1#!/usr/bin/env python3
2
3# Copyright (C) 2016 Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
4
5# This program is free software; you can redistribute it and/or modify
6# it under the terms of the GNU General Public License as published by
7# the Free Software Foundation; either version 2 of the License, or
8# (at your option) any later version.
9#
10# This program is distributed in the hope that it will be useful,
11# but WITHOUT ANY WARRANTY; without even the implied warranty of
12# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13# General Public License for more details.
14#
15# You should have received a copy of the GNU General Public License
16# along with this program; if not, write to the Free Software
17# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
19# TODO (improvements)
20# - support K,M,G size suffixes for threshold
21# - output CSV file in addition to stdout reporting
22
23import csv
24import argparse
25import sys
26
27
28def read_file_size_csv(inputf, detail=None):
29    """Extract package or file sizes from CSV file into size dictionary"""
30    sizes = {}
31    reader = csv.reader(inputf)
32
33    header = next(reader)
34    if header[0] != 'File name' or header[1] != 'Package name' or \
35       header[2] != 'File size' or header[3] != 'Package size':
36        print(("Input file %s does not contain the expected header. Are you "
37               "sure this file corresponds to the file-size-stats.csv "
38               "file created by 'make graph-size'?") % inputf.name)
39        sys.exit(1)
40
41    for row in reader:
42        if detail:
43            sizes[(row[0], row[1])] = int(row[2])
44        else:
45            sizes[(None, row[1])] = int(row[3])
46
47    return sizes
48
49
50def compare_sizes(old, new):
51    """Return delta/added/removed dictionaries based on two input size
52    dictionaries"""
53    delta = {}
54    oldkeys = set(old.keys())
55    newkeys = set(new.keys())
56
57    # packages/files in both
58    for entry in newkeys.intersection(oldkeys):
59        delta[entry] = ('', new[entry] - old[entry])
60    # packages/files only in new
61    for entry in newkeys.difference(oldkeys):
62        delta[entry] = ('added', new[entry])
63    # packages/files only in old
64    for entry in oldkeys.difference(newkeys):
65        delta[entry] = ('removed', -old[entry])
66
67    return delta
68
69
70def print_results(result, threshold):
71    """Print the given result dictionary sorted by size, ignoring any entries
72    below or equal to threshold"""
73
74    from six import iteritems
75    list_result = list(iteritems(result))
76    # result is a dictionary: (filename, pkgname) -> (flag, size difference)
77    # list_result is a list of tuples: ((filename, pkgname), (flag, size difference))
78    # filename may be None if no detail is requested.
79
80    maxpkgname = max(len(pkgname) for filename, pkgname in result)
81
82    for entry in sorted(list_result, key=lambda entry: entry[1][1]):
83        data = dict(
84            filename=entry[0][0],
85            pkgname=entry[0][1],
86            action=entry[1][0],
87            size=entry[1][1],
88            maxpkgname=maxpkgname,
89        )
90
91        if threshold is not None and abs(data['size']) <= threshold:
92            continue
93        if data['filename']:
94            print('{size:12d} {action:7s} {pkgname:{maxpkgname}s} {filename}'.format(**data))
95        else:
96            print('{size:12d} {action:7s} {pkgname}'.format(**data))
97
98
99# main #########################################################################
100
101description = """
102Compare rootfs size between Buildroot compilations, for example after changing
103configuration options or after switching to another Buildroot release.
104
105This script compares the file-size-stats.csv file generated by 'make graph-size'
106with the corresponding file from another Buildroot compilation.
107The size differences can be reported per package or per file.
108Size differences smaller or equal than a given threshold can be ignored.
109"""
110
111parser = argparse.ArgumentParser(description=description,
112                                 formatter_class=argparse.RawDescriptionHelpFormatter)
113
114parser.add_argument('-d', '--detail', action='store_true',
115                    help='''report differences for individual files rather than
116                            packages''')
117parser.add_argument('-t', '--threshold', type=int,
118                    help='''ignore size differences smaller or equal than this
119                            value (bytes)''')
120parser.add_argument('old_file_size_csv', type=argparse.FileType('r'),
121                    metavar='old-file-size-stats.csv',
122                    help="""old CSV file with file and package size statistics,
123                            generated by 'make graph-size'""")
124parser.add_argument('new_file_size_csv', type=argparse.FileType('r'),
125                    metavar='new-file-size-stats.csv',
126                    help='new CSV file with file and package size statistics')
127args = parser.parse_args()
128
129if args.detail:
130    keyword = 'file'
131else:
132    keyword = 'package'
133
134old_sizes = read_file_size_csv(args.old_file_size_csv, args.detail)
135new_sizes = read_file_size_csv(args.new_file_size_csv, args.detail)
136
137delta = compare_sizes(old_sizes, new_sizes)
138
139print('Size difference per %s (bytes), threshold = %s' % (keyword, args.threshold))
140print(80*'-')
141print_results(delta, args.threshold)
142print(80*'-')
143print_results({(None, 'TOTAL'): ('', sum(new_sizes.values()) - sum(old_sizes.values()))},
144              threshold=None)
145