tomwalters@0: #!/usr/bin/python
tomwalters@0: #
tomwalters@0: # Copyright (c) 2009 Google Inc. All rights reserved.
tomwalters@0: #
tomwalters@0: # Redistribution and use in source and binary forms, with or without
tomwalters@0: # modification, are permitted provided that the following conditions are
tomwalters@0: # met:
tomwalters@0: #
tomwalters@0: #    * Redistributions of source code must retain the above copyright
tomwalters@0: # notice, this list of conditions and the following disclaimer.
tomwalters@0: #    * Redistributions in binary form must reproduce the above
tomwalters@0: # copyright notice, this list of conditions and the following disclaimer
tomwalters@0: # in the documentation and/or other materials provided with the
tomwalters@0: # distribution.
tomwalters@0: #    * Neither the name of Google Inc. nor the names of its
tomwalters@0: # contributors may be used to endorse or promote products derived from
tomwalters@0: # this software without specific prior written permission.
tomwalters@0: #
tomwalters@0: # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
tomwalters@0: # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
tomwalters@0: # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
tomwalters@0: # A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
tomwalters@0: # OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
tomwalters@0: # SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
tomwalters@0: # LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
tomwalters@0: # DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
tomwalters@0: # THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
tomwalters@0: # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
tomwalters@0: # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
tomwalters@0: 
tomwalters@0: # Here are some issues that I've had people identify in my code during reviews,
tomwalters@0: # that I think are possible to flag automatically in a lint tool.  If these were
tomwalters@0: # caught by lint, it would save time both for myself and that of my reviewers.
tomwalters@0: # Most likely, some of these are beyond the scope of the current lint framework,
tomwalters@0: # but I think it is valuable to retain these wish-list items even if they cannot
tomwalters@0: # be immediately implemented.
tomwalters@0: #
tomwalters@0: #  Suggestions
tomwalters@0: #  -----------
tomwalters@0: #  - Check for no 'explicit' for multi-arg ctor
tomwalters@0: #  - Check for boolean assign RHS in parens
tomwalters@0: #  - Check for ctor initializer-list colon position and spacing
tomwalters@0: #  - Check that if there's a ctor, there should be a dtor
tomwalters@0: #  - Check accessors that return non-pointer member variables are
tomwalters@0: #    declared const
tomwalters@0: #  - Check accessors that return non-const pointer member vars are
tomwalters@0: #    *not* declared const
tomwalters@0: #  - Check for using public includes for testing
tomwalters@0: #  - Check for spaces between brackets in one-line inline method
tomwalters@0: #  - Check for no assert()
tomwalters@0: #  - Check for spaces surrounding operators
tomwalters@0: #  - Check for 0 in pointer context (should be NULL)
tomwalters@0: #  - Check for 0 in char context (should be '\0')
tomwalters@0: #  - Check for camel-case method name conventions for methods
tomwalters@0: #    that are not simple inline getters and setters
tomwalters@0: #  - Check that base classes have virtual destructors
tomwalters@0: #    put "  // namespace" after } that closes a namespace, with
tomwalters@0: #    namespace's name after 'namespace' if it is named.
tomwalters@0: #  - Do not indent namespace contents
tomwalters@0: #  - Avoid inlining non-trivial constructors in header files
tomwalters@0: #    include base/basictypes.h if DISALLOW_EVIL_CONSTRUCTORS is used
tomwalters@0: #  - Check for old-school (void) cast for call-sites of functions
tomwalters@0: #    ignored return value
tomwalters@0: #  - Check gUnit usage of anonymous namespace
tomwalters@0: #  - Check for class declaration order (typedefs, consts, enums,
tomwalters@0: #    ctor(s?), dtor, friend declarations, methods, member vars)
tomwalters@0: #
tomwalters@0: 
tomwalters@0: """Does google-lint on c++ files.
tomwalters@0: 
tomwalters@0: The goal of this script is to identify places in the code that *may*
tomwalters@0: be in non-compliance with google style.  It does not attempt to fix
tomwalters@0: up these problems -- the point is to educate.  It does also not
tomwalters@0: attempt to find all problems, or to ensure that everything it does
tomwalters@0: find is legitimately a problem.
tomwalters@0: 
tomwalters@0: In particular, we can get very confused by /* and // inside strings!
tomwalters@0: We do a small hack, which is to ignore //'s with "'s after them on the
tomwalters@0: same line, but it is far from perfect (in either direction).
tomwalters@0: """
tomwalters@0: 
tomwalters@0: import codecs
tomwalters@0: import getopt
tomwalters@0: import math  # for log
tomwalters@0: import os
tomwalters@0: import re
tomwalters@0: import sre_compile
tomwalters@0: import string
tomwalters@0: import sys
tomwalters@0: import unicodedata
tomwalters@0: 
tomwalters@0: 
tomwalters@0: _USAGE = """
tomwalters@0: Syntax: cpplint.py [--verbose=#] [--output=vs7] [--filter=-x,+y,...]
tomwalters@0:                    [--counting=total|toplevel|detailed]
tomwalters@0:         <file> [file] ...
tomwalters@0: 
tomwalters@0:   The style guidelines this tries to follow are those in
tomwalters@0:     http://google-styleguide.googlecode.com/svn/trunk/cppguide.xml
tomwalters@0: 
tomwalters@0:   Every problem is given a confidence score from 1-5, with 5 meaning we are
tomwalters@0:   certain of the problem, and 1 meaning it could be a legitimate construct.
tomwalters@0:   This will miss some errors, and is not a substitute for a code review.
tomwalters@0: 
tomwalters@0:   To prevent specific lines from being linted, add a '// NOLINT' comment to the
tomwalters@0:   end of the line.
tomwalters@0: 
tomwalters@0:   The files passed in will be linted; at least one file must be provided.
tomwalters@0:   Linted extensions are .cc, .cpp, and .h.  Other file types will be ignored.
tomwalters@0: 
tomwalters@0:   Flags:
tomwalters@0: 
tomwalters@0:     output=vs7
tomwalters@0:       By default, the output is formatted to ease emacs parsing.  Visual Studio
tomwalters@0:       compatible output (vs7) may also be used.  Other formats are unsupported.
tomwalters@0: 
tomwalters@0:     verbose=#
tomwalters@0:       Specify a number 0-5 to restrict errors to certain verbosity levels.
tomwalters@0: 
tomwalters@0:     filter=-x,+y,...
tomwalters@0:       Specify a comma-separated list of category-filters to apply: only
tomwalters@0:       error messages whose category names pass the filters will be printed.
tomwalters@0:       (Category names are printed with the message and look like
tomwalters@0:       "[whitespace/indent]".)  Filters are evaluated left to right.
tomwalters@0:       "-FOO" and "FOO" means "do not print categories that start with FOO".
tomwalters@0:       "+FOO" means "do print categories that start with FOO".
tomwalters@0: 
tomwalters@0:       Examples: --filter=-whitespace,+whitespace/braces
tomwalters@0:                 --filter=whitespace,runtime/printf,+runtime/printf_format
tomwalters@0:                 --filter=-,+build/include_what_you_use
tomwalters@0: 
tomwalters@0:       To see a list of all the categories used in cpplint, pass no arg:
tomwalters@0:          --filter=
tomwalters@0: 
tomwalters@0:     counting=total|toplevel|detailed
tomwalters@0:       The total number of errors found is always printed. If
tomwalters@0:       'toplevel' is provided, then the count of errors in each of
tomwalters@0:       the top-level categories like 'build' and 'whitespace' will
tomwalters@0:       also be printed. If 'detailed' is provided, then a count
tomwalters@0:       is provided for each category like 'build/class'.
tomwalters@0: """
tomwalters@0: 
tomwalters@0: # We categorize each error message we print.  Here are the categories.
tomwalters@0: # We want an explicit list so we can list them all in cpplint --filter=.
tomwalters@0: # If you add a new error message with a new category, add it to the list
tomwalters@0: # here!  cpplint_unittest.py should tell you if you forget to do this.
tomwalters@0: # \ used for clearer layout -- pylint: disable-msg=C6013
tomwalters@0: _ERROR_CATEGORIES = '''\
tomwalters@0:   build/class
tomwalters@0:   build/deprecated
tomwalters@0:   build/endif_comment
tomwalters@0:   build/forward_decl
tomwalters@0:   build/header_guard
tomwalters@0:   build/include
tomwalters@0:   build/include_alpha
tomwalters@0:   build/include_order
tomwalters@0:   build/include_what_you_use
tomwalters@0:   build/namespaces
tomwalters@0:   build/printf_format
tomwalters@0:   build/storage_class
tomwalters@0:   legal/copyright
tomwalters@0:   readability/braces
tomwalters@0:   readability/casting
tomwalters@0:   readability/check
tomwalters@0:   readability/constructors
tomwalters@0:   readability/fn_size
tomwalters@0:   readability/function
tomwalters@0:   readability/multiline_comment
tomwalters@0:   readability/multiline_string
tomwalters@0:   readability/streams
tomwalters@0:   readability/todo
tomwalters@0:   readability/utf8
tomwalters@0:   runtime/arrays
tomwalters@0:   runtime/casting
tomwalters@0:   runtime/explicit
tomwalters@0:   runtime/int
tomwalters@0:   runtime/init
tomwalters@0:   runtime/invalid_increment
tomwalters@0:   runtime/member_string_references
tomwalters@0:   runtime/memset
tomwalters@0:   runtime/operator
tomwalters@0:   runtime/printf
tomwalters@0:   runtime/printf_format
tomwalters@0:   runtime/references
tomwalters@0:   runtime/rtti
tomwalters@0:   runtime/sizeof
tomwalters@0:   runtime/string
tomwalters@0:   runtime/threadsafe_fn
tomwalters@0:   runtime/virtual
tomwalters@0:   whitespace/blank_line
tomwalters@0:   whitespace/braces
tomwalters@0:   whitespace/comma
tomwalters@0:   whitespace/comments
tomwalters@0:   whitespace/end_of_line
tomwalters@0:   whitespace/ending_newline
tomwalters@0:   whitespace/indent
tomwalters@0:   whitespace/labels
tomwalters@0:   whitespace/line_length
tomwalters@0:   whitespace/newline
tomwalters@0:   whitespace/operators
tomwalters@0:   whitespace/parens
tomwalters@0:   whitespace/semicolon
tomwalters@0:   whitespace/tab
tomwalters@0:   whitespace/todo
tomwalters@0: '''
tomwalters@0: 
tomwalters@0: # The default state of the category filter. This is overrided by the --filter=
tomwalters@0: # flag. By default all errors are on, so only add here categories that should be
tomwalters@0: # off by default (i.e., categories that must be enabled by the --filter= flags).
tomwalters@0: # All entries here should start with a '-' or '+', as in the --filter= flag.
tomwalters@0: _DEFAULT_FILTERS = [ '-build/include_alpha' ]
tomwalters@0: 
tomwalters@0: # We used to check for high-bit characters, but after much discussion we
tomwalters@0: # decided those were OK, as long as they were in UTF-8 and didn't represent
tomwalters@0: # hard-coded international strings, which belong in a seperate i18n file.
tomwalters@0: 
tomwalters@0: # Headers that we consider STL headers.
tomwalters@0: _STL_HEADERS = frozenset([
tomwalters@0:     'algobase.h', 'algorithm', 'alloc.h', 'bitset', 'deque', 'exception',
tomwalters@0:     'function.h', 'functional', 'hash_map', 'hash_map.h', 'hash_set',
tomwalters@0:     'hash_set.h', 'iterator', 'list', 'list.h', 'map', 'memory', 'pair.h',
tomwalters@0:     'pthread_alloc', 'queue', 'set', 'set.h', 'sstream', 'stack',
tomwalters@0:     'stl_alloc.h', 'stl_relops.h', 'type_traits.h',
tomwalters@0:     'utility', 'vector', 'vector.h',
tomwalters@0:     ])
tomwalters@0: 
tomwalters@0: 
tomwalters@0: # Non-STL C++ system headers.
tomwalters@0: _CPP_HEADERS = frozenset([
tomwalters@0:     'algo.h', 'builtinbuf.h', 'bvector.h', 'cassert', 'cctype',
tomwalters@0:     'cerrno', 'cfloat', 'ciso646', 'climits', 'clocale', 'cmath',
tomwalters@0:     'complex', 'complex.h', 'csetjmp', 'csignal', 'cstdarg', 'cstddef',
tomwalters@0:     'cstdio', 'cstdlib', 'cstring', 'ctime', 'cwchar', 'cwctype',
tomwalters@0:     'defalloc.h', 'deque.h', 'editbuf.h', 'exception', 'fstream',
tomwalters@0:     'fstream.h', 'hashtable.h', 'heap.h', 'indstream.h', 'iomanip',
tomwalters@0:     'iomanip.h', 'ios', 'iosfwd', 'iostream', 'iostream.h', 'istream.h',
tomwalters@0:     'iterator.h', 'limits', 'map.h', 'multimap.h', 'multiset.h',
tomwalters@0:     'numeric', 'ostream.h', 'parsestream.h', 'pfstream.h', 'PlotFile.h',
tomwalters@0:     'procbuf.h', 'pthread_alloc.h', 'rope', 'rope.h', 'ropeimpl.h',
tomwalters@0:     'SFile.h', 'slist', 'slist.h', 'stack.h', 'stdexcept',
tomwalters@0:     'stdiostream.h', 'streambuf.h', 'stream.h', 'strfile.h', 'string',
tomwalters@0:     'strstream', 'strstream.h', 'tempbuf.h', 'tree.h', 'typeinfo', 'valarray',
tomwalters@0:     ])
tomwalters@0: 
tomwalters@0: 
tomwalters@0: # Assertion macros.  These are defined in base/logging.h and
tomwalters@0: # testing/base/gunit.h.  Note that the _M versions need to come first
tomwalters@0: # for substring matching to work.
tomwalters@0: _CHECK_MACROS = [
tomwalters@0:     'DCHECK', 'CHECK',
tomwalters@0:     'EXPECT_TRUE_M', 'EXPECT_TRUE',
tomwalters@0:     'ASSERT_TRUE_M', 'ASSERT_TRUE',
tomwalters@0:     'EXPECT_FALSE_M', 'EXPECT_FALSE',
tomwalters@0:     'ASSERT_FALSE_M', 'ASSERT_FALSE',
tomwalters@0:     ]
tomwalters@0: 
tomwalters@0: # Replacement macros for CHECK/DCHECK/EXPECT_TRUE/EXPECT_FALSE
tomwalters@0: _CHECK_REPLACEMENT = dict([(m, {}) for m in _CHECK_MACROS])
tomwalters@0: 
tomwalters@0: for op, replacement in [('==', 'EQ'), ('!=', 'NE'),
tomwalters@0:                         ('>=', 'GE'), ('>', 'GT'),
tomwalters@0:                         ('<=', 'LE'), ('<', 'LT')]:
tomwalters@0:   _CHECK_REPLACEMENT['DCHECK'][op] = 'DCHECK_%s' % replacement
tomwalters@0:   _CHECK_REPLACEMENT['CHECK'][op] = 'CHECK_%s' % replacement
tomwalters@0:   _CHECK_REPLACEMENT['EXPECT_TRUE'][op] = 'EXPECT_%s' % replacement
tomwalters@0:   _CHECK_REPLACEMENT['ASSERT_TRUE'][op] = 'ASSERT_%s' % replacement
tomwalters@0:   _CHECK_REPLACEMENT['EXPECT_TRUE_M'][op] = 'EXPECT_%s_M' % replacement
tomwalters@0:   _CHECK_REPLACEMENT['ASSERT_TRUE_M'][op] = 'ASSERT_%s_M' % replacement
tomwalters@0: 
tomwalters@0: for op, inv_replacement in [('==', 'NE'), ('!=', 'EQ'),
tomwalters@0:                             ('>=', 'LT'), ('>', 'LE'),
tomwalters@0:                             ('<=', 'GT'), ('<', 'GE')]:
tomwalters@0:   _CHECK_REPLACEMENT['EXPECT_FALSE'][op] = 'EXPECT_%s' % inv_replacement
tomwalters@0:   _CHECK_REPLACEMENT['ASSERT_FALSE'][op] = 'ASSERT_%s' % inv_replacement
tomwalters@0:   _CHECK_REPLACEMENT['EXPECT_FALSE_M'][op] = 'EXPECT_%s_M' % inv_replacement
tomwalters@0:   _CHECK_REPLACEMENT['ASSERT_FALSE_M'][op] = 'ASSERT_%s_M' % inv_replacement
tomwalters@0: 
tomwalters@0: 
tomwalters@0: # These constants define types of headers for use with
tomwalters@0: # _IncludeState.CheckNextIncludeOrder().
tomwalters@0: _C_SYS_HEADER = 1
tomwalters@0: _CPP_SYS_HEADER = 2
tomwalters@0: _LIKELY_MY_HEADER = 3
tomwalters@0: _POSSIBLE_MY_HEADER = 4
tomwalters@0: _OTHER_HEADER = 5
tomwalters@0: 
tomwalters@0: 
tomwalters@0: _regexp_compile_cache = {}
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def Match(pattern, s):
tomwalters@0:   """Matches the string with the pattern, caching the compiled regexp."""
tomwalters@0:   # The regexp compilation caching is inlined in both Match and Search for
tomwalters@0:   # performance reasons; factoring it out into a separate function turns out
tomwalters@0:   # to be noticeably expensive.
tomwalters@0:   if not pattern in _regexp_compile_cache:
tomwalters@0:     _regexp_compile_cache[pattern] = sre_compile.compile(pattern)
tomwalters@0:   return _regexp_compile_cache[pattern].match(s)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def Search(pattern, s):
tomwalters@0:   """Searches the string for the pattern, caching the compiled regexp."""
tomwalters@0:   if not pattern in _regexp_compile_cache:
tomwalters@0:     _regexp_compile_cache[pattern] = sre_compile.compile(pattern)
tomwalters@0:   return _regexp_compile_cache[pattern].search(s)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _IncludeState(dict):
tomwalters@0:   """Tracks line numbers for includes, and the order in which includes appear.
tomwalters@0: 
tomwalters@0:   As a dict, an _IncludeState object serves as a mapping between include
tomwalters@0:   filename and line number on which that file was included.
tomwalters@0: 
tomwalters@0:   Call CheckNextIncludeOrder() once for each header in the file, passing
tomwalters@0:   in the type constants defined above. Calls in an illegal order will
tomwalters@0:   raise an _IncludeError with an appropriate error message.
tomwalters@0: 
tomwalters@0:   """
tomwalters@0:   # self._section will move monotonically through this set. If it ever
tomwalters@0:   # needs to move backwards, CheckNextIncludeOrder will raise an error.
tomwalters@0:   _INITIAL_SECTION = 0
tomwalters@0:   _MY_H_SECTION = 1
tomwalters@0:   _C_SECTION = 2
tomwalters@0:   _CPP_SECTION = 3
tomwalters@0:   _OTHER_H_SECTION = 4
tomwalters@0: 
tomwalters@0:   _TYPE_NAMES = {
tomwalters@0:       _C_SYS_HEADER: 'C system header',
tomwalters@0:       _CPP_SYS_HEADER: 'C++ system header',
tomwalters@0:       _LIKELY_MY_HEADER: 'header this file implements',
tomwalters@0:       _POSSIBLE_MY_HEADER: 'header this file may implement',
tomwalters@0:       _OTHER_HEADER: 'other header',
tomwalters@0:       }
tomwalters@0:   _SECTION_NAMES = {
tomwalters@0:       _INITIAL_SECTION: "... nothing. (This can't be an error.)",
tomwalters@0:       _MY_H_SECTION: 'a header this file implements',
tomwalters@0:       _C_SECTION: 'C system header',
tomwalters@0:       _CPP_SECTION: 'C++ system header',
tomwalters@0:       _OTHER_H_SECTION: 'other header',
tomwalters@0:       }
tomwalters@0: 
tomwalters@0:   def __init__(self):
tomwalters@0:     dict.__init__(self)
tomwalters@0:     # The name of the current section.
tomwalters@0:     self._section = self._INITIAL_SECTION
tomwalters@0:     # The path of last found header.
tomwalters@0:     self._last_header = ''
tomwalters@0: 
tomwalters@0:   def CanonicalizeAlphabeticalOrder(self, header_path):
tomwalters@0:     """Returns a path canonicalized for alphabetical comparisson.
tomwalters@0: 
tomwalters@0:     - replaces "-" with "_" so they both cmp the same.
tomwalters@0:     - removes '-inl' since we don't require them to be after the main header.
tomwalters@0:     - lowercase everything, just in case.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       header_path: Path to be canonicalized.
tomwalters@0: 
tomwalters@0:     Returns:
tomwalters@0:       Canonicalized path.
tomwalters@0:     """
tomwalters@0:     return header_path.replace('-inl.h', '.h').replace('-', '_').lower()
tomwalters@0: 
tomwalters@0:   def IsInAlphabeticalOrder(self, header_path):
tomwalters@0:     """Check if a header is in alphabetical order with the previous header.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       header_path: Header to be checked.
tomwalters@0: 
tomwalters@0:     Returns:
tomwalters@0:       Returns true if the header is in alphabetical order.
tomwalters@0:     """
tomwalters@0:     canonical_header = self.CanonicalizeAlphabeticalOrder(header_path)
tomwalters@0:     if self._last_header > canonical_header:
tomwalters@0:       return False
tomwalters@0:     self._last_header = canonical_header
tomwalters@0:     return True
tomwalters@0: 
tomwalters@0:   def CheckNextIncludeOrder(self, header_type):
tomwalters@0:     """Returns a non-empty error message if the next header is out of order.
tomwalters@0: 
tomwalters@0:     This function also updates the internal state to be ready to check
tomwalters@0:     the next include.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       header_type: One of the _XXX_HEADER constants defined above.
tomwalters@0: 
tomwalters@0:     Returns:
tomwalters@0:       The empty string if the header is in the right order, or an
tomwalters@0:       error message describing what's wrong.
tomwalters@0: 
tomwalters@0:     """
tomwalters@0:     error_message = ('Found %s after %s' %
tomwalters@0:                      (self._TYPE_NAMES[header_type],
tomwalters@0:                       self._SECTION_NAMES[self._section]))
tomwalters@0: 
tomwalters@0:     last_section = self._section
tomwalters@0: 
tomwalters@0:     if header_type == _C_SYS_HEADER:
tomwalters@0:       if self._section <= self._C_SECTION:
tomwalters@0:         self._section = self._C_SECTION
tomwalters@0:       else:
tomwalters@0:         self._last_header = ''
tomwalters@0:         return error_message
tomwalters@0:     elif header_type == _CPP_SYS_HEADER:
tomwalters@0:       if self._section <= self._CPP_SECTION:
tomwalters@0:         self._section = self._CPP_SECTION
tomwalters@0:       else:
tomwalters@0:         self._last_header = ''
tomwalters@0:         return error_message
tomwalters@0:     elif header_type == _LIKELY_MY_HEADER:
tomwalters@0:       if self._section <= self._MY_H_SECTION:
tomwalters@0:         self._section = self._MY_H_SECTION
tomwalters@0:       else:
tomwalters@0:         self._section = self._OTHER_H_SECTION
tomwalters@0:     elif header_type == _POSSIBLE_MY_HEADER:
tomwalters@0:       if self._section <= self._MY_H_SECTION:
tomwalters@0:         self._section = self._MY_H_SECTION
tomwalters@0:       else:
tomwalters@0:         # This will always be the fallback because we're not sure
tomwalters@0:         # enough that the header is associated with this file.
tomwalters@0:         self._section = self._OTHER_H_SECTION
tomwalters@0:     else:
tomwalters@0:       assert header_type == _OTHER_HEADER
tomwalters@0:       self._section = self._OTHER_H_SECTION
tomwalters@0: 
tomwalters@0:     if last_section != self._section:
tomwalters@0:       self._last_header = ''
tomwalters@0: 
tomwalters@0:     return ''
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _CppLintState(object):
tomwalters@0:   """Maintains module-wide state.."""
tomwalters@0: 
tomwalters@0:   def __init__(self):
tomwalters@0:     self.verbose_level = 1  # global setting.
tomwalters@0:     self.error_count = 0    # global count of reported errors
tomwalters@0:     # filters to apply when emitting error messages
tomwalters@0:     self.filters = _DEFAULT_FILTERS[:]
tomwalters@0:     self.counting = 'total'  # In what way are we counting errors?
tomwalters@0:     self.errors_by_category = {}  # string to int dict storing error counts
tomwalters@0: 
tomwalters@0:     # output format:
tomwalters@0:     # "emacs" - format that emacs can parse (default)
tomwalters@0:     # "vs7" - format that Microsoft Visual Studio 7 can parse
tomwalters@0:     self.output_format = 'emacs'
tomwalters@0: 
tomwalters@0:   def SetOutputFormat(self, output_format):
tomwalters@0:     """Sets the output format for errors."""
tomwalters@0:     self.output_format = output_format
tomwalters@0: 
tomwalters@0:   def SetVerboseLevel(self, level):
tomwalters@0:     """Sets the module's verbosity, and returns the previous setting."""
tomwalters@0:     last_verbose_level = self.verbose_level
tomwalters@0:     self.verbose_level = level
tomwalters@0:     return last_verbose_level
tomwalters@0: 
tomwalters@0:   def SetCountingStyle(self, counting_style):
tomwalters@0:     """Sets the module's counting options."""
tomwalters@0:     self.counting = counting_style
tomwalters@0: 
tomwalters@0:   def SetFilters(self, filters):
tomwalters@0:     """Sets the error-message filters.
tomwalters@0: 
tomwalters@0:     These filters are applied when deciding whether to emit a given
tomwalters@0:     error message.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       filters: A string of comma-separated filters (eg "+whitespace/indent").
tomwalters@0:                Each filter should start with + or -; else we die.
tomwalters@0: 
tomwalters@0:     Raises:
tomwalters@0:       ValueError: The comma-separated filters did not all start with '+' or '-'.
tomwalters@0:                   E.g. "-,+whitespace,-whitespace/indent,whitespace/badfilter"
tomwalters@0:     """
tomwalters@0:     # Default filters always have less priority than the flag ones.
tomwalters@0:     self.filters = _DEFAULT_FILTERS[:]
tomwalters@0:     for filt in filters.split(','):
tomwalters@0:       clean_filt = filt.strip()
tomwalters@0:       if clean_filt:
tomwalters@0:         self.filters.append(clean_filt)
tomwalters@0:     for filt in self.filters:
tomwalters@0:       if not (filt.startswith('+') or filt.startswith('-')):
tomwalters@0:         raise ValueError('Every filter in --filters must start with + or -'
tomwalters@0:                          ' (%s does not)' % filt)
tomwalters@0: 
tomwalters@0:   def ResetErrorCounts(self):
tomwalters@0:     """Sets the module's error statistic back to zero."""
tomwalters@0:     self.error_count = 0
tomwalters@0:     self.errors_by_category = {}
tomwalters@0: 
tomwalters@0:   def IncrementErrorCount(self, category):
tomwalters@0:     """Bumps the module's error statistic."""
tomwalters@0:     self.error_count += 1
tomwalters@0:     if self.counting in ('toplevel', 'detailed'):
tomwalters@0:       if self.counting != 'detailed':
tomwalters@0:         category = category.split('/')[0]
tomwalters@0:       if category not in self.errors_by_category:
tomwalters@0:         self.errors_by_category[category] = 0
tomwalters@0:       self.errors_by_category[category] += 1
tomwalters@0: 
tomwalters@0:   def PrintErrorCounts(self):
tomwalters@0:     """Print a summary of errors by category, and the total."""
tomwalters@0:     for category, count in self.errors_by_category.iteritems():
tomwalters@0:       sys.stderr.write('Category \'%s\' errors found: %d\n' %
tomwalters@0:                        (category, count))
tomwalters@0:     sys.stderr.write('Total errors found: %d\n' % self.error_count)
tomwalters@0: 
tomwalters@0: _cpplint_state = _CppLintState()
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _OutputFormat():
tomwalters@0:   """Gets the module's output format."""
tomwalters@0:   return _cpplint_state.output_format
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _SetOutputFormat(output_format):
tomwalters@0:   """Sets the module's output format."""
tomwalters@0:   _cpplint_state.SetOutputFormat(output_format)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _VerboseLevel():
tomwalters@0:   """Returns the module's verbosity setting."""
tomwalters@0:   return _cpplint_state.verbose_level
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _SetVerboseLevel(level):
tomwalters@0:   """Sets the module's verbosity, and returns the previous setting."""
tomwalters@0:   return _cpplint_state.SetVerboseLevel(level)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _SetCountingStyle(level):
tomwalters@0:   """Sets the module's counting options."""
tomwalters@0:   _cpplint_state.SetCountingStyle(level)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _Filters():
tomwalters@0:   """Returns the module's list of output filters, as a list."""
tomwalters@0:   return _cpplint_state.filters
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _SetFilters(filters):
tomwalters@0:   """Sets the module's error-message filters.
tomwalters@0: 
tomwalters@0:   These filters are applied when deciding whether to emit a given
tomwalters@0:   error message.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filters: A string of comma-separated filters (eg "whitespace/indent").
tomwalters@0:              Each filter should start with + or -; else we die.
tomwalters@0:   """
tomwalters@0:   _cpplint_state.SetFilters(filters)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _FunctionState(object):
tomwalters@0:   """Tracks current function name and the number of lines in its body."""
tomwalters@0: 
tomwalters@0:   _NORMAL_TRIGGER = 250  # for --v=0, 500 for --v=1, etc.
tomwalters@0:   _TEST_TRIGGER = 400    # about 50% more than _NORMAL_TRIGGER.
tomwalters@0: 
tomwalters@0:   def __init__(self):
tomwalters@0:     self.in_a_function = False
tomwalters@0:     self.lines_in_function = 0
tomwalters@0:     self.current_function = ''
tomwalters@0: 
tomwalters@0:   def Begin(self, function_name):
tomwalters@0:     """Start analyzing function body.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       function_name: The name of the function being tracked.
tomwalters@0:     """
tomwalters@0:     self.in_a_function = True
tomwalters@0:     self.lines_in_function = 0
tomwalters@0:     self.current_function = function_name
tomwalters@0: 
tomwalters@0:   def Count(self):
tomwalters@0:     """Count line in current function body."""
tomwalters@0:     if self.in_a_function:
tomwalters@0:       self.lines_in_function += 1
tomwalters@0: 
tomwalters@0:   def Check(self, error, filename, linenum):
tomwalters@0:     """Report if too many lines in function body.
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       error: The function to call with any errors found.
tomwalters@0:       filename: The name of the current file.
tomwalters@0:       linenum: The number of the line to check.
tomwalters@0:     """
tomwalters@0:     if Match(r'T(EST|est)', self.current_function):
tomwalters@0:       base_trigger = self._TEST_TRIGGER
tomwalters@0:     else:
tomwalters@0:       base_trigger = self._NORMAL_TRIGGER
tomwalters@0:     trigger = base_trigger * 2**_VerboseLevel()
tomwalters@0: 
tomwalters@0:     if self.lines_in_function > trigger:
tomwalters@0:       error_level = int(math.log(self.lines_in_function / base_trigger, 2))
tomwalters@0:       # 50 => 0, 100 => 1, 200 => 2, 400 => 3, 800 => 4, 1600 => 5, ...
tomwalters@0:       if error_level > 5:
tomwalters@0:         error_level = 5
tomwalters@0:       error(filename, linenum, 'readability/fn_size', error_level,
tomwalters@0:             'Small and focused functions are preferred:'
tomwalters@0:             ' %s has %d non-comment lines'
tomwalters@0:             ' (error triggered by exceeding %d lines).'  % (
tomwalters@0:                 self.current_function, self.lines_in_function, trigger))
tomwalters@0: 
tomwalters@0:   def End(self):
tomwalters@0:     """Stop analizing function body."""
tomwalters@0:     self.in_a_function = False
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _IncludeError(Exception):
tomwalters@0:   """Indicates a problem with the include order in a file."""
tomwalters@0:   pass
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class FileInfo:
tomwalters@0:   """Provides utility functions for filenames.
tomwalters@0: 
tomwalters@0:   FileInfo provides easy access to the components of a file's path
tomwalters@0:   relative to the project root.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   def __init__(self, filename):
tomwalters@0:     self._filename = filename
tomwalters@0: 
tomwalters@0:   def FullName(self):
tomwalters@0:     """Make Windows paths like Unix."""
tomwalters@0:     return os.path.abspath(self._filename).replace('\\', '/')
tomwalters@0: 
tomwalters@0:   def RepositoryName(self):
tomwalters@0:     """FullName after removing the local path to the repository.
tomwalters@0: 
tomwalters@0:     If we have a real absolute path name here we can try to do something smart:
tomwalters@0:     detecting the root of the checkout and truncating /path/to/checkout from
tomwalters@0:     the name so that we get header guards that don't include things like
tomwalters@0:     "C:\Documents and Settings\..." or "/home/username/..." in them and thus
tomwalters@0:     people on different computers who have checked the source out to different
tomwalters@0:     locations won't see bogus errors.
tomwalters@0:     """
tomwalters@0:     fullname = self.FullName()
tomwalters@0: 
tomwalters@0:     if os.path.exists(fullname):
tomwalters@0:       project_dir = os.path.dirname(fullname)
tomwalters@0: 
tomwalters@0:       if os.path.exists(os.path.join(project_dir, ".svn")):
tomwalters@0:         # If there's a .svn file in the current directory, we recursively look
tomwalters@0:         # up the directory tree for the top of the SVN checkout
tomwalters@0:         root_dir = project_dir
tomwalters@0:         one_up_dir = os.path.dirname(root_dir)
tomwalters@0:         while os.path.exists(os.path.join(one_up_dir, ".svn")):
tomwalters@0:           root_dir = os.path.dirname(root_dir)
tomwalters@0:           one_up_dir = os.path.dirname(one_up_dir)
tomwalters@0: 
tomwalters@0:         prefix = os.path.commonprefix([root_dir, project_dir])
tomwalters@0:         return fullname[len(prefix) + 1:]
tomwalters@0: 
tomwalters@0:       # Not SVN? Try to find a git or hg top level directory by searching up
tomwalters@0:       # from the current path.
tomwalters@0:       root_dir = os.path.dirname(fullname)
tomwalters@0:       while (root_dir != os.path.dirname(root_dir) and
tomwalters@0:              not os.path.exists(os.path.join(root_dir, ".git")) and
tomwalters@0:              not os.path.exists(os.path.join(root_dir, ".hg"))):
tomwalters@0:         root_dir = os.path.dirname(root_dir)
tomwalters@0:         if (os.path.exists(os.path.join(root_dir, ".git")) or
tomwalters@0:             os.path.exists(os.path.join(root_dir, ".hg"))):
tomwalters@0:           prefix = os.path.commonprefix([root_dir, project_dir])
tomwalters@0:           return fullname[len(prefix) + 1:]
tomwalters@0: 
tomwalters@0:     # Don't know what to do; header guard warnings may be wrong...
tomwalters@0:     return fullname
tomwalters@0: 
tomwalters@0:   def Split(self):
tomwalters@0:     """Splits the file into the directory, basename, and extension.
tomwalters@0: 
tomwalters@0:     For 'chrome/browser/browser.cc', Split() would
tomwalters@0:     return ('chrome/browser', 'browser', '.cc')
tomwalters@0: 
tomwalters@0:     Returns:
tomwalters@0:       A tuple of (directory, basename, extension).
tomwalters@0:     """
tomwalters@0: 
tomwalters@0:     googlename = self.RepositoryName()
tomwalters@0:     project, rest = os.path.split(googlename)
tomwalters@0:     return (project,) + os.path.splitext(rest)
tomwalters@0: 
tomwalters@0:   def BaseName(self):
tomwalters@0:     """File base name - text after the final slash, before the final period."""
tomwalters@0:     return self.Split()[1]
tomwalters@0: 
tomwalters@0:   def Extension(self):
tomwalters@0:     """File extension - text following the final period."""
tomwalters@0:     return self.Split()[2]
tomwalters@0: 
tomwalters@0:   def NoExtension(self):
tomwalters@0:     """File has no source file extension."""
tomwalters@0:     return '/'.join(self.Split()[0:2])
tomwalters@0: 
tomwalters@0:   def IsSource(self):
tomwalters@0:     """File has a source file extension."""
tomwalters@0:     return self.Extension()[1:] in ('c', 'cc', 'cpp', 'cxx')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _ShouldPrintError(category, confidence):
tomwalters@0:   """Returns true iff confidence >= verbose, and category passes filter."""
tomwalters@0:   # There are two ways we might decide not to print an error message:
tomwalters@0:   # the verbosity level isn't high enough, or the filters filter it out.
tomwalters@0:   if confidence < _cpplint_state.verbose_level:
tomwalters@0:     return False
tomwalters@0: 
tomwalters@0:   is_filtered = False
tomwalters@0:   for one_filter in _Filters():
tomwalters@0:     if one_filter.startswith('-'):
tomwalters@0:       if category.startswith(one_filter[1:]):
tomwalters@0:         is_filtered = True
tomwalters@0:     elif one_filter.startswith('+'):
tomwalters@0:       if category.startswith(one_filter[1:]):
tomwalters@0:         is_filtered = False
tomwalters@0:     else:
tomwalters@0:       assert False  # should have been checked for in SetFilter.
tomwalters@0:   if is_filtered:
tomwalters@0:     return False
tomwalters@0: 
tomwalters@0:   return True
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def Error(filename, linenum, category, confidence, message):
tomwalters@0:   """Logs the fact we've found a lint error.
tomwalters@0: 
tomwalters@0:   We log where the error was found, and also our confidence in the error,
tomwalters@0:   that is, how certain we are this is a legitimate style regression, and
tomwalters@0:   not a misidentification or a use that's sometimes justified.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the file containing the error.
tomwalters@0:     linenum: The number of the line containing the error.
tomwalters@0:     category: A string used to describe the "category" this bug
tomwalters@0:       falls under: "whitespace", say, or "runtime".  Categories
tomwalters@0:       may have a hierarchy separated by slashes: "whitespace/indent".
tomwalters@0:     confidence: A number from 1-5 representing a confidence score for
tomwalters@0:       the error, with 5 meaning that we are certain of the problem,
tomwalters@0:       and 1 meaning that it could be a legitimate construct.
tomwalters@0:     message: The error message.
tomwalters@0:   """
tomwalters@0:   # There are two ways we might decide not to print an error message:
tomwalters@0:   # the verbosity level isn't high enough, or the filters filter it out.
tomwalters@0:   if _ShouldPrintError(category, confidence):
tomwalters@0:     _cpplint_state.IncrementErrorCount(category)
tomwalters@0:     if _cpplint_state.output_format == 'vs7':
tomwalters@0:       sys.stderr.write('%s(%s):  %s  [%s] [%d]\n' % (
tomwalters@0:           filename, linenum, message, category, confidence))
tomwalters@0:     else:
tomwalters@0:       sys.stderr.write('%s:%s:  %s  [%s] [%d]\n' % (
tomwalters@0:           filename, linenum, message, category, confidence))
tomwalters@0: 
tomwalters@0: 
tomwalters@0: # Matches standard C++ escape esequences per 2.13.2.3 of the C++ standard.
tomwalters@0: _RE_PATTERN_CLEANSE_LINE_ESCAPES = re.compile(
tomwalters@0:     r'\\([abfnrtv?"\\\']|\d+|x[0-9a-fA-F]+)')
tomwalters@0: # Matches strings.  Escape codes should already be removed by ESCAPES.
tomwalters@0: _RE_PATTERN_CLEANSE_LINE_DOUBLE_QUOTES = re.compile(r'"[^"]*"')
tomwalters@0: # Matches characters.  Escape codes should already be removed by ESCAPES.
tomwalters@0: _RE_PATTERN_CLEANSE_LINE_SINGLE_QUOTES = re.compile(r"'.'")
tomwalters@0: # Matches multi-line C++ comments.
tomwalters@0: # This RE is a little bit more complicated than one might expect, because we
tomwalters@0: # have to take care of space removals tools so we can handle comments inside
tomwalters@0: # statements better.
tomwalters@0: # The current rule is: We only clear spaces from both sides when we're at the
tomwalters@0: # end of the line. Otherwise, we try to remove spaces from the right side,
tomwalters@0: # if this doesn't work we try on left side but only if there's a non-character
tomwalters@0: # on the right.
tomwalters@0: _RE_PATTERN_CLEANSE_LINE_C_COMMENTS = re.compile(
tomwalters@0:     r"""(\s*/\*.*\*/\s*$|
tomwalters@0:             /\*.*\*/\s+|
tomwalters@0:          \s+/\*.*\*/(?=\W)|
tomwalters@0:             /\*.*\*/)""", re.VERBOSE)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def IsCppString(line):
tomwalters@0:   """Does line terminate so, that the next symbol is in string constant.
tomwalters@0: 
tomwalters@0:   This function does not consider single-line nor multi-line comments.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     line: is a partial line of code starting from the 0..n.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     True, if next character appended to 'line' is inside a
tomwalters@0:     string constant.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   line = line.replace(r'\\', 'XX')  # after this, \\" does not match to \"
tomwalters@0:   return ((line.count('"') - line.count(r'\"') - line.count("'\"'")) & 1) == 1
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def FindNextMultiLineCommentStart(lines, lineix):
tomwalters@0:   """Find the beginning marker for a multiline comment."""
tomwalters@0:   while lineix < len(lines):
tomwalters@0:     if lines[lineix].strip().startswith('/*'):
tomwalters@0:       # Only return this marker if the comment goes beyond this line
tomwalters@0:       if lines[lineix].strip().find('*/', 2) < 0:
tomwalters@0:         return lineix
tomwalters@0:     lineix += 1
tomwalters@0:   return len(lines)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def FindNextMultiLineCommentEnd(lines, lineix):
tomwalters@0:   """We are inside a comment, find the end marker."""
tomwalters@0:   while lineix < len(lines):
tomwalters@0:     if lines[lineix].strip().endswith('*/'):
tomwalters@0:       return lineix
tomwalters@0:     lineix += 1
tomwalters@0:   return len(lines)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def RemoveMultiLineCommentsFromRange(lines, begin, end):
tomwalters@0:   """Clears a range of lines for multi-line comments."""
tomwalters@0:   # Having // dummy comments makes the lines non-empty, so we will not get
tomwalters@0:   # unnecessary blank line warnings later in the code.
tomwalters@0:   for i in range(begin, end):
tomwalters@0:     lines[i] = '// dummy'
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def RemoveMultiLineComments(filename, lines, error):
tomwalters@0:   """Removes multiline (c-style) comments from lines."""
tomwalters@0:   lineix = 0
tomwalters@0:   while lineix < len(lines):
tomwalters@0:     lineix_begin = FindNextMultiLineCommentStart(lines, lineix)
tomwalters@0:     if lineix_begin >= len(lines):
tomwalters@0:       return
tomwalters@0:     lineix_end = FindNextMultiLineCommentEnd(lines, lineix_begin)
tomwalters@0:     if lineix_end >= len(lines):
tomwalters@0:       error(filename, lineix_begin + 1, 'readability/multiline_comment', 5,
tomwalters@0:             'Could not find end of multi-line comment')
tomwalters@0:       return
tomwalters@0:     RemoveMultiLineCommentsFromRange(lines, lineix_begin, lineix_end + 1)
tomwalters@0:     lineix = lineix_end + 1
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CleanseComments(line):
tomwalters@0:   """Removes //-comments and single-line C-style /* */ comments.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     line: A line of C++ source.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     The line with single-line comments removed.
tomwalters@0:   """
tomwalters@0:   commentpos = line.find('//')
tomwalters@0:   if commentpos != -1 and not IsCppString(line[:commentpos]):
tomwalters@0:     line = line[:commentpos]
tomwalters@0:   # get rid of /* ... */
tomwalters@0:   return _RE_PATTERN_CLEANSE_LINE_C_COMMENTS.sub('', line)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class CleansedLines(object):
tomwalters@0:   """Holds 3 copies of all lines with different preprocessing applied to them.
tomwalters@0: 
tomwalters@0:   1) elided member contains lines without strings and comments,
tomwalters@0:   2) lines member contains lines without comments, and
tomwalters@0:   3) raw member contains all the lines without processing.
tomwalters@0:   All these three members are of <type 'list'>, and of the same length.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   def __init__(self, lines):
tomwalters@0:     self.elided = []
tomwalters@0:     self.lines = []
tomwalters@0:     self.raw_lines = lines
tomwalters@0:     self.num_lines = len(lines)
tomwalters@0:     for linenum in range(len(lines)):
tomwalters@0:       self.lines.append(CleanseComments(lines[linenum]))
tomwalters@0:       elided = self._CollapseStrings(lines[linenum])
tomwalters@0:       self.elided.append(CleanseComments(elided))
tomwalters@0: 
tomwalters@0:   def NumLines(self):
tomwalters@0:     """Returns the number of lines represented."""
tomwalters@0:     return self.num_lines
tomwalters@0: 
tomwalters@0:   @staticmethod
tomwalters@0:   def _CollapseStrings(elided):
tomwalters@0:     """Collapses strings and chars on a line to simple "" or '' blocks.
tomwalters@0: 
tomwalters@0:     We nix strings first so we're not fooled by text like '"http://"'
tomwalters@0: 
tomwalters@0:     Args:
tomwalters@0:       elided: The line being processed.
tomwalters@0: 
tomwalters@0:     Returns:
tomwalters@0:       The line with collapsed strings.
tomwalters@0:     """
tomwalters@0:     if not _RE_PATTERN_INCLUDE.match(elided):
tomwalters@0:       # Remove escaped characters first to make quote/single quote collapsing
tomwalters@0:       # basic.  Things that look like escaped characters shouldn't occur
tomwalters@0:       # outside of strings and chars.
tomwalters@0:       elided = _RE_PATTERN_CLEANSE_LINE_ESCAPES.sub('', elided)
tomwalters@0:       elided = _RE_PATTERN_CLEANSE_LINE_SINGLE_QUOTES.sub("''", elided)
tomwalters@0:       elided = _RE_PATTERN_CLEANSE_LINE_DOUBLE_QUOTES.sub('""', elided)
tomwalters@0:     return elided
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CloseExpression(clean_lines, linenum, pos):
tomwalters@0:   """If input points to ( or { or [, finds the position that closes it.
tomwalters@0: 
tomwalters@0:   If lines[linenum][pos] points to a '(' or '{' or '[', finds the the
tomwalters@0:   linenum/pos that correspond to the closing of the expression.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     pos: A position on the line.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     A tuple (line, linenum, pos) pointer *past* the closing brace, or
tomwalters@0:     (line, len(lines), -1) if we never find a close.  Note we ignore
tomwalters@0:     strings and comments when matching; and the line we return is the
tomwalters@0:     'cleansed' line at linenum.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0:   startchar = line[pos]
tomwalters@0:   if startchar not in '({[':
tomwalters@0:     return (line, clean_lines.NumLines(), -1)
tomwalters@0:   if startchar == '(': endchar = ')'
tomwalters@0:   if startchar == '[': endchar = ']'
tomwalters@0:   if startchar == '{': endchar = '}'
tomwalters@0: 
tomwalters@0:   num_open = line.count(startchar) - line.count(endchar)
tomwalters@0:   while linenum < clean_lines.NumLines() and num_open > 0:
tomwalters@0:     linenum += 1
tomwalters@0:     line = clean_lines.elided[linenum]
tomwalters@0:     num_open += line.count(startchar) - line.count(endchar)
tomwalters@0:   # OK, now find the endchar that actually got us back to even
tomwalters@0:   endpos = len(line)
tomwalters@0:   while num_open >= 0:
tomwalters@0:     endpos = line.rfind(')', 0, endpos)
tomwalters@0:     num_open -= 1                 # chopped off another )
tomwalters@0:   return (line, linenum, endpos + 1)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForCopyright(filename, lines, error):
tomwalters@0:   """Logs an error if no Copyright message appears at the top of the file."""
tomwalters@0: 
tomwalters@0:   # We'll say it should occur by line 10. Don't forget there's a
tomwalters@0:   # dummy line at the front.
tomwalters@0:   for line in xrange(1, min(len(lines), 11)):
tomwalters@0:     if re.search(r'Copyright', lines[line], re.I): break
tomwalters@0:   else:                       # means no copyright line was found
tomwalters@0:     error(filename, 0, 'legal/copyright', 5,
tomwalters@0:           'No copyright message found.  '
tomwalters@0:           'You should have a line: "Copyright [year] <Copyright Owner>"')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def GetHeaderGuardCPPVariable(filename):
tomwalters@0:   """Returns the CPP variable that should be used as a header guard.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of a C++ header file.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     The CPP variable that should be used as a header guard in the
tomwalters@0:     named file.
tomwalters@0: 
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   fileinfo = FileInfo(filename)
tomwalters@0:   return re.sub(r'[-./\s]', '_', fileinfo.RepositoryName()).upper() + '_'
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForHeaderGuard(filename, lines, error):
tomwalters@0:   """Checks that the file contains a header guard.
tomwalters@0: 
tomwalters@0:   Logs an error if no #ifndef header guard is present.  For other
tomwalters@0:   headers, checks that the full pathname is used.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the C++ header file.
tomwalters@0:     lines: An array of strings, each representing a line of the file.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   cppvar = GetHeaderGuardCPPVariable(filename)
tomwalters@0: 
tomwalters@0:   ifndef = None
tomwalters@0:   ifndef_linenum = 0
tomwalters@0:   define = None
tomwalters@0:   endif = None
tomwalters@0:   endif_linenum = 0
tomwalters@0:   for linenum, line in enumerate(lines):
tomwalters@0:     linesplit = line.split()
tomwalters@0:     if len(linesplit) >= 2:
tomwalters@0:       # find the first occurrence of #ifndef and #define, save arg
tomwalters@0:       if not ifndef and linesplit[0] == '#ifndef':
tomwalters@0:         # set ifndef to the header guard presented on the #ifndef line.
tomwalters@0:         ifndef = linesplit[1]
tomwalters@0:         ifndef_linenum = linenum
tomwalters@0:       if not define and linesplit[0] == '#define':
tomwalters@0:         define = linesplit[1]
tomwalters@0:     # find the last occurrence of #endif, save entire line
tomwalters@0:     if line.startswith('#endif'):
tomwalters@0:       endif = line
tomwalters@0:       endif_linenum = linenum
tomwalters@0: 
tomwalters@0:   if not ifndef or not define or ifndef != define:
tomwalters@0:     error(filename, 0, 'build/header_guard', 5,
tomwalters@0:           'No #ifndef header guard found, suggested CPP variable is: %s' %
tomwalters@0:           cppvar)
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # The guard should be PATH_FILE_H_, but we also allow PATH_FILE_H__
tomwalters@0:   # for backward compatibility.
tomwalters@0:   if ifndef != cppvar and not Search(r'\bNOLINT\b', lines[ifndef_linenum]):
tomwalters@0:     error_level = 0
tomwalters@0:     if ifndef != cppvar + '_':
tomwalters@0:       error_level = 5
tomwalters@0: 
tomwalters@0:     error(filename, ifndef_linenum, 'build/header_guard', error_level,
tomwalters@0:           '#ifndef header guard has wrong style, please use: %s' % cppvar)
tomwalters@0: 
tomwalters@0:   if (endif != ('#endif  // %s' % cppvar) and
tomwalters@0:       not Search(r'\bNOLINT\b', lines[endif_linenum])):
tomwalters@0:     error_level = 0
tomwalters@0:     if endif != ('#endif  // %s' % (cppvar + '_')):
tomwalters@0:       error_level = 5
tomwalters@0: 
tomwalters@0:     error(filename, endif_linenum, 'build/header_guard', error_level,
tomwalters@0:           '#endif line should be "#endif  // %s"' % cppvar)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForUnicodeReplacementCharacters(filename, lines, error):
tomwalters@0:   """Logs an error for each line containing Unicode replacement characters.
tomwalters@0: 
tomwalters@0:   These indicate that either the file contained invalid UTF-8 (likely)
tomwalters@0:   or Unicode replacement characters (which it shouldn't).  Note that
tomwalters@0:   it's possible for this to throw off line numbering if the invalid
tomwalters@0:   UTF-8 occurred adjacent to a newline.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     lines: An array of strings, each representing a line of the file.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   for linenum, line in enumerate(lines):
tomwalters@0:     if u'\ufffd' in line:
tomwalters@0:       error(filename, linenum, 'readability/utf8', 5,
tomwalters@0:             'Line contains invalid UTF-8 (or Unicode replacement character).')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForNewlineAtEOF(filename, lines, error):
tomwalters@0:   """Logs an error if there is no newline char at the end of the file.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     lines: An array of strings, each representing a line of the file.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   # The array lines() was created by adding two newlines to the
tomwalters@0:   # original file (go figure), then splitting on \n.
tomwalters@0:   # To verify that the file ends in \n, we just have to make sure the
tomwalters@0:   # last-but-two element of lines() exists and is empty.
tomwalters@0:   if len(lines) < 3 or lines[-2]:
tomwalters@0:     error(filename, len(lines) - 2, 'whitespace/ending_newline', 5,
tomwalters@0:           'Could not find a newline character at the end of the file.')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForMultilineCommentsAndStrings(filename, clean_lines, linenum, error):
tomwalters@0:   """Logs an error if we see /* ... */ or "..." that extend past one line.
tomwalters@0: 
tomwalters@0:   /* ... */ comments are legit inside macros, for one line.
tomwalters@0:   Otherwise, we prefer // comments, so it's ok to warn about the
tomwalters@0:   other.  Likewise, it's ok for strings to extend across multiple
tomwalters@0:   lines, as long as a line continuation character (backslash)
tomwalters@0:   terminates each line. Although not currently prohibited by the C++
tomwalters@0:   style guide, it's ugly and unnecessary. We don't do well with either
tomwalters@0:   in this lint program, so we warn about both.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0: 
tomwalters@0:   # Remove all \\ (escaped backslashes) from the line. They are OK, and the
tomwalters@0:   # second (escaped) slash may trigger later \" detection erroneously.
tomwalters@0:   line = line.replace('\\\\', '')
tomwalters@0: 
tomwalters@0:   if line.count('/*') > line.count('*/'):
tomwalters@0:     error(filename, linenum, 'readability/multiline_comment', 5,
tomwalters@0:           'Complex multi-line /*...*/-style comment found. '
tomwalters@0:           'Lint may give bogus warnings.  '
tomwalters@0:           'Consider replacing these with //-style comments, '
tomwalters@0:           'with #if 0...#endif, '
tomwalters@0:           'or with more clearly structured multi-line comments.')
tomwalters@0: 
tomwalters@0:   if (line.count('"') - line.count('\\"')) % 2:
tomwalters@0:     error(filename, linenum, 'readability/multiline_string', 5,
tomwalters@0:           'Multi-line string ("...") found.  This lint script doesn\'t '
tomwalters@0:           'do well with such strings, and may give bogus warnings.  They\'re '
tomwalters@0:           'ugly and unnecessary, and you should use concatenation instead".')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: threading_list = (
tomwalters@0:     ('asctime(', 'asctime_r('),
tomwalters@0:     ('ctime(', 'ctime_r('),
tomwalters@0:     ('getgrgid(', 'getgrgid_r('),
tomwalters@0:     ('getgrnam(', 'getgrnam_r('),
tomwalters@0:     ('getlogin(', 'getlogin_r('),
tomwalters@0:     ('getpwnam(', 'getpwnam_r('),
tomwalters@0:     ('getpwuid(', 'getpwuid_r('),
tomwalters@0:     ('gmtime(', 'gmtime_r('),
tomwalters@0:     ('localtime(', 'localtime_r('),
tomwalters@0:     ('rand(', 'rand_r('),
tomwalters@0:     ('readdir(', 'readdir_r('),
tomwalters@0:     ('strtok(', 'strtok_r('),
tomwalters@0:     ('ttyname(', 'ttyname_r('),
tomwalters@0:     )
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckPosixThreading(filename, clean_lines, linenum, error):
tomwalters@0:   """Checks for calls to thread-unsafe functions.
tomwalters@0: 
tomwalters@0:   Much code has been originally written without consideration of
tomwalters@0:   multi-threading. Also, engineers are relying on their old experience;
tomwalters@0:   they have learned posix before threading extensions were added. These
tomwalters@0:   tests guide the engineers to use thread-safe functions (when using
tomwalters@0:   posix directly).
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0:   for single_thread_function, multithread_safe_function in threading_list:
tomwalters@0:     ix = line.find(single_thread_function)
tomwalters@0:     # Comparisons made explicit for clarity -- pylint: disable-msg=C6403
tomwalters@0:     if ix >= 0 and (ix == 0 or (not line[ix - 1].isalnum() and
tomwalters@0:                                 line[ix - 1] not in ('_', '.', '>'))):
tomwalters@0:       error(filename, linenum, 'runtime/threadsafe_fn', 2,
tomwalters@0:             'Consider using ' + multithread_safe_function +
tomwalters@0:             '...) instead of ' + single_thread_function +
tomwalters@0:             '...) for improved thread safety.')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: # Matches invalid increment: *count++, which moves pointer instead of
tomwalters@0: # incrementing a value.
tomwalters@0: _RE_PATTERN_INVALID_INCREMENT = re.compile(
tomwalters@0:     r'^\s*\*\w+(\+\+|--);')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckInvalidIncrement(filename, clean_lines, linenum, error):
tomwalters@0:   """Checks for invalid increment *count++.
tomwalters@0: 
tomwalters@0:   For example following function:
tomwalters@0:   void increment_counter(int* count) {
tomwalters@0:     *count++;
tomwalters@0:   }
tomwalters@0:   is invalid, because it effectively does count++, moving pointer, and should
tomwalters@0:   be replaced with ++*count, (*count)++ or *count += 1.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0:   if _RE_PATTERN_INVALID_INCREMENT.match(line):
tomwalters@0:     error(filename, linenum, 'runtime/invalid_increment', 5,
tomwalters@0:           'Changing pointer instead of value (or unused value of operator*).')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _ClassInfo(object):
tomwalters@0:   """Stores information about a class."""
tomwalters@0: 
tomwalters@0:   def __init__(self, name, linenum):
tomwalters@0:     self.name = name
tomwalters@0:     self.linenum = linenum
tomwalters@0:     self.seen_open_brace = False
tomwalters@0:     self.is_derived = False
tomwalters@0:     self.virtual_method_linenumber = None
tomwalters@0:     self.has_virtual_destructor = False
tomwalters@0:     self.brace_depth = 0
tomwalters@0: 
tomwalters@0: 
tomwalters@0: class _ClassState(object):
tomwalters@0:   """Holds the current state of the parse relating to class declarations.
tomwalters@0: 
tomwalters@0:   It maintains a stack of _ClassInfos representing the parser's guess
tomwalters@0:   as to the current nesting of class declarations. The innermost class
tomwalters@0:   is at the top (back) of the stack. Typically, the stack will either
tomwalters@0:   be empty or have exactly one entry.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   def __init__(self):
tomwalters@0:     self.classinfo_stack = []
tomwalters@0: 
tomwalters@0:   def CheckFinished(self, filename, error):
tomwalters@0:     """Checks that all classes have been completely parsed.
tomwalters@0: 
tomwalters@0:     Call this when all lines in a file have been processed.
tomwalters@0:     Args:
tomwalters@0:       filename: The name of the current file.
tomwalters@0:       error: The function to call with any errors found.
tomwalters@0:     """
tomwalters@0:     if self.classinfo_stack:
tomwalters@0:       # Note: This test can result in false positives if #ifdef constructs
tomwalters@0:       # get in the way of brace matching. See the testBuildClass test in
tomwalters@0:       # cpplint_unittest.py for an example of this.
tomwalters@0:       error(filename, self.classinfo_stack[0].linenum, 'build/class', 5,
tomwalters@0:             'Failed to find complete declaration of class %s' %
tomwalters@0:             self.classinfo_stack[0].name)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForNonStandardConstructs(filename, clean_lines, linenum,
tomwalters@0:                                   class_state, error):
tomwalters@0:   """Logs an error if we see certain non-ANSI constructs ignored by gcc-2.
tomwalters@0: 
tomwalters@0:   Complain about several constructs which gcc-2 accepts, but which are
tomwalters@0:   not standard C++.  Warning about these in lint is one way to ease the
tomwalters@0:   transition to new compilers.
tomwalters@0:   - put storage class first (e.g. "static const" instead of "const static").
tomwalters@0:   - "%lld" instead of %qd" in printf-type functions.
tomwalters@0:   - "%1$d" is non-standard in printf-type functions.
tomwalters@0:   - "\%" is an undefined character escape sequence.
tomwalters@0:   - text after #endif is not allowed.
tomwalters@0:   - invalid inner-style forward declaration.
tomwalters@0:   - >? and <? operators, and their >?= and <?= cousins.
tomwalters@0:   - classes with virtual methods need virtual destructors (compiler warning
tomwalters@0:     available, but not turned on yet.)
tomwalters@0: 
tomwalters@0:   Additionally, check for constructor/destructor style violations and reference
tomwalters@0:   members, as it is very convenient to do so while checking for
tomwalters@0:   gcc-2 compliance.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     class_state: A _ClassState instance which maintains information about
tomwalters@0:                  the current stack of nested class declarations being parsed.
tomwalters@0:     error: A callable to which errors are reported, which takes 4 arguments:
tomwalters@0:            filename, line number, error level, and message
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   # Remove comments from the line, but leave in strings for now.
tomwalters@0:   line = clean_lines.lines[linenum]
tomwalters@0: 
tomwalters@0:   if Search(r'printf\s*\(.*".*%[-+ ]?\d*q', line):
tomwalters@0:     error(filename, linenum, 'runtime/printf_format', 3,
tomwalters@0:           '%q in format strings is deprecated.  Use %ll instead.')
tomwalters@0: 
tomwalters@0:   if Search(r'printf\s*\(.*".*%\d+\$', line):
tomwalters@0:     error(filename, linenum, 'runtime/printf_format', 2,
tomwalters@0:           '%N$ formats are unconventional.  Try rewriting to avoid them.')
tomwalters@0: 
tomwalters@0:   # Remove escaped backslashes before looking for undefined escapes.
tomwalters@0:   line = line.replace('\\\\', '')
tomwalters@0: 
tomwalters@0:   if Search(r'("|\').*\\(%|\[|\(|{)', line):
tomwalters@0:     error(filename, linenum, 'build/printf_format', 3,
tomwalters@0:           '%, [, (, and { are undefined character escapes.  Unescape them.')
tomwalters@0: 
tomwalters@0:   # For the rest, work with both comments and strings removed.
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0: 
tomwalters@0:   if Search(r'\b(const|volatile|void|char|short|int|long'
tomwalters@0:             r'|float|double|signed|unsigned'
tomwalters@0:             r'|schar|u?int8|u?int16|u?int32|u?int64)'
tomwalters@0:             r'\s+(auto|register|static|extern|typedef)\b',
tomwalters@0:             line):
tomwalters@0:     error(filename, linenum, 'build/storage_class', 5,
tomwalters@0:           'Storage class (static, extern, typedef, etc) should be first.')
tomwalters@0: 
tomwalters@0:   if Match(r'\s*#\s*endif\s*[^/\s]+', line):
tomwalters@0:     error(filename, linenum, 'build/endif_comment', 5,
tomwalters@0:           'Uncommented text after #endif is non-standard.  Use a comment.')
tomwalters@0: 
tomwalters@0:   if Match(r'\s*class\s+(\w+\s*::\s*)+\w+\s*;', line):
tomwalters@0:     error(filename, linenum, 'build/forward_decl', 5,
tomwalters@0:           'Inner-style forward declarations are invalid.  Remove this line.')
tomwalters@0: 
tomwalters@0:   if Search(r'(\w+|[+-]?\d+(\.\d*)?)\s*(<|>)\?=?\s*(\w+|[+-]?\d+)(\.\d*)?',
tomwalters@0:             line):
tomwalters@0:     error(filename, linenum, 'build/deprecated', 3,
tomwalters@0:           '>? and <? (max and min) operators are non-standard and deprecated.')
tomwalters@0: 
tomwalters@0:   if Search(r'^\s*const\s*string\s*&\s*\w+\s*;', line):
tomwalters@0:     # TODO(unknown): Could it be expanded safely to arbitrary references,
tomwalters@0:     # without triggering too many false positives? The first
tomwalters@0:     # attempt triggered 5 warnings for mostly benign code in the regtest, hence
tomwalters@0:     # the restriction.
tomwalters@0:     # Here's the original regexp, for the reference:
tomwalters@0:     # type_name = r'\w+((\s*::\s*\w+)|(\s*<\s*\w+?\s*>))?'
tomwalters@0:     # r'\s*const\s*' + type_name + '\s*&\s*\w+\s*;'
tomwalters@0:     error(filename, linenum, 'runtime/member_string_references', 2,
tomwalters@0:           'const string& members are dangerous. It is much better to use '
tomwalters@0:           'alternatives, such as pointers or simple constants.')
tomwalters@0: 
tomwalters@0:   # Track class entry and exit, and attempt to find cases within the
tomwalters@0:   # class declaration that don't meet the C++ style
tomwalters@0:   # guidelines. Tracking is very dependent on the code matching Google
tomwalters@0:   # style guidelines, but it seems to perform well enough in testing
tomwalters@0:   # to be a worthwhile addition to the checks.
tomwalters@0:   classinfo_stack = class_state.classinfo_stack
tomwalters@0:   # Look for a class declaration
tomwalters@0:   class_decl_match = Match(
tomwalters@0:       r'\s*(template\s*<[\w\s<>,:]*>\s*)?(class|struct)\s+(\w+(::\w+)*)', line)
tomwalters@0:   if class_decl_match:
tomwalters@0:     classinfo_stack.append(_ClassInfo(class_decl_match.group(3), linenum))
tomwalters@0: 
tomwalters@0:   # Everything else in this function uses the top of the stack if it's
tomwalters@0:   # not empty.
tomwalters@0:   if not classinfo_stack:
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   classinfo = classinfo_stack[-1]
tomwalters@0: 
tomwalters@0:   # If the opening brace hasn't been seen look for it and also
tomwalters@0:   # parent class declarations.
tomwalters@0:   if not classinfo.seen_open_brace:
tomwalters@0:     # If the line has a ';' in it, assume it's a forward declaration or
tomwalters@0:     # a single-line class declaration, which we won't process.
tomwalters@0:     if line.find(';') != -1:
tomwalters@0:       classinfo_stack.pop()
tomwalters@0:       return
tomwalters@0:     classinfo.seen_open_brace = (line.find('{') != -1)
tomwalters@0:     # Look for a bare ':'
tomwalters@0:     if Search('(^|[^:]):($|[^:])', line):
tomwalters@0:       classinfo.is_derived = True
tomwalters@0:     if not classinfo.seen_open_brace:
tomwalters@0:       return  # Everything else in this function is for after open brace
tomwalters@0: 
tomwalters@0:   # The class may have been declared with namespace or classname qualifiers.
tomwalters@0:   # The constructor and destructor will not have those qualifiers.
tomwalters@0:   base_classname = classinfo.name.split('::')[-1]
tomwalters@0: 
tomwalters@0:   # Look for single-argument constructors that aren't marked explicit.
tomwalters@0:   # Technically a valid construct, but against style.
tomwalters@0:   args = Match(r'(?<!explicit)\s+%s\s*\(([^,()]+)\)'
tomwalters@0:                % re.escape(base_classname),
tomwalters@0:                line)
tomwalters@0:   if (args and
tomwalters@0:       args.group(1) != 'void' and
tomwalters@0:       not Match(r'(const\s+)?%s\s*&' % re.escape(base_classname),
tomwalters@0:                 args.group(1).strip())):
tomwalters@0:     error(filename, linenum, 'runtime/explicit', 5,
tomwalters@0:           'Single-argument constructors should be marked explicit.')
tomwalters@0: 
tomwalters@0:   # Look for methods declared virtual.
tomwalters@0:   if Search(r'\bvirtual\b', line):
tomwalters@0:     classinfo.virtual_method_linenumber = linenum
tomwalters@0:     # Only look for a destructor declaration on the same line. It would
tomwalters@0:     # be extremely unlikely for the destructor declaration to occupy
tomwalters@0:     # more than one line.
tomwalters@0:     if Search(r'~%s\s*\(' % base_classname, line):
tomwalters@0:       classinfo.has_virtual_destructor = True
tomwalters@0: 
tomwalters@0:   # Look for class end.
tomwalters@0:   brace_depth = classinfo.brace_depth
tomwalters@0:   brace_depth = brace_depth + line.count('{') - line.count('}')
tomwalters@0:   if brace_depth <= 0:
tomwalters@0:     classinfo = classinfo_stack.pop()
tomwalters@0:     # Try to detect missing virtual destructor declarations.
tomwalters@0:     # For now, only warn if a non-derived class with virtual methods lacks
tomwalters@0:     # a virtual destructor. This is to make it less likely that people will
tomwalters@0:     # declare derived virtual destructors without declaring the base
tomwalters@0:     # destructor virtual.
tomwalters@0:     if ((classinfo.virtual_method_linenumber is not None) and
tomwalters@0:         (not classinfo.has_virtual_destructor) and
tomwalters@0:         (not classinfo.is_derived)):  # Only warn for base classes
tomwalters@0:       error(filename, classinfo.linenum, 'runtime/virtual', 4,
tomwalters@0:             'The class %s probably needs a virtual destructor due to '
tomwalters@0:             'having virtual method(s), one declared at line %d.'
tomwalters@0:             % (classinfo.name, classinfo.virtual_method_linenumber))
tomwalters@0:   else:
tomwalters@0:     classinfo.brace_depth = brace_depth
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckSpacingForFunctionCall(filename, line, linenum, error):
tomwalters@0:   """Checks for the correctness of various spacing around function calls.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     line: The text of the line to check.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   # Since function calls often occur inside if/for/while/switch
tomwalters@0:   # expressions - which have their own, more liberal conventions - we
tomwalters@0:   # first see if we should be looking inside such an expression for a
tomwalters@0:   # function call, to which we can apply more strict standards.
tomwalters@0:   fncall = line    # if there's no control flow construct, look at whole line
tomwalters@0:   for pattern in (r'\bif\s*\((.*)\)\s*{',
tomwalters@0:                   r'\bfor\s*\((.*)\)\s*{',
tomwalters@0:                   r'\bwhile\s*\((.*)\)\s*[{;]',
tomwalters@0:                   r'\bswitch\s*\((.*)\)\s*{'):
tomwalters@0:     match = Search(pattern, line)
tomwalters@0:     if match:
tomwalters@0:       fncall = match.group(1)    # look inside the parens for function calls
tomwalters@0:       break
tomwalters@0: 
tomwalters@0:   # Except in if/for/while/switch, there should never be space
tomwalters@0:   # immediately inside parens (eg "f( 3, 4 )").  We make an exception
tomwalters@0:   # for nested parens ( (a+b) + c ).  Likewise, there should never be
tomwalters@0:   # a space before a ( when it's a function argument.  I assume it's a
tomwalters@0:   # function argument when the char before the whitespace is legal in
tomwalters@0:   # a function name (alnum + _) and we're not starting a macro. Also ignore
tomwalters@0:   # pointers and references to arrays and functions coz they're too tricky:
tomwalters@0:   # we use a very simple way to recognize these:
tomwalters@0:   # " (something)(maybe-something)" or
tomwalters@0:   # " (something)(maybe-something," or
tomwalters@0:   # " (something)[something]"
tomwalters@0:   # Note that we assume the contents of [] to be short enough that
tomwalters@0:   # they'll never need to wrap.
tomwalters@0:   if (  # Ignore control structures.
tomwalters@0:       not Search(r'\b(if|for|while|switch|return|delete)\b', fncall) and
tomwalters@0:       # Ignore pointers/references to functions.
tomwalters@0:       not Search(r' \([^)]+\)\([^)]*(\)|,$)', fncall) and
tomwalters@0:       # Ignore pointers/references to arrays.
tomwalters@0:       not Search(r' \([^)]+\)\[[^\]]+\]', fncall)):
tomwalters@0:     if Search(r'\w\s*\(\s(?!\s*\\$)', fncall):      # a ( used for a fn call
tomwalters@0:       error(filename, linenum, 'whitespace/parens', 4,
tomwalters@0:             'Extra space after ( in function call')
tomwalters@0:     elif Search(r'\(\s+(?!(\s*\\)|\()', fncall):
tomwalters@0:       error(filename, linenum, 'whitespace/parens', 2,
tomwalters@0:             'Extra space after (')
tomwalters@0:     if (Search(r'\w\s+\(', fncall) and
tomwalters@0:         not Search(r'#\s*define|typedef', fncall)):
tomwalters@0:       error(filename, linenum, 'whitespace/parens', 4,
tomwalters@0:             'Extra space before ( in function call')
tomwalters@0:     # If the ) is followed only by a newline or a { + newline, assume it's
tomwalters@0:     # part of a control statement (if/while/etc), and don't complain
tomwalters@0:     if Search(r'[^)]\s+\)\s*[^{\s]', fncall):
tomwalters@0:       error(filename, linenum, 'whitespace/parens', 2,
tomwalters@0:             'Extra space before )')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def IsBlankLine(line):
tomwalters@0:   """Returns true if the given line is blank.
tomwalters@0: 
tomwalters@0:   We consider a line to be blank if the line is empty or consists of
tomwalters@0:   only white spaces.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     line: A line of a string.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     True, if the given line is blank.
tomwalters@0:   """
tomwalters@0:   return not line or line.isspace()
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForFunctionLengths(filename, clean_lines, linenum,
tomwalters@0:                             function_state, error):
tomwalters@0:   """Reports for long function bodies.
tomwalters@0: 
tomwalters@0:   For an overview why this is done, see:
tomwalters@0:   http://google-styleguide.googlecode.com/svn/trunk/cppguide.xml#Write_Short_Functions
tomwalters@0: 
tomwalters@0:   Uses a simplistic algorithm assuming other style guidelines
tomwalters@0:   (especially spacing) are followed.
tomwalters@0:   Only checks unindented functions, so class members are unchecked.
tomwalters@0:   Trivial bodies are unchecked, so constructors with huge initializer lists
tomwalters@0:   may be missed.
tomwalters@0:   Blank/comment lines are not counted so as to avoid encouraging the removal
tomwalters@0:   of vertical space and commments just to get through a lint check.
tomwalters@0:   NOLINT *on the last line of a function* disables this check.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     function_state: Current function name and lines in body so far.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   lines = clean_lines.lines
tomwalters@0:   line = lines[linenum]
tomwalters@0:   raw = clean_lines.raw_lines
tomwalters@0:   raw_line = raw[linenum]
tomwalters@0:   joined_line = ''
tomwalters@0: 
tomwalters@0:   starting_func = False
tomwalters@0:   regexp = r'(\w(\w|::|\*|\&|\s)*)\('  # decls * & space::name( ...
tomwalters@0:   match_result = Match(regexp, line)
tomwalters@0:   if match_result:
tomwalters@0:     # If the name is all caps and underscores, figure it's a macro and
tomwalters@0:     # ignore it, unless it's TEST or TEST_F.
tomwalters@0:     function_name = match_result.group(1).split()[-1]
tomwalters@0:     if function_name == 'TEST' or function_name == 'TEST_F' or (
tomwalters@0:         not Match(r'[A-Z_]+$', function_name)):
tomwalters@0:       starting_func = True
tomwalters@0: 
tomwalters@0:   if starting_func:
tomwalters@0:     body_found = False
tomwalters@0:     for start_linenum in xrange(linenum, clean_lines.NumLines()):
tomwalters@0:       start_line = lines[start_linenum]
tomwalters@0:       joined_line += ' ' + start_line.lstrip()
tomwalters@0:       if Search(r'(;|})', start_line):  # Declarations and trivial functions
tomwalters@0:         body_found = True
tomwalters@0:         break                              # ... ignore
tomwalters@0:       elif Search(r'{', start_line):
tomwalters@0:         body_found = True
tomwalters@0:         function = Search(r'((\w|:)*)\(', line).group(1)
tomwalters@0:         if Match(r'TEST', function):    # Handle TEST... macros
tomwalters@0:           parameter_regexp = Search(r'(\(.*\))', joined_line)
tomwalters@0:           if parameter_regexp:             # Ignore bad syntax
tomwalters@0:             function += parameter_regexp.group(1)
tomwalters@0:         else:
tomwalters@0:           function += '()'
tomwalters@0:         function_state.Begin(function)
tomwalters@0:         break
tomwalters@0:     if not body_found:
tomwalters@0:       # No body for the function (or evidence of a non-function) was found.
tomwalters@0:       error(filename, linenum, 'readability/fn_size', 5,
tomwalters@0:             'Lint failed to find start of function body.')
tomwalters@0:   elif Match(r'^\}\s*$', line):  # function end
tomwalters@0:     if not Search(r'\bNOLINT\b', raw_line):
tomwalters@0:       function_state.Check(error, filename, linenum)
tomwalters@0:     function_state.End()
tomwalters@0:   elif not Match(r'^\s*$', line):
tomwalters@0:     function_state.Count()  # Count non-blank/non-comment lines.
tomwalters@0: 
tomwalters@0: 
tomwalters@0: _RE_PATTERN_TODO = re.compile(r'^//(\s*)TODO(\(.+?\))?:?(\s|$)?')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckComment(comment, filename, linenum, error):
tomwalters@0:   """Checks for common mistakes in TODO comments.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     comment: The text of the comment from the line in question.
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   match = _RE_PATTERN_TODO.match(comment)
tomwalters@0:   if match:
tomwalters@0:     # One whitespace is correct; zero whitespace is handled elsewhere.
tomwalters@0:     leading_whitespace = match.group(1)
tomwalters@0:     if len(leading_whitespace) > 1:
tomwalters@0:       error(filename, linenum, 'whitespace/todo', 2,
tomwalters@0:             'Too many spaces before TODO')
tomwalters@0: 
tomwalters@0:     username = match.group(2)
tomwalters@0:     if not username:
tomwalters@0:       error(filename, linenum, 'readability/todo', 2,
tomwalters@0:             'Missing username in TODO; it should look like '
tomwalters@0:             '"// TODO(my_username): Stuff."')
tomwalters@0: 
tomwalters@0:     middle_whitespace = match.group(3)
tomwalters@0:     # Comparisons made explicit for correctness -- pylint: disable-msg=C6403
tomwalters@0:     if middle_whitespace != ' ' and middle_whitespace != '':
tomwalters@0:       error(filename, linenum, 'whitespace/todo', 2,
tomwalters@0:             'TODO(my_username) should be followed by a space')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckSpacing(filename, clean_lines, linenum, error):
tomwalters@0:   """Checks for the correctness of various spacing issues in the code.
tomwalters@0: 
tomwalters@0:   Things we check for: spaces around operators, spaces after
tomwalters@0:   if/for/while/switch, no spaces around parens in function calls, two
tomwalters@0:   spaces between code and comment, don't start a block with a blank
tomwalters@0:   line, don't end a function with a blank line, don't have too many
tomwalters@0:   blank lines in a row.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   raw = clean_lines.raw_lines
tomwalters@0:   line = raw[linenum]
tomwalters@0: 
tomwalters@0:   # Before nixing comments, check if the line is blank for no good
tomwalters@0:   # reason.  This includes the first line after a block is opened, and
tomwalters@0:   # blank lines at the end of a function (ie, right before a line like '}'
tomwalters@0:   if IsBlankLine(line):
tomwalters@0:     elided = clean_lines.elided
tomwalters@0:     prev_line = elided[linenum - 1]
tomwalters@0:     prevbrace = prev_line.rfind('{')
tomwalters@0:     # TODO(unknown): Don't complain if line before blank line, and line after,
tomwalters@0:     #                both start with alnums and are indented the same amount.
tomwalters@0:     #                This ignores whitespace at the start of a namespace block
tomwalters@0:     #                because those are not usually indented.
tomwalters@0:     if (prevbrace != -1 and prev_line[prevbrace:].find('}') == -1
tomwalters@0:         and prev_line[:prevbrace].find('namespace') == -1):
tomwalters@0:       # OK, we have a blank line at the start of a code block.  Before we
tomwalters@0:       # complain, we check if it is an exception to the rule: The previous
tomwalters@0:       # non-empty line has the paramters of a function header that are indented
tomwalters@0:       # 4 spaces (because they did not fit in a 80 column line when placed on
tomwalters@0:       # the same line as the function name).  We also check for the case where
tomwalters@0:       # the previous line is indented 6 spaces, which may happen when the
tomwalters@0:       # initializers of a constructor do not fit into a 80 column line.
tomwalters@0:       exception = False
tomwalters@0:       if Match(r' {6}\w', prev_line):  # Initializer list?
tomwalters@0:         # We are looking for the opening column of initializer list, which
tomwalters@0:         # should be indented 4 spaces to cause 6 space indentation afterwards.
tomwalters@0:         search_position = linenum-2
tomwalters@0:         while (search_position >= 0
tomwalters@0:                and Match(r' {6}\w', elided[search_position])):
tomwalters@0:           search_position -= 1
tomwalters@0:         exception = (search_position >= 0
tomwalters@0:                      and elided[search_position][:5] == '    :')
tomwalters@0:       else:
tomwalters@0:         # Search for the function arguments or an initializer list.  We use a
tomwalters@0:         # simple heuristic here: If the line is indented 4 spaces; and we have a
tomwalters@0:         # closing paren, without the opening paren, followed by an opening brace
tomwalters@0:         # or colon (for initializer lists) we assume that it is the last line of
tomwalters@0:         # a function header.  If we have a colon indented 4 spaces, it is an
tomwalters@0:         # initializer list.
tomwalters@0:         exception = (Match(r' {4}\w[^\(]*\)\s*(const\s*)?(\{\s*$|:)',
tomwalters@0:                            prev_line)
tomwalters@0:                      or Match(r' {4}:', prev_line))
tomwalters@0: 
tomwalters@0:       if not exception:
tomwalters@0:         error(filename, linenum, 'whitespace/blank_line', 2,
tomwalters@0:               'Blank line at the start of a code block.  Is this needed?')
tomwalters@0:     # This doesn't ignore whitespace at the end of a namespace block
tomwalters@0:     # because that is too hard without pairing open/close braces;
tomwalters@0:     # however, a special exception is made for namespace closing
tomwalters@0:     # brackets which have a comment containing "namespace".
tomwalters@0:     #
tomwalters@0:     # Also, ignore blank lines at the end of a block in a long if-else
tomwalters@0:     # chain, like this:
tomwalters@0:     #   if (condition1) {
tomwalters@0:     #     // Something followed by a blank line
tomwalters@0:     #
tomwalters@0:     #   } else if (condition2) {
tomwalters@0:     #     // Something else
tomwalters@0:     #   }
tomwalters@0:     if linenum + 1 < clean_lines.NumLines():
tomwalters@0:       next_line = raw[linenum + 1]
tomwalters@0:       if (next_line
tomwalters@0:           and Match(r'\s*}', next_line)
tomwalters@0:           and next_line.find('namespace') == -1
tomwalters@0:           and next_line.find('} else ') == -1):
tomwalters@0:         error(filename, linenum, 'whitespace/blank_line', 3,
tomwalters@0:               'Blank line at the end of a code block.  Is this needed?')
tomwalters@0: 
tomwalters@0:   # Next, we complain if there's a comment too near the text
tomwalters@0:   commentpos = line.find('//')
tomwalters@0:   if commentpos != -1:
tomwalters@0:     # Check if the // may be in quotes.  If so, ignore it
tomwalters@0:     # Comparisons made explicit for clarity -- pylint: disable-msg=C6403
tomwalters@0:     if (line.count('"', 0, commentpos) -
tomwalters@0:         line.count('\\"', 0, commentpos)) % 2 == 0:   # not in quotes
tomwalters@0:       # Allow one space for new scopes, two spaces otherwise:
tomwalters@0:       if (not Match(r'^\s*{ //', line) and
tomwalters@0:           ((commentpos >= 1 and
tomwalters@0:             line[commentpos-1] not in string.whitespace) or
tomwalters@0:            (commentpos >= 2 and
tomwalters@0:             line[commentpos-2] not in string.whitespace))):
tomwalters@0:         error(filename, linenum, 'whitespace/comments', 2,
tomwalters@0:               'At least two spaces is best between code and comments')
tomwalters@0:       # There should always be a space between the // and the comment
tomwalters@0:       commentend = commentpos + 2
tomwalters@0:       if commentend < len(line) and not line[commentend] == ' ':
tomwalters@0:         # but some lines are exceptions -- e.g. if they're big
tomwalters@0:         # comment delimiters like:
tomwalters@0:         # //----------------------------------------------------------
tomwalters@0:         # or they begin with multiple slashes followed by a space:
tomwalters@0:         # //////// Header comment
tomwalters@0:         match = (Search(r'[=/-]{4,}\s*$', line[commentend:]) or
tomwalters@0:                  Search(r'^/+ ', line[commentend:]))
tomwalters@0:         if not match:
tomwalters@0:           error(filename, linenum, 'whitespace/comments', 4,
tomwalters@0:                 'Should have a space between // and comment')
tomwalters@0:       CheckComment(line[commentpos:], filename, linenum, error)
tomwalters@0: 
tomwalters@0:   line = clean_lines.elided[linenum]  # get rid of comments and strings
tomwalters@0: 
tomwalters@0:   # Don't try to do spacing checks for operator methods
tomwalters@0:   line = re.sub(r'operator(==|!=|<|<<|<=|>=|>>|>)\(', 'operator\(', line)
tomwalters@0: 
tomwalters@0:   # We allow no-spaces around = within an if: "if ( (a=Foo()) == 0 )".
tomwalters@0:   # Otherwise not.  Note we only check for non-spaces on *both* sides;
tomwalters@0:   # sometimes people put non-spaces on one side when aligning ='s among
tomwalters@0:   # many lines (not that this is behavior that I approve of...)
tomwalters@0:   if Search(r'[\w.]=[\w.]', line) and not Search(r'\b(if|while) ', line):
tomwalters@0:     error(filename, linenum, 'whitespace/operators', 4,
tomwalters@0:           'Missing spaces around =')
tomwalters@0: 
tomwalters@0:   # It's ok not to have spaces around binary operators like + - * /, but if
tomwalters@0:   # there's too little whitespace, we get concerned.  It's hard to tell,
tomwalters@0:   # though, so we punt on this one for now.  TODO.
tomwalters@0: 
tomwalters@0:   # You should always have whitespace around binary operators.
tomwalters@0:   # Alas, we can't test < or > because they're legitimately used sans spaces
tomwalters@0:   # (a->b, vector<int> a).  The only time we can tell is a < with no >, and
tomwalters@0:   # only if it's not template params list spilling into the next line.
tomwalters@0:   match = Search(r'[^<>=!\s](==|!=|<=|>=)[^<>=!\s]', line)
tomwalters@0:   if not match:
tomwalters@0:     # Note that while it seems that the '<[^<]*' term in the following
tomwalters@0:     # regexp could be simplified to '<.*', which would indeed match
tomwalters@0:     # the same class of strings, the [^<] means that searching for the
tomwalters@0:     # regexp takes linear rather than quadratic time.
tomwalters@0:     if not Search(r'<[^<]*,\s*$', line):  # template params spill
tomwalters@0:       match = Search(r'[^<>=!\s](<)[^<>=!\s]([^>]|->)*$', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'whitespace/operators', 3,
tomwalters@0:           'Missing spaces around %s' % match.group(1))
tomwalters@0:   # We allow no-spaces around << and >> when used like this: 10<<20, but
tomwalters@0:   # not otherwise (particularly, not when used as streams)
tomwalters@0:   match = Search(r'[^0-9\s](<<|>>)[^0-9\s]', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'whitespace/operators', 3,
tomwalters@0:           'Missing spaces around %s' % match.group(1))
tomwalters@0: 
tomwalters@0:   # There shouldn't be space around unary operators
tomwalters@0:   match = Search(r'(!\s|~\s|[\s]--[\s;]|[\s]\+\+[\s;])', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'whitespace/operators', 4,
tomwalters@0:           'Extra space for operator %s' % match.group(1))
tomwalters@0: 
tomwalters@0:   # A pet peeve of mine: no spaces after an if, while, switch, or for
tomwalters@0:   match = Search(r' (if\(|for\(|while\(|switch\()', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'whitespace/parens', 5,
tomwalters@0:           'Missing space before ( in %s' % match.group(1))
tomwalters@0: 
tomwalters@0:   # For if/for/while/switch, the left and right parens should be
tomwalters@0:   # consistent about how many spaces are inside the parens, and
tomwalters@0:   # there should either be zero or one spaces inside the parens.
tomwalters@0:   # We don't want: "if ( foo)" or "if ( foo   )".
tomwalters@0:   # Exception: "for ( ; foo; bar)" and "for (foo; bar; )" are allowed.
tomwalters@0:   match = Search(r'\b(if|for|while|switch)\s*'
tomwalters@0:                  r'\(([ ]*)(.).*[^ ]+([ ]*)\)\s*{\s*$',
tomwalters@0:                  line)
tomwalters@0:   if match:
tomwalters@0:     if len(match.group(2)) != len(match.group(4)):
tomwalters@0:       if not (match.group(3) == ';' and
tomwalters@0:               len(match.group(2)) == 1 + len(match.group(4)) or
tomwalters@0:               not match.group(2) and Search(r'\bfor\s*\(.*; \)', line)):
tomwalters@0:         error(filename, linenum, 'whitespace/parens', 5,
tomwalters@0:               'Mismatching spaces inside () in %s' % match.group(1))
tomwalters@0:     if not len(match.group(2)) in [0, 1]:
tomwalters@0:       error(filename, linenum, 'whitespace/parens', 5,
tomwalters@0:             'Should have zero or one spaces inside ( and ) in %s' %
tomwalters@0:             match.group(1))
tomwalters@0: 
tomwalters@0:   # You should always have a space after a comma (either as fn arg or operator)
tomwalters@0:   if Search(r',[^\s]', line):
tomwalters@0:     error(filename, linenum, 'whitespace/comma', 3,
tomwalters@0:           'Missing space after ,')
tomwalters@0: 
tomwalters@0:   # Next we will look for issues with function calls.
tomwalters@0:   CheckSpacingForFunctionCall(filename, line, linenum, error)
tomwalters@0: 
tomwalters@0:   # Except after an opening paren, you should have spaces before your braces.
tomwalters@0:   # And since you should never have braces at the beginning of a line, this is
tomwalters@0:   # an easy test.
tomwalters@0:   if Search(r'[^ (]{', line):
tomwalters@0:     error(filename, linenum, 'whitespace/braces', 5,
tomwalters@0:           'Missing space before {')
tomwalters@0: 
tomwalters@0:   # Make sure '} else {' has spaces.
tomwalters@0:   if Search(r'}else', line):
tomwalters@0:     error(filename, linenum, 'whitespace/braces', 5,
tomwalters@0:           'Missing space before else')
tomwalters@0: 
tomwalters@0:   # You shouldn't have spaces before your brackets, except maybe after
tomwalters@0:   # 'delete []' or 'new char * []'.
tomwalters@0:   if Search(r'\w\s+\[', line) and not Search(r'delete\s+\[', line):
tomwalters@0:     error(filename, linenum, 'whitespace/braces', 5,
tomwalters@0:           'Extra space before [')
tomwalters@0: 
tomwalters@0:   # You shouldn't have a space before a semicolon at the end of the line.
tomwalters@0:   # There's a special case for "for" since the style guide allows space before
tomwalters@0:   # the semicolon there.
tomwalters@0:   if Search(r':\s*;\s*$', line):
tomwalters@0:     error(filename, linenum, 'whitespace/semicolon', 5,
tomwalters@0:           'Semicolon defining empty statement. Use { } instead.')
tomwalters@0:   elif Search(r'^\s*;\s*$', line):
tomwalters@0:     error(filename, linenum, 'whitespace/semicolon', 5,
tomwalters@0:           'Line contains only semicolon. If this should be an empty statement, '
tomwalters@0:           'use { } instead.')
tomwalters@0:   elif (Search(r'\s+;\s*$', line) and
tomwalters@0:         not Search(r'\bfor\b', line)):
tomwalters@0:     error(filename, linenum, 'whitespace/semicolon', 5,
tomwalters@0:           'Extra space before last semicolon. If this should be an empty '
tomwalters@0:           'statement, use { } instead.')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def GetPreviousNonBlankLine(clean_lines, linenum):
tomwalters@0:   """Return the most recent non-blank line and its line number.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     clean_lines: A CleansedLines instance containing the file contents.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     A tuple with two elements.  The first element is the contents of the last
tomwalters@0:     non-blank line before the current line, or the empty string if this is the
tomwalters@0:     first non-blank line.  The second is the line number of that line, or -1
tomwalters@0:     if this is the first non-blank line.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   prevlinenum = linenum - 1
tomwalters@0:   while prevlinenum >= 0:
tomwalters@0:     prevline = clean_lines.elided[prevlinenum]
tomwalters@0:     if not IsBlankLine(prevline):     # if not a blank line...
tomwalters@0:       return (prevline, prevlinenum)
tomwalters@0:     prevlinenum -= 1
tomwalters@0:   return ('', -1)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckBraces(filename, clean_lines, linenum, error):
tomwalters@0:   """Looks for misplaced braces (e.g. at the end of line).
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   line = clean_lines.elided[linenum]        # get rid of comments and strings
tomwalters@0: 
tomwalters@0:   if Match(r'\s*{\s*$', line):
tomwalters@0:     # We allow an open brace to start a line in the case where someone
tomwalters@0:     # is using braces in a block to explicitly create a new scope,
tomwalters@0:     # which is commonly used to control the lifetime of
tomwalters@0:     # stack-allocated variables.  We don't detect this perfectly: we
tomwalters@0:     # just don't complain if the last non-whitespace character on the
tomwalters@0:     # previous non-blank line is ';', ':', '{', or '}'.
tomwalters@0:     prevline = GetPreviousNonBlankLine(clean_lines, linenum)[0]
tomwalters@0:     if not Search(r'[;:}{]\s*$', prevline):
tomwalters@0:       error(filename, linenum, 'whitespace/braces', 4,
tomwalters@0:             '{ should almost always be at the end of the previous line')
tomwalters@0: 
tomwalters@0:   # An else clause should be on the same line as the preceding closing brace.
tomwalters@0:   if Match(r'\s*else\s*', line):
tomwalters@0:     prevline = GetPreviousNonBlankLine(clean_lines, linenum)[0]
tomwalters@0:     if Match(r'\s*}\s*$', prevline):
tomwalters@0:       error(filename, linenum, 'whitespace/newline', 4,
tomwalters@0:             'An else should appear on the same line as the preceding }')
tomwalters@0: 
tomwalters@0:   # If braces come on one side of an else, they should be on both.
tomwalters@0:   # However, we have to worry about "else if" that spans multiple lines!
tomwalters@0:   if Search(r'}\s*else[^{]*$', line) or Match(r'[^}]*else\s*{', line):
tomwalters@0:     if Search(r'}\s*else if([^{]*)$', line):       # could be multi-line if
tomwalters@0:       # find the ( after the if
tomwalters@0:       pos = line.find('else if')
tomwalters@0:       pos = line.find('(', pos)
tomwalters@0:       if pos > 0:
tomwalters@0:         (endline, _, endpos) = CloseExpression(clean_lines, linenum, pos)
tomwalters@0:         if endline[endpos:].find('{') == -1:    # must be brace after if
tomwalters@0:           error(filename, linenum, 'readability/braces', 5,
tomwalters@0:                 'If an else has a brace on one side, it should have it on both')
tomwalters@0:     else:            # common case: else not followed by a multi-line if
tomwalters@0:       error(filename, linenum, 'readability/braces', 5,
tomwalters@0:             'If an else has a brace on one side, it should have it on both')
tomwalters@0: 
tomwalters@0:   # Likewise, an else should never have the else clause on the same line
tomwalters@0:   if Search(r'\belse [^\s{]', line) and not Search(r'\belse if\b', line):
tomwalters@0:     error(filename, linenum, 'whitespace/newline', 4,
tomwalters@0:           'Else clause should never be on same line as else (use 2 lines)')
tomwalters@0: 
tomwalters@0:   # In the same way, a do/while should never be on one line
tomwalters@0:   if Match(r'\s*do [^\s{]', line):
tomwalters@0:     error(filename, linenum, 'whitespace/newline', 4,
tomwalters@0:           'do/while clauses should not be on a single line')
tomwalters@0: 
tomwalters@0:   # Braces shouldn't be followed by a ; unless they're defining a struct
tomwalters@0:   # or initializing an array.
tomwalters@0:   # We can't tell in general, but we can for some common cases.
tomwalters@0:   prevlinenum = linenum
tomwalters@0:   while True:
tomwalters@0:     (prevline, prevlinenum) = GetPreviousNonBlankLine(clean_lines, prevlinenum)
tomwalters@0:     if Match(r'\s+{.*}\s*;', line) and not prevline.count(';'):
tomwalters@0:       line = prevline + line
tomwalters@0:     else:
tomwalters@0:       break
tomwalters@0:   if (Search(r'{.*}\s*;', line) and
tomwalters@0:       line.count('{') == line.count('}') and
tomwalters@0:       not Search(r'struct|class|enum|\s*=\s*{', line)):
tomwalters@0:     error(filename, linenum, 'readability/braces', 4,
tomwalters@0:           "You don't need a ; after a }")
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def ReplaceableCheck(operator, macro, line):
tomwalters@0:   """Determine whether a basic CHECK can be replaced with a more specific one.
tomwalters@0: 
tomwalters@0:   For example suggest using CHECK_EQ instead of CHECK(a == b) and
tomwalters@0:   similarly for CHECK_GE, CHECK_GT, CHECK_LE, CHECK_LT, CHECK_NE.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     operator: The C++ operator used in the CHECK.
tomwalters@0:     macro: The CHECK or EXPECT macro being called.
tomwalters@0:     line: The current source line.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     True if the CHECK can be replaced with a more specific one.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   # This matches decimal and hex integers, strings, and chars (in that order).
tomwalters@0:   match_constant = r'([-+]?(\d+|0[xX][0-9a-fA-F]+)[lLuU]{0,3}|".*"|\'.*\')'
tomwalters@0: 
tomwalters@0:   # Expression to match two sides of the operator with something that
tomwalters@0:   # looks like a literal, since CHECK(x == iterator) won't compile.
tomwalters@0:   # This means we can't catch all the cases where a more specific
tomwalters@0:   # CHECK is possible, but it's less annoying than dealing with
tomwalters@0:   # extraneous warnings.
tomwalters@0:   match_this = (r'\s*' + macro + r'\((\s*' +
tomwalters@0:                 match_constant + r'\s*' + operator + r'[^<>].*|'
tomwalters@0:                 r'.*[^<>]' + operator + r'\s*' + match_constant +
tomwalters@0:                 r'\s*\))')
tomwalters@0: 
tomwalters@0:   # Don't complain about CHECK(x == NULL) or similar because
tomwalters@0:   # CHECK_EQ(x, NULL) won't compile (requires a cast).
tomwalters@0:   # Also, don't complain about more complex boolean expressions
tomwalters@0:   # involving && or || such as CHECK(a == b || c == d).
tomwalters@0:   return Match(match_this, line) and not Search(r'NULL|&&|\|\|', line)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckCheck(filename, clean_lines, linenum, error):
tomwalters@0:   """Checks the use of CHECK and EXPECT macros.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   # Decide the set of replacement macros that should be suggested
tomwalters@0:   raw_lines = clean_lines.raw_lines
tomwalters@0:   current_macro = ''
tomwalters@0:   for macro in _CHECK_MACROS:
tomwalters@0:     if raw_lines[linenum].find(macro) >= 0:
tomwalters@0:       current_macro = macro
tomwalters@0:       break
tomwalters@0:   if not current_macro:
tomwalters@0:     # Don't waste time here if line doesn't contain 'CHECK' or 'EXPECT'
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   line = clean_lines.elided[linenum]        # get rid of comments and strings
tomwalters@0: 
tomwalters@0:   # Encourage replacing plain CHECKs with CHECK_EQ/CHECK_NE/etc.
tomwalters@0:   for operator in ['==', '!=', '>=', '>', '<=', '<']:
tomwalters@0:     if ReplaceableCheck(operator, current_macro, line):
tomwalters@0:       error(filename, linenum, 'readability/check', 2,
tomwalters@0:             'Consider using %s instead of %s(a %s b)' % (
tomwalters@0:                 _CHECK_REPLACEMENT[current_macro][operator],
tomwalters@0:                 current_macro, operator))
tomwalters@0:       break
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def GetLineWidth(line):
tomwalters@0:   """Determines the width of the line in column positions.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     line: A string, which may be a Unicode string.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     The width of the line in column positions, accounting for Unicode
tomwalters@0:     combining characters and wide characters.
tomwalters@0:   """
tomwalters@0:   if isinstance(line, unicode):
tomwalters@0:     width = 0
tomwalters@0:     for c in unicodedata.normalize('NFC', line):
tomwalters@0:       if unicodedata.east_asian_width(c) in ('W', 'F'):
tomwalters@0:         width += 2
tomwalters@0:       elif not unicodedata.combining(c):
tomwalters@0:         width += 1
tomwalters@0:     return width
tomwalters@0:   else:
tomwalters@0:     return len(line)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckStyle(filename, clean_lines, linenum, file_extension, error):
tomwalters@0:   """Checks rules from the 'C++ style rules' section of cppguide.html.
tomwalters@0: 
tomwalters@0:   Most of these rules are hard to test (naming, comment style), but we
tomwalters@0:   do what we can.  In particular we check for 2-space indents, line lengths,
tomwalters@0:   tab usage, spaces inside code, etc.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     file_extension: The extension (without the dot) of the filename.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   raw_lines = clean_lines.raw_lines
tomwalters@0:   line = raw_lines[linenum]
tomwalters@0: 
tomwalters@0:   if line.find('\t') != -1:
tomwalters@0:     error(filename, linenum, 'whitespace/tab', 1,
tomwalters@0:           'Tab found; better to use spaces')
tomwalters@0: 
tomwalters@0:   # One or three blank spaces at the beginning of the line is weird; it's
tomwalters@0:   # hard to reconcile that with 2-space indents.
tomwalters@0:   # NOTE: here are the conditions rob pike used for his tests.  Mine aren't
tomwalters@0:   # as sophisticated, but it may be worth becoming so:  RLENGTH==initial_spaces
tomwalters@0:   # if(RLENGTH > 20) complain = 0;
tomwalters@0:   # if(match($0, " +(error|private|public|protected):")) complain = 0;
tomwalters@0:   # if(match(prev, "&& *$")) complain = 0;
tomwalters@0:   # if(match(prev, "\\|\\| *$")) complain = 0;
tomwalters@0:   # if(match(prev, "[\",=><] *$")) complain = 0;
tomwalters@0:   # if(match($0, " <<")) complain = 0;
tomwalters@0:   # if(match(prev, " +for \\(")) complain = 0;
tomwalters@0:   # if(prevodd && match(prevprev, " +for \\(")) complain = 0;
tomwalters@0:   initial_spaces = 0
tomwalters@0:   cleansed_line = clean_lines.elided[linenum]
tomwalters@0:   while initial_spaces < len(line) and line[initial_spaces] == ' ':
tomwalters@0:     initial_spaces += 1
tomwalters@0:   if line and line[-1].isspace():
tomwalters@0:     error(filename, linenum, 'whitespace/end_of_line', 4,
tomwalters@0:           'Line ends in whitespace.  Consider deleting these extra spaces.')
tomwalters@0:   # There are certain situations we allow one space, notably for labels
tomwalters@0:   elif ((initial_spaces == 1 or initial_spaces == 3) and
tomwalters@0:         not Match(r'\s*\w+\s*:\s*$', cleansed_line)):
tomwalters@0:     error(filename, linenum, 'whitespace/indent', 3,
tomwalters@0:           'Weird number of spaces at line-start.  '
tomwalters@0:           'Are you using a 2-space indent?')
tomwalters@0:   # Labels should always be indented at least one space.
tomwalters@0:   elif not initial_spaces and line[:2] != '//' and Search(r'[^:]:\s*$',
tomwalters@0:                                                           line):
tomwalters@0:     error(filename, linenum, 'whitespace/labels', 4,
tomwalters@0:           'Labels should always be indented at least one space.  '
tomwalters@0:           'If this is a member-initializer list in a constructor, '
tomwalters@0:           'the colon should be on the line after the definition header.')
tomwalters@0: 
tomwalters@0:   # Check if the line is a header guard.
tomwalters@0:   is_header_guard = False
tomwalters@0:   if file_extension == 'h':
tomwalters@0:     cppvar = GetHeaderGuardCPPVariable(filename)
tomwalters@0:     if (line.startswith('#ifndef %s' % cppvar) or
tomwalters@0:         line.startswith('#define %s' % cppvar) or
tomwalters@0:         line.startswith('#endif  // %s' % cppvar)):
tomwalters@0:       is_header_guard = True
tomwalters@0:   # #include lines and header guards can be long, since there's no clean way to
tomwalters@0:   # split them.
tomwalters@0:   #
tomwalters@0:   # URLs can be long too.  It's possible to split these, but it makes them
tomwalters@0:   # harder to cut&paste.
tomwalters@0:   if (not line.startswith('#include') and not is_header_guard and
tomwalters@0:       not Match(r'^\s*//.*http(s?)://\S*$', line)):
tomwalters@0:     line_width = GetLineWidth(line)
tomwalters@0:     if line_width > 100:
tomwalters@0:       error(filename, linenum, 'whitespace/line_length', 4,
tomwalters@0:             'Lines should very rarely be longer than 100 characters')
tomwalters@0:     elif line_width > 80:
tomwalters@0:       error(filename, linenum, 'whitespace/line_length', 2,
tomwalters@0:             'Lines should be <= 80 characters long')
tomwalters@0: 
tomwalters@0:   if (cleansed_line.count(';') > 1 and
tomwalters@0:       # for loops are allowed two ;'s (and may run over two lines).
tomwalters@0:       cleansed_line.find('for') == -1 and
tomwalters@0:       (GetPreviousNonBlankLine(clean_lines, linenum)[0].find('for') == -1 or
tomwalters@0:        GetPreviousNonBlankLine(clean_lines, linenum)[0].find(';') != -1) and
tomwalters@0:       # It's ok to have many commands in a switch case that fits in 1 line
tomwalters@0:       not ((cleansed_line.find('case ') != -1 or
tomwalters@0:             cleansed_line.find('default:') != -1) and
tomwalters@0:            cleansed_line.find('break;') != -1)):
tomwalters@0:     error(filename, linenum, 'whitespace/newline', 4,
tomwalters@0:           'More than one command on the same line')
tomwalters@0: 
tomwalters@0:   # Some more style checks
tomwalters@0:   CheckBraces(filename, clean_lines, linenum, error)
tomwalters@0:   CheckSpacing(filename, clean_lines, linenum, error)
tomwalters@0:   CheckCheck(filename, clean_lines, linenum, error)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: _RE_PATTERN_INCLUDE_NEW_STYLE = re.compile(r'#include +"[^/]+\.h"')
tomwalters@0: _RE_PATTERN_INCLUDE = re.compile(r'^\s*#\s*include\s*([<"])([^>"]*)[>"].*$')
tomwalters@0: # Matches the first component of a filename delimited by -s and _s. That is:
tomwalters@0: #  _RE_FIRST_COMPONENT.match('foo').group(0) == 'foo'
tomwalters@0: #  _RE_FIRST_COMPONENT.match('foo.cc').group(0) == 'foo'
tomwalters@0: #  _RE_FIRST_COMPONENT.match('foo-bar_baz.cc').group(0) == 'foo'
tomwalters@0: #  _RE_FIRST_COMPONENT.match('foo_bar-baz.cc').group(0) == 'foo'
tomwalters@0: _RE_FIRST_COMPONENT = re.compile(r'^[^-_.]+')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _DropCommonSuffixes(filename):
tomwalters@0:   """Drops common suffixes like _test.cc or -inl.h from filename.
tomwalters@0: 
tomwalters@0:   For example:
tomwalters@0:     >>> _DropCommonSuffixes('foo/foo-inl.h')
tomwalters@0:     'foo/foo'
tomwalters@0:     >>> _DropCommonSuffixes('foo/bar/foo.cc')
tomwalters@0:     'foo/bar/foo'
tomwalters@0:     >>> _DropCommonSuffixes('foo/foo_internal.h')
tomwalters@0:     'foo/foo'
tomwalters@0:     >>> _DropCommonSuffixes('foo/foo_unusualinternal.h')
tomwalters@0:     'foo/foo_unusualinternal'
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The input filename.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     The filename with the common suffix removed.
tomwalters@0:   """
tomwalters@0:   for suffix in ('test.cc', 'regtest.cc', 'unittest.cc',
tomwalters@0:                  'inl.h', 'impl.h', 'internal.h'):
tomwalters@0:     if (filename.endswith(suffix) and len(filename) > len(suffix) and
tomwalters@0:         filename[-len(suffix) - 1] in ('-', '_')):
tomwalters@0:       return filename[:-len(suffix) - 1]
tomwalters@0:   return os.path.splitext(filename)[0]
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _IsTestFilename(filename):
tomwalters@0:   """Determines if the given filename has a suffix that identifies it as a test.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The input filename.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     True if 'filename' looks like a test, False otherwise.
tomwalters@0:   """
tomwalters@0:   if (filename.endswith('_test.cc') or
tomwalters@0:       filename.endswith('_unittest.cc') or
tomwalters@0:       filename.endswith('_regtest.cc')):
tomwalters@0:     return True
tomwalters@0:   else:
tomwalters@0:     return False
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def _ClassifyInclude(fileinfo, include, is_system):
tomwalters@0:   """Figures out what kind of header 'include' is.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     fileinfo: The current file cpplint is running over. A FileInfo instance.
tomwalters@0:     include: The path to a #included file.
tomwalters@0:     is_system: True if the #include used <> rather than "".
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     One of the _XXX_HEADER constants.
tomwalters@0: 
tomwalters@0:   For example:
tomwalters@0:     >>> _ClassifyInclude(FileInfo('foo/foo.cc'), 'stdio.h', True)
tomwalters@0:     _C_SYS_HEADER
tomwalters@0:     >>> _ClassifyInclude(FileInfo('foo/foo.cc'), 'string', True)
tomwalters@0:     _CPP_SYS_HEADER
tomwalters@0:     >>> _ClassifyInclude(FileInfo('foo/foo.cc'), 'foo/foo.h', False)
tomwalters@0:     _LIKELY_MY_HEADER
tomwalters@0:     >>> _ClassifyInclude(FileInfo('foo/foo_unknown_extension.cc'),
tomwalters@0:     ...                  'bar/foo_other_ext.h', False)
tomwalters@0:     _POSSIBLE_MY_HEADER
tomwalters@0:     >>> _ClassifyInclude(FileInfo('foo/foo.cc'), 'foo/bar.h', False)
tomwalters@0:     _OTHER_HEADER
tomwalters@0:   """
tomwalters@0:   # This is a list of all standard c++ header files, except
tomwalters@0:   # those already checked for above.
tomwalters@0:   is_stl_h = include in _STL_HEADERS
tomwalters@0:   is_cpp_h = is_stl_h or include in _CPP_HEADERS
tomwalters@0: 
tomwalters@0:   if is_system:
tomwalters@0:     if is_cpp_h:
tomwalters@0:       return _CPP_SYS_HEADER
tomwalters@0:     else:
tomwalters@0:       return _C_SYS_HEADER
tomwalters@0: 
tomwalters@0:   # If the target file and the include we're checking share a
tomwalters@0:   # basename when we drop common extensions, and the include
tomwalters@0:   # lives in . , then it's likely to be owned by the target file.
tomwalters@0:   target_dir, target_base = (
tomwalters@0:       os.path.split(_DropCommonSuffixes(fileinfo.RepositoryName())))
tomwalters@0:   include_dir, include_base = os.path.split(_DropCommonSuffixes(include))
tomwalters@0:   if target_base == include_base and (
tomwalters@0:       include_dir == target_dir or
tomwalters@0:       include_dir == os.path.normpath(target_dir + '/../public')):
tomwalters@0:     return _LIKELY_MY_HEADER
tomwalters@0: 
tomwalters@0:   # If the target and include share some initial basename
tomwalters@0:   # component, it's possible the target is implementing the
tomwalters@0:   # include, so it's allowed to be first, but we'll never
tomwalters@0:   # complain if it's not there.
tomwalters@0:   target_first_component = _RE_FIRST_COMPONENT.match(target_base)
tomwalters@0:   include_first_component = _RE_FIRST_COMPONENT.match(include_base)
tomwalters@0:   if (target_first_component and include_first_component and
tomwalters@0:       target_first_component.group(0) ==
tomwalters@0:       include_first_component.group(0)):
tomwalters@0:     return _POSSIBLE_MY_HEADER
tomwalters@0: 
tomwalters@0:   return _OTHER_HEADER
tomwalters@0: 
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckIncludeLine(filename, clean_lines, linenum, include_state, error):
tomwalters@0:   """Check rules that are applicable to #include lines.
tomwalters@0: 
tomwalters@0:   Strings on #include lines are NOT removed from elided line, to make
tomwalters@0:   certain tasks easier. However, to prevent false positives, checks
tomwalters@0:   applicable to #include lines in CheckLanguage must be put here.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     include_state: An _IncludeState instance in which the headers are inserted.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   fileinfo = FileInfo(filename)
tomwalters@0: 
tomwalters@0:   line = clean_lines.lines[linenum]
tomwalters@0: 
tomwalters@0:   # "include" should use the new style "foo/bar.h" instead of just "bar.h"
tomwalters@0:   if _RE_PATTERN_INCLUDE_NEW_STYLE.search(line):
tomwalters@0:     error(filename, linenum, 'build/include', 4,
tomwalters@0:           'Include the directory when naming .h files')
tomwalters@0: 
tomwalters@0:   # we shouldn't include a file more than once. actually, there are a
tomwalters@0:   # handful of instances where doing so is okay, but in general it's
tomwalters@0:   # not.
tomwalters@0:   match = _RE_PATTERN_INCLUDE.search(line)
tomwalters@0:   if match:
tomwalters@0:     include = match.group(2)
tomwalters@0:     is_system = (match.group(1) == '<')
tomwalters@0:     if include in include_state:
tomwalters@0:       error(filename, linenum, 'build/include', 4,
tomwalters@0:             '"%s" already included at %s:%s' %
tomwalters@0:             (include, filename, include_state[include]))
tomwalters@0:     else:
tomwalters@0:       include_state[include] = linenum
tomwalters@0: 
tomwalters@0:       # We want to ensure that headers appear in the right order:
tomwalters@0:       # 1) for foo.cc, foo.h  (preferred location)
tomwalters@0:       # 2) c system files
tomwalters@0:       # 3) cpp system files
tomwalters@0:       # 4) for foo.cc, foo.h  (deprecated location)
tomwalters@0:       # 5) other google headers
tomwalters@0:       #
tomwalters@0:       # We classify each include statement as one of those 5 types
tomwalters@0:       # using a number of techniques. The include_state object keeps
tomwalters@0:       # track of the highest type seen, and complains if we see a
tomwalters@0:       # lower type after that.
tomwalters@0:       error_message = include_state.CheckNextIncludeOrder(
tomwalters@0:           _ClassifyInclude(fileinfo, include, is_system))
tomwalters@0:       if error_message:
tomwalters@0:         error(filename, linenum, 'build/include_order', 4,
tomwalters@0:               '%s. Should be: %s.h, c system, c++ system, other.' %
tomwalters@0:               (error_message, fileinfo.BaseName()))
tomwalters@0:       if not include_state.IsInAlphabeticalOrder(include):
tomwalters@0:         error(filename, linenum, 'build/include_alpha', 4,
tomwalters@0:               'Include "%s" not in alphabetical order' % include)
tomwalters@0: 
tomwalters@0:   # Look for any of the stream classes that are part of standard C++.
tomwalters@0:   match = _RE_PATTERN_INCLUDE.match(line)
tomwalters@0:   if match:
tomwalters@0:     include = match.group(2)
tomwalters@0:     if Match(r'(f|ind|io|i|o|parse|pf|stdio|str|)?stream$', include):
tomwalters@0:       # Many unit tests use cout, so we exempt them.
tomwalters@0:       if not _IsTestFilename(filename):
tomwalters@0:         error(filename, linenum, 'readability/streams', 3,
tomwalters@0:               'Streams are highly discouraged.')
tomwalters@0: 
tomwalters@0: def CheckLanguage(filename, clean_lines, linenum, file_extension, include_state,
tomwalters@0:                   error):
tomwalters@0:   """Checks rules from the 'C++ language rules' section of cppguide.html.
tomwalters@0: 
tomwalters@0:   Some of these rules are hard to test (function overloading, using
tomwalters@0:   uint32 inappropriately), but we do the best we can.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     file_extension: The extension (without the dot) of the filename.
tomwalters@0:     include_state: An _IncludeState instance in which the headers are inserted.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   # If the line is empty or consists of entirely a comment, no need to
tomwalters@0:   # check it.
tomwalters@0:   line = clean_lines.elided[linenum]
tomwalters@0:   if not line:
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   match = _RE_PATTERN_INCLUDE.search(line)
tomwalters@0:   if match:
tomwalters@0:     CheckIncludeLine(filename, clean_lines, linenum, include_state, error)
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # Create an extended_line, which is the concatenation of the current and
tomwalters@0:   # next lines, for more effective checking of code that may span more than one
tomwalters@0:   # line.
tomwalters@0:   if linenum + 1 < clean_lines.NumLines():
tomwalters@0:     extended_line = line + clean_lines.elided[linenum + 1]
tomwalters@0:   else:
tomwalters@0:     extended_line = line
tomwalters@0: 
tomwalters@0:   # Make Windows paths like Unix.
tomwalters@0:   fullname = os.path.abspath(filename).replace('\\', '/')
tomwalters@0: 
tomwalters@0:   # TODO(unknown): figure out if they're using default arguments in fn proto.
tomwalters@0: 
tomwalters@0:   # Check for non-const references in functions.  This is tricky because &
tomwalters@0:   # is also used to take the address of something.  We allow <> for templates,
tomwalters@0:   # (ignoring whatever is between the braces) and : for classes.
tomwalters@0:   # These are complicated re's.  They try to capture the following:
tomwalters@0:   # paren (for fn-prototype start), typename, &, varname.  For the const
tomwalters@0:   # version, we're willing for const to be before typename or after
tomwalters@0:   # Don't check the implemention on same line.
tomwalters@0:   fnline = line.split('{', 1)[0]
tomwalters@0:   if (len(re.findall(r'\([^()]*\b(?:[\w:]|<[^()]*>)+(\s?&|&\s?)\w+', fnline)) >
tomwalters@0:       len(re.findall(r'\([^()]*\bconst\s+(?:typename\s+)?(?:struct\s+)?'
tomwalters@0:                      r'(?:[\w:]|<[^()]*>)+(\s?&|&\s?)\w+', fnline)) +
tomwalters@0:       len(re.findall(r'\([^()]*\b(?:[\w:]|<[^()]*>)+\s+const(\s?&|&\s?)[\w]+',
tomwalters@0:                      fnline))):
tomwalters@0: 
tomwalters@0:     # We allow non-const references in a few standard places, like functions
tomwalters@0:     # called "swap()" or iostream operators like "<<" or ">>".
tomwalters@0:     if not Search(
tomwalters@0:         r'(swap|Swap|operator[<>][<>])\s*\(\s*(?:[\w:]|<.*>)+\s*&',
tomwalters@0:         fnline):
tomwalters@0:       error(filename, linenum, 'runtime/references', 2,
tomwalters@0:             'Is this a non-const reference? '
tomwalters@0:             'If so, make const or use a pointer.')
tomwalters@0: 
tomwalters@0:   # Check to see if they're using an conversion function cast.
tomwalters@0:   # I just try to capture the most common basic types, though there are more.
tomwalters@0:   # Parameterless conversion functions, such as bool(), are allowed as they are
tomwalters@0:   # probably a member operator declaration or default constructor.
tomwalters@0:   match = Search(
tomwalters@0:       r'(\bnew\s+)?\b'  # Grab 'new' operator, if it's there
tomwalters@0:       r'(int|float|double|bool|char|int32|uint32|int64|uint64)\([^)]', line)
tomwalters@0:   if match:
tomwalters@0:     # gMock methods are defined using some variant of MOCK_METHODx(name, type)
tomwalters@0:     # where type may be float(), int(string), etc.  Without context they are
tomwalters@0:     # virtually indistinguishable from int(x) casts.
tomwalters@0:     if (match.group(1) is None and  # If new operator, then this isn't a cast
tomwalters@0:         not Match(r'^\s*MOCK_(CONST_)?METHOD\d+(_T)?\(', line)):
tomwalters@0:       error(filename, linenum, 'readability/casting', 4,
tomwalters@0:             'Using deprecated casting style.  '
tomwalters@0:             'Use static_cast<%s>(...) instead' %
tomwalters@0:             match.group(2))
tomwalters@0: 
tomwalters@0:   CheckCStyleCast(filename, linenum, line, clean_lines.raw_lines[linenum],
tomwalters@0:                   'static_cast',
tomwalters@0:                   r'\((int|float|double|bool|char|u?int(16|32|64))\)',
tomwalters@0:                   error)
tomwalters@0:   # This doesn't catch all cases.  Consider (const char * const)"hello".
tomwalters@0:   CheckCStyleCast(filename, linenum, line, clean_lines.raw_lines[linenum],
tomwalters@0:                   'reinterpret_cast', r'\((\w+\s?\*+\s?)\)', error)
tomwalters@0: 
tomwalters@0:   # In addition, we look for people taking the address of a cast.  This
tomwalters@0:   # is dangerous -- casts can assign to temporaries, so the pointer doesn't
tomwalters@0:   # point where you think.
tomwalters@0:   if Search(
tomwalters@0:       r'(&\([^)]+\)[\w(])|(&(static|dynamic|reinterpret)_cast\b)', line):
tomwalters@0:     error(filename, linenum, 'runtime/casting', 4,
tomwalters@0:           ('Are you taking an address of a cast?  '
tomwalters@0:            'This is dangerous: could be a temp var.  '
tomwalters@0:            'Take the address before doing the cast, rather than after'))
tomwalters@0: 
tomwalters@0:   # Check for people declaring static/global STL strings at the top level.
tomwalters@0:   # This is dangerous because the C++ language does not guarantee that
tomwalters@0:   # globals with constructors are initialized before the first access.
tomwalters@0:   match = Match(
tomwalters@0:       r'((?:|static +)(?:|const +))string +([a-zA-Z0-9_:]+)\b(.*)',
tomwalters@0:       line)
tomwalters@0:   # Make sure it's not a function.
tomwalters@0:   # Function template specialization looks like: "string foo<Type>(...".
tomwalters@0:   # Class template definitions look like: "string Foo<Type>::Method(...".
tomwalters@0:   if match and not Match(r'\s*(<.*>)?(::[a-zA-Z0-9_]+)?\s*\(([^"]|$)',
tomwalters@0:                          match.group(3)):
tomwalters@0:     error(filename, linenum, 'runtime/string', 4,
tomwalters@0:           'For a static/global string constant, use a C style string instead: '
tomwalters@0:           '"%schar %s[]".' %
tomwalters@0:           (match.group(1), match.group(2)))
tomwalters@0: 
tomwalters@0:   # Check that we're not using RTTI outside of testing code.
tomwalters@0:   if Search(r'\bdynamic_cast<', line) and not _IsTestFilename(filename):
tomwalters@0:     error(filename, linenum, 'runtime/rtti', 5,
tomwalters@0:           'Do not use dynamic_cast<>.  If you need to cast within a class '
tomwalters@0:           "hierarchy, use static_cast<> to upcast.  Google doesn't support "
tomwalters@0:           'RTTI.')
tomwalters@0: 
tomwalters@0:   if Search(r'\b([A-Za-z0-9_]*_)\(\1\)', line):
tomwalters@0:     error(filename, linenum, 'runtime/init', 4,
tomwalters@0:           'You seem to be initializing a member variable with itself.')
tomwalters@0: 
tomwalters@0:   if file_extension == 'h':
tomwalters@0:     # TODO(unknown): check that 1-arg constructors are explicit.
tomwalters@0:     #                How to tell it's a constructor?
tomwalters@0:     #                (handled in CheckForNonStandardConstructs for now)
tomwalters@0:     # TODO(unknown): check that classes have DISALLOW_EVIL_CONSTRUCTORS
tomwalters@0:     #                (level 1 error)
tomwalters@0:     pass
tomwalters@0: 
tomwalters@0:   # Check if people are using the verboten C basic types.  The only exception
tomwalters@0:   # we regularly allow is "unsigned short port" for port.
tomwalters@0:   if Search(r'\bshort port\b', line):
tomwalters@0:     if not Search(r'\bunsigned short port\b', line):
tomwalters@0:       error(filename, linenum, 'runtime/int', 4,
tomwalters@0:             'Use "unsigned short" for ports, not "short"')
tomwalters@0:   else:
tomwalters@0:     match = Search(r'\b(short|long(?! +double)|long long)\b', line)
tomwalters@0:     if match:
tomwalters@0:       error(filename, linenum, 'runtime/int', 4,
tomwalters@0:             'Use int16/int64/etc, rather than the C type %s' % match.group(1))
tomwalters@0: 
tomwalters@0:   # When snprintf is used, the second argument shouldn't be a literal.
tomwalters@0:   match = Search(r'snprintf\s*\(([^,]*),\s*([0-9]*)\s*,', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'runtime/printf', 3,
tomwalters@0:           'If you can, use sizeof(%s) instead of %s as the 2nd arg '
tomwalters@0:           'to snprintf.' % (match.group(1), match.group(2)))
tomwalters@0: 
tomwalters@0:   # Check if some verboten C functions are being used.
tomwalters@0:   if Search(r'\bsprintf\b', line):
tomwalters@0:     error(filename, linenum, 'runtime/printf', 5,
tomwalters@0:           'Never use sprintf.  Use snprintf instead.')
tomwalters@0:   match = Search(r'\b(strcpy|strcat)\b', line)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'runtime/printf', 4,
tomwalters@0:           'Almost always, snprintf is better than %s' % match.group(1))
tomwalters@0: 
tomwalters@0:   if Search(r'\bsscanf\b', line):
tomwalters@0:     error(filename, linenum, 'runtime/printf', 1,
tomwalters@0:           'sscanf can be ok, but is slow and can overflow buffers.')
tomwalters@0: 
tomwalters@0:   # Check if some verboten operator overloading is going on
tomwalters@0:   # TODO(unknown): catch out-of-line unary operator&:
tomwalters@0:   #   class X {};
tomwalters@0:   #   int operator&(const X& x) { return 42; }  // unary operator&
tomwalters@0:   # The trick is it's hard to tell apart from binary operator&:
tomwalters@0:   #   class Y { int operator&(const Y& x) { return 23; } }; // binary operator&
tomwalters@0:   if Search(r'\boperator\s*&\s*\(\s*\)', line):
tomwalters@0:     error(filename, linenum, 'runtime/operator', 4,
tomwalters@0:           'Unary operator& is dangerous.  Do not use it.')
tomwalters@0: 
tomwalters@0:   # Check for suspicious usage of "if" like
tomwalters@0:   # } if (a == b) {
tomwalters@0:   if Search(r'\}\s*if\s*\(', line):
tomwalters@0:     error(filename, linenum, 'readability/braces', 4,
tomwalters@0:           'Did you mean "else if"? If not, start a new line for "if".')
tomwalters@0: 
tomwalters@0:   # Check for potential format string bugs like printf(foo).
tomwalters@0:   # We constrain the pattern not to pick things like DocidForPrintf(foo).
tomwalters@0:   # Not perfect but it can catch printf(foo.c_str()) and printf(foo->c_str())
tomwalters@0:   match = re.search(r'\b((?:string)?printf)\s*\(([\w.\->()]+)\)', line, re.I)
tomwalters@0:   if match:
tomwalters@0:     error(filename, linenum, 'runtime/printf', 4,
tomwalters@0:           'Potential format string bug. Do %s("%%s", %s) instead.'
tomwalters@0:           % (match.group(1), match.group(2)))
tomwalters@0: 
tomwalters@0:   # Check for potential memset bugs like memset(buf, sizeof(buf), 0).
tomwalters@0:   match = Search(r'memset\s*\(([^,]*),\s*([^,]*),\s*0\s*\)', line)
tomwalters@0:   if match and not Match(r"^''|-?[0-9]+|0x[0-9A-Fa-f]$", match.group(2)):
tomwalters@0:     error(filename, linenum, 'runtime/memset', 4,
tomwalters@0:           'Did you mean "memset(%s, 0, %s)"?'
tomwalters@0:           % (match.group(1), match.group(2)))
tomwalters@0: 
tomwalters@0:   if Search(r'\busing namespace\b', line):
tomwalters@0:     error(filename, linenum, 'build/namespaces', 5,
tomwalters@0:           'Do not use namespace using-directives.  '
tomwalters@0:           'Use using-declarations instead.')
tomwalters@0: 
tomwalters@0:   # Detect variable-length arrays.
tomwalters@0:   match = Match(r'\s*(.+::)?(\w+) [a-z]\w*\[(.+)];', line)
tomwalters@0:   if (match and match.group(2) != 'return' and match.group(2) != 'delete' and
tomwalters@0:       match.group(3).find(']') == -1):
tomwalters@0:     # Split the size using space and arithmetic operators as delimiters.
tomwalters@0:     # If any of the resulting tokens are not compile time constants then
tomwalters@0:     # report the error.
tomwalters@0:     tokens = re.split(r'\s|\+|\-|\*|\/|<<|>>]', match.group(3))
tomwalters@0:     is_const = True
tomwalters@0:     skip_next = False
tomwalters@0:     for tok in tokens:
tomwalters@0:       if skip_next:
tomwalters@0:         skip_next = False
tomwalters@0:         continue
tomwalters@0: 
tomwalters@0:       if Search(r'sizeof\(.+\)', tok): continue
tomwalters@0:       if Search(r'arraysize\(\w+\)', tok): continue
tomwalters@0: 
tomwalters@0:       tok = tok.lstrip('(')
tomwalters@0:       tok = tok.rstrip(')')
tomwalters@0:       if not tok: continue
tomwalters@0:       if Match(r'\d+', tok): continue
tomwalters@0:       if Match(r'0[xX][0-9a-fA-F]+', tok): continue
tomwalters@0:       if Match(r'k[A-Z0-9]\w*', tok): continue
tomwalters@0:       if Match(r'(.+::)?k[A-Z0-9]\w*', tok): continue
tomwalters@0:       if Match(r'(.+::)?[A-Z][A-Z0-9_]*', tok): continue
tomwalters@0:       # A catch all for tricky sizeof cases, including 'sizeof expression',
tomwalters@0:       # 'sizeof(*type)', 'sizeof(const type)', 'sizeof(struct StructName)'
tomwalters@0:       # requires skipping the next token becasue we split on ' ' and '*'.
tomwalters@0:       if tok.startswith('sizeof'):
tomwalters@0:         skip_next = True
tomwalters@0:         continue
tomwalters@0:       is_const = False
tomwalters@0:       break
tomwalters@0:     if not is_const:
tomwalters@0:       error(filename, linenum, 'runtime/arrays', 1,
tomwalters@0:             'Do not use variable-length arrays.  Use an appropriately named '
tomwalters@0:             "('k' followed by CamelCase) compile-time constant for the size.")
tomwalters@0: 
tomwalters@0:   # If DISALLOW_EVIL_CONSTRUCTORS, DISALLOW_COPY_AND_ASSIGN, or
tomwalters@0:   # DISALLOW_IMPLICIT_CONSTRUCTORS is present, then it should be the last thing
tomwalters@0:   # in the class declaration.
tomwalters@0:   match = Match(
tomwalters@0:       (r'\s*'
tomwalters@0:        r'(DISALLOW_(EVIL_CONSTRUCTORS|COPY_AND_ASSIGN|IMPLICIT_CONSTRUCTORS))'
tomwalters@0:        r'\(.*\);$'),
tomwalters@0:       line)
tomwalters@0:   if match and linenum + 1 < clean_lines.NumLines():
tomwalters@0:     next_line = clean_lines.elided[linenum + 1]
tomwalters@0:     if not Search(r'^\s*};', next_line):
tomwalters@0:       error(filename, linenum, 'readability/constructors', 3,
tomwalters@0:             match.group(1) + ' should be the last thing in the class')
tomwalters@0: 
tomwalters@0:   # Check for use of unnamed namespaces in header files.  Registration
tomwalters@0:   # macros are typically OK, so we allow use of "namespace {" on lines
tomwalters@0:   # that end with backslashes.
tomwalters@0:   if (file_extension == 'h'
tomwalters@0:       and Search(r'\bnamespace\s*{', line)
tomwalters@0:       and line[-1] != '\\'):
tomwalters@0:     error(filename, linenum, 'build/namespaces', 4,
tomwalters@0:           'Do not use unnamed namespaces in header files.  See '
tomwalters@0:           'http://google-styleguide.googlecode.com/svn/trunk/cppguide.xml#Namespaces'
tomwalters@0:           ' for more information.')
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckCStyleCast(filename, linenum, line, raw_line, cast_type, pattern,
tomwalters@0:                     error):
tomwalters@0:   """Checks for a C-style cast by looking for the pattern.
tomwalters@0: 
tomwalters@0:   This also handles sizeof(type) warnings, due to similarity of content.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     linenum: The number of the line to check.
tomwalters@0:     line: The line of code to check.
tomwalters@0:     raw_line: The raw line of code to check, with comments.
tomwalters@0:     cast_type: The string for the C++ cast to recommend.  This is either
tomwalters@0:       reinterpret_cast or static_cast, depending.
tomwalters@0:     pattern: The regular expression used to find C-style casts.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:   """
tomwalters@0:   match = Search(pattern, line)
tomwalters@0:   if not match:
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # e.g., sizeof(int)
tomwalters@0:   sizeof_match = Match(r'.*sizeof\s*$', line[0:match.start(1) - 1])
tomwalters@0:   if sizeof_match:
tomwalters@0:     error(filename, linenum, 'runtime/sizeof', 1,
tomwalters@0:           'Using sizeof(type).  Use sizeof(varname) instead if possible')
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   remainder = line[match.end(0):]
tomwalters@0: 
tomwalters@0:   # The close paren is for function pointers as arguments to a function.
tomwalters@0:   # eg, void foo(void (*bar)(int));
tomwalters@0:   # The semicolon check is a more basic function check; also possibly a
tomwalters@0:   # function pointer typedef.
tomwalters@0:   # eg, void foo(int); or void foo(int) const;
tomwalters@0:   # The equals check is for function pointer assignment.
tomwalters@0:   # eg, void *(*foo)(int) = ...
tomwalters@0:   #
tomwalters@0:   # Right now, this will only catch cases where there's a single argument, and
tomwalters@0:   # it's unnamed.  It should probably be expanded to check for multiple
tomwalters@0:   # arguments with some unnamed.
tomwalters@0:   function_match = Match(r'\s*(\)|=|(const)?\s*(;|\{|throw\(\)))', remainder)
tomwalters@0:   if function_match:
tomwalters@0:     if (not function_match.group(3) or
tomwalters@0:         function_match.group(3) == ';' or
tomwalters@0:         raw_line.find('/*') < 0):
tomwalters@0:       error(filename, linenum, 'readability/function', 3,
tomwalters@0:             'All parameters should be named in a function')
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # At this point, all that should be left is actual casts.
tomwalters@0:   error(filename, linenum, 'readability/casting', 4,
tomwalters@0:         'Using C-style cast.  Use %s<%s>(...) instead' %
tomwalters@0:         (cast_type, match.group(1)))
tomwalters@0: 
tomwalters@0: 
tomwalters@0: _HEADERS_CONTAINING_TEMPLATES = (
tomwalters@0:     ('<deque>', ('deque',)),
tomwalters@0:     ('<functional>', ('unary_function', 'binary_function',
tomwalters@0:                       'plus', 'minus', 'multiplies', 'divides', 'modulus',
tomwalters@0:                       'negate',
tomwalters@0:                       'equal_to', 'not_equal_to', 'greater', 'less',
tomwalters@0:                       'greater_equal', 'less_equal',
tomwalters@0:                       'logical_and', 'logical_or', 'logical_not',
tomwalters@0:                       'unary_negate', 'not1', 'binary_negate', 'not2',
tomwalters@0:                       'bind1st', 'bind2nd',
tomwalters@0:                       'pointer_to_unary_function',
tomwalters@0:                       'pointer_to_binary_function',
tomwalters@0:                       'ptr_fun',
tomwalters@0:                       'mem_fun_t', 'mem_fun', 'mem_fun1_t', 'mem_fun1_ref_t',
tomwalters@0:                       'mem_fun_ref_t',
tomwalters@0:                       'const_mem_fun_t', 'const_mem_fun1_t',
tomwalters@0:                       'const_mem_fun_ref_t', 'const_mem_fun1_ref_t',
tomwalters@0:                       'mem_fun_ref',
tomwalters@0:                      )),
tomwalters@0:     ('<limits>', ('numeric_limits',)),
tomwalters@0:     ('<list>', ('list',)),
tomwalters@0:     ('<map>', ('map', 'multimap',)),
tomwalters@0:     ('<memory>', ('allocator',)),
tomwalters@0:     ('<queue>', ('queue', 'priority_queue',)),
tomwalters@0:     ('<set>', ('set', 'multiset',)),
tomwalters@0:     ('<stack>', ('stack',)),
tomwalters@0:     ('<string>', ('char_traits', 'basic_string',)),
tomwalters@0:     ('<utility>', ('pair',)),
tomwalters@0:     ('<vector>', ('vector',)),
tomwalters@0: 
tomwalters@0:     # gcc extensions.
tomwalters@0:     # Note: std::hash is their hash, ::hash is our hash
tomwalters@0:     ('<hash_map>', ('hash_map', 'hash_multimap',)),
tomwalters@0:     ('<hash_set>', ('hash_set', 'hash_multiset',)),
tomwalters@0:     ('<slist>', ('slist',)),
tomwalters@0:     )
tomwalters@0: 
tomwalters@0: _HEADERS_ACCEPTED_BUT_NOT_PROMOTED = {
tomwalters@0:     # We can trust with reasonable confidence that map gives us pair<>, too.
tomwalters@0:     'pair<>': ('map', 'multimap', 'hash_map', 'hash_multimap')
tomwalters@0: }
tomwalters@0: 
tomwalters@0: _RE_PATTERN_STRING = re.compile(r'\bstring\b')
tomwalters@0: 
tomwalters@0: _re_pattern_algorithm_header = []
tomwalters@0: for _template in ('copy', 'max', 'min', 'min_element', 'sort', 'swap',
tomwalters@0:                   'transform'):
tomwalters@0:   # Match max<type>(..., ...), max(..., ...), but not foo->max, foo.max or
tomwalters@0:   # type::max().
tomwalters@0:   _re_pattern_algorithm_header.append(
tomwalters@0:       (re.compile(r'[^>.]\b' + _template + r'(<.*?>)?\([^\)]'),
tomwalters@0:        _template,
tomwalters@0:        '<algorithm>'))
tomwalters@0: 
tomwalters@0: _re_pattern_templates = []
tomwalters@0: for _header, _templates in _HEADERS_CONTAINING_TEMPLATES:
tomwalters@0:   for _template in _templates:
tomwalters@0:     _re_pattern_templates.append(
tomwalters@0:         (re.compile(r'(\<|\b)' + _template + r'\s*\<'),
tomwalters@0:          _template + '<>',
tomwalters@0:          _header))
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def FilesBelongToSameModule(filename_cc, filename_h):
tomwalters@0:   """Check if these two filenames belong to the same module.
tomwalters@0: 
tomwalters@0:   The concept of a 'module' here is a as follows:
tomwalters@0:   foo.h, foo-inl.h, foo.cc, foo_test.cc and foo_unittest.cc belong to the
tomwalters@0:   same 'module' if they are in the same directory.
tomwalters@0:   some/path/public/xyzzy and some/path/internal/xyzzy are also considered
tomwalters@0:   to belong to the same module here.
tomwalters@0: 
tomwalters@0:   If the filename_cc contains a longer path than the filename_h, for example,
tomwalters@0:   '/absolute/path/to/base/sysinfo.cc', and this file would include
tomwalters@0:   'base/sysinfo.h', this function also produces the prefix needed to open the
tomwalters@0:   header. This is used by the caller of this function to more robustly open the
tomwalters@0:   header file. We don't have access to the real include paths in this context,
tomwalters@0:   so we need this guesswork here.
tomwalters@0: 
tomwalters@0:   Known bugs: tools/base/bar.cc and base/bar.h belong to the same module
tomwalters@0:   according to this implementation. Because of this, this function gives
tomwalters@0:   some false positives. This should be sufficiently rare in practice.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename_cc: is the path for the .cc file
tomwalters@0:     filename_h: is the path for the header path
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     Tuple with a bool and a string:
tomwalters@0:     bool: True if filename_cc and filename_h belong to the same module.
tomwalters@0:     string: the additional prefix needed to open the header file.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   if not filename_cc.endswith('.cc'):
tomwalters@0:     return (False, '')
tomwalters@0:   filename_cc = filename_cc[:-len('.cc')]
tomwalters@0:   if filename_cc.endswith('_unittest'):
tomwalters@0:     filename_cc = filename_cc[:-len('_unittest')]
tomwalters@0:   elif filename_cc.endswith('_test'):
tomwalters@0:     filename_cc = filename_cc[:-len('_test')]
tomwalters@0:   filename_cc = filename_cc.replace('/public/', '/')
tomwalters@0:   filename_cc = filename_cc.replace('/internal/', '/')
tomwalters@0: 
tomwalters@0:   if not filename_h.endswith('.h'):
tomwalters@0:     return (False, '')
tomwalters@0:   filename_h = filename_h[:-len('.h')]
tomwalters@0:   if filename_h.endswith('-inl'):
tomwalters@0:     filename_h = filename_h[:-len('-inl')]
tomwalters@0:   filename_h = filename_h.replace('/public/', '/')
tomwalters@0:   filename_h = filename_h.replace('/internal/', '/')
tomwalters@0: 
tomwalters@0:   files_belong_to_same_module = filename_cc.endswith(filename_h)
tomwalters@0:   common_path = ''
tomwalters@0:   if files_belong_to_same_module:
tomwalters@0:     common_path = filename_cc[:-len(filename_h)]
tomwalters@0:   return files_belong_to_same_module, common_path
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def UpdateIncludeState(filename, include_state, io=codecs):
tomwalters@0:   """Fill up the include_state with new includes found from the file.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: the name of the header to read.
tomwalters@0:     include_state: an _IncludeState instance in which the headers are inserted.
tomwalters@0:     io: The io factory to use to read the file. Provided for testability.
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     True if a header was succesfully added. False otherwise.
tomwalters@0:   """
tomwalters@0:   headerfile = None
tomwalters@0:   try:
tomwalters@0:     headerfile = io.open(filename, 'r', 'utf8', 'replace')
tomwalters@0:   except IOError:
tomwalters@0:     return False
tomwalters@0:   linenum = 0
tomwalters@0:   for line in headerfile:
tomwalters@0:     linenum += 1
tomwalters@0:     clean_line = CleanseComments(line)
tomwalters@0:     match = _RE_PATTERN_INCLUDE.search(clean_line)
tomwalters@0:     if match:
tomwalters@0:       include = match.group(2)
tomwalters@0:       # The value formatting is cute, but not really used right now.
tomwalters@0:       # What matters here is that the key is in include_state.
tomwalters@0:       include_state.setdefault(include, '%s:%d' % (filename, linenum))
tomwalters@0:   return True
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def CheckForIncludeWhatYouUse(filename, clean_lines, include_state, error,
tomwalters@0:                               io=codecs):
tomwalters@0:   """Reports for missing stl includes.
tomwalters@0: 
tomwalters@0:   This function will output warnings to make sure you are including the headers
tomwalters@0:   necessary for the stl containers and functions that you use. We only give one
tomwalters@0:   reason to include a header. For example, if you use both equal_to<> and
tomwalters@0:   less<> in a .h file, only one (the latter in the file) of these will be
tomwalters@0:   reported as a reason to include the <functional>.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the current file.
tomwalters@0:     clean_lines: A CleansedLines instance containing the file.
tomwalters@0:     include_state: An _IncludeState instance.
tomwalters@0:     error: The function to call with any errors found.
tomwalters@0:     io: The IO factory to use to read the header file. Provided for unittest
tomwalters@0:         injection.
tomwalters@0:   """
tomwalters@0:   required = {}  # A map of header name to linenumber and the template entity.
tomwalters@0:                  # Example of required: { '<functional>': (1219, 'less<>') }
tomwalters@0: 
tomwalters@0:   for linenum in xrange(clean_lines.NumLines()):
tomwalters@0:     line = clean_lines.elided[linenum]
tomwalters@0:     if not line or line[0] == '#':
tomwalters@0:       continue
tomwalters@0: 
tomwalters@0:     # String is special -- it is a non-templatized type in STL.
tomwalters@0:     if _RE_PATTERN_STRING.search(line):
tomwalters@0:       required['<string>'] = (linenum, 'string')
tomwalters@0: 
tomwalters@0:     for pattern, template, header in _re_pattern_algorithm_header:
tomwalters@0:       if pattern.search(line):
tomwalters@0:         required[header] = (linenum, template)
tomwalters@0: 
tomwalters@0:     # The following function is just a speed up, no semantics are changed.
tomwalters@0:     if not '<' in line:  # Reduces the cpu time usage by skipping lines.
tomwalters@0:       continue
tomwalters@0: 
tomwalters@0:     for pattern, template, header in _re_pattern_templates:
tomwalters@0:       if pattern.search(line):
tomwalters@0:         required[header] = (linenum, template)
tomwalters@0: 
tomwalters@0:   # The policy is that if you #include something in foo.h you don't need to
tomwalters@0:   # include it again in foo.cc. Here, we will look at possible includes.
tomwalters@0:   # Let's copy the include_state so it is only messed up within this function.
tomwalters@0:   include_state = include_state.copy()
tomwalters@0: 
tomwalters@0:   # Did we find the header for this file (if any) and succesfully load it?
tomwalters@0:   header_found = False
tomwalters@0: 
tomwalters@0:   # Use the absolute path so that matching works properly.
tomwalters@0:   abs_filename = os.path.abspath(filename)
tomwalters@0: 
tomwalters@0:   # For Emacs's flymake.
tomwalters@0:   # If cpplint is invoked from Emacs's flymake, a temporary file is generated
tomwalters@0:   # by flymake and that file name might end with '_flymake.cc'. In that case,
tomwalters@0:   # restore original file name here so that the corresponding header file can be
tomwalters@0:   # found.
tomwalters@0:   # e.g. If the file name is 'foo_flymake.cc', we should search for 'foo.h'
tomwalters@0:   # instead of 'foo_flymake.h'
tomwalters@0:   emacs_flymake_suffix = '_flymake.cc'
tomwalters@0:   if abs_filename.endswith(emacs_flymake_suffix):
tomwalters@0:     abs_filename = abs_filename[:-len(emacs_flymake_suffix)] + '.cc'
tomwalters@0: 
tomwalters@0:   # include_state is modified during iteration, so we iterate over a copy of
tomwalters@0:   # the keys.
tomwalters@0:   for header in include_state.keys():  #NOLINT
tomwalters@0:     (same_module, common_path) = FilesBelongToSameModule(abs_filename, header)
tomwalters@0:     fullpath = common_path + header
tomwalters@0:     if same_module and UpdateIncludeState(fullpath, include_state, io):
tomwalters@0:       header_found = True
tomwalters@0: 
tomwalters@0:   # If we can't find the header file for a .cc, assume it's because we don't
tomwalters@0:   # know where to look. In that case we'll give up as we're not sure they
tomwalters@0:   # didn't include it in the .h file.
tomwalters@0:   # TODO(unknown): Do a better job of finding .h files so we are confident that
tomwalters@0:   # not having the .h file means there isn't one.
tomwalters@0:   if filename.endswith('.cc') and not header_found:
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # All the lines have been processed, report the errors found.
tomwalters@0:   for required_header_unstripped in required:
tomwalters@0:     template = required[required_header_unstripped][1]
tomwalters@0:     if template in _HEADERS_ACCEPTED_BUT_NOT_PROMOTED:
tomwalters@0:       headers = _HEADERS_ACCEPTED_BUT_NOT_PROMOTED[template]
tomwalters@0:       if [True for header in headers if header in include_state]:
tomwalters@0:         continue
tomwalters@0:     if required_header_unstripped.strip('<>"') not in include_state:
tomwalters@0:       error(filename, required[required_header_unstripped][0],
tomwalters@0:             'build/include_what_you_use', 4,
tomwalters@0:             'Add #include ' + required_header_unstripped + ' for ' + template)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def ProcessLine(filename, file_extension,
tomwalters@0:                 clean_lines, line, include_state, function_state,
tomwalters@0:                 class_state, error):
tomwalters@0:   """Processes a single line in the file.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: Filename of the file that is being processed.
tomwalters@0:     file_extension: The extension (dot not included) of the file.
tomwalters@0:     clean_lines: An array of strings, each representing a line of the file,
tomwalters@0:                  with comments stripped.
tomwalters@0:     line: Number of line being processed.
tomwalters@0:     include_state: An _IncludeState instance in which the headers are inserted.
tomwalters@0:     function_state: A _FunctionState instance which counts function lines, etc.
tomwalters@0:     class_state: A _ClassState instance which maintains information about
tomwalters@0:                  the current stack of nested class declarations being parsed.
tomwalters@0:     error: A callable to which errors are reported, which takes 4 arguments:
tomwalters@0:            filename, line number, error level, and message
tomwalters@0: 
tomwalters@0:   """
tomwalters@0:   raw_lines = clean_lines.raw_lines
tomwalters@0:   CheckForFunctionLengths(filename, clean_lines, line, function_state, error)
tomwalters@0:   if Search(r'\bNOLINT\b', raw_lines[line]):  # ignore nolint lines
tomwalters@0:     return
tomwalters@0:   CheckForMultilineCommentsAndStrings(filename, clean_lines, line, error)
tomwalters@0:   CheckStyle(filename, clean_lines, line, file_extension, error)
tomwalters@0:   CheckLanguage(filename, clean_lines, line, file_extension, include_state,
tomwalters@0:                 error)
tomwalters@0:   CheckForNonStandardConstructs(filename, clean_lines, line,
tomwalters@0:                                 class_state, error)
tomwalters@0:   CheckPosixThreading(filename, clean_lines, line, error)
tomwalters@0:   CheckInvalidIncrement(filename, clean_lines, line, error)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def ProcessFileData(filename, file_extension, lines, error):
tomwalters@0:   """Performs lint checks and reports any errors to the given error function.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: Filename of the file that is being processed.
tomwalters@0:     file_extension: The extension (dot not included) of the file.
tomwalters@0:     lines: An array of strings, each representing a line of the file, with the
tomwalters@0:            last element being empty if the file is termined with a newline.
tomwalters@0:     error: A callable to which errors are reported, which takes 4 arguments:
tomwalters@0:   """
tomwalters@0:   lines = (['// marker so line numbers and indices both start at 1'] + lines +
tomwalters@0:            ['// marker so line numbers end in a known way'])
tomwalters@0: 
tomwalters@0:   include_state = _IncludeState()
tomwalters@0:   function_state = _FunctionState()
tomwalters@0:   class_state = _ClassState()
tomwalters@0: 
tomwalters@0:   CheckForCopyright(filename, lines, error)
tomwalters@0: 
tomwalters@0:   if file_extension == 'h':
tomwalters@0:     CheckForHeaderGuard(filename, lines, error)
tomwalters@0: 
tomwalters@0:   RemoveMultiLineComments(filename, lines, error)
tomwalters@0:   clean_lines = CleansedLines(lines)
tomwalters@0:   for line in xrange(clean_lines.NumLines()):
tomwalters@0:     ProcessLine(filename, file_extension, clean_lines, line,
tomwalters@0:                 include_state, function_state, class_state, error)
tomwalters@0:   class_state.CheckFinished(filename, error)
tomwalters@0: 
tomwalters@0:   CheckForIncludeWhatYouUse(filename, clean_lines, include_state, error)
tomwalters@0: 
tomwalters@0:   # We check here rather than inside ProcessLine so that we see raw
tomwalters@0:   # lines rather than "cleaned" lines.
tomwalters@0:   CheckForUnicodeReplacementCharacters(filename, lines, error)
tomwalters@0: 
tomwalters@0:   CheckForNewlineAtEOF(filename, lines, error)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def ProcessFile(filename, vlevel):
tomwalters@0:   """Does google-lint on a single file.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     filename: The name of the file to parse.
tomwalters@0: 
tomwalters@0:     vlevel: The level of errors to report.  Every error of confidence
tomwalters@0:     >= verbose_level will be reported.  0 is a good default.
tomwalters@0:   """
tomwalters@0: 
tomwalters@0:   _SetVerboseLevel(vlevel)
tomwalters@0: 
tomwalters@0:   try:
tomwalters@0:     # Support the UNIX convention of using "-" for stdin.  Note that
tomwalters@0:     # we are not opening the file with universal newline support
tomwalters@0:     # (which codecs doesn't support anyway), so the resulting lines do
tomwalters@0:     # contain trailing '\r' characters if we are reading a file that
tomwalters@0:     # has CRLF endings.
tomwalters@0:     # If after the split a trailing '\r' is present, it is removed
tomwalters@0:     # below. If it is not expected to be present (i.e. os.linesep !=
tomwalters@0:     # '\r\n' as in Windows), a warning is issued below if this file
tomwalters@0:     # is processed.
tomwalters@0: 
tomwalters@0:     if filename == '-':
tomwalters@0:       lines = codecs.StreamReaderWriter(sys.stdin,
tomwalters@0:                                         codecs.getreader('utf8'),
tomwalters@0:                                         codecs.getwriter('utf8'),
tomwalters@0:                                         'replace').read().split('\n')
tomwalters@0:     else:
tomwalters@0:       lines = codecs.open(filename, 'r', 'utf8', 'replace').read().split('\n')
tomwalters@0: 
tomwalters@0:     carriage_return_found = False
tomwalters@0:     # Remove trailing '\r'.
tomwalters@0:     for linenum in range(len(lines)):
tomwalters@0:       if lines[linenum].endswith('\r'):
tomwalters@0:         lines[linenum] = lines[linenum].rstrip('\r')
tomwalters@0:         carriage_return_found = True
tomwalters@0: 
tomwalters@0:   except IOError:
tomwalters@0:     sys.stderr.write(
tomwalters@0:         "Skipping input '%s': Can't open for reading\n" % filename)
tomwalters@0:     return
tomwalters@0: 
tomwalters@0:   # Note, if no dot is found, this will give the entire filename as the ext.
tomwalters@0:   file_extension = filename[filename.rfind('.') + 1:]
tomwalters@0: 
tomwalters@0:   # When reading from stdin, the extension is unknown, so no cpplint tests
tomwalters@0:   # should rely on the extension.
tomwalters@0:   if (filename != '-' and file_extension != 'cc' and file_extension != 'h'
tomwalters@0:       and file_extension != 'cpp'):
tomwalters@0:     sys.stderr.write('Ignoring %s; not a .cc or .h file\n' % filename)
tomwalters@0:   else:
tomwalters@0:     ProcessFileData(filename, file_extension, lines, Error)
tomwalters@0:     if carriage_return_found and os.linesep != '\r\n':
tomwalters@0:       # Use 0 for linenum since outputing only one error for potentially
tomwalters@0:       # several lines.
tomwalters@0:       Error(filename, 0, 'whitespace/newline', 1,
tomwalters@0:             'One or more unexpected \\r (^M) found;'
tomwalters@0:             'better to use only a \\n')
tomwalters@0: 
tomwalters@0:   sys.stderr.write('Done processing %s\n' % filename)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def PrintUsage(message):
tomwalters@0:   """Prints a brief usage string and exits, optionally with an error message.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     message: The optional error message.
tomwalters@0:   """
tomwalters@0:   sys.stderr.write(_USAGE)
tomwalters@0:   if message:
tomwalters@0:     sys.exit('\nFATAL ERROR: ' + message)
tomwalters@0:   else:
tomwalters@0:     sys.exit(1)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def PrintCategories():
tomwalters@0:   """Prints a list of all the error-categories used by error messages.
tomwalters@0: 
tomwalters@0:   These are the categories used to filter messages via --filter.
tomwalters@0:   """
tomwalters@0:   sys.stderr.write(_ERROR_CATEGORIES)
tomwalters@0:   sys.exit(0)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def ParseArguments(args):
tomwalters@0:   """Parses the command line arguments.
tomwalters@0: 
tomwalters@0:   This may set the output format and verbosity level as side-effects.
tomwalters@0: 
tomwalters@0:   Args:
tomwalters@0:     args: The command line arguments:
tomwalters@0: 
tomwalters@0:   Returns:
tomwalters@0:     The list of filenames to lint.
tomwalters@0:   """
tomwalters@0:   try:
tomwalters@0:     (opts, filenames) = getopt.getopt(args, '', ['help', 'output=', 'verbose=',
tomwalters@0:                                                  'counting=',
tomwalters@0:                                                  'filter='])
tomwalters@0:   except getopt.GetoptError:
tomwalters@0:     PrintUsage('Invalid arguments.')
tomwalters@0: 
tomwalters@0:   verbosity = _VerboseLevel()
tomwalters@0:   output_format = _OutputFormat()
tomwalters@0:   filters = ''
tomwalters@0:   counting_style = ''
tomwalters@0: 
tomwalters@0:   for (opt, val) in opts:
tomwalters@0:     if opt == '--help':
tomwalters@0:       PrintUsage(None)
tomwalters@0:     elif opt == '--output':
tomwalters@0:       if not val in ('emacs', 'vs7'):
tomwalters@0:         PrintUsage('The only allowed output formats are emacs and vs7.')
tomwalters@0:       output_format = val
tomwalters@0:     elif opt == '--verbose':
tomwalters@0:       verbosity = int(val)
tomwalters@0:     elif opt == '--filter':
tomwalters@0:       filters = val
tomwalters@0:       if not filters:
tomwalters@0:         PrintCategories()
tomwalters@0:     elif opt == '--counting':
tomwalters@0:       if val not in ('total', 'toplevel', 'detailed'):
tomwalters@0:         PrintUsage('Valid counting options are total, toplevel, and detailed')
tomwalters@0:       counting_style = val
tomwalters@0: 
tomwalters@0:   if not filenames:
tomwalters@0:     PrintUsage('No files were specified.')
tomwalters@0: 
tomwalters@0:   _SetOutputFormat(output_format)
tomwalters@0:   _SetVerboseLevel(verbosity)
tomwalters@0:   _SetFilters(filters)
tomwalters@0:   _SetCountingStyle(counting_style)
tomwalters@0: 
tomwalters@0:   return filenames
tomwalters@0: 
tomwalters@0: 
tomwalters@0: def main():
tomwalters@0:   filenames = ParseArguments(sys.argv[1:])
tomwalters@0: 
tomwalters@0:   # Change stderr to write with replacement characters so we don't die
tomwalters@0:   # if we try to print something containing non-ASCII characters.
tomwalters@0:   sys.stderr = codecs.StreamReaderWriter(sys.stderr,
tomwalters@0:                                          codecs.getreader('utf8'),
tomwalters@0:                                          codecs.getwriter('utf8'),
tomwalters@0:                                          'replace')
tomwalters@0: 
tomwalters@0:   _cpplint_state.ResetErrorCounts()
tomwalters@0:   for filename in filenames:
tomwalters@0:     ProcessFile(filename, _cpplint_state.verbose_level)
tomwalters@0:   _cpplint_state.PrintErrorCounts()
tomwalters@0: 
tomwalters@0:   sys.exit(_cpplint_state.error_count > 0)
tomwalters@0: 
tomwalters@0: 
tomwalters@0: if __name__ == '__main__':
tomwalters@0:   main()