I am late convert to Python 3. I am trying to process output from a REST api for protein sequences using urllib.
In legacy python I could use:
self.seq_fileobj = urllib2.urlopen("http://www.uniprot.org/uniprot/{}.fasta".format(uniprot_id))
self.seq_header = self.seq_fileobj.next()
print "Read in sequence information for {}.".format(self.seq_header[:-1])
self.sequence = [achar for a_line in self.seq_fileobj for achar in a_line if achar != "\n"]
print("Sequence:{}\n".format("".join(self.sequence)))
For the same section of code in python 3, I use:
context = ssl._create_unverified_context()
self.seq_fileobj = urllib.request.urlopen("https://www.uniprot.org/uniprot/{}.fasta".format(uniprot_id),context=context)
self.seq_header = next(self.seq_fileobj)
print("Read in sequence information for {}.".format(self.seq_header.rstrip()))
self.b_sequence = [str(achar).encode('utf-8') for a_line in self.seq_fileobj for achar in a_line]
self.sequence = [chr(int(x)) for x in self.b_sequence]
I have read a little about string encoding and decoding to modify my list comprehension for python 3:
self.b_sequence = [str(achar).encode('utf-8') for a_line in self.seq_fileobj for achar in a_line]
self.sequence = [chr(int(x)) for x in self.b_sequence]
Although my code is working- is this the best way to achieve this result where I go from an array of bytes of ascii characters encoded with utf-8 to their resulting strings?. The chr(int(x)) bit is what seems un pythonic to me and I fear I may be missing something.
You don't need to convert the bytes to strings on a character-to-character basis. Since you want to strip out the newline characters, you can instead read the entire file as bytes, convert the bytes to strings with the decode method (which defaults to the utf-8 encoding as you are using) and remove the newline characters using the str.replace method:
self.sequence = list(self.seq_fileobj.read().decode().replace('\n', ''))