def ResolveObject(self, path):
opath = path
if not isinstance(path, bytes):
path = path.encode(self.chm_encoding)
ans = CHMFile.ResolveObject(self, path)
if ans[0] != chmlib.CHM_RESOLVE_SUCCESS and not isinstance(opath, bytes):
path = opath.encode('utf-8')
ans = CHMFile.ResolveObject(self, path)
return ans
def _get_contents(self, path):
"""return html contents of file `path' in chm archive."""
obj = CHMFile.ResolveObject(self, path)
if obj[0] != 0:
return None
html = CHMFile.RetrieveObject(self, obj[1])
return html[1]
def ResolveObject(self, path):
# filenames are utf-8 encoded in the chm index as far as I can
# determine, see https://tika.apache.org/1.11/api/org/apache/tika/parser/chm/accessor/ChmPmgiHeader.html
if not isinstance(path, bytes):
path = path.encode('utf-8')
return CHMFile.ResolveObject(self, path)