lang/python31: Backport fox for Python issue #8168

- Backport Python issue #8168 [1]:

    python3 py_compile does not ignore UTF-8 BOM characters

This causes installation (during bytecode compilation) errors for Python
ports with sources that contain BOM characters [2]

The issue was fixed [3] in the default branch at the time (3.2) but was
not backported to 3.1.

Since Python 3.1 is now in security-fix-only mode (no new features or bug
fixes), backporting is required.

[1] http://bugs.python.org/issue8168
[2] http://www.freebsd.org/cgi/query-pr.cgi?pr=ports/186034
[3] http://hg.python.org/cpython/rev/e15a8a476494/

PR:		ports/186034
Reported by:	Mark Andrews <marka at isc dot org>

1 files changed, 49 insertions, 0 deletions

diff --git a/lang/python31/files/patch-Lib__py_compile.py b/lang/python31/files/patch-Lib__py_compile.py
new file mode 100644
index 000000000000..1eaaca1e5697
--- /dev/null
+++ b/lang/python31/files/patch-Lib__py_compile.py
@@ -0,0 +1,49 @@
+# Description: python3 py_compile does not ignore UTF-8 BOM characters 
+# Issue:  http://bugs.python.org/issue8168
+# Commit: http://hg.python.org/cpython/rev/e15a8a476494/
+# PR:     ports/186034
+
+--- ./Lib/py_compile.py.orig	2014-02-21 23:28:42.491208180 +1100
++++ ./Lib/py_compile.py	2014-02-21 23:29:22.052513709 +1100
+@@ -7,8 +7,8 @@
+ import imp
+ import marshal
+ import os
+-import re
+ import sys
++import tokenize
+ import traceback
+ 
+ MAGIC = imp.get_magic()
+@@ -78,21 +78,6 @@
+                    (x >> 16) & 0xff,
+                    (x >> 24) & 0xff]))
+ 
+-def read_encoding(file, default):
+-    """Read the first two lines of the file looking for coding: xyzzy."""
+-    f = open(file, "rb")
+-    try:
+-        for i in range(2):
+-            line = f.readline()
+-            if not line:
+-                break
+-            m = re.match(br".*\bcoding:\s*(\S+)\b", line)
+-            if m:
+-                return m.group(1).decode("ascii")
+-        return default
+-    finally:
+-        f.close()
+-
+ def compile(file, cfile=None, dfile=None, doraise=False):
+     """Byte-compile one Python source file to Python bytecode.
+ 
+@@ -128,7 +113,8 @@
+     directories).
+ 
+     """
+-    encoding = read_encoding(file, "utf-8")
++    with open(file, "rb") as f:
++        encoding = tokenize.detect_encoding(f.readline)[0]
+     f = open(file, 'U', encoding=encoding)
+     try:
+         timestamp = int(os.fstat(f.fileno()).st_mtime)