Search is not available for this dataset
text
stringlengths
75
104k
def parse_table(table, flatten=True, footer=False): """Parses a table from sports-reference sites into a pandas dataframe. :param table: the PyQuery object representing the HTML table :param flatten: if True, flattens relative URLs to IDs. otherwise, leaves all fields as text without cleaning. :param footer: If True, returns the summary/footer of the page. Recommended to use this with flatten=False. Defaults to False. :returns: pd.DataFrame """ if not len(table): return pd.DataFrame() # get columns columns = [c.attrib['data-stat'] for c in table('thead tr:not([class]) th[data-stat]')] # get data rows = list(table('tbody tr' if not footer else 'tfoot tr') .not_('.thead, .stat_total, .stat_average').items()) data = [ [flatten_links(td) if flatten else td.text() for td in row.items('th,td')] for row in rows ] # make DataFrame df = pd.DataFrame(data, columns=columns, dtype='float') # add has_class columns allClasses = set( cls for row in rows if row.attr['class'] for cls in row.attr['class'].split() ) for cls in allClasses: df['has_class_' + cls] = [ bool(row.attr['class'] and cls in row.attr['class'].split()) for row in rows ] # cleaning the DataFrame df.drop(['ranker', 'Xxx', 'Yyy', 'Zzz'], axis=1, inplace=True, errors='ignore') # year_id -> year (as int) if 'year_id' in df.columns: df.rename(columns={'year_id': 'year'}, inplace=True) if flatten: df.year = df.year.fillna(method='ffill') df['year'] = df.year.map(lambda s: str(s)[:4]).astype(int) # pos -> position if 'pos' in df.columns: df.rename(columns={'pos': 'position'}, inplace=True) # boxscore_word, game_date -> boxscore_id and separate into Y, M, D columns for bs_id_col in ('boxscore_word', 'game_date', 'box_score_text'): if bs_id_col in df.columns: df.rename(columns={bs_id_col: 'boxscore_id'}, inplace=True) break # ignore *, +, and other characters used to note things df.replace(re.compile(r'[\*\+\u2605]', re.U), '', inplace=True) for col in df.columns: if hasattr(df[col], 'str'): df[col] = df[col].str.strip() # player -> player_id and/or player_name if 'player' in df.columns: if flatten: df.rename(columns={'player': 'player_id'}, inplace=True) # when flattening, keep a column for names player_names = parse_table(table, flatten=False)['player_name'] df['player_name'] = player_names else: df.rename(columns={'player': 'player_name'}, inplace=True) # team, team_name -> team_id for team_col in ('team', 'team_name'): if team_col in df.columns: # first, get rid of faulty rows df = df.loc[~df[team_col].isin(['XXX'])] if flatten: df.rename(columns={team_col: 'team_id'}, inplace=True) # season -> int if 'season' in df.columns and flatten: df['season'] = df['season'].astype(int) # handle date_game columns (different types) if 'date_game' in df.columns and flatten: date_re = r'month=(?P<month>\d+)&day=(?P<day>\d+)&year=(?P<year>\d+)' date_df = df['date_game'].str.extract(date_re, expand=True) if date_df.notnull().all(axis=1).any(): df = pd.concat((df, date_df), axis=1) else: df.rename(columns={'date_game': 'boxscore_id'}, inplace=True) # game_location -> is_home if 'game_location' in df.columns and flatten: df['game_location'] = df['game_location'].isnull() df.rename(columns={'game_location': 'is_home'}, inplace=True) # mp: (min:sec) -> float(min + sec / 60), notes -> NaN, new column if 'mp' in df.columns and df.dtypes['mp'] == object and flatten: mp_df = df['mp'].str.extract( r'(?P<m>\d+):(?P<s>\d+)', expand=True).astype(float) no_match = mp_df.isnull().all(axis=1) if no_match.any(): df.loc[no_match, 'note'] = df.loc[no_match, 'mp'] df['mp'] = mp_df['m'] + mp_df['s'] / 60 # converts number-y things to floats def convert_to_float(val): # percentages: (number%) -> float(number * 0.01) m = re.search(r'([-\.\d]+)\%', val if isinstance(val, basestring) else str(val), re.U) try: if m: return float(m.group(1)) / 100 if m else val if m: return int(m.group(1)) + int(m.group(2)) / 60 except ValueError: return val # salaries: $ABC,DEF,GHI -> float(ABCDEFGHI) m = re.search(r'\$[\d,]+', val if isinstance(val, basestring) else str(val), re.U) try: if m: return float(re.sub(r'\$|,', '', val)) except Exception: return val # generally try to coerce to float, unless it's an int or bool try: if isinstance(val, (int, bool)): return val else: return float(val) except Exception: return val if flatten: df = df.applymap(convert_to_float) df = df.loc[df.astype(bool).any(axis=1)] return df
def parse_info_table(table): """Parses an info table, like the "Game Info" table or the "Officials" table on the PFR Boxscore page. Keys are lower case and have spaces/special characters converted to underscores. :table: PyQuery object representing the HTML table. :returns: A dictionary representing the information. """ ret = {} for tr in list(table('tr').not_('.thead').items()): th, td = list(tr('th, td').items()) key = th.text().lower() key = re.sub(r'\W', '_', key) val = sportsref.utils.flatten_links(td) ret[key] = val return ret
def flatten_links(td, _recurse=False): """Flattens relative URLs within text of a table cell to IDs and returns the result. :td: the PyQuery object for the HTML to convert :returns: the string with the links flattened to IDs """ # helper function to flatten individual strings/links def _flatten_node(c): if isinstance(c, basestring): return c.strip() elif 'href' in c.attrib: c_id = rel_url_to_id(c.attrib['href']) return c_id if c_id else c.text_content().strip() else: return flatten_links(pq(c), _recurse=True) # if there's no text, just return None if td is None or not td.text(): return '' if _recurse else None td.remove('span.note') return ''.join(_flatten_node(c) for c in td.contents())
def rel_url_to_id(url): """Converts a relative URL to a unique ID. Here, 'ID' refers generally to the unique ID for a given 'type' that a given datum has. For example, 'BradTo00' is Tom Brady's player ID - this corresponds to his relative URL, '/players/B/BradTo00.htm'. Similarly, '201409070dal' refers to the boxscore of the SF @ DAL game on 09/07/14. Supported types: * player/... * boxscores/... * teams/... * years/... * leagues/... * awards/... * coaches/... * officials/... * schools/... * schools/high_schools.cgi?id=... :returns: ID associated with the given relative URL. """ yearRegex = r'.*/years/(\d{4}).*|.*/gamelog/(\d{4}).*' playerRegex = r'.*/players/(?:\w/)?(.+?)(?:/|\.html?)' boxscoresRegex = r'.*/boxscores/(.+?)\.html?' teamRegex = r'.*/teams/(\w{3})/.*' coachRegex = r'.*/coaches/(.+?)\.html?' stadiumRegex = r'.*/stadiums/(.+?)\.html?' refRegex = r'.*/officials/(.+?r)\.html?' collegeRegex = r'.*/schools/(\S+?)/.*|.*college=([^&]+)' hsRegex = r'.*/schools/high_schools\.cgi\?id=([^\&]{8})' bsDateRegex = r'.*/boxscores/index\.f?cgi\?(month=\d+&day=\d+&year=\d+)' leagueRegex = r'.*/leagues/(.*_\d{4}).*' awardRegex = r'.*/awards/(.+)\.htm' regexes = [ yearRegex, playerRegex, boxscoresRegex, teamRegex, coachRegex, stadiumRegex, refRegex, collegeRegex, hsRegex, bsDateRegex, leagueRegex, awardRegex, ] for regex in regexes: match = re.match(regex, url, re.I) if match: return [_f for _f in match.groups() if _f][0] # things we don't want to match but don't want to print a WARNING if any( url.startswith(s) for s in ( '/play-index/', ) ): return url print('WARNING. NO MATCH WAS FOUND FOR "{}"'.format(url)) return url
def PlayerSeasonFinder(**kwargs): """ Docstring will be filled in by __init__.py """ if 'offset' not in kwargs: kwargs['offset'] = 0 playerSeasons = [] while True: querystring = _kwargs_to_qs(**kwargs) url = '{}?{}'.format(PSF_URL, querystring) if kwargs.get('verbose', False): print(url) html = utils.get_html(url) doc = pq(html) table = doc('table#results') df = utils.parse_table(table) if df.empty: break thisSeason = list(zip(df.player_id, df.year)) playerSeasons.extend(thisSeason) if doc('*:contains("Next Page")'): kwargs['offset'] += 100 else: break return playerSeasons
def _kwargs_to_qs(**kwargs): """Converts kwargs given to PSF to a querystring. :returns: the querystring. """ # start with defaults inpOptDef = inputs_options_defaults() opts = { name: dct['value'] for name, dct in inpOptDef.items() } # clean up keys and values for k, v in kwargs.items(): del kwargs[k] # bool => 'Y'|'N' if isinstance(v, bool): kwargs[k] = 'Y' if v else 'N' # tm, team => team_id elif k.lower() in ('tm', 'team'): kwargs['team_id'] = v # yr, year, yrs, years => year_min, year_max elif k.lower() in ('yr', 'year', 'yrs', 'years'): if isinstance(v, collections.Iterable): lst = list(v) kwargs['year_min'] = min(lst) kwargs['year_max'] = max(lst) elif isinstance(v, basestring): v = list(map(int, v.split(','))) kwargs['year_min'] = min(v) kwargs['year_max'] = max(v) else: kwargs['year_min'] = v kwargs['year_max'] = v # pos, position, positions => pos[] elif k.lower() in ('pos', 'position', 'positions'): if isinstance(v, basestring): v = v.split(',') elif not isinstance(v, collections.Iterable): v = [v] kwargs['pos[]'] = v # draft_pos, ... => draft_pos[] elif k.lower() in ( 'draft_pos', 'draftpos', 'draftposition', 'draftpositions', 'draft_position', 'draft_positions' ): if isinstance(v, basestring): v = v.split(',') elif not isinstance(v, collections.Iterable): v = [v] kwargs['draft_pos[]'] = v # if not one of these cases, put it back in kwargs else: kwargs[k] = v # update based on kwargs for k, v in kwargs.items(): # if overwriting a default, overwrite it (with a list so the # opts -> querystring list comp works) if k in opts or k in ('pos[]', 'draft_pos[]'): # if multiple values separated by commas, split em if isinstance(v, basestring): v = v.split(',') # otherwise, make sure it's a list elif not isinstance(v, collections.Iterable): v = [v] # then, add list of values to the querystring dict *opts* opts[k] = v if 'draft' in k: opts['draft'] = [1] opts['request'] = [1] opts['offset'] = [kwargs.get('offset', 0)] qs = '&'.join( '{}={}'.format(urllib.parse.quote_plus(name), val) for name, vals in sorted(opts.items()) for val in vals ) return qs
def _Streamer__read_process(self, path, read_size, cbuf, stop, barrier, cyclic, offset, read_skip, sync): """ Main function for the processes that read from the HDF5 file. :param self: A reference to the streamer object that created these processes. :param path: The HDF5 path to the node to be read from. :param read_size: The length of the block along the outer dimension to read. :param cbuf: The circular buffer to place read elements into. :param stop: The Event that signals the process to stop reading. :param barrier: The Barrier that synchonises read cycles. :param cyclic: True if the process should read cyclically. :param offset: Offset into the dataset that this process should start reading at. :param read_skip: How many element to skip on each iteration. :param sync: GuardSynchonizer to order writes to the buffer. :return: Nothing """ # Multi-process access to HDF5 seems to behave better there are no top level imports of PyTables. import tables as tb h5_file = tb.open_file(self.filename, 'r', **self.h5_kw_args) ary = h5_file.get_node(path) i = offset while not stop.is_set(): vals = ary[i:i + read_size] # If the read goes off the end of the dataset, then wrap to the start. if i + read_size > len(ary): vals = np.concatenate([vals, ary[0:read_size - len(vals)]]) if sync is None: # If no ordering is requested, then just write to the next available space in the buffer. with cbuf.put_direct() as put_ary: put_ary[:] = vals else: # Otherwise, use the sync object to ensure that writes occur in the order provided by i. # So i = 0 will write first, then i = block_size, then i = 2*block_size, etc... # The sync object has two ordered barriers so that acquisition and release of the buffer spaces # are synchronized in order, but the actual writing to the buffer can happen simultaneously. # If only one barrier were used, writing to the buffer would be linearised. with sync.do(cbuf.put_direct(), i, (i+read_size) % len(ary)) as put_ary: put_ary[:] = vals i += read_skip if cyclic: # If the next iteration is past the end of the dataset, wrap it around. if i >= len(ary): i %= len(ary) barrier.wait() else: # But if cyclic mode is disabled, break the loop as the work is now done. if i + read_size > len(ary): break
def wait(self): """Wait until all processes have reached the barrier.""" with self.cvar: self.count.value += 1 self.cvar.notify_all() while self.count.value < self.n_procs: self.cvar.wait()
def wait(self): """Wait until all processes have reached the barrier.""" self.barrier_A.wait() # The current barrier (barrier_A) is switched with the reserve barrier. # This is because the current barrier cannot be safely reset until the reserve barrier has been passed. self.barrier_A, self.barrier_B = self.barrier_B, self.barrier_A self.barrier_A.reset()
def wait(self, index, next_index=None): """ Block until it is the turn indicated by index. :param index: :param next_index: Set the index to this value after finishing. Releases the process waiting on next_index. Defaults to incrementing index by 1. :return: """ return OrderedBarrier.Guard(self, index, index+1 if next_index is None else next_index)
def do(self, guard, index, next_index): """ Create a guard that requires the resource guard to be entered and exited based on the order provided by index. :param guard: The context manager for the resource. :param index: The order to wait for. :param next_index: The next index to release. :return: """ return GuardSynchronizer.Guard(self, guard, index, next_index)
def put(self, v): """ Put an unsigned integer into the queue. This method always assumes that there is space in the queue. ( In the circular buffer, this is guaranteed by the implementation ) :param v: The item to insert. Must be >= 0, as -2 is used to signal a queue close. :return: """ if v is QueueClosed: v = -2 else: assert(v >= 0) with self.cvar: assert(self.size.value < len(self.vals)) head = (self.tail.value + self.size.value) % len(self.vals) self.vals[head] = v self.size.value += 1 self.cvar.notify()
def get(self): """ Fetch the next item in the queue. Blocks until an item is ready. :return: The next unsigned integer in the queue. """ with self.cvar: while True: if self.size.value > 0: rval = self.vals[self.tail.value] self.tail.value = (self.tail.value + 1) % len(self.vals) self.size.value -= 1 if rval == -2: return QueueClosed assert(rval >= 0) return rval self.cvar.wait()
def put_direct(self): """ Allows direct access to the buffer element. Blocks until there is room to write into the buffer. :return: A guard object that returns the buffer element. """ # Once the guard is released, write_idx will be placed into read_queue. return self.Guard(self.read_queue, self.arys, self.__put_idx)
def get_direct(self): """ Allows direct access to the buffer element. Blocks until there is data that can be read. :return: A guard object that returns the buffer element. """ read_idx = self.__get_idx() if read_idx is QueueClosed: return QueueClosed # Once the guard is released, read_idx will be placed into write_queue. return self.Guard(self.write_queue, self.arys, lambda: read_idx)
def close(self): """Close the queue, signalling that no more data can be put into the queue.""" self.read_queue.put(QueueClosed) self.write_queue.put(QueueClosed)
def __get_batch(self, path, length, last=False): """ Get a block of data from the node at path. :param path: The path to the node to read from. :param length: The length along the outer dimension to read. :param last: True if the remainder elements should be read. :return: A copy of the requested block of data as a numpy array. """ import tables h5_file = tables.open_file(self.filename, 'r') h5_node = h5_file.get_node(path) if len(h5_node) == 0: raise Exception("Cannot read from empty dataset.") # If the length isn't specified, then fall back to default values. if length is None: chunkshape = h5_node.chunkshape # If the array isn't chunked, then try to make the block close to 128KB. if chunkshape is None: default_length = 128*2**10//h5_node[0].nbytes # Divides by one row of the dataset. length = min(h5_node.shape[0], default_length) # If it is chunked, then use the chunkshape for best performance. else: length = chunkshape[0] if last: example = h5_node[length*(len(h5_node)//length):].copy() else: example = h5_node[:length].copy() h5_file.close() return example
def get_remainder(self, path, block_size): """ Get the remainder elements. These elements will not be read in the direct queue access cyclic=False mode. :param path: The HDF5 path to the dataset to be read. :param block_size: The block size is used to calculate which elements will remain. :return: A copy of the remainder elements as a numpy array. """ return self.__get_batch(path, length=block_size, last=True)
def get_queue(self, path, n_procs=4, read_ahead=None, cyclic=False, block_size=None, ordered=False): """ Get a queue that allows direct access to the internal buffer. If the dataset to be read is chunked, the block_size should be a multiple of the chunk size to maximise performance. In this case it is best to leave it to the default. When cyclic=False, and block_size does not divide the dataset evenly, the remainder elements will not be returned by the queue. When cyclic=True, the remainder elements will be part of a block that wraps around the end and includes element from the beginning of the dataset. By default, blocks are returned in the order in which they become available. The ordered option will force blocks to be returned in on-disk order. :param path: The HDF5 path to the dataset that should be read. :param n_procs: The number of background processes used to read the datset in parallel. :param read_ahead: The number of blocks to allocate in the internal buffer. :param cyclic: True if the queue should wrap at the end of the dataset. :param block_size: The size along the outer dimension of the blocks to be read. Defaults to a multiple of the chunk size, or to a 128KB sized block if the dataset is not chunked. :param ordered: Force the reader return data in on-disk order. May result in performance penalty. :return: A queue object that allows access to the internal buffer. """ # Get a block_size length of elements from the dataset to serve as a template for creating the buffer. # If block_size=None, then get_batch calculates an appropriate block size. example = self.__get_batch(path, block_size) block_size = example.shape[0] if read_ahead is None: # 2x No. of processes for writing, 1 extra for reading. read_ahead = 2*n_procs + 1 cbuf = SharedCircBuf(read_ahead, example) stop = multiprocessing.Event() barrier = Barrier(n_procs) # If ordering has been requested, create a synchronizer. sync = GuardSynchronizer() if ordered else None procs = [] for i in range(n_procs): # Each process is offset in the dataset by i*block_size # The skip length is set to n_procs*block_size so that no block is read by 2 processes. process = multiprocessing.Process(target=_Streamer__read_process, args=( self, path, block_size, cbuf, stop, barrier, cyclic, i * block_size, n_procs * block_size, sync )) process.daemon = True process.start() procs.append(process) # If the queue is not cyclic, then the cessation of reading data needs to be monitored. if not cyclic: # This closure defines a background thread that waits until all processes have finished. # At this point, all data from the dataset has been read, and the buffer is closed. def monitor(): for p in procs: p.join() cbuf.close() monitor_thread = threading.Thread(target=monitor) monitor_thread.daemon = True monitor_thread.start() return Streamer.Queue(cbuf, stop, block_size)
def get_generator(self, path, *args, **kw_args): """ Get a generator that allows convenient access to the streamed data. Elements from the dataset are returned from the generator one row at a time. Unlike the direct access queue, this generator also returns the remainder elements. Additional arguments are forwarded to get_queue. See the get_queue method for documentation of these parameters. :param path: :return: A generator that iterates over the rows in the dataset. """ q = self.get_queue(path=path, *args, **kw_args) try: # This generator just implements a standard access pattern for the direct access queue. for guard in q.iter(): with guard as batch: batch_copy = batch.copy() for row in batch_copy: yield row last_batch = self.get_remainder(path, q.block_size) for row in last_batch: yield row finally: q.close()
def parse(ifp, pb_cls, **kwargs): """Parse a stream. Args: ifp (string or file-like object): input stream. pb_cls (protobuf.message.Message.__class__): The class object of the protobuf message type encoded in the stream. """ mode = 'rb' if isinstance(ifp, str): istream = open(ifp, mode=mode, **kwargs) else: istream = open(fileobj=ifp, mode=mode, **kwargs) with istream: for data in istream: pb_obj = pb_cls() pb_obj.ParseFromString(data) yield pb_obj
def dump(ofp, *pb_objs, **kwargs): """Write to a stream. Args: ofp (string or file-like object): output stream. pb_objs (*protobuf.message.Message): list of protobuf message objects to be written. """ mode = 'wb' if isinstance(ofp, str): ostream = open(ofp, mode=mode, **kwargs) else: ostream = open(fileobj=ofp, mode=mode, **kwargs) with ostream: ostream.write(*pb_objs)
def _read_varint(self): """Read a varint from file, parse it, and return the decoded integer. """ buff = self._fd.read(1) if buff == b'': return 0 while (bytearray(buff)[-1] & 0x80) >> 7 == 1: # while the MSB is 1 new_byte = self._fd.read(1) if new_byte == b'': raise EOFError('unexpected EOF.') buff += new_byte varint, _ = decodeVarint(buff, 0) return varint
def _get_objs(self): """A generator yielding all protobuf object data in the file. It is the main parser of the stream encoding. """ while True: count = self._read_varint() if count == 0: break # Read a group containing `count` number of objects. for _ in range(count): size = self._read_varint() if size == 0: raise EOFError('unexpected EOF.') # Read an object from the object group. yield self._fd.read(size) if self._group_delim: yield self._delimiter() if self._delimiter is not None \ else None
def close(self): """Close the stream.""" self.flush() if self._myfd is not None: self._myfd.close() self._myfd = None
def write(self, *pb2_obj): """Write a group of one or more protobuf objects to the file. Multiple object groups can be written by calling this method several times before closing stream or exiting the runtime context. The input protobuf objects get buffered and will be written down when the number of buffered objects exceed the `self._buffer_size`. Args: pb2_obj (*protobuf.message.Message): list of protobuf messages. """ base = len(self._write_buff) for idx, obj in enumerate(pb2_obj): if self._buffer_size > 0 and \ (idx + base) != 0 and \ (idx + base) % self._buffer_size == 0: self.flush() self._write_buff.append(obj) if self._buffer_size == 0: self.flush()
def flush(self): """Write down buffer to the file.""" if not self.is_output(): return count = len(self._write_buff) if count == 0: return encodeVarint(self._fd.write, count, True) for obj in self._write_buff: obj_str = obj.SerializeToString() encodeVarint(self._fd.write, len(obj_str), True) self._fd.write(obj_str) self._write_buff = []
def get_game_dir(self, username=False): """Returns joined game directory path relative to Steamapps""" if not self.common and not username: raise RuntimeError("Can't determine this game's directory without username") if self.common: subdir = "common" else: subdir = "username" subsubdir = self.dir if WIN32 or CYGWIN: subsubdir = subsubdir.lower() return os.path.join(subdir, subsubdir)
def _get_MAP_spikes(F, c_hat, theta, dt, tol=1E-6, maxiter=100, verbosity=0): """ Used internally by deconvolve to compute the maximum a posteriori spike train for a given set of fluorescence traces and model parameters. See the documentation for deconvolve for the meaning of the arguments Returns: n_hat_best, c_hat_best, LL_best """ npix, nt = F.shape sigma, alpha, beta, lamb, gamma = theta # we project everything onto the alpha mask so that we only ever have to # deal with 1D vector norms alpha_ss = np.dot(alpha, alpha) c = np.dot(alpha, F) - np.dot(alpha, beta) # used for computing the LL and gradient scale_var = 1. / (2 * sigma * sigma) lD = lamb * dt # used for computing the gradient (M.T.dot(lamb * dt)) grad_lnprior = np.zeros(nt, dtype=DTYPE) grad_lnprior[1:] = lD grad_lnprior[:-1] -= lD * gamma # initialize the weight of the barrier term to 1 z = 1. # initial estimate of spike probabilities n_hat = c_hat[1:] - gamma * c_hat[:-1] # assert not np.any(n_hat < 0), "spike probabilities < 0" # (actual - predicted) fluorescence res = c - alpha_ss * c_hat # best overall posterior log-likelihood of the fluorescence LL_best = _post_LL(n_hat, res, scale_var, lD, z) LL_barrier = LL_best nloop1 = 0 terminate_interior = False # in the outer loop we'll progressively reduce the weight of the barrier # term and check the interior point termination criteria while not terminate_interior: nloop2 = 0 terminate_barrier = False # converge for this barrier weight while not terminate_barrier: # by projecting everything onto alpha, we reduce this to a 1D # vector norm res = c - alpha_ss * c_hat # compute direction of newton step d = _direction(n_hat, res, alpha_ss, gamma, scale_var, grad_lnprior, z) terminate_linesearch = False # find the largest step we can take in direction d without # violating the non-negativity constraint on n_hat s_upper_bnd = -n_hat / (d[1:] - gamma * d[:-1]) # we are only interested in positive step sizes feasible = (s_upper_bnd > 0) if np.any(feasible): # largest allowable step size is 1. s = min(1., 0.999 * np.min(s_upper_bnd[feasible])) else: # if there is no step size that will keep n_hat >= 0, just # reduce the barrier weight and try again terminate_linesearch = True terminate_barrier = True if verbosity >= 2: print("skipping linesearch: no positive step size will " "keep n_hat >= 0") nloop3 = 0 # backtracking line search for the largest step size that increases # the posterior log-likelihood of the fluorescence while not terminate_linesearch: # update estimated calcium c_hat_line = c_hat + (s * d) # update spike probabilities n_hat_line = c_hat_line[1:] - gamma * c_hat_line[:-1] # assert not np.any(n_hat_line < 0), "spike probabilities < 0" # (actual - predicted) fluorescence res = c - alpha_ss * c_hat_line # compute the new posterior log-likelihood LL_line = _post_LL(n_hat_line, res, scale_var, lD, z) # assert not np.any(np.isnan(LL1)), "nan LL" if verbosity >= 2: print('spikes: iter=%3i, %3i, %3i; z=%-10.4g; s=%-10.4g;' ' LL=%-10.4g' % (nloop1, nloop2, nloop3, z, s, LL_line)) # if the step size gets too small without making any progress, # we terminate the linesearch and reduce the barrier weight if s < S_TOL: if verbosity >= 2: print('--> terminated linesearch: s < %.3g on %i ' 'iterations' % (S_TOL, nloop3)) terminate_linesearch = True terminate_barrier = True # only update c_hat & LL if LL improved if LL_line > LL_barrier: LL_barrier, n_hat, c_hat = LL_line, n_hat_line, c_hat_line terminate_linesearch = True # reduce the step size else: s /= S_FAC nloop3 += 1 # if d gets too small, reduce the barrier weight if (np.linalg.norm(d) < D_TOL): terminate_barrier = True nloop2 += 1 # only test for convergence if we were actually able to enter the # linesearch if nloop3: delta_LL = -(LL_barrier - LL_best) / LL_best LL_best = LL_barrier if (delta_LL < tol): terminate_interior = True elif z < Z_TOL: if verbosity >= 2: print('MAP spike train failed to converge before z < %.3g' % Z_TOL) terminate_interior = True elif nloop1 > maxiter: if verbosity >= 2: print('MAP spike train failed to converge within maxiter (%i)' % maxiter) terminate_interior = True # increment the outer loop counter, reduce the barrier weight nloop1 += 1 z /= Z_FAC return n_hat, c_hat, LL_best
def trisolve(dl, d, du, b, inplace=False): """ The tridiagonal matrix (Thomas) algorithm for solving tridiagonal systems of equations: a_{i}x_{i-1} + b_{i}x_{i} + c_{i}x_{i+1} = y_{i} in matrix form: Mx = b TDMA is O(n), whereas standard Gaussian elimination is O(n^3). Arguments: ----------- dl: (n - 1,) vector the lower diagonal of M d: (n,) vector the main diagonal of M du: (n - 1,) vector the upper diagonal of M b: (n,) vector the result of Mx inplace: if True, and if d and b are both float64 vectors, they will be modified in place (may be faster) Returns: ----------- x: (n,) vector the solution to Mx = b References: ----------- http://en.wikipedia.org/wiki/Tridiagonal_matrix_algorithm http://www.netlib.org/lapack/explore-html/d1/db3/dgtsv_8f.html """ if (dl.shape[0] != du.shape[0] or (d.shape[0] != dl.shape[0] + 1) or d.shape[0] != b.shape[0]): raise ValueError('Invalid diagonal shapes') bshape_in = b.shape rtype = np.result_type(dl, d, du, b) if not inplace: # force a copy dl = np.array(dl, dtype=rtype, copy=True, order='F') d = np.array(d, dtype=rtype, copy=True, order='F') du = np.array(du, dtype=rtype, copy=True, order='F') b = np.array(b, dtype=rtype, copy=True, order='F') # this may also force copies if arrays have inconsistent types / incorrect # order dl, d, du, b = (np.array(v, dtype=rtype, copy=False, order='F') for v in (dl, d, du, b)) # use the LAPACK implementation _lapack_trisolve(dl, d, du, b, rtype) return b.reshape(bshape_in)
def from_web_element(self, web_element): """ Store reference to a WebElement instance representing the element on the DOM. Use it when an instance of WebElement has already been created (e.g. as the result of find_element) and you want to create a UIComponent out of it without evaluating it from the locator again. Returns an instance of the class. """ if isinstance(web_element, WebElement) is not True: raise TypeError("web_element parameter is not of type WebElement.") self._web_element = web_element return self
def locate(self): """ Lazily locates the element on the DOM if the WebElement instance is not available already. Returns a WebElement object. It also caches the element if caching has been set through cache(). """ if self._web_element: return self._web_element else: locator_type, locator_value = self.__locator element = self.driver.find_element(by=locator_type, value=locator_value) self._cache_web_element(element) # cache the element if allowed return element
def input_text_with_keyboard_emulation(self, text): """ Works around the problem of emulating user interactions with text inputs. Emulates a key-down action on the first char of the input. This way, implementations which require key-down event to trigger auto-suggest are testable. Then the chains sends the rest of the text and releases the key. """ ActionChains(self.driver).key_down(text).key_up(Keys.CONTROL).perform()
def make_fake_movie(nframes, mask_shape=(64, 64), mask_center=None, bg_intensity=0.1, mask_sigma=10, dt=0.02, rate=1.0, tau=1., sigma=0.001, seed=None): """ Generate 2D fake fluorescence movie Arguments: --------------------------------------------------------------------------- nframes: number of timebins to simulate mask_shape: tuple (nrows, ncols), shape of a single movie frame mask_center: tuple (x, y), pixel coords of cell center bg_intensity: scalar, amplitude of (static) baseline fluorescence mask_sigma: scalar, standard deviation of Gaussian mask dt: timestep (s) rate: mean spike rate (Hz) tau: time constant of decay in calcium concentration (s) sigma: SD of additive noise on fluorescence seed: Seed for RNG Returns: --------------------------------------------------------------------------- F: fluorescence [npixels, nframes] c: calcium concentration [nframes,] n: spike train [nframes,] theta: tuple of true model parameters: (sigma, alpha, beta, lambda, gamma) """ gen = np.random.RandomState(seed) # poisson spikes n = gen.poisson(rate * dt, size=nframes) # internal calcium dynamics gamma = np.exp(-dt / tau) c = signal.lfilter(np.r_[1], np.r_[1, -gamma], n, axis=0) # pixel weights (sum == 1) nr, nc = mask_shape npix = nr * nc if mask_center is None: mask_center = (nc // 2., nr // 2.) a, b = mask_center y, x = np.ogrid[:nr, :nc] xs = (x - a) ** 2. ys = (y - b) ** 2. twoss = 2. * mask_sigma ** 2. alpha = np.exp(-1 * ((xs / twoss) + (ys / twoss))).ravel() alpha /= alpha.sum() # background fluorescence beta = gen.randn(npix) * bg_intensity # firing rate (spike probability per sec) lamb = rate # spatially & temporally white noise epsilon = gen.randn(npix, nframes) * sigma # simulated fluorescence F = c[None, :] * alpha[:, None] + beta[:, None] + epsilon theta = (sigma, alpha, beta, lamb, gamma) return F, c, n, theta
def evaluate_traits(self): """ Evaluates traits and returns a list containing the description of traits which are not true. Notice that if LAZY_EVALUATION is set to False all traits are evaluated before returning. Use this option only for debugging purposes. """ return_value = [] for trait in self.traits: if not trait.condition(): if not self.traits_eager_evaluation: return [trait.description] else: return_value.append(trait.description) return return_value
def until_condition(self, condition, condition_description): """ Waits until conditions is True or returns a non-None value. If any of the trait is still not present after timeout, raises a TimeoutException. """ end_time = time.time() + self._timeout count = 1 while True: try: if not hasattr(condition, '__call__'): raise TypeError("condition is not callable") value = condition() if type(value) is bool and value is not False: return value elif type(value) is not bool and value is not None: return value else: logger.debug("#" + str(count) + " - wait until " + condition_description) # pragma: no cover except self._ignored_exceptions as ex: logger.debug("Captured {0} : {1}".format(str(ex.__class__).replace("<type '", "").replace("'>", ""), str(ex))) # pragma: no cover time.sleep(self._poll) count += 1 if time.time() > end_time: # pragma: no cover break raise TimeoutException( msg="condition <" + condition_description + "> was not true after " + str(self._timeout) + " seconds.")
def until_traits_are_present(self, element_with_traits): """ Waits until all traits are present. If any of the traits is still not present after timeout, raises a TimeoutException. """ end_time = time.time() + self._timeout count = 1 missing_traits_descriptions = None while True: missing_traits_descriptions = [] try: missing_traits_descriptions = element_with_traits.evaluate_traits() if len(missing_traits_descriptions) == 0: return True else: logger.debug("#{0} - wait until all traits are present: <{1}>".format(str(count), '> <'.join( missing_traits_descriptions))) except self._ignored_exceptions as ex: # pragma: no cover logger.debug("Captured {0}: {1}".format(str(ex.__class__).replace("<type '", "").replace("'>", ""), str(ex))) # pragma: no cover pass # pragma: no cover time.sleep(self._poll) count += 1 if time.time() > end_time: break raise TimeoutException( msg="conditions " + '<' + '> <'.join(missing_traits_descriptions) + '>' + " not true after " + str( self._timeout) + " seconds.")
def with_ignored_exceptions(self, *ignored_exceptions): """ Set a list of exceptions that should be ignored inside the wait loop. """ for exception in ignored_exceptions: self._ignored_exceptions = self._ignored_exceptions + (exception,) return self
def s2h(ss): """convert seconds to a pretty "d hh:mm:ss.s" format""" mm, ss = divmod(ss, 60) hh, mm = divmod(mm, 60) dd, hh = divmod(hh, 24) tstr = "%02i:%04.1f" % (mm, ss) if hh > 0: tstr = ("%02i:" % hh) + tstr if dd > 0: tstr = ("%id " % dd) + tstr return tstr
def exec_command(self, domain, function, operator, value=None): """ Write a command to the receiver and read the value it returns. The receiver will always return a value, also when setting a value. """ if operator in CMDS[domain][function]['supported_operators']: if operator is '=' and value is None: raise ValueError('No value provided') if value is None: cmd = ''.join([CMDS[domain][function]['cmd'], operator]) else: cmd = ''.join( [CMDS[domain][function]['cmd'], operator, str(value)]) else: raise ValueError('Invalid operator provided %s' % operator) if not self.ser.is_open: self.ser.open() try: self.lock.acquire() self.ser.write(''.join(['\r', cmd, '\r']).encode('utf-8')) time.sleep(0.1) # not sure why, but otherwise it is not ready yet to do the read. msg = self.ser.read(self.ser.in_waiting) try: msg = msg.decode()[1:-1] msg = msg.split('=')[1] return msg except IndexError: pass finally: self.lock.release()
def main_volume(self, operator, value=None): """ Execute Main.Volume. Returns int """ try: res = int(self.exec_command('main', 'volume', operator, value)) return res except (ValueError, TypeError): pass return None
def main_source(self, operator, value=None): """ Execute Main.Source. Returns int """ try: source = int(self.exec_command('main', 'source', operator, value)) return source except (ValueError, TypeError): pass return None
def _send(self, message, read_reply=False): """Send a command string to the amplifier.""" sock = None for tries in range(0, 3): try: sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM) sock.connect((self._host, self.PORT)) break except (ConnectionError, BrokenPipeError): if tries == 3: print("socket connect failed.") return sleep(0.1) sock.send(codecs.decode(message, 'hex_codec')) if read_reply: sleep(0.1) reply = '' tries = 0 max_tries = 20 while len(reply) < len(message) and tries < max_tries: try: reply += codecs.encode(sock.recv(self.BUFFERSIZE), 'hex')\ .decode("utf-8") except (ConnectionError, BrokenPipeError): pass tries += 1 sock.close() if tries >= max_tries: return return reply sock.close()
def status(self): """ Return the status of the device. Returns a dictionary with keys 'volume' (int 0-200) , 'power' (bool), 'muted' (bool) and 'source' (str). """ nad_reply = self._send(self.POLL_VOLUME + self.POLL_POWER + self.POLL_MUTED + self.POLL_SOURCE, read_reply=True) if nad_reply is None: return # split reply into parts of 10 characters num_chars = 10 nad_status = [nad_reply[i:i + num_chars] for i in range(0, len(nad_reply), num_chars)] return {'volume': int(nad_status[0][-2:], 16), 'power': nad_status[1][-2:] == '01', 'muted': nad_status[2][-2:] == '01', 'source': self.SOURCES_REVERSED[nad_status[3][-2:]]}
def power_off(self): """Power the device off.""" status = self.status() if status['power']: # Setting power off when it is already off can cause hangs self._send(self.CMD_POWERSAVE + self.CMD_OFF)
def power_on(self): """Power the device on.""" status = self.status() if not status['power']: self._send(self.CMD_ON, read_reply=True) sleep(0.5)
def set_volume(self, volume): """Set volume level of the device. Accepts integer values 0-200.""" if 0 <= volume <= 200: volume = format(volume, "02x") # Convert to hex self._send(self.CMD_VOLUME + volume)
def select_source(self, source): """Select a source from the list of sources.""" status = self.status() if status['power']: # Changing source when off may hang NAD7050 if status['source'] != source: # Setting the source to the current source will hang the NAD7050 if source in self.SOURCES: self._send(self.CMD_SOURCE + self.SOURCES[source], read_reply=True)
def exec_command(self, domain, function, operator, value=None): """ Write a command to the receiver and read the value it returns. """ if operator in CMDS[domain][function]['supported_operators']: if operator is '=' and value is None: raise ValueError('No value provided') if value is None: cmd = ''.join([CMDS[domain][function]['cmd'], operator]) else: cmd = ''.join( [CMDS[domain][function]['cmd'], operator, str(value)]) else: raise ValueError('Invalid operator provided %s' % operator) if self._open_connection(): # For telnet the first \r / \n is recommended only self.telnet.write((''.join(['\r', cmd, '\n']).encode())) # Could raise eg. socket.error, UnicodeError, let the client handle it # Test 3 x buffer is completely empty # With the default timeout that means a delay at # about 3+ seconds loop = 3 while loop: msg = self.telnet.read_until('\n'.encode(), self.timeout) # Could raise eg. EOFError, UnicodeError, let the client handle it if msg == "": # Nothing in buffer loop -= 1 continue msg = msg.decode().strip('\r\n') # Could raise eg. UnicodeError, let the client handle it #print("NAD reponded with '%s'" % msg) # Wait for the response that equals the requested domain.function if msg.strip().split('=')[0].lower() == '.'.join([domain, function]).lower(): # b'Main.Volume=-12\r will return -12 return msg.strip().split('=')[1] raise RuntimeError('Failed to read response') raise RuntimeError('Failed to open connection')
def deobfuscate(request, key, juice=None): """ Deobfuscates the URL and returns HttpResponse from source view. SEO juice is mostly ignored as it is intended for display purposes only. """ try: url = decrypt(str(key), settings.UNFRIENDLY_SECRET, settings.UNFRIENDLY_IV, checksum=settings.UNFRIENDLY_ENFORCE_CHECKSUM) except (CheckSumError, InvalidKeyError): return HttpResponseNotFound() try: url = url.decode('utf-8') except UnicodeDecodeError: return HttpResponseNotFound() url_parts = urlparse(unquote(url)) path = url_parts.path query = url_parts.query try: view, args, kwargs = resolve(path) except Resolver404: return HttpResponseNotFound() # fix-up the environ object environ = request.environ.copy() environ['PATH_INFO'] = path[len(environ['SCRIPT_NAME']):] environ['QUERY_STRING'] = query # init a new request patched_request = request.__class__(environ) # copy over any missing request attributes - this feels hackish missing_items = set(dir(request)) - set(dir(patched_request)) while missing_items: missing_item = missing_items.pop() patched_request.__setattr__(missing_item, request.__getattribute__(missing_item)) # mark this request as obfuscated patched_request.META['obfuscated'] = True response = view(patched_request, *args, **kwargs) # offer up a friendlier juice-powered filename if downloaded if juice and not response.has_header('Content-Disposition'): response['Content-Disposition'] = 'inline; filename=%s' % juice return response
def _lazysecret(secret, blocksize=32, padding='}'): """Pads secret if not legal AES block size (16, 24, 32)""" if not len(secret) in (16, 24, 32): return secret + (blocksize - len(secret)) * padding return secret
def _crc(plaintext): """Generates crc32. Modulo keep the value within int range.""" if not isinstance(plaintext, six.binary_type): plaintext = six.b(plaintext) return (zlib.crc32(plaintext) % 2147483647) & 0xffffffff
def encrypt(plaintext, secret, inital_vector, checksum=True, lazy=True): """Encrypts plaintext with secret plaintext - content to encrypt secret - secret to encrypt plaintext inital_vector - initial vector lazy - pad secret if less than legal blocksize (default: True) checksum - attach crc32 byte encoded (default: True) returns ciphertext """ if not isinstance(plaintext, six.binary_type): plaintext = six.b(plaintext) secret = _lazysecret(secret) if lazy else secret encobj = AES.new(secret, AES.MODE_CFB, inital_vector) if checksum: packed = _pack_crc(plaintext) plaintext += base64.urlsafe_b64encode(packed) encoded = base64.urlsafe_b64encode(encobj.encrypt(plaintext)) if isinstance(plaintext, six.binary_type): encoded = encoded.decode() return encoded.replace('=', '')
def decrypt(ciphertext, secret, inital_vector, checksum=True, lazy=True): """Decrypts ciphertext with secret ciphertext - encrypted content to decrypt secret - secret to decrypt ciphertext inital_vector - initial vector lazy - pad secret if less than legal blocksize (default: True) checksum - verify crc32 byte encoded checksum (default: True) returns plaintext """ secret = _lazysecret(secret) if lazy else secret encobj = AES.new(secret, AES.MODE_CFB, inital_vector) try: padded = ciphertext + ('=' * (len(ciphertext) % 4)) decoded = base64.urlsafe_b64decode(str(padded)) plaintext = encobj.decrypt(decoded) except (TypeError, binascii.Error): raise InvalidKeyError("invalid key") if checksum: try: crc, plaintext = (base64.urlsafe_b64decode( plaintext[-8:]), plaintext[:-8]) except (TypeError, binascii.Error): raise CheckSumError("checksum mismatch") if not crc == _pack_crc(plaintext): raise CheckSumError("checksum mismatch") return plaintext
def obfuscate(value, juice=None): """ Template filter that obfuscates whatever text it is applied to. The text is supposed to be a URL, but it will obfuscate anything. Usage: Extremely unfriendly URL: {{ "/my-secret-path/"|obfuscate }} Include some SEO juice: {{ "/my-secret-path/"|obfuscate:"some SEO friendly text" }} """ if not settings.UNFRIENDLY_ENABLE_FILTER: return value kwargs = { 'key': encrypt(value, settings.UNFRIENDLY_SECRET, settings.UNFRIENDLY_IV, checksum=settings.UNFRIENDLY_ENFORCE_CHECKSUM), } if juice: kwargs['juice'] = slugify(juice) return reverse('unfriendly-deobfuscate', kwargs=kwargs)
def missing_schema(self,html,song_name): ''' It will print the list of songs that can be downloaded ''' #html=self.get_html_response(url) soup=BeautifulSoup(html) name=' '.join(song_name) print '%s not found'%name print "But you can download any of the following songs :" a_list=soup.findAll('a','touch') for x in xrange(len(a_list)-1): r=a_list[x] p=str(r) q=re.sub(r'<a.*/>|<span.*">|</span>|</a>|<a.*html">|<font.*">|</font>','',p) print q
def list_of_all_href(self,html): ''' It will return all hyper links found in the mr-jatt page for download ''' soup=BeautifulSoup(html) links=[] a_list=soup.findAll('a','touch') for x in xrange(len(a_list)-1): link = a_list[x].get('href') name = a_list[x] name = str(name) name=re.sub(r'<a.*/>|<span.*">|</span>|</a>|<a.*html">|<font.*">|</font>','',name) name=re.sub(r'^[0-9]+\.','',name) links.append([link,name]) #quit() return links
def check_if_song_name(self,html): ''' Returns true if user entered artist or movie name ''' soup=BeautifulSoup(html) a_list=soup.findAll('a','touch') #print a_list text=[str(x) for x in a_list] text=''.join(text) text=text.lower() string1='download in 48 kbps' string2='download in 128 kbps' string3='download in 320 kbps' href='' if string3 in text: #print 'Downloading in 320 kbps' href=a_list[2].get('href') elif string2 in text: #print 'Downloading in 128 kbps' href=a_list[1].get('href') elif string1 in text: #print 'Downloading in 48 kbps' href=a_list[0].get('href') else: return (True,'nothing') return (False,href)
def Parse(self,url,song_name,flag): ''' It will the resource URL if song is found, Otherwise it will return the list of songs that can be downloaded ''' file_download=FileDownload() html=file_download.get_html_response(url) if flag == False: soup=BeautifulSoup(html) a_list=soup.findAll('a','touch') #print a_list text=[str(x) for x in a_list] text=''.join(text) text=text.lower() string1='download in 48 kbps' string2='download in 128 kbps' string3='download in 320 kbps' href='' if string3 in text: print 'Downloading in 320 kbps' href=a_list[2].get('href') elif string2 in text: print 'Downloading in 128 kbps' href=a_list[1].get('href') elif string1 in text: print 'Downloading in 48 kbps' href=a_list[0].get('href') else: self.missing_schema(html,song_name) quit() return href else: x,href=self.check_if_song_name(html) links = [] if x==True: links=self.list_of_all_href(html) else: file_download=FileDownload() file_download.file_download_cross_platform(href) quit() return links
def google_url(self,song_name,website): ''' It will return the google url to be searched''' name='+'.join(song_name) prefix='https://www.google.co.in/search?q=' website=website.split(" ") suffix='+'.join(website) url=prefix+name+suffix #print url return url
def parse_google(self,html): '''It will parse google html response and return the first url ''' soup = BeautifulSoup(html) href=soup.find('div','g').find('a').get('href') href_list=href.split('&') download_url=href_list[0] download_url=download_url.strip() download_url=download_url.replace('/url?q=','') return download_url
def Parse(self,song_name,website): ''' song_name is a list of strings website is a string It will return the url from where music file needs to be downloaded ''' url_to_be_parsed=self.google_url(song_name,website) file_download=FileDownload() html=file_download.get_html_response(url_to_be_parsed) website_url=self.parse_google(html) return website_url
def get_html_response(self,url): '''It will download the html page specified by url and return the html response ''' print "Downloading page %s .."%url try: response=requests.get(url,timeout=50) except requests.exceptions.SSLError: try: response=requests.get(url,verify=False,timeout=50) except requests.exceptions.RequestException as e: print e quit() except requests.exceptions.RequestException as e: print e quit() return response.content
def file_download_using_requests(self,url): '''It will download file specified by url using requests module''' file_name=url.split('/')[-1] if os.path.exists(os.path.join(os.getcwd(),file_name)): print 'File already exists' return #print 'Downloading file %s '%file_name #print 'Downloading from %s'%url try: r=requests.get(url,stream=True,timeout=200) except requests.exceptions.SSLError: try: response=requests.get(url,stream=True,verify=False,timeout=200) except requests.exceptions.RequestException as e: print e quit() except requests.exceptions.RequestException as e: print e quit() chunk_size = 1024 total_size = int(r.headers['Content-Length']) total_chunks = total_size/chunk_size file_iterable = r.iter_content(chunk_size = chunk_size) tqdm_iter = tqdm(iterable = file_iterable,total = total_chunks,unit = 'KB', leave = False ) with open(file_name,'wb') as f: for data in tqdm_iter: f.write(data) #total_size=float(r.headers['Content-Length'])/(1024*1024) '''print 'Total size of file to be downloaded %.2f MB '%total_size total_downloaded_size=0.0 with open(file_name,'wb') as f: for chunk in r.iter_content(chunk_size=1*1024*1024): if chunk: size_of_chunk=float(len(chunk))/(1024*1024) total_downloaded_size+=size_of_chunk print '{0:.0%} Downloaded'.format(total_downloaded_size/total_size) f.write(chunk)''' print 'Downloaded file %s '%file_name
def file_download_using_wget(self,url): '''It will download file specified by url using wget utility of linux ''' file_name=url.split('/')[-1] print 'Downloading file %s '%file_name command='wget -c --read-timeout=50 --tries=3 -q --show-progress --no-check-certificate ' url='"'+url+'"' command=command+url os.system(command)
def main(): """Main CLI entrypoint.""" #print VERSION from commands.download import Download options = docopt(__doc__, version=VERSION) #print "You reached here" #print options print "working." p=Download(options) p.run()
def ReadingBloomFilter(filename, want_lock=False): """ Create a read-only bloom filter with an upperbound of (num_elements, max_fp_prob) as a specification and using filename as the backing datastore. """ with open('{}.desc'.format(filename), 'r') as descriptor: num_elements = int(descriptor.readline()) max_fp_prob = float(descriptor.readline()) ignore_case = int(descriptor.readline()) return _hydra.BloomFilter.getFilter( num_elements, max_fp_prob, filename=filename, ignore_case=ignore_case, read_only=True, want_lock=want_lock)
def WritingBloomFilter(num_elements, max_fp_prob, filename=None, ignore_case=False, want_lock=False, fdatasync_on_close=True): """ Create a read/write bloom filter with an upperbound of (num_elements, max_fp_prob) as a specification and using filename as the backing datastore. """ new_filter = _hydra.BloomFilter.getFilter( num_elements, max_fp_prob, filename=filename, ignore_case=ignore_case, read_only=False, want_lock=want_lock, fdatasync_on_close=fdatasync_on_close) if filename: with open('{}.desc'.format(filename), 'w') as descriptor: descriptor.write("{}\n".format(num_elements)) descriptor.write("{:0.8f}\n".format(max_fp_prob)) descriptor.write("{:d}\n".format(ignore_case)) return new_filter
def findStationCodesByCity(city_name, token): """Lookup AQI database for station codes in a given city.""" req = requests.get( API_ENDPOINT_SEARCH, params={ 'token': token, 'keyword': city_name }) if req.status_code == 200 and req.json()["status"] == "ok": return [result["uid"] for result in req.json()["data"]] else: return []
def get_location_observation(lat, lng, token): """Lookup observations by geo coordinates.""" req = requests.get( API_ENDPOINT_GEO % (lat, lng), params={ 'token': token }) if req.status_code == 200 and req.json()["status"] == "ok": return parse_observation_response(req.json()["data"]) return {}
def parse_observation_response(json): """Decode AQICN observation response JSON into python object.""" logging.debug(json) iaqi = json['iaqi'] result = { 'idx': json['idx'], 'city': json.get('city', ''), 'aqi': json['aqi'], 'dominentpol': json.get("dominentpol", ''), 'time': json['time']['s'], 'iaqi': [{'p': item, 'v': iaqi[item]['v']} for item in iaqi] } return result
def get_station_observation(station_code, token): """Request station data for a specific station identified by code. A language parameter can also be specified to translate location information (default: "en") """ req = requests.get( API_ENDPOINT_OBS % (station_code), params={ 'token': token }) if req.status_code == 200 and req.json()['status'] == "ok": return parse_observation_response(req.json()['data']) else: return {}
def search_paths(self): """The list of logical paths which are used to search for an asset. This property makes sense only if the attributes was created with logical path. It is assumed that the logical path can be a directory containing a file named ``index`` with the same suffix. Example:: >>> attrs = AssetAttributes(environment, 'js/app.js') >>> attrs.search_paths ['js/app.js', 'js/app/index.js'] >>> attrs = AssetAttributes(environment, 'js/app/index.js') >>> attrs.search_paths ['js/models/index.js'] """ paths = [self.path] if os.path.basename(self.path_without_suffix) != 'index': path = os.path.join(self.path_without_suffix, 'index') paths.append(path + ''.join(self.suffix)) return paths
def path_without_suffix(self): """The relative path to asset without suffix. Example:: >>> attrs = AssetAttributes(environment, 'js/app.js') >>> attrs.path_without_suffix 'js/app' """ if self.suffix: return self.path[:-len(''.join(self.suffix))] return self.path
def logical_path(self): """The logical path to asset. Example:: >>> attrs = AssetAttributes(environment, 'js/models.js.coffee') >>> attrs.logical_path 'js/models.js' """ format_extension = self.format_extension or self.compiler_format_extension if format_extension is None: return self.path return self.path_without_suffix + format_extension
def extensions(self): """The list of asset extensions. Example:: >>> attrs = AssetAttributes(environment, 'js/models.js.coffee') >>> attrs.extensions ['.js', '.coffee'] >>> attrs = AssetAttributes(environment, 'js/lib/external.min.js.coffee') >>> attrs.format_extension ['.min', '.js', '.coffee'] """ return re.findall(r'\.[^.]+', os.path.basename(self.path))
def format_extension(self): """The format extension of asset. Example:: >>> attrs = AssetAttributes(environment, 'js/models.js.coffee') >>> attrs.format_extension '.js' >>> attrs = AssetAttributes(environment, 'js/lib/external.min.js.coffee') >>> attrs.format_extension '.js' """ for extension in reversed(self.extensions): compiler = self.environment.compilers.get(extension) if not compiler and self.environment.mimetypes.get(extension): return extension
def unknown_extensions(self): """The list of unknown extensions, which are actually parts of asset filename. Example:: >>> attrs = AssetAttributes(environment, 'js/lib-2.0.min.js') >>> attrs.suffix ['.0', '.min'] """ unknown_extensions = [] for extension in self.extensions: compiler = self.environment.compilers.get(extension) if compiler or self.environment.mimetypes.get(extension): return unknown_extensions unknown_extensions.append(extension) return unknown_extensions
def compiler_extensions(self): """The list of compiler extensions. Example:: >>> attrs = AssetAttributes(environment, 'js/lib/external.min.js.coffee') >>> attrs.compiler_extensions ['.coffee'] """ try: index = self.extensions.index(self.format_extension) except ValueError: index = 0 extensions = self.extensions[index:] return [e for e in extensions if self.environment.compilers.get(e)]
def compilers(self): """The list of compilers used to build asset.""" return [self.environment.compilers.get(e) for e in self.compiler_extensions]
def processors(self): """The list of all processors (preprocessors, compilers, postprocessors) used to build asset. """ return self.preprocessors + list(reversed(self.compilers)) + self.postprocessors
def mimetype(self): """MIME type of the asset.""" return (self.environment.mimetypes.get(self.format_extension) or self.compiler_mimetype or 'application/octet-stream')
def compiler_mimetype(self): """Implicit MIME type of the asset by its compilers.""" for compiler in reversed(self.compilers): if compiler.result_mimetype: return compiler.result_mimetype return None
def compiler_format_extension(self): """Implicit format extension on the asset by its compilers.""" for extension, mimetype in self.environment.mimetypes.items(): if mimetype == self.compiler_mimetype: return extension return None
def register(self, mimetype, processor): """Register passed `processor` for passed `mimetype`.""" if mimetype not in self or processor not in self[mimetype]: self.setdefault(mimetype, []).append(processor)
def unregister(self, mimetype, processor): """Remove passed `processor` for passed `mimetype`. If processor for this MIME type does not found in the registry, nothing happens. """ if mimetype in self and processor in self[mimetype]: self[mimetype].remove(processor)
def register_defaults(self): """Register :class:`~gears.processors.DirectivesProcessor` as a preprocessor for `text/css` and `application/javascript` MIME types. """ self.register('text/css', DirectivesProcessor.as_handler()) self.register('application/javascript', DirectivesProcessor.as_handler())
def suffixes(self): """The registry for supported suffixes of assets. It is built from MIME types and compilers registries, and is cached at the first call. See :class:`~gears.environment.Suffixes` for more information. """ if not hasattr(self, '_suffixes'): suffixes = Suffixes() for extension, mimetype in self.mimetypes.items(): suffixes.register(extension, root=True, mimetype=mimetype) for extension, compiler in self.compilers.items(): suffixes.register(extension, to=compiler.result_mimetype) self._suffixes = suffixes return self._suffixes
def paths(self): """The list of search paths. It is built from registered finders, which has ``paths`` property. Can be useful for compilers to resolve internal dependencies. """ if not hasattr(self, '_paths'): paths = [] for finder in self.finders: if hasattr(finder, 'paths'): paths.extend(finder.paths) self._paths = paths return self._paths
def register_defaults(self): """Register default compilers, preprocessors and MIME types.""" self.mimetypes.register_defaults() self.preprocessors.register_defaults() self.postprocessors.register_defaults()
def register_entry_points(self, exclude=()): """Allow Gears plugins to inject themselves to the environment. For example, if your plugin's package contains such ``entry_points`` definition in ``setup.py``, ``gears_plugin.register`` function will be called with current environment during ``register_entry_points`` call:: entry_points = { 'gears': [ 'register = gears_plugin:register', ], } Here is an example of such function:: def register(environment): assets_dir = os.path.join(os.path.dirname(__file__), 'assets') assets_dir = os.path.absolute_path(assets_dir) environment.register(FileSystemFinder([assets_dir])) If you want to disable this behavior for some plugins, list their packages using ``exclude`` argument:: environment.register_entry_points(exclude=['plugin']) """ for entry_point in iter_entry_points('gears', 'register'): if entry_point.module_name not in exclude: register = entry_point.load() register(self)
def find(self, item, logical=False): """Find files using :attr:`finders` registry. The ``item`` parameter can be an instance of :class:`~gears.asset_attributes.AssetAttributes` class, a path to the asset or a logical path to the asset. If ``item`` is a logical path, `logical` parameter must be set to ``True``. Returns a tuple with :class:`~gears.asset_attributes.AssetAttributes` instance for found file path as first item, and absolute path to this file as second item. If nothing is found, :class:`gears.exceptions.FileNotFound` exception is rased. """ if isinstance(item, AssetAttributes): for path in item.search_paths: try: return self.find(path, logical) except FileNotFound: continue raise FileNotFound(item.path) if logical: asset_attributes = AssetAttributes(self, item) suffixes = self.suffixes.find(asset_attributes.mimetype) if not suffixes: return self.find(item) path = asset_attributes.path_without_suffix for suffix in suffixes: try: return self.find(path + suffix) except FileNotFound: continue else: for finder in self.finders: try: absolute_path = finder.find(item) except FileNotFound: continue return AssetAttributes(self, item), absolute_path raise FileNotFound(item)
def list(self, path, mimetype=None): """Yield two-tuples for all files found in the directory given by ``path`` parameter. Result can be filtered by the second parameter, ``mimetype``, that must be a MIME type of assets compiled source code. Each tuple has :class:`~gears.asset_attributes.AssetAttributes` instance for found file path as first item, and absolute path to this file as second item. Usage example:: # Yield all files from 'js/templates' directory. environment.list('js/templates/*') # Yield only files that are in 'js/templates' directory and have # 'application/javascript' MIME type of compiled source code. environment.list('js/templates/*', mimetype='application/javascript') """ basename_pattern = os.path.basename(path) if path.endswith('**'): paths = [path] else: paths = AssetAttributes(self, path).search_paths paths = map(lambda p: p if p.endswith('*') else p + '*', paths) results = unique(self._list_paths(paths), lambda x: x[0]) for logical_path, absolute_path in results: asset_attributes = AssetAttributes(self, logical_path) if mimetype is not None and asset_attributes.mimetype != mimetype: continue basename = os.path.basename(asset_attributes.path_without_suffix) if not fnmatch(basename, basename_pattern) and basename != 'index': continue yield asset_attributes, absolute_path
def save(self): """Save handled public assets to :attr:`root` directory.""" for asset_attributes, absolute_path in self.list('**'): logical_path = os.path.normpath(asset_attributes.logical_path) check_asset = build_asset(self, logical_path, check=True) if check_asset.is_public: asset = build_asset(self, logical_path) source = bytes(asset) self.save_file(logical_path, source, asset.gzippable) if self.fingerprinting: self.save_file(asset.hexdigest_path, source, asset.gzippable) self.manifest.files[logical_path] = asset.hexdigest_path self.manifest.dump()
def PopulateForm(self): """ +-----------------------------------------------------------------------+ | +--- splitter ------------------------------------------------------+ | | | +-- list widget--------------+ +- IdaSettingsView -------------+ | | | | | | | | | | | | | - plugin name | | | | | | | | - plugin name | | | | | | | | - plugin name | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | +----------------------------+ +-------------------------------+ | | | +-------------------------------------------------------------------+ | +-----------------------------------------------------------------------+ """ hbox = QtWidgets.QHBoxLayout(self.parent) self._splitter = QtWidgets.QSplitter(QtCore.Qt.Horizontal) self._plugin_list = QtWidgets.QListWidget() plugin_names = set([]) for scope, fn in (("idb", ida_settings.IDASettings.get_idb_plugin_names), ("directory", ida_settings.IDASettings.get_directory_plugin_names), ("user", ida_settings.IDASettings.get_user_plugin_names), ("system", ida_settings.IDASettings.get_system_plugin_names)): for plugin_name in fn(): plugin_names.add(plugin_name) for plugin_name in plugin_names: self._plugin_list.addItem(plugin_name) self._splitter.addWidget(self._plugin_list) hbox.addWidget(self._splitter) self.parent.setLayout(hbox) self._plugin_list.currentItemChanged.connect(self._handle_plugin_changed)
def as_handler(cls, **initkwargs): """Converts the class into an actual handler function that can be used when registering different types of processors in :class:`~gears.environment.Environment` class instance. The arguments passed to :meth:`as_handler` are forwarded to the constructor of the class. """ @wraps(cls, updated=()) def handler(asset, *args, **kwargs): return handler.handler_class(**initkwargs)(asset, *args, **kwargs) handler.handler_class = cls handler.supports_check_mode = cls.supports_check_mode return handler
def run(self, input): """Runs :attr:`executable` with ``input`` as stdin. :class:`AssetHandlerError` exception is raised, if execution is failed, otherwise stdout is returned. """ p = self.get_process() output, errors = p.communicate(input=input.encode('utf-8')) if p.returncode != 0: raise AssetHandlerError(errors) return output.decode('utf-8')
def get_process(self): """Returns :class:`subprocess.Popen` instance with args from :meth:`get_args` result and piped stdin, stdout and stderr. """ return Popen(self.get_args(), stdin=PIPE, stdout=PIPE, stderr=PIPE)
def import_qtcore(): """ This nasty piece of code is here to force the loading of IDA's Qt bindings. Without it, Python attempts to load PySide from the site-packages directory, and failing, as it does not play nicely with IDA. via: github.com/tmr232/Cute """ has_ida = False try: # if we're running under IDA, # then we'll use IDA's Qt bindings import idaapi has_ida = True except ImportError: # not running under IDA, # so use default Qt installation has_ida = False if has_ida: old_path = sys.path[:] try: ida_python_path = os.path.dirname(idaapi.__file__) sys.path.insert(0, ida_python_path) if idaapi.IDA_SDK_VERSION >= 690: from PyQt5 import QtCore return QtCore else: from PySide import QtCore return QtCore finally: sys.path = old_path else: try: from PyQt5 import QtCore return QtCore except ImportError: pass try: from PySide import QtCore return QtCore except ImportError: pass raise ImportError("No module named PySide or PyQt")
def get_meta_netnode(): """ Get the netnode used to store settings metadata in the current IDB. Note that this implicitly uses the open IDB via the idc iterface. """ node_name = "$ {org:s}.{application:s}".format( org=IDA_SETTINGS_ORGANIZATION, application=IDA_SETTINGS_APPLICATION) return netnode.Netnode(node_name)