OLD | NEW |
(Empty) | |
| 1 #!/usr/bin/env python |
| 2 # Copyright 2016 The Chromium Authors. All rights reserved. |
| 3 # Use of this source code is governed by a BSD-style license that can be |
| 4 # found in the LICENSE file. |
| 5 |
| 6 """This script will search through the target folder specified and try to find |
| 7 duplicate includes from h and cc files, and remove them from the cc files. The |
| 8 current/working directory needs to be chromium_checkout/src/ when this tool is |
| 9 run. |
| 10 |
| 11 Usage: remove_duplicate_includes.py --dry-run components/foo components/bar |
| 12 """ |
| 13 |
| 14 import argparse |
| 15 import collections |
| 16 import logging |
| 17 import os |
| 18 import re |
| 19 import sys |
| 20 |
| 21 # This could be generalized if desired, and moved to command line arguments. |
| 22 H_FILE_SUFFIX = '.h' |
| 23 CC_FILE_SUFFIX = '.cc' |
| 24 |
| 25 # The \s should allow us to ignore any whitespace and only focus on the group |
| 26 # captured when comparing between files. |
| 27 INCLUDE_REGEX = re.compile('^\s*(#include\s+[\"<](.*?)[\">])\s*$') |
| 28 |
| 29 def HasSuffix(file_name, suffix): |
| 30 return os.path.splitext(file_name)[1] == suffix |
| 31 |
| 32 def IsEmpty(line): |
| 33 return not line.strip() |
| 34 |
| 35 def FindIncludeSet(input_lines, h_path_to_include_set, cc_file_name): |
| 36 """Finds and returns the corresponding include set for the given .cc file. |
| 37 |
| 38 This is done by finding the first include in the file and then trying to look |
| 39 up an .h file in the passed in map. If not present, then None is returned |
| 40 immediately. |
| 41 """ |
| 42 for line in input_lines: |
| 43 match = INCLUDE_REGEX.search(line) |
| 44 # The first include match should be the corresponding .h file, else skip. |
| 45 if match: |
| 46 h_file_path = os.path.join(os.getcwd(), match.group(2)) |
| 47 if h_file_path not in h_path_to_include_set: |
| 48 print 'First include did not match to a known .h file, skipping ' + \ |
| 49 cc_file_name + ', line: ' + match.group(1) |
| 50 return None |
| 51 return h_path_to_include_set[h_file_path] |
| 52 |
| 53 def WithoutDuplicates(input_lines, include_set, cc_file_name): |
| 54 """Checks every input line and sees if we can remove it based on the contents |
| 55 of the given include set. |
| 56 |
| 57 Returns what the new contents of the file should be. |
| 58 """ |
| 59 output_lines = [] |
| 60 # When a section of includes are completely removed, we want to remove the |
| 61 # trailing empty as well. |
| 62 lastCopiedLineWasEmpty = False |
| 63 lastLineWasOmitted = False |
| 64 for line in input_lines: |
| 65 match = INCLUDE_REGEX.search(line) |
| 66 if match and match.group(2) in include_set: |
| 67 print 'Removed ' + match.group(1) + ' from ' + cc_file_name |
| 68 lastLineWasOmitted = True |
| 69 elif lastCopiedLineWasEmpty and lastLineWasOmitted and IsEmpty(line): |
| 70 print 'Removed empty line from ' + cc_file_name |
| 71 lastLineWasOmitted = True |
| 72 else: |
| 73 lastCopiedLineWasEmpty = IsEmpty(line) |
| 74 lastLineWasOmitted = False |
| 75 output_lines.append(line) |
| 76 return output_lines |
| 77 |
| 78 def main(): |
| 79 parser = argparse.ArgumentParser() |
| 80 parser.add_argument('--dry-run', action='store_true', |
| 81 help='Does not actually remove lines when specified.') |
| 82 parser.add_argument('targets', nargs='+', |
| 83 help='Relative path to folders to search for duplicate includes in.') |
| 84 args = parser.parse_args() |
| 85 |
| 86 # A map of header file paths to the includes they contain. |
| 87 h_path_to_include_set = {} |
| 88 |
| 89 # Simply collects the path of all cc files present. |
| 90 cc_file_path_set = set() |
| 91 |
| 92 for relative_root in args.targets: |
| 93 absolute_root = os.path.join(os.getcwd(), relative_root) |
| 94 for dir_path, dir_name_list, file_name_list in os.walk(absolute_root): |
| 95 for file_name in file_name_list: |
| 96 file_path = os.path.join(dir_path, file_name) |
| 97 if HasSuffix(file_name, H_FILE_SUFFIX): |
| 98 # By manually adding the set instead of using defaultdict we can avoid |
| 99 # warning about missing .h files when the .h file has no includes. |
| 100 h_path_to_include_set[file_path] = set() |
| 101 with open(file_path) as fh: |
| 102 for line in fh: |
| 103 match = INCLUDE_REGEX.search(line) |
| 104 if match: |
| 105 h_path_to_include_set[file_path].add(match.group(2)) |
| 106 elif HasSuffix(file_name, CC_FILE_SUFFIX): |
| 107 cc_file_path_set.add(file_path) |
| 108 |
| 109 for cc_file_path in cc_file_path_set: |
| 110 cc_file_name = os.path.basename(cc_file_path) |
| 111 with open(cc_file_path, 'r' if args.dry_run else 'r+') as fh: |
| 112 # Read out all lines and reset file position to allow overwriting. |
| 113 input_lines = fh.readlines() |
| 114 fh.seek(0) |
| 115 include_set = FindIncludeSet(input_lines, h_path_to_include_set, |
| 116 cc_file_name) |
| 117 if include_set: |
| 118 output_lines = WithoutDuplicates(input_lines, include_set, cc_file_name) |
| 119 if not args.dry_run: |
| 120 fh.writelines(output_lines) |
| 121 fh.truncate() |
| 122 |
| 123 if __name__ == '__main__': |
| 124 sys.exit(main()) |
OLD | NEW |