Properly handle unicode format strings for Python 2. Closes issue #296.

5 files changed, 53 insertions, 6 deletions

diff --git a/ChangeLog b/ChangeLog
index 26f417e..6ae9ce5 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -33,6 +33,9 @@ ChangeLog for Pylint
     * Don't emit 'attribute-defined-outside-init' if the attribute
       was set by a function call in a defining method. Closes issue #192.
 
+    * Properly handle unicode format strings for Python 2.
+      Closes issue #296.
+
 2014-07-26  --  1.3.0
 
     * Allow hanging continued indentation for implicitly concatenated
diff --git a/checkers/strings.py b/checkers/strings.py
index 714a1a2..5efbc90 100644
--- a/checkers/strings.py
+++ b/checkers/strings.py
@@ -21,6 +21,7 @@
 import sys
 import tokenize
 import string
+import numbers
 
 import astroid
 
@@ -122,7 +123,7 @@ if _PY3K:
 else:
     def _field_iterator_convertor(iterator):
         for is_attr, key in iterator:
-            if not isinstance(key, str):
+            if isinstance(key, numbers.Number):
                 yield is_attr, int(key)
             else:
                 yield is_attr, key
@@ -174,7 +175,7 @@ def parse_format_method_string(format_string):
             manual_pos_arg += 1
         elif name:
             keyname, fielditerator = split_format_field_names(name)
-            if not isinstance(keyname, str):
+            if isinstance(keyname, numbers.Number):
                 # In Python 2 it will return long which will lead
                 # to different output between 2 and 3
                 keyname = int(keyname)
@@ -358,9 +359,9 @@ class StringMethodsChecker(BaseChecker):
             return
 
         manual_fields = set(field[0] for field in fields
-                            if isinstance(field[0], int))
+                            if isinstance(field[0], numbers.Number))
         named_fields = set(field[0] for field in fields
-                           if isinstance(field[0], str))
+                           if isinstance(field[0], basestring))
         if num_args and manual_pos:
             self.add_message('format-combined-specification',
                              node=node)
@@ -384,7 +385,9 @@ class StringMethodsChecker(BaseChecker):
             # num_args can be 0 if manual_pos is not.
             num_args = num_args or manual_pos
             if positional or num_args:
-                if named or any(True for field in named_fields if field == ''):
+                empty = any(True for field in named_fields
+                            if field == '')
+                if named or empty:
                     # Verify the required number of positional arguments
                     # only if the .format got at least one keyword argument.
                     # This means that the format strings accepts both
@@ -421,7 +424,7 @@ class StringMethodsChecker(BaseChecker):
                 # to 0. It will not be present in `named`, so use the value
                 # 0 for it.
                 key = 0
-            if isinstance(key, int):
+            if isinstance(key, numbers.Number):
                 try:
                     argname = utils.get_argument_from_call(node, key)
                 except utils.NoSuchArgumentError:
diff --git a/test/functional/string_formatting_py27.py b/test/functional/string_formatting_py27.py
new file mode 100644
index 0000000..d6aeae7
--- /dev/null
+++ b/test/functional/string_formatting_py27.py
@@ -0,0 +1,23 @@
+"""test for Python 2 string formatting error

+"""

+from __future__ import unicode_literals

+# pylint: disable=line-too-long

+__revision__ = 1

+

+def pprint_bad():

+    """Test string format """

+    "{{}}".format(1) # [too-many-format-args]

+    "{} {".format() # [bad-format-string]

+    "{} }".format() # [bad-format-string]

+    "{0} {}".format(1, 2) # [format-combined-specification]

+    # +1: [missing-format-argument-key, unused-format-string-argument]

+    "{a} {b}".format(a=1, c=2)

+    "{} {a}".format(1, 2) # [missing-format-argument-key]

+    "{} {}".format(1) # [too-few-format-args]

+    "{} {}".format(1, 2, 3) # [too-many-format-args]

+    # +1: [missing-format-argument-key,missing-format-argument-key,missing-format-argument-key]

+    "{a} {b} {c}".format()

+    "{} {}".format(a=1, b=2) # [too-few-format-args]

+    # +1: [missing-format-argument-key, missing-format-argument-key]

+    "{a} {b}".format(1, 2)

+

diff --git a/test/functional/string_formatting_py27.rc b/test/functional/string_formatting_py27.rc
new file mode 100644
index 0000000..80170b7
--- /dev/null
+++ b/test/functional/string_formatting_py27.rc
@@ -0,0 +1,3 @@
+[testoptions]
+min_pyver=2.7

+max_pyver=3.0
\ No newline at end of file
diff --git a/test/functional/string_formatting_py27.txt b/test/functional/string_formatting_py27.txt
new file mode 100644
index 0000000..47f21de
--- /dev/null
+++ b/test/functional/string_formatting_py27.txt
@@ -0,0 +1,15 @@
+too-many-format-args:9:pprint_bad:Too many arguments for format string
+bad-format-string:10:pprint_bad:Invalid format string
+bad-format-string:11:pprint_bad:Invalid format string
+format-combined-specification:12:pprint_bad:Format string contains both automatic field numbering and manual field specification
+missing-format-argument-key:14:pprint_bad:Missing keyword argument u'b' for format string
+unused-format-string-argument:14:pprint_bad:Unused format argument 'c'
+missing-format-argument-key:15:pprint_bad:Missing keyword argument u'a' for format string
+too-few-format-args:16:pprint_bad:Not enough arguments for format string
+too-many-format-args:17:pprint_bad:Too many arguments for format string
+missing-format-argument-key:19:pprint_bad:Missing keyword argument u'a' for format string
+missing-format-argument-key:19:pprint_bad:Missing keyword argument u'b' for format string
+missing-format-argument-key:19:pprint_bad:Missing keyword argument u'c' for format string
+too-few-format-args:20:pprint_bad:Not enough arguments for format string
+missing-format-argument-key:22:pprint_bad:Missing keyword argument u'a' for format string
+missing-format-argument-key:22:pprint_bad:Missing keyword argument u'b' for format string