def _load_from_cache(cache_file_name, logger):
"""
Load from cache file
"""
try:
if not os.path.exists(cache_file_name):
return None, None, None
logger.debug(" Loading taxa data from cache")
data_file = open(cache_file_name, 'rb')
#
# Is file version correct
#
latest_version, errmsg = check_cache_file_version(
data_file, _FILE_VERSION_MAJ)
if not latest_version:
logger.warning(errmsg)
return None, None, None
#
# taxon data
#
taxon_id_to_scientific_name = marshal.load(data_file)
taxon_id_to_common_name = marshal.load(data_file)
taxon_id_to_parent = marshal.load(data_file)
return (taxon_id_to_scientific_name, taxon_id_to_common_name,
taxon_id_to_parent)
except:
raise
return None, None, None
def _load_from_cache (cache_file_name, logger, data_to_load = None):
"""
Load tree data from cache file
data_to_load is a list or set of the following values
"alignment", "node", "node_to_parent" which determines what data to load
usually just node or node_to_parent is useful.
"""
start_time = time.time()
try:
if not os.path.exists(cache_file_name):
return None, None, None
logger.debug(" Loading protein tree from cache")
logger.info(" Loading cache file = %s" % cache_file_name)
data_file = open(cache_file_name, 'rb')
#
# Is file version correct
#
latest_version, errmsg = check_cache_file_version (data_file, _FILE_VERSION_MAJ)
if not latest_version:
logger.warning(errmsg)
return None, None, None
taxon_id_to_species_data = cPickle.load(data_file)
taxon_ids_to_ortholog_species_pair = cPickle.load(data_file)
return taxon_id_to_species_data, taxon_ids_to_ortholog_species_pair
except:
raise
return None, None
def _load_feature_types_from_cache(cache_file_name, logger):
"""
Load features from cache file
"""
try:
if not os.path.exists(cache_file_name):
return None
logger.debug(" Loading feature types from cache")
data_file = open(cache_file_name, 'rb')
#
# Is file version correct
#
latest_version, errmsg = check_cache_file_version(
data_file, _FILE_VERSION_MAJ)
if not latest_version:
logger.warning(errmsg)
return None
#
# Ignore feature type data
#
cnt_protein_features = marshal.load(data_file)
for i in range(cnt_protein_features):
name = marshal.load(data_file)
t_protein_feature.load(data_file)
#
# Retrieve features by type
#
file_pos_by_section = read_directory_of_sections(data_file)
return sorted(file_pos_by_section.keys(), key=str.lower)
except:
raise
return None
def _load_from_cache (cache_file_name, logger, data_to_load = None):
"""
Load tree data from cache file
data_to_load is a list or set of the following values
"alignment", "node", "node_to_parent" which determines what data to load
usually just node or node_to_parent is useful.
"""
start_time = time.time()
try:
if not os.path.exists(cache_file_name):
return None, None, None
logger.debug(" Loading protein tree from cache")
logger.info(" Loading cache file = %s" % cache_file_name)
data_file = open(cache_file_name, 'rb')
#
# Is file version correct
#
latest_version, errmsg = check_cache_file_version (data_file, _FILE_VERSION_MAJ)
if not latest_version:
logger.warning(errmsg)
return None, None, None
#
# Retrieve protein tree data in sections
#
file_pos_by_section = read_directory_of_sections (data_file)
#
# Read all data if none specified
#
if data_to_load == None:
data_to_load = file_pos_by_section.keys()
load_method = dict()
load_method["alignment"] = load_dict_of_objects, t_prot_tree_alignment
load_method["node"] = load_dict_of_objects, t_prot_tree_node
load_method["node_to_parent"] = (lambda d, ignore: marshal.load(d)), dict
data = dict()
for section in file_pos_by_section:
#
# initialise to empty data
#
data[section] = dict()
if section not in data_to_load:
continue
#
# read data
#
data_file.seek(file_pos_by_section[section], os.SEEK_SET)
load_func, object_type = load_method[section]
data[section] = load_func(data_file, object_type)
# log what data have been read?
#
end_time = time.time()
logger.info(" Loaded tree data in %ds" % (end_time - start_time))
# prot_node_id_to_alignments, prot_node_id_to_nodes, prot_node_id_to_parent
return [data[k] for k in sorted(data.keys())]
except:
raise
return None, None, None
def _load_features_from_cache(cache_file_name,
logger,
filter_feature_types=None):
"""
Load features from cache file
"""
start_time = time.time()
try:
if not os.path.exists(cache_file_name):
logger.warning("The cache file %s does not exist" %
cache_file_name)
return None, None
logger.debug(" Loading features from cache")
logger.info(" Loading cache file = %s" % cache_file_name)
data_file = open(cache_file_name, 'rb')
#
# Is file version correct
#
latest_version, errmsg = check_cache_file_version(
data_file, _FILE_VERSION_MAJ)
if not latest_version:
logger.warning(errmsg)
return None, None
#
# Retrieve feature type data
#
protein_features = dict()
cnt_protein_features = marshal.load(data_file)
for i in range(cnt_protein_features):
name = marshal.load(data_file)
protein_features[name] = t_protein_feature.load(data_file)
#
# Retrieve features by type
#
file_pos_by_section = read_directory_of_sections(data_file)
#
# Read all features if none specified
#
if filter_feature_types == None:
filter_feature_types = file_pos_by_section.keys()
matches_by_type = defaultdict(lambda: defaultdict(list))
cnt_all_matches = 0
#
# have case-independent matches
#
filter_feature_types_lc = map(str.lower, filter_feature_types)
feature_type_lc_to_orig = dict(
zip(map(str.lower, file_pos_by_section.keys()),
file_pos_by_section.keys()))
for feature_type_lc in filter_feature_types_lc:
if feature_type_lc not in feature_type_lc_to_orig:
continue
feature_type = feature_type_lc_to_orig[feature_type_lc]
data_file.seek(file_pos_by_section[feature_type], os.SEEK_SET)
#
# read feature
#
#
cnt_prot_ids = marshal.load(data_file)
for i in xrange(cnt_prot_ids):
prot_id = marshal.load(data_file)
cnt_matches = marshal.load(data_file)
for i in range(cnt_matches):
matches_by_type[feature_type][prot_id].append(
t_protein_feature_match.load(data_file))
cnt_all_matches += cnt_matches
#
# log what features have been read?
#
end_time = time.time()
logger.info(" Loaded %d features in %ds" %
(cnt_all_matches, end_time - start_time))
return protein_features, matches_by_type
except:
raise
return None, None