Fix TypeError in prepare_tag_prefix on non-ASCII tag prefixes (Python 3)

[bysiber]

Summary

prepare_tag_prefix crashes with a TypeError when given a tag prefix containing non-ASCII characters (e.g., Unicode) in Python 3.

Problem

In Python 3, iterating over bytes yields int values, not single-byte bytes objects. The percent-encoding loop in prepare_tag_prefix calls ord(ch) on these integer values, which raises TypeError:

data = ch.encode('utf-8')
for ch in data:
    chunks.append('%%%02X' % ord(ch))  # TypeError: ord() expected string of length 1, but int found

The sibling method prepare_tag (just below on line ~607) handles this correctly with % ch instead of % ord(ch).

Reproduction

import yaml, io
from yaml.emitter import Emitter

stream = io.StringIO()
e = Emitter(stream)
e.prepare_tag_prefix('tag:ü,2002:')
# TypeError: ord() expected string of length 1, but int found

Fix

Drop the redundant ord() call to match the working prepare_tag implementation:

chunks.append('%%%02X' % ch)