bpo-31672: Fix string.Template accidentally matched non-ASCII identifiers (GH-3872)

Pattern `[a-z]` with `IGNORECASE` flag can match to some non-ASCII characters.

Straightforward solution for this is using `IGNORECASE | ASCII` flag.
But users may subclass `Template` and override only `idpattern`. So we want to
avoid changing `Template.flags`.

So this commit uses local flag `-i` for `idpattern` and change `[a-z]` to `[a-zA-Z]`.
(cherry picked from commit b22273ec5d1992b0cbe078b887427ae9977dfb78)

diff --git a/Doc/library/string.rst b/Doc/library/string.rst
index a0977b6461339429952926be9a2054a2c7440a88..7a9fcc38bbded80ab5c6750e6f44317c4e44ee95 100644 (file)
--- a/Doc/library/string.rst
+++ b/Doc/library/string.rst
@@ -746,8 +746,18 @@ to parse template strings.  To do this, you can override these class attributes:
 
 * *idpattern* -- This is the regular expression describing the pattern for
   non-braced placeholders (the braces will be added automatically as
-  appropriate).  The default value is the regular expression
-  ``[_a-z][_a-z0-9]*``.
+  appropriate). The default value is the regular expression
+  ``(?-i:[_a-zA-Z][_a-zA-Z0-9]*)``.
+
+  .. note::
+
+     Since default *flags* is ``re.IGNORECASE``, pattern ``[a-z]`` can match
+     with some non-ASCII characters. That's why we use local ``-i`` flag here.
+
+     While *flags* is kept to ``re.IGNORECASE`` for backward compatibility,
+     you can override it to ``0`` or ``re.IGNORECASE | re.ASCII`` when
+     subclassing.
+
 
 * *flags* -- The regular expression flags that will be applied when compiling
   the regular expression used for recognizing substitutions.  The default value

diff --git a/Lib/string.py b/Lib/string.py
index c9020076437e999c23829caec193b4b9aa7bfb88..670c1951a8a82070425a5754c4666ed4d9baa921 100644 (file)
--- a/Lib/string.py
+++ b/Lib/string.py
@@ -78,7 +78,11 @@ class Template(metaclass=_TemplateMetaclass):
     """A string class for supporting $-substitutions."""
 
     delimiter = '$'
-    idpattern = r'[_a-z][_a-z0-9]*'
+    # r'[a-z]' matches to non-ASCII letters when used with IGNORECASE,
+    # but without ASCII flag.  We can't add re.ASCII to flags because of
+    # backward compatibility.  So we use local -i flag and [a-zA-Z] pattern.
+    # See https://bugs.python.org/issue31672
+    idpattern = r'(?-i:[_a-zA-Z][_a-zA-Z0-9]*)'
     flags = _re.IGNORECASE
 
     def __init__(self, template):

diff --git a/Lib/test/test_string.py b/Lib/test/test_string.py
index 70439f85c89555ef08dca3d7c566f896f2b967cb..8db23e76c1c6ad5a2e7b61b2491e0f2736898cf7 100644 (file)
--- a/Lib/test/test_string.py
+++ b/Lib/test/test_string.py
@@ -271,6 +271,12 @@ class TestTemplate(unittest.TestCase):
         raises(ValueError, s.substitute, dict(who='tim'))
         s = Template('$who likes $100')
         raises(ValueError, s.substitute, dict(who='tim'))
+        # Template.idpattern should match to only ASCII characters.
+        # https://bugs.python.org/issue31672
+        s = Template("$who likes $\u0131")  # (DOTLESS I)
+        raises(ValueError, s.substitute, dict(who='tim'))
+        s = Template("$who likes $\u0130")  # (LATIN CAPITAL LETTER I WITH DOT ABOVE)
+        raises(ValueError, s.substitute, dict(who='tim'))
 
     def test_idpattern_override(self):
         class PathPattern(Template):

diff --git a/Misc/NEWS.d/next/Library/2017-10-12-02-47-16.bpo-31672.DaOkVd.rst b/Misc/NEWS.d/next/Library/2017-10-12-02-47-16.bpo-31672.DaOkVd.rst
new file mode 100644 (file)
index 0000000..b8de1f3
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2017-10-12-02-47-16.bpo-31672.DaOkVd.rst
@@ -0,0 +1,2 @@
+``idpattern`` in ``string.Template`` matched some non-ASCII characters. Now
+it uses ``-i`` regular expression local flag to avoid non-ASCII characters.