def __init__(self, debug = 0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = [ "input_file_name", "num_int_metabolites", \
"num_reactions", "metabolite_line", "unbalanced_metabolite", \
"num_rows", "num_cols", "irreversible_vector", \
"branch_metabolite", "non_branch_metabolite", \
"stoichiometric_tag", "kernel_tag", "subsets_tag", \
"reduced_system_tag", "convex_basis_tag", \
"conservation_relations_tag", "elementary_modes_tag", \
"reaction", "enzyme", "matrix_row", "sum_is_constant_line", \
"end_stochiometric", "end_kernel", "end_subsets", \
"end_reduced_system", "end_convex_basis", \
"end_conservation_relations", "end_elementary_modes" ]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names( metatool_format.metatool_record,
self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def __init__(self, debug = 0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = [ "input_file_name", "num_int_metabolites", \
"num_reactions", "metabolite_line", "unbalanced_metabolite", \
"num_rows", "num_cols", "irreversible_vector", \
"branch_metabolite", "non_branch_metabolite", \
"stoichiometric_tag", "kernel_tag", "subsets_tag", \
"reduced_system_tag", "convex_basis_tag", \
"conservation_relations_tag", "elementary_modes_tag", \
"reaction", "enzyme", "matrix_row", "sum_is_constant_line", \
"end_stochiometric", "end_kernel", "end_subsets", \
"end_reduced_system", "end_convex_basis", \
"end_conservation_relations", "end_elementary_modes" ]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names( metatool_format.metatool_record,
self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def index(handle, index_fn=None):
"""index(handle[, index_fn]) -> list of (PMID, MedlineID, start, end)
Index a Medline XML file. Returns where the records are, as
offsets from the beginning of the handle. index_fn is a callback
function with parameters (PMID, MedlineID, start, end) and is
called as soon as each record is indexes.
"""
# Find the correct format to parse the data.
data = handle.read(1000)
format_module = choose_format(data)
handle = _SavedDataHandle(handle, data)
format = format_module.format
wanted = ["MedlineCitation", "PMID", "MedlineID"]
format = Martel.select_names(format, wanted)
# Create an indexer that will save all the index information and
# call index_fn if appropriate.
indexes = []
def citation_fn(pmid, medline_id, start, end,
indexes=indexes, index_fn=index_fn):
if index_fn is not None:
index_fn(pmid, medline_id, start, end)
indexes.append((pmid, medline_id, start, end))
indexer = _IndexerHandler(citation_fn)
# Create the parser and parse the results.
parser = format.make_parser(debug_level=0)
parser.setContentHandler(indexer)
parser.setErrorHandler(handler.ErrorHandler())
parser.parseFile(handle)
return indexes
def __init__(self, debug = 0):
self.interest_tags = ["primer_name", "amplifier",
"amplifier_sequence", "amplifier_length",
"end_record"]
expression = Martel.select_names(primersearch_format.record,
self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def __init__(self, debug=0):
self.interest_tags = [
"primer_name", "amplifier", "amplifier_sequence",
"amplifier_length", "end_record"
]
expression = Martel.select_names(primersearch_format.record,
self.interest_tags)
self._parser = expression.make_parser(debug_level=debug)
def __init__(self, debug=0):
self.interest_tags = [
"comments", "single_primer_line", "start_primer", "product_size",
"forward_start", "forward_length", "forward_tm", "forward_gc",
"forward_seq", "reverse_start", "reverse_length", "reverse_tm",
"reverse_gc", "reverse_seq", "internal_start", "internal_length",
"internal_tm", "internal_gc", "internal_seq", "end_record"
]
expression = Martel.select_names(primer3_format.record,
self.interest_tags)
self._parser = expression.make_parser(debug_level=debug)
def __init__(self, debug = 0):
self.interest_tags = ["comments", "single_primer_line",
"start_primer", "product_size",
"forward_start", "forward_length",
"forward_tm", "forward_gc", "forward_seq",
"reverse_start", "reverse_length",
"reverse_tm", "reverse_gc", "reverse_seq",
"internal_start", "internal_length",
"internal_tm", "internal_gc", "internal_seq",
"end_record"]
expression = Martel.select_names(primer3_format.record,
self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def make_iterator(self, tag="record", select_names=None, debug_level=0):
"""S.make_iterator([tag][, select_names][, debug_level]) -> iterator"""
if select_names is not None:
select_names = list(select_names)
select_names.sort()
key = tuple(select_names), debug_level
else:
key = None, debug_level
if not self._iterator_cache.has_key(key):
import Martel
exp = self.expression
if select_names is not None:
exp = Martel.select_names(exp, select_names)
p = exp.make_iterator(tag, debug_level = debug_level)
self._iterator_cache[key] = p
return self._iterator_cache[key].copy()
def make_parser(self, select_names=None, debug_level=0):
"""S.make_parser([select_names][, debug_level]) -> parser"""
if select_names is not None:
select_names = list(select_names)
select_names.sort()
key = tuple(select_names), debug_level
else:
key = None, debug_level
if not self._parser_cache.has_key(key):
import Martel
exp = self.expression
if select_names is not None:
exp = Martel.select_names(exp, select_names)
p = exp.make_parser(debug_level = debug_level)
self._parser_cache[key] = p
return self._parser_cache[key].copy()
def __init__(self, debug = 0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = ["comment", "title_line", "sequence" ]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names(intelligenetics_format.intelligenetics_record, self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def __init__(self, debug=0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = ["comment", "title_line", "sequence"]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names(
intelligenetics_format.intelligenetics_record, self.interest_tags)
self._parser = expression.make_parser(debug_level=debug)
def __init__(self, debug = 0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = [ 'header_line', 'system_line', 'substance_multiline', \
'reactor_multiline', 'include_line' ]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names( ecell_format.ecell_record, self.interest_tags)
self._parser = expression.make_parser(debug_level = debug)
def __init__(self, debug_level = 0):
"""Initialize the scanner by setting up our caches.
Creating the parser takes a long time, so we want to cache it
to reduce parsing time.
Arguments:
o debug - The level of debugging that the parser should
display. Level 0 is no debugging, Level 2 displays the most
debugging info (but is much slower). See Martel documentation
for more info on this.
"""
# a listing of all tags we are interested in scanning for
# in the MartelParser
self.interest_tags = [ "cd_tag", \
"description_tag", \
"status_tag", \
"source_tag", \
"date_tag", \
"taxonomy_tag", \
"aligned_tag", \
"representative_tag", \
"range_tag", \
"sequence_tag", \
"description_contents_multiline", \
"status_contents_multiline", \
"source_contents_multiline", \
"date_contents_multiline", \
"reference_contents_multiline", \
"taxonomy_contents_multiline", \
"aligned_contents_multiline", \
"representative_contents_multiline", \
"range_contents_multiline", \
"cd_contents_multiline", \
"sequence_contents_multiline", \
"table_entry" ]
# make a parser that returns only the tags we are interested in
expression = Martel.select_names( cdd_format.cdd_record, self.interest_tags)
self._parser = expression.make_parser(debug_level )
def _fixup_sp_pattern(exp):
import re
import Martel
exp = Martel.select_names(exp, (Std.dbxref_dbname.tag,Std.dbxref_dbid.tag))
e = exp._find_groups(Std.dbxref_dbname.tag)
assert len(e) == 1
e = e[0]
e.name = "dbname"
dbstyle = e.attrs["style"]
e.attrs = {}
e = exp._find_groups(Std.dbxref_dbid.tag)
assert len(e) == 2
e[0].name = "primary_dbid"
primary_type = e[0].attrs["type"]
e[0].attrs = {}
e[1].name = "secondary_dbid"
secondary_type = e[1].attrs["type"]
e[1].attrs = {}
pattern = str(exp) + "$"
pat = re.compile(pattern)
return pat, dbstyle, primary_type, secondary_type
print "acc Start"
self.m_accession=1
if name=="pname":
self.m_pname=1
if name=="bioformat:sequence":
self.m_seq=1
if name=="bioformat:sequence_block":
self.seq=''
self.m_seq_block=1
def endElement(self, name):
if name == "accession":
print "acc End"
self.m_accession=0
if name == "pname":
self.m_pname=0
if name=="bioformat:sequence":
self.m_seq=0
if name=="bioformat:sequence_block":
#self.outf.write(self.seq+'\n')
self.m_seq_block=0
if __name__=="__main__":
exp = Martel.select_names(embl.format, ("record", "accession", "pname", "bioformat:sequence",))
parser=embl.format_expression.make_parser()
#outf=open(sys.argv[2],'w')
parser.setContentHandler(my_handler(sys.argv[2]))
parser.parse(sys.argv[1])
AC_block +
DT_created +
DT_seq_update +
DT_ann_update +
Martel.Opt(DE_block) +
Martel.Opt(GN_block) +
Martel.Opt(OS_block) +
Martel.Opt(OG_block) +
Martel.Opt(OC_block) +
Martel.Group("OX_block", Martel.NullOp()) +
Martel.Group("reference_block", Martel.Rep(reference)) +
comment +
Martel.Opt(DR_block) +
Martel.Opt(KW_block) +
Martel.Opt(feature_block) +
sequence +
end,
{"format": "swissprot/38"})
format_expression = Martel.Group("dataset", Martel.Rep1(record),
{"format": "swissprot/38"})
format = Martel.ParseRecords("dataset", {"format": "swissprot/38"},
record, RecordReader.EndsWith, ("//\n",) )
if __name__ == "__main__":
exp = Martel.select_names(format, ("entry_name", "sequence"))
parser = exp.make_parser()
parser.parseFile(open("/home/dalke/ftps/swissprot/sprot38.dat"))
