Merged revisions 79138-79139,79306 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/trunk

................
  r79138 | benjamin.peterson | 2010-03-20 11:16:44 -0500 (Sat, 20 Mar 2010) | 1 line

  document exitfunc fixer
................
  r79139 | benjamin.peterson | 2010-03-20 11:17:37 -0500 (Sat, 20 Mar 2010) | 1 line

  wrap
................
  r79306 | benjamin.peterson | 2010-03-22 17:40:06 -0500 (Mon, 22 Mar 2010) | 21 lines

  Merged revisions 79077,79137,79304-79305 via svnmerge from
  svn+ssh://pythondev@svn.python.org/sandbox/trunk/2to3/lib2to3

  ........
    r79077 | benjamin.peterson | 2010-03-18 18:05:29 -0500 (Thu, 18 Mar 2010) | 1 line

    port detect_encoding improvements from py3k
  ........
    r79137 | benjamin.peterson | 2010-03-20 11:12:53 -0500 (Sat, 20 Mar 2010) | 1 line

    add a fixer for setting sys.exitfunc #2356
  ........
    r79304 | benjamin.peterson | 2010-03-22 17:20:22 -0500 (Mon, 22 Mar 2010) | 1 line

    fix test_parser when it's run in a path with spaces #7666
  ........
    r79305 | benjamin.peterson | 2010-03-22 17:27:07 -0500 (Mon, 22 Mar 2010) | 1 line

    normalize whitespace
  ........
................

4 files changed, 170 insertions, 10 deletions

diff --git a/Lib/lib2to3/fixes/fix_exitfunc.py b/Lib/lib2to3/fixes/fix_exitfunc.py
new file mode 100644
index 0000000..b29b464
--- /dev/null
+++ b/Lib/lib2to3/fixes/fix_exitfunc.py
@@ -0,0 +1,70 @@
+"""
+Convert use of sys.exitfunc to use the atexit module.
+"""
+
+# Author: Benjamin Peterson
+
+from lib2to3 import pytree, fixer_base
+from lib2to3.fixer_util import Name, Attr, Call, Comma, Newline, syms
+
+
+class FixExitfunc(fixer_base.BaseFix):
+
+    PATTERN = """
+              (
+                  sys_import=import_name<'import'
+                      ('sys'
+                      |
+                      dotted_as_names< (any ',')* 'sys' (',' any)* >
+                      )
+                  >
+              |
+                  expr_stmt<
+                      power< 'sys' trailer< '.' 'exitfunc' > >
+                  '=' func=any >
+              )
+              """
+
+    def __init__(self, *args):
+        super(FixExitfunc, self).__init__(*args)
+
+    def start_tree(self, tree, filename):
+        super(FixExitfunc, self).start_tree(tree, filename)
+        self.sys_import = None
+
+    def transform(self, node, results):
+        # First, find a the sys import. We'll just hope it's global scope.
+        if "sys_import" in results:
+            if self.sys_import is None:
+                self.sys_import = results["sys_import"]
+            return
+
+        func = results["func"].clone()
+        func.prefix = ""
+        register = pytree.Node(syms.power,
+                               Attr(Name("atexit"), Name("register"))
+                               )
+        call = Call(register, [func], node.prefix)
+        node.replace(call)
+
+        if self.sys_import is None:
+            # That's interesting.
+            self.warning(node, "Can't find sys import; Please add an atexit "
+                             "import at the top of your file.")
+            return
+
+        # Now add an atexit import after the sys import.
+        names = self.sys_import.children[1]
+        if names.type == syms.dotted_as_names:
+            names.append_child(Comma())
+            names.append_child(Name("atexit", " "))
+        else:
+            containing_stmt = self.sys_import.parent
+            position = containing_stmt.children.index(self.sys_import)
+            stmt_container = containing_stmt.parent
+            new_import = pytree.Node("import_name",
+                              [Name("import"), Name("atexit", " ")]
+                              )
+            new = pytree.Node("simple_stmt", [new_import])
+            containing_stmt.insert_child(position + 1, Newline())
+            containing_stmt.insert_child(position + 2, new)
diff --git a/Lib/lib2to3/pgen2/tokenize.py b/Lib/lib2to3/pgen2/tokenize.py
index e0f45f9..521e7b7 100644
--- a/Lib/lib2to3/pgen2/tokenize.py
+++ b/Lib/lib2to3/pgen2/tokenize.py
@@ -253,14 +253,16 @@ def detect_encoding(readline):
     in.
 
     It detects the encoding from the presence of a utf-8 bom or an encoding
-    cookie as specified in pep-0263. If both a bom and a cookie are present,
-    but disagree, a SyntaxError will be raised. If the encoding cookie is an
-    invalid charset, raise a SyntaxError.
+    cookie as specified in pep-0263. If both a bom and a cookie are present, but
+    disagree, a SyntaxError will be raised. If the encoding cookie is an invalid
+    charset, raise a SyntaxError.  Note that if a utf-8 bom is found,
+    'utf-8-sig' is returned.
 
     If no encoding is specified, then the default of 'utf-8' will be returned.
     """
     bom_found = False
     encoding = None
+    default = 'utf-8'
     def read_or_stop():
         try:
             return readline()
@@ -287,17 +289,16 @@ def detect_encoding(readline):
             if codec.name != 'utf-8':
                 # This behaviour mimics the Python interpreter
                 raise SyntaxError('encoding problem: utf-8')
-            else:
-                # Allow it to be properly encoded and decoded.
-                encoding = 'utf-8-sig'
+            encoding += '-sig'
         return encoding
 
     first = read_or_stop()
     if first.startswith(BOM_UTF8):
         bom_found = True
         first = first[3:]
+        default = 'utf-8-sig'
     if not first:
-        return 'utf-8', []
+        return default, []
 
     encoding = find_cookie(first)
     if encoding:
@@ -305,13 +306,13 @@ def detect_encoding(readline):
 
     second = read_or_stop()
     if not second:
-        return 'utf-8', [first]
+        return default, [first]
 
     encoding = find_cookie(second)
     if encoding:
         return encoding, [first, second]
 
-    return 'utf-8', [first, second]
+    return default, [first, second]
 
 def untokenize(iterable):
     """Transform tokens back into Python source code.
diff --git a/Lib/lib2to3/tests/test_fixers.py b/Lib/lib2to3/tests/test_fixers.py
index 417ed2f..4b636f8 100644
--- a/Lib/lib2to3/tests/test_fixers.py
+++ b/Lib/lib2to3/tests/test_fixers.py
@@ -4285,3 +4285,91 @@ class Test_operator(FixerTestCase):
     def test_bare_sequenceIncludes(self):
         s = "sequenceIncludes(x, y)"
         self.warns_unchanged(s, "You should use operator.contains here.")
+
+
+class Test_exitfunc(FixerTestCase):
+
+    fixer = "exitfunc"
+
+    def test_simple(self):
+        b = """
+            import sys
+            sys.exitfunc = my_atexit
+            """
+        a = """
+            import sys
+            import atexit
+            atexit.register(my_atexit)
+            """
+        self.check(b, a)
+
+    def test_names_import(self):
+        b = """
+            import sys, crumbs
+            sys.exitfunc = my_func
+            """
+        a = """
+            import sys, crumbs, atexit
+            atexit.register(my_func)
+            """
+        self.check(b, a)
+
+    def test_complex_expression(self):
+        b = """
+            import sys
+            sys.exitfunc = do(d)/a()+complex(f=23, g=23)*expression
+            """
+        a = """
+            import sys
+            import atexit
+            atexit.register(do(d)/a()+complex(f=23, g=23)*expression)
+            """
+        self.check(b, a)
+
+    def test_comments(self):
+        b = """
+            import sys # Foo
+            sys.exitfunc = f # Blah
+            """
+        a = """
+            import sys
+            import atexit # Foo
+            atexit.register(f) # Blah
+            """
+        self.check(b, a)
+
+        b = """
+            import apples, sys, crumbs, larry # Pleasant comments
+            sys.exitfunc = func
+            """
+        a = """
+            import apples, sys, crumbs, larry, atexit # Pleasant comments
+            atexit.register(func)
+            """
+        self.check(b, a)
+
+    def test_in_a_function(self):
+        b = """
+            import sys
+            def f():
+                sys.exitfunc = func
+            """
+        a = """
+            import sys
+            import atexit
+            def f():
+                atexit.register(func)
+             """
+        self.check(b, a)
+
+    def test_no_sys_import(self):
+        b = """sys.exitfunc = f"""
+        a = """atexit.register(f)"""
+        msg = ("Can't find sys import; Please add an atexit import at the "
+            "top of your file.")
+        self.warns(b, a, msg)
+
+
+    def test_unchanged(self):
+        s = """f(sys.exitfunc)"""
+        self.unchanged(s)
diff --git a/Lib/lib2to3/tests/test_parser.py b/Lib/lib2to3/tests/test_parser.py
index 92fe522..6d0c51b 100644
--- a/Lib/lib2to3/tests/test_parser.py
+++ b/Lib/lib2to3/tests/test_parser.py
@@ -209,6 +209,7 @@ def diff(fn, result, encoding):
     finally:
         f.close()
     try:
-        return os.system("diff -u %r @" % fn)
+        fn = fn.replace('"', '\\"')
+        return os.system('diff -u "%s" @' % fn)
     finally:
         os.remove("@")