|
# http://pyrocko.org - GPLv3 # # The * Pyrocko Developers, 21st Century # ---|P------/S----------~Lg----------
# TODO remove debugging stuff # from matplotlib import pyplot as plt # # class SFig(object): # def __init__(self): # self.line = 0 # # def draw(self, eles, color='black'): # if not isinstance(eles, list): # eles = [eles] # # for ele in eles: # plt.plot( # [ele.tmin, ele.tmax], # [self.line, self.line], color=color, alpha=0.5) # plt.plot( # [ele.tmin, ele.tmax], # [self.line, self.line], 'o', color=color, alpha=0.5, ms=2) # # def draw_span(self, tmin, tmax, color='orange'): # plt.plot( # [tmin, tmax], # [self.line, self.line], color=color) # # def next(self): # self.line += 1 # # def show(self): # plt.show() # # def set_tspan(self, tmin, tmax): # plt.xlim(tmin, tmax) # # # f = SFig()
global g_icount
'station': 4, 'channel': 5, 'waveform': 6, 'waveform_promise': 6}
cfill2 = list(cfill) cfill2[3] = '[*]' return cfill, cfill2
groups = defaultdict(list) for channel in channels: codes = channel.codes gcodes = codes[:-1] + (codes[-1][:-1],) groups[gcodes].append(channel)
return groups
list_of_args = [channel._get_pyrocko_station_args() for channel in group] args = util.consistency_merge(list_of_args + extra_args) from pyrocko import model as pmodel return pmodel.Station( network=args[0], station=args[1], location=args[2], lat=args[3], lon=args[4], elevation=args[5], depth=args[6], channels=[ch.get_pyrocko_channel() for ch in group])
# f.next() # f.draw_span(tmin, tmax, 'purple') # f.next()
# f.draw_span(tmin_a, tmax_a, 'cyan') assert tmin_a < tmax_a data.append((tmin_a, 1)) data.append((tmax_a, -1))
# f.next()
# for tmin, tmax in gaps: # f.draw_span(tmin, tmax, 'pink') # # f.next()
''' Database backed file selection.
:param database: :py:class:`Database` object or path to database :param str persistent: if given a name, create a persistent selection
By default, a temporary table in the database is created to hold the names of the files in the selection. This table is only visible inside the application which created it. If a name is given to ``persistent``, a named selection is created, which is visible also in other applications using the same database. Paths of files can be added to the selection using the :py:meth:`add` method. '''
raise error.SquirrelError( 'invalid persistent selection name: %s' % persistent)
else:
'db': 'main' if self._persistent else 'temp', 'file_states': self.name + '_file_states', 'bulkinsert': self.name + '_bulkinsert'}
''' CREATE TABLE IF NOT EXISTS %(db)s.%(file_states)s ( file_id integer PRIMARY KEY, file_state integer, kind_mask integer, format text) ''')))
''' CREATE INDEX IF NOT EXISTS %(db)s.%(file_states)s_index_file_state ON %(file_states)s (file_state) '''))
else:
''' Get the database to which this selection belongs.
:returns: :py:class:`Database` object '''
''' Destroy the tables assoctiated with this selection. ''' 'DROP TABLE %(db)s.%(file_states)s'))
''' Update modification time of file without initiating reindexing.
Useful to prolong validity period of data with expiration date. '''
raise FileLoadError( 'Silent update for file "%s" failed: size has changed.' % file_path)
UPDATE files SET mtime = ? WHERE path = ? '''
self, file_paths, kind_mask=model.g_kind_mask_all, format='detect'):
''' Add files to the selection.
:param file_paths: Paths to files to be added to the selection. :type file_paths: iterator yielding ``str`` objects '''
file_paths = [file_paths]
# short non-iterator file_paths: can do without temp table
''' INSERT OR IGNORE INTO files VALUES (NULL, ?, NULL, NULL, NULL) ''', ((x,) for x in file_paths))
''' DELETE FROM %(db)s.%(file_states)s WHERE file_id IN ( SELECT files.file_id FROM files WHERE files.path == ? ) AND kind_mask != ? OR format != ? '''), ( (path, kind_mask, format) for path in file_paths))
''' INSERT OR IGNORE INTO %(db)s.%(file_states)s SELECT files.file_id, 0, ?, ? FROM files WHERE files.path = ? '''), ((kind_mask, format, path) for path in file_paths))
''' UPDATE %(db)s.%(file_states)s SET file_state = 1 WHERE file_id IN ( SELECT files.file_id FROM files WHERE files.path == ? ) AND file_state != 0 '''), ((path,) for path in file_paths))
''' CREATE TEMP TABLE temp.%(bulkinsert)s (path text) '''))
'INSERT INTO temp.%(bulkinsert)s VALUES (?)'), ((x,) for x in file_paths))
''' INSERT OR IGNORE INTO files SELECT NULL, path, NULL, NULL, NULL FROM temp.%(bulkinsert)s '''))
''' DELETE FROM %(db)s.%(file_states)s WHERE file_id IN ( SELECT files.file_id FROM temp.%(bulkinsert)s INNER JOIN files ON temp.%(bulkinsert)s.path == files.path) AND kind_mask != ? OR format != ? '''), (kind_mask, format))
''' INSERT OR IGNORE INTO %(db)s.%(file_states)s SELECT files.file_id, 0, ?, ? FROM temp.%(bulkinsert)s INNER JOIN files ON temp.%(bulkinsert)s.path == files.path '''), (kind_mask, format))
''' UPDATE %(db)s.%(file_states)s SET file_state = 1 WHERE file_id IN ( SELECT files.file_id FROM temp.%(bulkinsert)s INNER JOIN files ON temp.%(bulkinsert)s.path == files.path) AND file_state != 0 '''))
'DROP TABLE temp.%(bulkinsert)s'))
''' Remove files from the selection.
:param file_paths: Paths to files to be removed from the selection. :type file_paths: ``list`` of ``str`` ''' file_paths = [file_paths]
''' DELETE FROM %(db)s.%(file_states)s WHERE %(db)s.%(file_states)s.file_id IN (SELECT files.file_id FROM files WHERE files.path == ?) '''), ((path,) for path in file_paths))
''' Set file states to "known" (2). ''' ''' UPDATE %(db)s.%(file_states)s SET file_state = 2 WHERE file_state < 2 '''))
''' Set file states to "known" (2). ''' ''' UPDATE %(db)s.%(file_states)s SET file_state = 1 '''))
''' Get content inventory of all files in selection.
:param: skip_unchanged: if ``True`` only inventory of modified files is yielded (:py:meth:`_flag_modified` must be called beforehand).
This generator yields tuples ``((format, path), nuts)`` where ``path`` is the path to the file, ``format`` is the format assignation or ``'detect'`` and ``nuts`` is a list of :py:class:`pyrocko.squirrel.Nut` objects representing the contents of the file. '''
WHERE %(db)s.%(file_states)s.file_state == 0 ''' else:
SELECT %(db)s.%(file_states)s.format, files.path, files.format, files.mtime, files.size, nuts.file_segment, nuts.file_element, kind_codes.kind_id, kind_codes.codes, nuts.tmin_seconds, nuts.tmin_offset, nuts.tmax_seconds, nuts.tmax_offset, nuts.deltat FROM %(db)s.%(file_states)s LEFT OUTER JOIN files ON %(db)s.%(file_states)s.file_id = files.file_id LEFT OUTER JOIN nuts ON files.file_id = nuts.file_id LEFT OUTER JOIN kind_codes ON nuts.kind_codes_id == kind_codes.kind_codes_id ''' + where + ''' ORDER BY %(db)s.%(file_states)s.file_id ''')
sql = self._sql(''' SELECT files.path FROM %(db)s.%(file_states)s INNER JOIN files ON files.file_id = %(db)s.%(file_states)s.file_id ORDER BY %(db)s.%(file_states)s.file_id ''')
for values in self._conn.execute(sql): yield values[0]
''' Mark files which have been modified.
:param check: if ``True`` query modification times of known files on disk. If ``False``, only flag unknown files.
Assumes file state is 0 for newly added files, 1 for files added again to the selection (forces check), or 2 for all others (no checking is done for those).
Sets file state to 0 for unknown or modified files, 2 for known and not modified files. '''
UPDATE %(db)s.%(file_states)s SET file_state = 0 WHERE ( SELECT mtime FROM files WHERE files.file_id == %(db)s.%(file_states)s.file_id) IS NULL AND file_state == 1 ''')
UPDATE %(db)s.%(file_states)s SET file_state = 2 WHERE file_state == 1 ''')
SELECT files.file_id, files.path, files.format, files.mtime, files.size FROM %(db)s.%(file_states)s INNER JOIN files ON %(db)s.%(file_states)s.file_id == files.file_id WHERE %(db)s.%(file_states)s.file_state == 1 ORDER BY %(db)s.%(file_states)s.file_id ''')
size_db) in self._conn.execute(sql):
except io.UnknownFormat: continue
else:
# could better use callback function here...
UPDATE %(db)s.%(file_states)s SET file_state = ? WHERE file_id = ? ''')
return list(self.iter_paths())
''' Container to hold statistics about contents available through a squirrel. '''
help='number of files in selection') help='number of index nuts in selection') Tuple.T(content_t=String.T()), help='available code sequences in selection, e.g. ' '(agency, network, station, location) for stations nuts.') String.T(), help='available content types in selection') help='aggregated file size of files is selection') String.T(), Dict.T(Tuple.T(content_t=String.T()), Int.T()), help='breakdown of how many nuts of any content type and code ' 'sequence are available in selection, ``counts[kind][codes]``') optional=True, help='earliest start time of all nuts in selection') optional=True, help='latest end time of all nuts in selection')
def __str__(self): kind_counts = dict( (kind, sum(self.counts[kind].values())) for kind in self.kinds)
codes = ['.'.join(x) for x in self.codes]
if len(codes) > 20: codes = codes[:10] \ + ['[%i more]' % (len(codes) - 20)] \ + codes[-10:]
scodes = '\n' + util.ewrap(codes, indent=' ') if codes else '<none>' stmin = util.tts(self.tmin) if self.tmin is not None else '<none>' stmax = util.tts(self.tmax) if self.tmax is not None else '<none>'
s = ''' available codes: %s number of files: %i total size of known files: %s number of index nuts: %i available nut kinds: %s time span of indexed contents: %s - %s''' % ( scodes, self.nfiles, util.human_bytesize(self.total_size), self.nnuts, ', '.join('%s: %i' % ( kind, kind_counts[kind]) for kind in sorted(self.kinds)), stmin, stmax)
return s
''' Prompt, lazy, indexing, caching, dynamic seismological dataset access.
:param database: :py:class:`Database` object or path to database :param str persistent: if given a name, create a persistent selection
By default, temporary tables are created in the attached database to hold the names of the files in the selection as well as various indices and counters. These tables are only visible inside the application which created it. If a name is given to ``persistent``, a named selection is created, which is visible also in other applications using the same database. Paths of files can be added to the selection using the :py:meth:`add` method. '''
self, env=None, database=None, cache_path=None, persistent=None):
'waveform': cache.ContentCache(), 'default': cache.ContentCache()}
'nuts': self.name + '_nuts', 'kind_codes_count': self.name + '_kind_codes_count'})
''' CREATE TABLE IF NOT EXISTS %(db)s.%(nuts)s ( nut_id integer PRIMARY KEY, file_id integer, file_segment integer, file_element integer, kind_id integer, kind_codes_id integer, tmin_seconds integer, tmin_offset float, tmax_seconds integer, tmax_offset float, deltat float, kscale integer) ''')))
''' CREATE UNIQUE INDEX IF NOT EXISTS %(db)s.%(nuts)s_file_element ON %(nuts)s (file_id, file_segment, file_element) '''))
''' CREATE TABLE IF NOT EXISTS %(db)s.%(kind_codes_count)s ( kind_codes_id integer PRIMARY KEY, count integer) ''')))
''' CREATE INDEX IF NOT EXISTS %(db)s.%(nuts)s_index_file_id ON %(nuts)s (file_id) '''))
''' CREATE INDEX IF NOT EXISTS %(db)s.%(nuts)s_index_tmin_seconds ON %(nuts)s (tmin_seconds) '''))
''' CREATE INDEX IF NOT EXISTS %(db)s.%(nuts)s_index_tmax_seconds ON %(nuts)s (tmax_seconds) '''))
''' CREATE INDEX IF NOT EXISTS %(db)s.%(nuts)s_index_kscale ON %(nuts)s (kind_id, kscale, tmin_seconds) '''))
''' CREATE TRIGGER IF NOT EXISTS %(db)s.%(nuts)s_delete_nuts BEFORE DELETE ON main.files FOR EACH ROW BEGIN DELETE FROM %(nuts)s WHERE file_id == old.file_id; END '''))
# trigger only on size to make silent update of mtime possible ''' CREATE TRIGGER IF NOT EXISTS %(db)s.%(nuts)s_delete_nuts2 BEFORE UPDATE OF size ON main.files FOR EACH ROW BEGIN DELETE FROM %(nuts)s WHERE file_id == old.file_id; END '''))
''' CREATE TRIGGER IF NOT EXISTS %(db)s.%(file_states)s_delete_files BEFORE DELETE ON %(db)s.%(file_states)s FOR EACH ROW BEGIN DELETE FROM %(nuts)s WHERE file_id == old.file_id; END '''))
''' CREATE TRIGGER IF NOT EXISTS %(db)s.%(nuts)s_inc_kind_codes BEFORE INSERT ON %(nuts)s FOR EACH ROW BEGIN INSERT OR IGNORE INTO %(kind_codes_count)s VALUES (new.kind_codes_id, 0); UPDATE %(kind_codes_count)s SET count = count + 1 WHERE new.kind_codes_id == %(kind_codes_count)s.kind_codes_id; END '''))
''' CREATE TRIGGER IF NOT EXISTS %(db)s.%(nuts)s_dec_kind_codes BEFORE DELETE ON %(nuts)s FOR EACH ROW BEGIN UPDATE %(kind_codes_count)s SET count = count - 1 WHERE old.kind_codes_id == %(kind_codes_count)s.kind_codes_id; END '''))
'''Delete database tables associated with this squirrel.'''
DROP TRIGGER %(db)s.%(nuts)s_delete_nuts; DROP TRIGGER %(db)s.%(nuts)s_delete_nuts2; DROP TRIGGER %(db)s.%(file_states)s_delete_files; DROP TRIGGER %(db)s.%(nuts)s_inc_kind_codes; DROP TRIGGER %(db)s.%(nuts)s_dec_kind_codes; DROP TABLE %(db)s.%(nuts)s; DROP TABLE %(db)s.%(kind_codes_count)s; '''.strip().splitlines():
stream = sys.stdout
table_names = [table_names]
'selection_file_states', 'selection_nuts', 'selection_kind_codes_count', 'files', 'nuts', 'kind_codes', 'kind_codes_count']
'selection_file_states': '%(db)s.%(file_states)s', 'selection_nuts': '%(db)s.%(nuts)s', 'selection_kind_codes_count': '%(db)s.%(kind_codes_count)s', 'files': 'files', 'nuts': 'nuts', 'kind_codes': 'kind_codes', 'kind_codes_count': 'kind_codes_count'}
m[table_name] % self._names, stream=stream)
else:
''' Add files to the selection.
:param file_paths: iterator yielding paths to files or directories to be added to the selection. Recurses into directories given. If given a `str`, it is treated as a single path to be added. :param kinds: if given, allowed content types to be made available through the squirrel selection. :type kinds: ``list`` of ``str`` :param format: file format identifier or ``'detect'`` for auto-detection :type format: str
Complexity: O(log N) '''
self, self.iter_expand_dirs(file_paths), kind_mask, format)
''' Add content which is not backed by files.
Stores to the main database and the selection.
If ``virtual_file_paths`` are given, this prevents creating a temp list of the nuts while aggregating the file paths for the selection. '''
virtual_file_paths = [virtual_file_paths]
nuts_add = [] virtual_file_paths = set() for nut in nuts: virtual_file_paths.add(nut.file_path) nuts_add.append(nut) else:
self, content=[], skip_unchanged=True, check=check):
''' INSERT INTO %(db)s.%(nuts)s SELECT NULL, nuts.file_id, nuts.file_segment, nuts.file_element, nuts.kind_id, nuts.kind_codes_id, nuts.tmin_seconds, nuts.tmin_offset, nuts.tmax_seconds, nuts.tmax_offset, nuts.deltat, nuts.kscale FROM %(db)s.%(file_states)s INNER JOIN nuts ON %(db)s.%(file_states)s.file_id == nuts.file_id INNER JOIN kind_codes ON nuts.kind_codes_id == kind_codes.kind_codes_id WHERE %(db)s.%(file_states)s.file_state != 2 AND (((1 << kind_codes.kind_id) & %(db)s.%(file_states)s.kind_mask) != 0) '''))
''' Add remote resource. '''
''' Add FDSN site for transparent remote data access. '''
''' Add online catalog for transparent event data access. '''
self, obj=None, tmin=None, tmax=None, time=None, codes=None):
tmin = time tmax = time
tmin = tmin if tmin is not None else obj.tmin tmax = tmax if tmax is not None else obj.tmax codes = codes if codes is not None else obj.codes
codes = tuple(codes.split('.'))
self, kind=None, tmin=None, tmax=None, codes=None, naiv=False):
''' Iterate content matching given contraints.
:param kind: ``str``, content kind to extract or sequence of such :param tmin: timestamp, start time of interval :param tmax: timestamp, end time of interval :param codes: tuple of str, pattern of content codes to be matched
Complexity: O(log N) for the time selection part due to heavy use of database indexes.
Yields :py:class:`pyrocko.squirrel.Nut` objects representing the intersecting content.
Query time span is treated as a half-open interval ``[tmin, tmax)``. However, if ``tmin`` equals ``tmax``, the edge logics are modified to closed-interval so that content intersecting with the time instant ``t = tmin = tmax`` is returned (otherwise nothing would be returned as ``[t, t)`` never matches anything).
Content time spans are also treated as half open intervals, e.g. content span ``[0, 1)`` is matched by query span ``[0, 1)`` but not by ``[-1, 0)`` or ``[1, 2)``. Also here, logics are modified to closed-interval when the content time span is an empty interval, i.e. to indicate a time instant. E.g. time instant 0 is matched by ``[0, 1)`` but not by ``[-1, 0)`` or ``[1, 2)``. '''
tmin = self.get_time_span()[0] tmax = self.get_time_span()[1] + 1.0
else:
(%(db)s.%(nuts)s.kind_id = ? AND %(db)s.%(nuts)s.kscale == ? AND %(db)s.%(nuts)s.tmin_seconds BETWEEN ? AND ?) ''') (to_kind_id(kind), kscale, tmin_seconds - tscale - 1, tmax_seconds + 1))
else: (%(db)s.%(nuts)s.kind_id == ? AND %(db)s.%(nuts)s.kscale == ? AND %(db)s.%(nuts)s.tmin_seconds <= ?) ''')
(to_kind_id(kind), kscale, tmax_seconds + 1))
pats = codes_patterns_for_kind(kind, codes) extra_cond.append( ' ( %s ) ' % ' OR '.join( ('kind_codes.codes GLOB ?',) * len(pats))) args.extend(separator.join(pat) for pat in pats)
SELECT files.path, files.format, files.mtime, files.size, %(db)s.%(nuts)s.file_segment, %(db)s.%(nuts)s.file_element, kind_codes.kind_id, kind_codes.codes, %(db)s.%(nuts)s.tmin_seconds, %(db)s.%(nuts)s.tmin_offset, %(db)s.%(nuts)s.tmax_seconds, %(db)s.%(nuts)s.tmax_offset, %(db)s.%(nuts)s.deltat FROM files INNER JOIN %(db)s.%(nuts)s ON files.file_id == %(db)s.%(nuts)s.file_id INNER JOIN kind_codes ON %(db)s.%(nuts)s.kind_codes_id == kind_codes.kind_codes_id ''')
else: or (nut.tmin == nut.tmax and tmin == nut.tmin):
else: or (nut.tmin == nut.tmax and tmin <= nut.tmin < tmax):
self, kind, tmin=None, tmax=None, codes=None, file_path=None):
(%(db)s.%(nuts)s.kind_id = ? AND %(db)s.%(nuts)s.kscale == ? AND %(db)s.%(nuts)s.tmin_seconds BETWEEN ? AND ?) ''') (to_kind_id(kind), kscale, tmin_seconds - tscale - 1, tmax_seconds + 1))
else: (%(db)s.%(nuts)s.kind_id == ? AND %(db)s.%(nuts)s.kscale == ? AND %(db)s.%(nuts)s.tmin_seconds <= ?) ''')
(to_kind_id(kind), kscale, tmax_seconds + 1))
' ( %s ) ' % ' OR '.join( ('kind_codes.codes GLOB ?',) * len(pats)))
SELECT %(db)s.%(nuts)s.nut_id, %(db)s.%(nuts)s.tmin_seconds, %(db)s.%(nuts)s.tmin_offset, %(db)s.%(nuts)s.tmax_seconds, %(db)s.%(nuts)s.tmax_offset, %(db)s.%(nuts)s.deltat FROM files INNER JOIN %(db)s.%(nuts)s ON files.file_id == %(db)s.%(nuts)s.file_id INNER JOIN kind_codes ON %(db)s.%(nuts)s.kind_codes_id == kind_codes.kind_codes_id WHERE ( ''' + ' OR '.join(tmin_cond) + ''' ) AND ''' + ' AND '.join(extra_cond))
nut_tmax_seconds, nut_tmax_offset, nut_deltat = row
nut_tmin_seconds, nut_tmin_offset, nut_deltat) nut_tmax_seconds, nut_tmax_offset, nut_deltat)
nut_tmin_seconds, nut_tmin_offset, tmin_seconds, tmin_offset, model.tscale_to_kscale( tmin_seconds - nut_tmin_seconds), nut_id))
tmax_seconds, tmax_offset, nut_tmax_seconds, nut_tmax_offset, model.tscale_to_kscale( nut_tmax_seconds - tmax_seconds), nut_id))
INSERT INTO %(db)s.%(nuts)s ( file_id, file_segment, file_element, kind_id, kind_codes_id, deltat, tmin_seconds, tmin_offset, tmax_seconds, tmax_offset, kscale ) SELECT file_id, file_segment, file_element, kind_id, kind_codes_id, deltat, ?, ?, ?, ?, ? FROM %(db)s.%(nuts)s WHERE nut_id == ? '''
''' Get time interval over all content in selection.
Complexity O(1), independent of number of nuts
:returns: (tmin, tmax) ''' SELECT MIN(tmin_seconds + tmin_offset) FROM %(db)s.%(nuts)s WHERE tmin_seconds == (SELECT MIN(tmin_seconds) FROM %(db)s.%(nuts)s) ''')
SELECT MAX(tmax_seconds + tmax_offset) FROM %(db)s.%(nuts)s WHERE tmax_seconds == (SELECT MAX(tmax_seconds) FROM %(db)s.%(nuts)s) ''')
''' Iterate over content types available in selection.
:param codes: if given, get kinds only for selected codes identifier
Complexity: O(1), independent of number of nuts '''
codes=codes, kind_codes_count='%(db)s.%(kind_codes_count)s' % self._names)
''' Iterate over content identifier code sequences available in selection.
:param kind: if given, get codes only for a given content type :type kind: ``str``
Complexity: O(1), independent of number of nuts ''' kind=kind, kind_codes_count='%(db)s.%(kind_codes_count)s' % self._names)
''' Iterate over number of occurrences of any (kind, codes) combination.
:param kind: if given, get counts only for selected content type
Yields tuples ``((kind, codes), count)``
Complexity: O(1), independent of number of nuts ''' kind=kind, kind_codes_count='%(db)s.%(kind_codes_count)s' % self._names)
''' Get content types available in selection.
:param codes: if given, get kinds only for selected codes identifier
Complexity: O(1), independent of number of nuts
:returns: sorted list of available content types '''
''' Get identifier code sequences available in selection.
:param kind: if given, get codes only for selected content type
Complexity: O(1), independent of number of nuts
:returns: sorted list of available codes as tuples of strings '''
''' Get number of occurrences of any (kind, codes) combination.
:param kind: if given, get codes only for selected content type
Complexity: O(1), independent of number of nuts
:returns: ``dict`` with ``counts[kind][codes] or ``counts[codes]`` if kind is not ``None`` '''
else:
''' Update inventory of remote content for a given selection.
This function triggers all attached remote sources, to check for updates in the metadata. The sources will only submit queries when their expiration date has passed, or if the selection spans into previously unseen times or areas. '''
''' Get number of files in selection. '''
''' Get number of nuts in selection. '''
''' Get aggregated file size available in selection. '''
SELECT SUM(files.size) FROM %(db)s.%(file_states)s INNER JOIN files ON %(db)s.%(file_states)s.file_id = files.file_id ''')
''' Get statistics on contents available through this selection. '''
nfiles=self.get_nfiles(), nnuts=self.get_nnuts(), kinds=self.get_kinds(), codes=self.get_codes(), total_size=self.get_total_size(), counts=self.get_counts(), tmin=tmin, tmax=tmax)
''' Get and possibly load full content for a given index entry from file.
Loads the actual content objects (channel, station, waveform, ...) from file. For efficiency sibling content (all stuff in the same file segment) will also be loaded as a side effect. The loaded contents are cached in the squirrel object. '''
nut.file_path, segment=nut.file_segment, format=nut.file_format, database=self._database):
except KeyError: raise error.NotAvailable( 'Unable to retrieve content: %s, %s, %s, %s' % nut.key)
self._content_caches.keys() if cache is None else [cache]):
self._content_caches.keys() if cache is None else [cache]):
'Multiple entries matching codes %s' % '.'.join(codes.split(separator)))
self.iter_nuts('station', *args), key=lambda nut: nut.dkey)
self.iter_nuts('channel', *args), key=lambda nut: nut.dkey)
self.iter_nuts('event', *args), key=lambda nut: nut.dkey)
# f.draw_span(tmin, tmax) # f.next()
# f.draw(waveforms, 'red') # f.draw(promises, 'gray') # # f.next()
if isinstance(x, tuple): return tuple(tts(e) for e in x) elif isinstance(x, list): return list(tts(e) for e in x) else: return util.time_to_str(x)
max(tmin, promise.tmin), min(tmax, promise.tmax), promise.deltat):
WaveformOrder( source_id=promise.file_path, codes=tuple(promise.codes.split(separator)), tmin=block_tmin, tmax=block_tmax, deltat=promise.deltat, gaps=gaps(waveforms_avail, block_tmin, block_tmax)))
# f.draw(orders_noop, 'green') # f.draw(orders, 'blue') # f.next()
logger.info( 'Waveform orders already satisified with cached/local data: ' '%i (%i)' % (len(order_keys_noop), len(orders_noop)))
(source_id, i) for (i, source_id) in enumerate(source_ids))
key=lambda order: source_priority[order.source_id])
'Waveform orders standing for download: %i (%i)' % (len(order_groups), len(orders)))
'waveform_promise', order.tmin, order.tmax, codes=order.codes, file_path=order.source_id)
pass
# f.draw(order, 'black')
split_promise(order)
# TODO: parallelize this loop self, by_source_id[source_id], success=success, error_permanent=split_promise, error_temporary=noop)
# f.next()
self.iter_nuts('waveform', *args), key=lambda nut: nut.dkey)
nuts = self.get_waveform_nuts(*args, **kwargs) # self.check_duplicates(nuts) return [self.get_content(nut, 'waveform') for nut in nuts]
name=cargs[0], azimuth=cargs[1], dip=cargs[2]))
network=sargs[0], station=sargs[1], location=sargs[2], lat=sargs[3], lon=sargs[4], elevation=sargs[5], depth=sargs[6] or 0.0, channels=pchannels))
def pile(self):
self.pile.snuffle()
gather(codes) for codes in self.iter_codes(kind) if selector is None or selector(codes))
def __str__(self): return str(self.get_stats())
''' Container to hold statistics about contents cached in meta-information db. '''
help='number of files in database') help='number of index nuts in database') Tuple.T(content_t=String.T()), help='available code sequences in database, e.g. ' '(agency, network, station, location) for stations nuts.') String.T(), help='available content types in database') help='aggregated file size of files referenced in database') String.T(), Dict.T(Tuple.T(content_t=String.T()), Int.T()), help='breakdown of how many nuts of any content type and code ' 'sequence are available in database, ``counts[kind][codes]``')
def __str__(self): kind_counts = dict( (kind, sum(self.counts[kind].values())) for kind in self.kinds)
codes = ['.'.join(x) for x in self.codes]
if len(codes) > 20: codes = codes[:10] \ + ['[%i more]' % (len(codes) - 20)] \ + codes[-10:]
scodes = '\n' + util.ewrap(codes, indent=' ') if codes else '<none>'
s = ''' available codes: %s number of files: %i total size of known files: %s number of index nuts: %i available nut kinds: %s''' % ( scodes, self.nfiles, util.human_bytesize(self.total_size), self.nnuts, ', '.join('%s: %i' % ( kind, kind_counts[kind]) for kind in sorted(self.kinds)))
return s
''' Shared meta-information database used by squirrel. '''
except sqlite3.OperationalError: raise error.SquirrelError( 'Cannot connect to database: %s' % database_path)
self._conn.set_trace_callback(self._log_statement)
logger.debug(statement)
c.execute( ''' SELECT name FROM sqlite_master WHERE type = 'table' AND name = '{files}' '''))): return
'''PRAGMA recursive_triggers = true''')
''' CREATE TABLE IF NOT EXISTS files ( file_id integer PRIMARY KEY, path text, format text, mtime float, size integer) '''))
''' CREATE UNIQUE INDEX IF NOT EXISTS index_files_file_path ON files (path) ''')
''' CREATE TABLE IF NOT EXISTS nuts ( nut_id integer PRIMARY KEY AUTOINCREMENT, file_id integer, file_segment integer, file_element integer, kind_id integer, kind_codes_id integer, tmin_seconds integer, tmin_offset float, tmax_seconds integer, tmax_offset float, deltat float, kscale integer) '''))
''' CREATE UNIQUE INDEX IF NOT EXISTS index_nuts_file_element ON nuts (file_id, file_segment, file_element) ''')
''' CREATE TABLE IF NOT EXISTS kind_codes ( kind_codes_id integer PRIMARY KEY, kind_id integer, codes text) '''))
''' CREATE UNIQUE INDEX IF NOT EXISTS index_kind_codes ON kind_codes (kind_id, codes) ''')
''' CREATE TABLE IF NOT EXISTS kind_codes_count ( kind_codes_id integer PRIMARY KEY, count integer) '''))
''' CREATE INDEX IF NOT EXISTS index_nuts_file_id ON nuts (file_id) ''')
''' CREATE TRIGGER IF NOT EXISTS delete_nuts_on_delete_file BEFORE DELETE ON files FOR EACH ROW BEGIN DELETE FROM nuts where file_id == old.file_id; END ''')
# trigger only on size to make silent update of mtime possible ''' CREATE TRIGGER IF NOT EXISTS delete_nuts_on_update_file BEFORE UPDATE OF size ON files FOR EACH ROW BEGIN DELETE FROM nuts where file_id == old.file_id; END ''')
''' CREATE TRIGGER IF NOT EXISTS increment_kind_codes BEFORE INSERT ON nuts FOR EACH ROW BEGIN INSERT OR IGNORE INTO kind_codes_count VALUES (new.kind_codes_id, 0); UPDATE kind_codes_count SET count = count + 1 WHERE new.kind_codes_id == kind_codes_id; END ''')
''' CREATE TRIGGER IF NOT EXISTS decrement_kind_codes BEFORE DELETE ON nuts FOR EACH ROW BEGIN UPDATE kind_codes_count SET count = count - 1 WHERE old.kind_codes_id == kind_codes_id; END ''')
''' Store or update content meta-information.
Given ``nuts`` are assumed to represent an up-to-date and complete inventory of a set of files. Any old information about these files is first pruned from the database (via database triggers). If such content is part of a live selection, it is also removed there. Then the new content meta-information is inserted into the main database. The content is not automatically inserted into the live selections again. It is in the responsibility of the selection object to perform this step. '''
return
nut.file_path, nut.file_format, nut.file_mtime, nut.file_size))
'INSERT OR IGNORE INTO files VALUES (NULL,?,?,?,?)', files)
'''UPDATE files SET format = ?, mtime = ?, size = ? WHERE path == ? ''', ((x[1], x[2], x[3], x[0]) for x in files))
'INSERT OR IGNORE INTO kind_codes VALUES (NULL,?,?)', kind_codes)
''' INSERT INTO nuts VALUES (NULL, ( SELECT file_id FROM files WHERE path == ? ),?,?,?, ( SELECT kind_codes_id FROM kind_codes WHERE kind_id == ? AND codes == ? ), ?,?,?,?,?,?) ''', ((nut.file_path, nut.file_segment, nut.file_element, nut.kind_id, nut.kind_id, nut.codes, nut.tmin_seconds, nut.tmin_offset, nut.tmax_seconds, nut.tmax_offset, nut.deltat, nut.kscale) for nut in nuts))
SELECT files.path, files.format, files.mtime, files.size, nuts.file_segment, nuts.file_element, kind_codes.kind_id, kind_codes.codes, nuts.tmin_seconds, nuts.tmin_offset, nuts.tmax_seconds, nuts.tmax_offset, nuts.deltat FROM files INNER JOIN nuts ON files.file_id = nuts.file_id INNER JOIN kind_codes ON nuts.kind_codes_id == kind_codes.kind_codes_id WHERE path == ? '''
for row in self._conn.execute(sql, (path,))]
SELECT files.path, files.format, files.mtime, files.size, nuts.file_segment, nuts.file_element, kind_codes.kind_id, kind_codes.codes, nuts.tmin_seconds, nuts.tmin_offset, nuts.tmax_seconds, nuts.tmax_offset, nuts.deltat FROM files INNER JOIN nuts ON files.file_id == nuts.file_id INNER JOIN kind_codes ON nuts.kind_codes_id == kind_codes.kind_codes_id '''
return None
''' Prune content meta-inforamation about a given file.
All content pieces belonging to file ``path`` are removed from the main database and any attached live selections (via database triggers). '''
self._conn.execute( 'DELETE FROM files WHERE path = ?', (path,))
''' Prune information associated with a given file, but keep the file path.
This method is called when reading a file failed. File attributes, format, size and modification time are set to NULL. File content meta-information is removed from the database and any attached live selections (via database triggers). '''
''' UPDATE files SET format = NULL, mtime = NULL, size = NULL WHERE path = ? ''', (path,))
SELECT kind_codes.kind_id, kind_codes.codes, %(kind_codes_count)s.count FROM %(kind_codes_count)s INNER JOIN kind_codes ON %(kind_codes_count)s.kind_codes_id == kind_codes.kind_codes_id WHERE %(kind_codes_count)s.count > 0 ''' + sel + ''' ''') % {'kind_codes_count': kind_codes_count}
SELECT DISTINCT kind_codes.codes FROM %(kind_codes_count)s INNER JOIN kind_codes ON %(kind_codes_count)s.kind_codes_id == kind_codes.kind_codes_id WHERE %(kind_codes_count)s.count > 0 ''' + sel + ''' ORDER BY kind_codes.codes ''') % {'kind_codes_count': kind_codes_count}
SELECT DISTINCT kind_codes.kind_id FROM %(kind_codes_count)s INNER JOIN kind_codes ON %(kind_codes_count)s.kind_codes_id == kind_codes.kind_codes_id WHERE %(kind_codes_count)s.count > 0 ''' + sel + ''' ORDER BY kind_codes.kind_id ''') % {'kind_codes_count': kind_codes_count}
else:
SELECT SUM(files.size) FROM files '''
nfiles=self.get_nfiles(), nnuts=self.get_nnuts(), kinds=self.get_kinds(), codes=self.get_codes(), counts=self.get_counts(), total_size=self.get_total_size())
def __str__(self): return str(self.get_stats())
'files', 'nuts', 'kind_codes', 'kind_codes_count']:
stream = sys.stdout
def __repr__(self): return self
max((len(repr(x)) if x is not None else 0) for x in col) for col in zip(*tab)]
(repr(x).ljust(wid) if x is not None else ''.ljust(wid, '-')) for (x, wid) in zip(row, widths)))
'Selection', 'Squirrel', 'SquirrelStats', 'Database', 'DatabaseStats', ] |