Fix issue 2522. locale.format now checks that it is passed

exactly one pattern, which avoids mysterious errors where it
had seemed to fail to do localization.

2 files changed, 29 insertions, 8 deletions

diff --git a/Lib/locale.py b/Lib/locale.py
index 163b044..777bb03 100644
--- a/Lib/locale.py
+++ b/Lib/locale.py
@@ -11,7 +11,11 @@
 
 """
 
-import sys, encodings, encodings.aliases
+import sys
+import encodings
+import encodings.aliases
+import re
+import operator
 import functools
 
 # Try importing the _locale module.
@@ -166,6 +170,9 @@ def _strip_padding(s, amount):
         amount -= 1
     return s[lpos:rpos+1]
 
+_percent_re = re.compile(r'%(?:\((?P<key>.*?)\))?'
+                         r'(?P<modifiers>[-#0-9 +*.hlL]*?)[eEfFgGdiouxXcrs%]')
+
 def format(percent, value, grouping=False, monetary=False, *additional):
     """Returns the locale-aware substitution of a %? specifier
     (percent).
@@ -173,9 +180,13 @@ def format(percent, value, grouping=False, monetary=False, *additional):
     additional is for format strings which contain one or more
     '*' modifiers."""
     # this is only for one-percent-specifier strings and this should be checked
-    if percent[0] != '%':
-        raise ValueError("format() must be given exactly one %char "
-                         "format specifier")
+    match = _percent_re.match(percent)
+    if not match or len(match.group())!= len(percent):
+        raise ValueError(("format() must be given exactly one %%char "
+                         "format specifier, %s not valid") % repr(percent))
+    return _format(percent, value, grouping, monetary, *additional)
+
+def _format(percent, value, grouping=False, monetary=False, *additional):
     if additional:
         formatted = percent % ((value,) + additional)
     else:
@@ -199,10 +210,6 @@ def format(percent, value, grouping=False, monetary=False, *additional):
             formatted = _strip_padding(formatted, seps)
     return formatted
 
-import re, operator
-_percent_re = re.compile(r'%(?:\((?P<key>.*?)\))?'
-                         r'(?P<modifiers>[-#0-9 +*.hlL]*?)[eEfFgGdiouxXcrs%]')
-
 def format_string(f, val, grouping=False):
     """Formats a string in the same way that the % formatting would use,
     but takes the current locale into account.
diff --git a/Lib/test/test_locale.py b/Lib/test/test_locale.py
index 180f403..d9d3cd9 100644
--- a/Lib/test/test_locale.py
+++ b/Lib/test/test_locale.py
@@ -221,6 +221,19 @@ class EnUSNumberFormatting(BaseFormattingTest):
                 (self.sep, self.sep))
 
 
+class TestFormatPatternArg(unittest.TestCase):
+    # Test handling of pattern argument of format
+
+    def test_onlyOnePattern(self):
+        # Issue 2522: accept exactly one % pattern, and no extra chars.
+        self.assertRaises(ValueError, locale.format, "%f\n", 'foo')
+        self.assertRaises(ValueError, locale.format, "%f\r", 'foo')
+        self.assertRaises(ValueError, locale.format, "%f\r\n", 'foo')
+        self.assertRaises(ValueError, locale.format, " %f", 'foo')
+        self.assertRaises(ValueError, locale.format, "%fg", 'foo')
+        self.assertRaises(ValueError, locale.format, "%^g", 'foo')
+
+
 class TestNumberFormatting(BaseLocalizedTest, EnUSNumberFormatting):
     # Test number formatting with a real English locale.
 
@@ -351,6 +364,7 @@ class TestMiscellaneous(unittest.TestCase):
 def test_main():
     tests = [
         TestMiscellaneous,
+        TestFormatPatternArg,
         TestEnUSNumberFormatting,
         TestCNumberFormatting,
         TestFrFRNumberFormatting,