tools/binary_size/binary_size_utils.py

   1 # Copyright 2014 The Chromium Authors. All rights reserved.
   2 # Use of this source code is governed by a BSD-style license that can be
   3 # found in the LICENSE file.
   4
   5 """Common utilities for tools that deal with binary size information.
   6 """
   7
   8 import logging
   9 import re
  10
  11
  12 def ParseNm(nm_lines):
  13   """Parse nm output, returning data for all relevant (to binary size)
  14   symbols and ignoring the rest.
  15
  16   Args:
  17       nm_lines: an iterable over lines of nm output.
  18
  19   Yields:
  20       (symbol name, symbol type, symbol size, source file path).
  21
  22       Path may be None if nm couldn't figure out the source file.
  23   """
  24
  25   # Match lines with size, symbol, optional location, optional discriminator
  26   sym_re = re.compile(r'^([0-9a-f]{8,}) ' # address (8+ hex digits)
  27                       '([0-9a-f]{8,}) ' # size (8+ hex digits)
  28                       '(.) ' # symbol type, one character
  29                       '([^\t]+)' # symbol name, separated from next by tab
  30                       '(?:\t(.*):[\d\?]+)?.*$') # location
  31   # Match lines with addr but no size.
  32   addr_re = re.compile(r'^[0-9a-f]{8,} (.) ([^\t]+)(?:\t.*)?$')
  33   # Match lines that don't have an address at all -- typically external symbols.
  34   noaddr_re = re.compile(r'^ {8,} (.) (.*)$')
  35   # Match lines with no symbol name, only addr and type
  36   addr_only_re = re.compile(r'^[0-9a-f]{8,} (.)$')
  37
  38   for line in nm_lines:
  39     line = line.rstrip()
  40     match = sym_re.match(line)
  41     if match:
  42       address, size, sym_type, sym = match.groups()[0:4]
  43       size = int(size, 16)
  44       if sym_type in ('B', 'b'):
  45         continue  # skip all BSS for now.
  46       path = match.group(5)
  47       yield sym, sym_type, size, path, address
  48       continue
  49     match = addr_re.match(line)
  50     if match:
  51       # sym_type, sym = match.groups()[0:2]
  52       continue  # No size == we don't care.
  53     match = noaddr_re.match(line)
  54     if match:
  55       sym_type, sym = match.groups()
  56       if sym_type in ('U', 'w'):
  57         continue  # external or weak symbol
  58     match = addr_only_re.match(line)
  59     if match:
  60       continue  # Nothing to do.
  61
  62
  63     # If we reach this part of the loop, there was something in the
  64     # line that we didn't expect or recognize.
  65     logging.warning('nm output parser failed to parse: %s', repr(line))