def parse(self, source, parser=None):
"""
Overrides ElementTree.ElementTree.parse() to validate and strip the
the OFX header before feeding the body tags to custom
TreeBuilder subclass (below) for parsing into Element instances.
"""
# If our source doesn't follow the file API...
if not hasattr(source, 'read'):
# ...try to interpret it as a file
try:
source = open(source, 'rb')
except OSError:
msg = "Can't read source '{}'".format(source)
raise ParseError(msg)
header = OFXHeader.parse(source)
source = source.read()
source = source.decode(header.codec)
# Cut a parser instance
parser = parser or TreeBuilder
parser = parser()
# Then parse tag soup into tree of Elements
parser.feed(source)
# ElementTree.TreeBuilder.close() returns the root.
# Follow ElementTree API and stash as self._root (so all normal
# ElementTree methods e.g. find() work normally on our subclass).
self._root = parser.close()
def _read(self, source):
"""
"""
# If our source doesn't follow the file API, try to interpret it
# as a file path
if not hasattr(source, 'read'):
try:
source = open(source, 'rb')
except OSError:
msg = "Can't read source '{}'".format(source)
raise ParseError(msg)
header = OFXHeader.parse(source)
source = source.read()
# Decode source stream according to the CHARSET declared by OFX header
source = source.decode(header.codec)
return header, source