 4
answer.append(pdb)
return answer
def _get_ids(self):
id_vals = []
years = range(1988, 2017)
for year in years:
soup = _BeautifulSoup(
_urlopen(
'http://www.legislation.gov.uk/ukpga/{0}'.format(year)))
n_results = _re.search('has returned ([0-9]+) results',
soup.text.lower()).group(1)
id_vals += [
str(year) + '_' + str(i) for i in range(1,
int(n_results) + 1)
]
return id_vals
def _get_data(self, publication_id):
max_attempts = 10
attempts = 0
xml_content = None
soup = None
while attempts < max_attempts:
search_id = _re.sub('_', '/', publication_id)
try:
xml_content = _urlopen(
'http://www.legislation.gov.uk/ukpga/{0}/data.xml'.format(
search_id)).read()
soup = _BeautifulSoup(xml_content, 'xml')
break
except:
attempts += 1
if 'amendment' in soup.title.text.lower():
amend = True
else:
amend = False
if 'repeal' in soup.title.text.lower():
repeal = True
else:
repeal = False
if soup.EnactmentDate is not None:
date = soup.EnactmentDate['Date']
elif soup.PrimaryPrelims is not None:
date = soup.PrimaryPrelims['RestrictStartDate']
else:
date = None
print 'warning! No date found.'
meta = _format_meta_entry(country=u'united_kingdom',
title=soup.title.text,
id=publication_id,
date=date,
type=u'annual',
xml=xml_content,
amendment=amend,
repealed=repeal)
return meta
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
url = str(filepath_or_buffer)
req = _urlopen(url)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
return reader, encoding, compression
if _is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# Convert pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
if isinstance(filepath_or_buffer, (compat.string_types,
compat.binary_type,
mmap.mmap)):
return _expand_user(filepath_or_buffer), None, compression
if not is_file_like(filepath_or_buffer):
msg = "Invalid file path or buffer object type: {_type}"
raise ValueError(msg.format(_type=type(filepath_or_buffer)))
return filepath_or_buffer, None, compression
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
url = str(filepath_or_buffer)
req = _urlopen(url)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
return reader, encoding, compression
if _is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# Convert pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
if isinstance(filepath_or_buffer, (compat.string_types,
compat.binary_type,
mmap.mmap)):
return _expand_user(filepath_or_buffer), None, compression
if not is_file_like(filepath_or_buffer):
msg = "Invalid file path or buffer object type: {_type}"
raise ValueError(msg.format(_type=type(filepath_or_buffer)))
return filepath_or_buffer, None, compression
def get_filepath_or_buffer(filepath_or_buffer, encoding=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compat.PY3: # pragma: no cover
if encoding:
errors = 'strict'
else:
errors = 'replace'
encoding = 'utf-8'
out = StringIO(req.read().decode(encoding, errors))
else:
encoding = None
out = req
return out, encoding
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
conn = boto.connect_s3()
b = conn.get_bucket(parsed_url.netloc)
k = boto.s3.key.Key(b)
k.key = parsed_url.path
filepath_or_buffer = StringIO(k.get_contents_as_string())
return filepath_or_buffer, None
return filepath_or_buffer, None
def get_filepath_or_buffer(filepath_or_buffer, encoding=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath, or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compat.PY3: # pragma: no cover
if encoding:
errors = 'strict'
else:
errors = 'replace'
encoding = 'utf-8'
out = StringIO(req.read().decode(encoding, errors))
else:
encoding = None
out = req
return out, encoding
if _is_s3_url(filepath_or_buffer):
try:
import boto
except:
raise ImportError("boto is required to handle s3 files")
# Assuming AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
# are environment variables
parsed_url = parse_url(filepath_or_buffer)
conn = boto.connect_s3()
b = conn.get_bucket(parsed_url.netloc)
k = boto.s3.key.Key(b)
k.key = parsed_url.path
filepath_or_buffer = StringIO(k.get_contents_as_string())
return filepath_or_buffer, None
return filepath_or_buffer, None
def _update_register(self):
print 'Getting agencies from the federal register...'
url_base = 'https://www.federalregister.gov/agencies'
soup = _BeautifulSoup(_urlopen(url_base))
links = soup.find_all(lambda x: x.name == 'li' and x.has_attr('data-filter-live') and not x.has_attr('class'))
agency_dic = {}
for link in links:
agency = link.find('a')
name_short = self._preprocess_name(agency.string)
agency_dic[name_short] = {'html': agency,
'url': agency['href'],
'name_full': agency.string,
'source': 'federal_register'}
print agency.string
self.agency_dictionary.update(agency_dic)
def get_filepath_or_buffer(filepath_or_buffer,
encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compression == 'infer':
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
compression = 'gzip'
else:
compression = None
# cat on the compression to the tuple returned by the function
to_return = (
list(maybe_read_encoded_stream(req, encoding, compression)) +
[compression])
return tuple(to_return)
if _is_s3_url(filepath_or_buffer):
from pandas.io.s3 import get_filepath_or_buffer
return get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# It is a pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
return _expand_user(filepath_or_buffer), None, compression
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer
passthru otherwise.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
req = _urlopen(str(filepath_or_buffer))
if compression == 'infer':
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
compression = 'gzip'
else:
compression = None
# cat on the compression to the tuple returned by the function
to_return = (list(maybe_read_encoded_stream(req, encoding,
compression)) +
[compression])
return tuple(to_return)
if _is_s3_url(filepath_or_buffer):
from pandas.io.s3 import get_filepath_or_buffer
return get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# It is a pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
return _expand_user(filepath_or_buffer), None, compression
def get_filepath_or_buffer(filepath_or_buffer,
encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
url = str(filepath_or_buffer)
req = _urlopen(url)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
return reader, encoding, compression
if _is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# It is a pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
return _expand_user(filepath_or_buffer), None, compression
def get_filepath_or_buffer(filepath_or_buffer, encoding=None,
compression=None):
"""
If the filepath_or_buffer is a url, translate and return the buffer.
Otherwise passthrough.
Parameters
----------
filepath_or_buffer : a url, filepath (str, py.path.local or pathlib.Path),
or buffer
encoding : the encoding to use to decode py3 bytes, default is 'utf-8'
Returns
-------
a filepath_or_buffer, the encoding, the compression
"""
if _is_url(filepath_or_buffer):
url = str(filepath_or_buffer)
req = _urlopen(url)
content_encoding = req.headers.get('Content-Encoding', None)
if content_encoding == 'gzip':
# Override compression based on Content-Encoding header
compression = 'gzip'
reader = BytesIO(req.read())
return reader, encoding, compression
if _is_s3_url(filepath_or_buffer):
from pandas.io import s3
return s3.get_filepath_or_buffer(filepath_or_buffer,
encoding=encoding,
compression=compression)
# It is a pathlib.Path/py.path.local or string
filepath_or_buffer = _stringify_path(filepath_or_buffer)
return _expand_user(filepath_or_buffer), None, compression
def get_newest_version(timeout=5, _url=__GLCREATE_CURRENT_VERSION_URL__):
"""
Returns the version of GraphLab Create currently available from turi.com.
Will raise an exception if we are unable to reach the turi.com servers.
Parameters
----------
timeout: int
How many seconds to wait for the remote server to respond
url: string
The URL to go to to check the current version.
Returns
-------
version : str
The version number of the current graphlab create.
"""
request = _urlopen(url=_url, timeout=timeout)
version = request.read()
if version:
version = version.decode()
__LOGGER__.debug("current_version read %s" % version)
return version
def urlopen(*args, **kwargs):
return closing(_urlopen(*args, **kwargs))
def urlopen(f):
return closing(_urlopen(f))
def _get_ids(self):
""" Note structure here is a little different than later classes - metadata and IDs retrieved at same time. """
def get_meta(bill_content):
""" Get metadata search results based on a given URL. """
title = bill_content.BillTitle.find(name='Title',
language='en').text
if 'amend' in title.lower():
amend = True
else:
amend = False
# rarely, no published version of a bill is available
publication_tags = [
t for t in bill_content.find_all('Publication')
if t.find(name='Title', language='en').text == 'Royal Assent'
]
if len(publication_tags) == 1:
publication_id = publication_tags[0]['id']
else:
publication_id = None
# all other metadata appear to be consistently present
date = bill_content.Events.LastMajorStageEvent.Event['date']
session = bill_content.ParliamentSession['parliamentNumber']
subtype = bill_content.BillType.find(name='Title',
language='en').text
sponsor = bill_content.SponsorAffiliation.Person.FullName.text
sponsor_party = bill_content.SponsorAffiliation.PoliticalParty.find(
name='Title', language='en').text
majority_party = bill_content.PrimeMinister.PoliticalParty.find(
name='Title', language='en').text
committee_tags = bill_content.find_all(name='Committee',
accronym=True)
committee_names = [t['accronym'] for t in committee_tags]
committee_data = {
c: committee_names.count(c)
for c in set(committee_names)
}
metadata = _format_meta_entry(country=u'canada',
title=title,
id=publication_id,
date=date,
session=session,
type=u'annual',
subtype=subtype,
amendment=amend,
sponsor=sponsor,
sponsor_party=sponsor_party,
majority_party=majority_party,
hearings=committee_data)
return metadata
base_url = 'http://www.parl.gc.ca{0}'
bill_types = [
'/LegisInfo/Result.aspx?BillType=Senate%20Government%20Bill' +
'&BillStatus=RoyalAssentGiven&Language=E&Mode=1',
'/LegisInfo/Result.aspx?BillType=Private%20Member%E2%80%99s%20Bill'
+ '&BillStatus=RoyalAssentGiven&Language=E&Mode=1',
'/LegisInfo/Result.aspx?BillType=House%20Government%20Bill' +
'&BillStatus=RoyalAssentGiven&Language=E&Mode=1',
'/LegisInfo/Result.aspx?BillType=Senate%20Public%20Bill'
]
searches = []
for bill_type in bill_types:
search_content = _BeautifulSoup(
_urlopen(base_url.format(bill_type)))
sessions = [
_re.sub('&Page=1', '&download=xml', tag['href'])
for tag in search_content.find_all('a') if _re.search(
'[0-9]{2}-[0-9]\s*\([0-9]+\)', tag.text) is not None
]
searches += sessions
id_vals = []
for s in searches:
url = base_url.format(s)
content = _BeautifulSoup(_urlopen(url).read(), features='xml')
bills = content.find_all('Bill')
for bill in bills:
meta = get_meta(bill)
if meta['id'] not in self.log_data['Annual']['Canada']:
id_vals.append(meta['id'])
self.data[meta['id']] = meta
return id_vals
def _get_data(self, publication_id):
import bs4
search_term = _re.sub('_', '/', publication_id)
text_soup = None
text_content = None
try:
text_url = 'https://www.congress.gov/bill/{0}/text'.format(
search_term)
text_soup = _BeautifulSoup(_urlopen(text_url))
except:
pass
if text_soup is not None:
if text_soup.find('pre') is not None:
text_content = str(text_soup.find('pre'))
else:
text_content = str(
text_soup.find('table',
attrs={'class': 'lbexTableStyleEnr'}))
meta_url = 'https://www.congress.gov/bill/{0}/all-info'.format(
search_term)
meta_soup = _BeautifulSoup(_urlopen(meta_url))
title = _re.search(
': (.*)',
meta_soup.find('meta', attrs={'name': 'dc.title'})['content'])
if title is not None:
title = title.group(1)
date = meta_soup.find('meta', attrs={'name': 'dc.date'})['content']
sponsor = meta_soup.find('meta', attrs={'name': 'dc.creator'})
if sponsor is not None:
sponsor = sponsor['content']
sponsor_party = _re.search(sponsor + ' \[([A-Z])', meta_soup.text)
if sponsor_party is not None:
sponsor_party = sponsor_party.group(1)
else:
sponsor_party = None
cosponsors = [
tag.text for tag in meta_soup.find_all('a', href=True)
if 'member/' in tag['href'] and sponsor not in tag.text
]
policy_area = _re.search('Policy Area:\s*(.*)', meta_soup.text)
if policy_area is not None:
policy_area = policy_area.group(1)
committee_entries = meta_soup.find_all('tr', class_='committee')
referred = [entry.find('th').text for entry in committee_entries]
hearings_held = []
for entry in committee_entries:
committee_name = entry.find('th').text
actions = [entry.find_all('td')[1].text]
entry = entry.next_sibling
while type(entry) == bs4.element.Tag and (
'class' not in entry.attrs
or 'committee' not in entry['class']):
actions.append(entry.find_all('td')[1].text)
entry = entry.next_sibling
if type(entry) == bs4.element.NavigableString:
break
hearings = [action for action in actions if 'Hearing' in action]
hearings_held += [committee_name] * len(hearings)
if 'amend' in title:
amendment = True
else:
amendment = False
if 'resolution' in publication_id:
subtype = u'resolution'
else:
subtype = u'law'
meta = _format_meta_entry(country=u'united_states',
title=title,
id=publication_id,
date=date,
type=u'annual',
subtype=subtype,
amendment=amendment,
sponsor=sponsor,
sponsor_party=sponsor_party,
cosponsors=cosponsors,
referred=referred,
hearings=hearings_held,
policy_area=policy_area,
html=text_content)
return meta
def get_html(html_link):
html_response = _urlopen(html_link)
html_data = html_response.read()
return html_data
def retrieve_pdb_file(self,pdb_code, obsolete=0, compression=None,
uncompress=None, pdir=None):
""" Retrieves a PDB structure file from the PDB server and
stores it in a local file tree.
The PDB structure is returned as a single string.
If obsolete==1, the file will be saved in a special file tree.
If uncompress is specified, a system utility will decompress the .gz
archive. Otherwise, Python gzip utility will handle it.
compression does nothing, as all archives are already in .gz format
@param pdir: put the file in this directory (default: create a PDB-style directory tree)
@type pdir: string
@return: filename
@rtype: string
"""
# Alert the user about deprecated parameters
if compression is not None:
warnings.warn("PDB file servers now only host .gz archives: "
"the compression parameter will not do anything"
, BiopythonDeprecationWarning)
if uncompress is not None:
warnings.warn("Decompression is handled with the gzip module: "
"the uncompression parameter will not do anything"
, BiopythonDeprecationWarning)
# Get the structure
code=pdb_code.lower()
filename="pdb%s.ent.gz"%code
if not obsolete:
url=(self.pdb_server+
'/pub/pdb/data/structures/divided/pdb/%s/pdb%s.ent.gz'
% (code[1:3],code))
else:
url=(self.pdb_server+
'/pub/pdb/data/structures/obsolete/pdb/%s/pdb%s.ent.gz'
% (code[1:3],code))
# In which dir to put the pdb file?
if pdir is None:
if self.flat_tree:
if not obsolete:
path=self.local_pdb
else:
path=self.obsolete_pdb
else:
# Put in PDB-style directory tree
if not obsolete:
path=os.path.join(self.local_pdb, code[1:3])
else:
path=os.path.join(self.obsolete_pdb,code[1:3])
else:
# Put in specified directory
path=pdir
if not os.access(path,os.F_OK):
os.makedirs(path)
filename=os.path.join(path, filename)
# the final uncompressed file
final_file=os.path.join(path, "pdb%s.ent" % code)
# Skip download if the file already exists
if not self.overwrite:
if os.path.exists(final_file):
print "Structure exists: '%s' " % final_file
return final_file
# Retrieve the file
print "Downloading PDB structure '%s'..." % pdb_code
lines = _urlopen(url).read()
open(filename,'wb').write(lines)
# Uncompress the file
gz = gzip.open(filename, 'rb')
out = open(final_file, 'wb')
out.writelines(gz.read())
gz.close()
out.close()
os.remove(filename)
return final_file
def urlopen(*args, **kwargs):
with closing(_urlopen(*args, **kwargs)) as f:
yield f
def urlopen(id):
key = choice(API_KEY)
url = URL%(id.strip(), key)
return _urlopen(url, timeout=60).read()
def urlopen(*args, **kwargs):
with closing(_urlopen(*args, **kwargs)) as f:
yield f
def retrieve_pdb_file(self, pdb_code, obsolete=0, compression=None, uncompress=None, pdir=None):
""" Retrieves a PDB structure file from the PDB server and
stores it in a local file tree.
The PDB structure is returned as a single string.
If obsolete==1, the file will be saved in a special file tree.
If uncompress is specified, a system utility will decompress the .gz
archive. Otherwise, Python gzip utility will handle it.
compression does nothing, as all archives are already in .gz format
@param pdir: put the file in this directory (default: create a PDB-style directory tree)
@type pdir: string
@return: filename
@rtype: string
"""
# Alert the user about deprecated parameters
if compression is not None:
warnings.warn(
"PDB file servers now only host .gz archives: " "the compression parameter will not do anything",
BiopythonDeprecationWarning,
)
if uncompress is not None:
warnings.warn(
"Decompression is handled with the gzip module: " "the uncompression parameter will not do anything",
BiopythonDeprecationWarning,
)
# Get the structure
code = pdb_code.lower()
filename = "pdb%s.ent.gz" % code
if not obsolete:
url = self.pdb_server + "/pub/pdb/data/structures/divided/pdb/%s/pdb%s.ent.gz" % (code[1:3], code)
else:
url = self.pdb_server + "/pub/pdb/data/structures/obsolete/pdb/%s/pdb%s.ent.gz" % (code[1:3], code)
# In which dir to put the pdb file?
if pdir is None:
if self.flat_tree:
if not obsolete:
path = self.local_pdb
else:
path = self.obsolete_pdb
else:
# Put in PDB-style directory tree
if not obsolete:
path = os.path.join(self.local_pdb, code[1:3])
else:
path = os.path.join(self.obsolete_pdb, code[1:3])
else:
# Put in specified directory
path = pdir
if not os.access(path, os.F_OK):
os.makedirs(path)
filename = os.path.join(path, filename)
# the final uncompressed file
final_file = os.path.join(path, "pdb%s.ent" % code)
# Skip download if the file already exists
if not self.overwrite:
if os.path.exists(final_file):
print "Structure exists: '%s' " % final_file
return final_file
# Retrieve the file
print "Downloading PDB structure '%s'..." % pdb_code
lines = _urlopen(url).read()
open(filename, "wb").write(lines)
# Uncompress the file
gz = gzip.open(filename, "rb")
out = open(final_file, "wb")
out.writelines(gz.read())
gz.close()
out.close()
os.remove(filename)
return final_file
