[PATCH 2 of 2] encoding: extend test cases for utf8b

[Matt Mackall]

# HG changeset patch
# User Matt Mackall <mpm at selenic.com>
# Date 1446506253 21600
#      Mon Nov 02 17:17:33 2015 -0600
# Node ID a0664cf9d8d620ff82c5ec9954d7f58279c72dd8
# Parent  6bee6f327de32755da038193f453aa6bed6810c7
encoding: extend test cases for utf8b

This adds a round-trip helper and a few tests of streams that could
cause synchronization problems in the encoder.

diff -r 6bee6f327de3 -r a0664cf9d8d6 mercurial/encoding.py
--- a/mercurial/encoding.py	Mon Nov 02 17:16:16 2015 -0600
+++ b/mercurial/encoding.py	Mon Nov 02 17:17:33 2015 -0600
@@ -479,11 +479,17 @@
     is a round-trip process for strings like filenames, but metadata
     that's was passed through tolocal will remain in UTF-8.
 
+    >>> roundtrip = lambda x: fromutf8b(toutf8b(x)) == x
     >>> m = "\\xc3\\xa9\\x99abcd"
-    >>> n = toutf8b(m)
-    >>> n
+    >>> toutf8b(m)
     '\\xc3\\xa9\\xed\\xb2\\x99abcd'
-    >>> fromutf8b(n) == m
+    >>> roundtrip(m)
+    True
+    >>> roundtrip("\\xc2\\xc2\\x80")
+    True
+    >>> roundtrip("\\xef\\xbf\\xbd")
+    True
+    >>> roundtrip("\\xef\\xef\\xbf\\xbd")
     True
     '''