1 #!/usr/bin/env perl 1 #!/usr/bin/env perl
2 # SPDX-License-Identifier: GPL-2.0 <<
3 # 2 #
4 # Clean a patch file -- or directory of patch 3 # Clean a patch file -- or directory of patch files -- of stealth whitespace.
5 # WARNING: this can be a highly destructive op 4 # WARNING: this can be a highly destructive operation. Use with caution.
6 # 5 #
7 6
8 use warnings; 7 use warnings;
9 use bytes; 8 use bytes;
10 use File::Basename; 9 use File::Basename;
11 10
12 # Default options 11 # Default options
13 $max_width = 79; 12 $max_width = 79;
14 13
15 # Clean up space-tab sequences, either by remo 14 # Clean up space-tab sequences, either by removing spaces or
16 # replacing them with tabs. 15 # replacing them with tabs.
17 sub clean_space_tabs($) 16 sub clean_space_tabs($)
18 { 17 {
19 no bytes; # Tab alignmen 18 no bytes; # Tab alignment depends on characters
20 19
21 my($li) = @_; 20 my($li) = @_;
22 my($lo) = ''; 21 my($lo) = '';
23 my $pos = 0; 22 my $pos = 0;
24 my $nsp = 0; 23 my $nsp = 0;
25 my($i, $c); 24 my($i, $c);
26 25
27 for ($i = 0; $i < length($li); $i++) { 26 for ($i = 0; $i < length($li); $i++) {
28 $c = substr($li, $i, 1); 27 $c = substr($li, $i, 1);
29 if ($c eq "\t") { 28 if ($c eq "\t") {
30 my $npos = ($pos+$nsp+8) & ~7; 29 my $npos = ($pos+$nsp+8) & ~7;
31 my $ntab = ($npos >> 3) - ($pos >> 30 my $ntab = ($npos >> 3) - ($pos >> 3);
32 $lo .= "\t" x $ntab; 31 $lo .= "\t" x $ntab;
33 $pos = $npos; 32 $pos = $npos;
34 $nsp = 0; 33 $nsp = 0;
35 } elsif ($c eq "\n" || $c eq "\r") { 34 } elsif ($c eq "\n" || $c eq "\r") {
36 $lo .= " " x $nsp; 35 $lo .= " " x $nsp;
37 $pos += $nsp; 36 $pos += $nsp;
38 $nsp = 0; 37 $nsp = 0;
39 $lo .= $c; 38 $lo .= $c;
40 $pos = 0; 39 $pos = 0;
41 } elsif ($c eq " ") { 40 } elsif ($c eq " ") {
42 $nsp++; 41 $nsp++;
43 } else { 42 } else {
44 $lo .= " " x $nsp; 43 $lo .= " " x $nsp;
45 $pos += $nsp; 44 $pos += $nsp;
46 $nsp = 0; 45 $nsp = 0;
47 $lo .= $c; 46 $lo .= $c;
48 $pos++; 47 $pos++;
49 } 48 }
50 } 49 }
51 $lo .= " " x $nsp; 50 $lo .= " " x $nsp;
52 return $lo; 51 return $lo;
53 } 52 }
54 53
55 # Compute the visual width of a string 54 # Compute the visual width of a string
56 sub strwidth($) { 55 sub strwidth($) {
57 no bytes; # Tab alignmen 56 no bytes; # Tab alignment depends on characters
58 57
59 my($li) = @_; 58 my($li) = @_;
60 my($c, $i); 59 my($c, $i);
61 my $pos = 0; 60 my $pos = 0;
62 my $mlen = 0; 61 my $mlen = 0;
63 62
64 for ($i = 0; $i < length($li); $i++) { 63 for ($i = 0; $i < length($li); $i++) {
65 $c = substr($li,$i,1); 64 $c = substr($li,$i,1);
66 if ($c eq "\t") { 65 if ($c eq "\t") {
67 $pos = ($pos+8) & ~7; 66 $pos = ($pos+8) & ~7;
68 } elsif ($c eq "\n") { 67 } elsif ($c eq "\n") {
69 $mlen = $pos if ($pos > $mlen); 68 $mlen = $pos if ($pos > $mlen);
70 $pos = 0; 69 $pos = 0;
71 } else { 70 } else {
72 $pos++; 71 $pos++;
73 } 72 }
74 } 73 }
75 74
76 $mlen = $pos if ($pos > $mlen); 75 $mlen = $pos if ($pos > $mlen);
77 return $mlen; 76 return $mlen;
78 } 77 }
79 78
80 $name = basename($0); 79 $name = basename($0);
81 80
82 @files = (); 81 @files = ();
83 82
84 while (defined($a = shift(@ARGV))) { 83 while (defined($a = shift(@ARGV))) {
85 if ($a =~ /^-/) { 84 if ($a =~ /^-/) {
86 if ($a eq '-width' || $a eq '-w') { 85 if ($a eq '-width' || $a eq '-w') {
87 $max_width = shift(@ARGV)+0; 86 $max_width = shift(@ARGV)+0;
88 } else { 87 } else {
89 print STDERR "Usage: $name [-width 88 print STDERR "Usage: $name [-width #] files...\n";
90 exit 1; 89 exit 1;
91 } 90 }
92 } else { 91 } else {
93 push(@files, $a); 92 push(@files, $a);
94 } 93 }
95 } 94 }
96 95
97 foreach $f ( @files ) { 96 foreach $f ( @files ) {
98 print STDERR "$name: $f\n"; 97 print STDERR "$name: $f\n";
99 98
100 if (! -f $f) { 99 if (! -f $f) {
101 print STDERR "$f: not a file\n"; 100 print STDERR "$f: not a file\n";
102 next; 101 next;
103 } 102 }
104 103
105 if (!open(FILE, '+<', $f)) { 104 if (!open(FILE, '+<', $f)) {
106 print STDERR "$name: Cannot open file: 105 print STDERR "$name: Cannot open file: $f: $!\n";
107 next; 106 next;
108 } 107 }
109 108
110 binmode FILE; 109 binmode FILE;
111 110
112 # First, verify that it is not a binary fi 111 # First, verify that it is not a binary file; consider any file
113 # with a zero byte to be a binary file. I 112 # with a zero byte to be a binary file. Is there any better, or
114 # additional, heuristic that should be app 113 # additional, heuristic that should be applied?
115 $is_binary = 0; 114 $is_binary = 0;
116 115
117 while (read(FILE, $data, 65536) > 0) { 116 while (read(FILE, $data, 65536) > 0) {
118 if ($data =~ /\0/) { 117 if ($data =~ /\0/) {
119 $is_binary = 1; 118 $is_binary = 1;
120 last; 119 last;
121 } 120 }
122 } 121 }
123 122
124 if ($is_binary) { 123 if ($is_binary) {
125 print STDERR "$name: $f: binary file\n 124 print STDERR "$name: $f: binary file\n";
126 next; 125 next;
127 } 126 }
128 127
129 seek(FILE, 0, 0); 128 seek(FILE, 0, 0);
130 129
131 $in_bytes = 0; 130 $in_bytes = 0;
132 $out_bytes = 0; 131 $out_bytes = 0;
133 $lineno = 0; 132 $lineno = 0;
134 133
135 @lines = (); 134 @lines = ();
136 135
137 $in_hunk = 0; 136 $in_hunk = 0;
138 $err = 0; 137 $err = 0;
139 138
140 while ( defined($line = <FILE>) ) { 139 while ( defined($line = <FILE>) ) {
141 $lineno++; 140 $lineno++;
142 $in_bytes += length($line); 141 $in_bytes += length($line);
143 142
144 if (!$in_hunk) { 143 if (!$in_hunk) {
145 if ($line =~ 144 if ($line =~
146 /^\@\@\s+\-([0-9]+),([0-9]+)\s 145 /^\@\@\s+\-([0-9]+),([0-9]+)\s+\+([0-9]+),([0-9]+)\s\@\@/) {
147 $minus_lines = $2; 146 $minus_lines = $2;
148 $plus_lines = $4; 147 $plus_lines = $4;
149 if ($minus_lines || $plus_line 148 if ($minus_lines || $plus_lines) {
150 $in_hunk = 1; 149 $in_hunk = 1;
151 @hunk_lines = ($line); 150 @hunk_lines = ($line);
152 } 151 }
153 } else { 152 } else {
154 push(@lines, $line); 153 push(@lines, $line);
155 $out_bytes += length($line); 154 $out_bytes += length($line);
156 } 155 }
157 } else { 156 } else {
158 # We're in a hunk 157 # We're in a hunk
159 158
160 if ($line =~ /^\+/) { 159 if ($line =~ /^\+/) {
161 $plus_lines--; 160 $plus_lines--;
162 161
163 $text = substr($line, 1); 162 $text = substr($line, 1);
164 $text =~ s/[ \t\r]*$//; 163 $text =~ s/[ \t\r]*$//; # Remove trailing spaces
165 $text = clean_space_tabs($text 164 $text = clean_space_tabs($text);
166 165
167 $l_width = strwidth($text); 166 $l_width = strwidth($text);
168 if ($max_width && $l_width > $ 167 if ($max_width && $l_width > $max_width) {
169 print STDERR 168 print STDERR
170 "$f:$lineno: adds line 169 "$f:$lineno: adds line exceeds $max_width ",
171 "characters ($l_width) 170 "characters ($l_width)\n";
172 } 171 }
173 172
174 push(@hunk_lines, '+'.$text); 173 push(@hunk_lines, '+'.$text);
175 } elsif ($line =~ /^\-/) { 174 } elsif ($line =~ /^\-/) {
176 $minus_lines--; 175 $minus_lines--;
177 push(@hunk_lines, $line); 176 push(@hunk_lines, $line);
178 } elsif ($line =~ /^ /) { 177 } elsif ($line =~ /^ /) {
179 $plus_lines--; 178 $plus_lines--;
180 $minus_lines--; 179 $minus_lines--;
181 push(@hunk_lines, $line); 180 push(@hunk_lines, $line);
182 } else { 181 } else {
183 print STDERR "$name: $f: malfo 182 print STDERR "$name: $f: malformed patch\n";
184 $err = 1; 183 $err = 1;
185 last; 184 last;
186 } 185 }
187 186
188 if ($plus_lines < 0 || $minus_line 187 if ($plus_lines < 0 || $minus_lines < 0) {
189 print STDERR "$name: $f: malfo 188 print STDERR "$name: $f: malformed patch\n";
190 $err = 1; 189 $err = 1;
191 last; 190 last;
192 } elsif ($plus_lines == 0 && $minu 191 } elsif ($plus_lines == 0 && $minus_lines == 0) {
193 # End of a hunk. Process this 192 # End of a hunk. Process this hunk.
194 my $i; 193 my $i;
195 my $l; 194 my $l;
196 my @h = (); 195 my @h = ();
197 my $adj = 0; 196 my $adj = 0;
198 my $done = 0; 197 my $done = 0;
199 198
200 for ($i = scalar(@hunk_lines)- 199 for ($i = scalar(@hunk_lines)-1; $i > 0; $i--) {
201 $l = $hunk_lines[$i]; 200 $l = $hunk_lines[$i];
202 if (!$done && $l eq "+\n") 201 if (!$done && $l eq "+\n") {
203 $adj++; # Skip this li 202 $adj++; # Skip this line
204 } elsif ($l =~ /^[ +]/) { 203 } elsif ($l =~ /^[ +]/) {
205 $done = 1; 204 $done = 1;
206 unshift(@h, $l); 205 unshift(@h, $l);
207 } else { 206 } else {
208 unshift(@h, $l); 207 unshift(@h, $l);
209 } 208 }
210 } 209 }
211 210
212 $l = $hunk_lines[0]; # Hunk h 211 $l = $hunk_lines[0]; # Hunk header
213 undef @hunk_lines; # Free m 212 undef @hunk_lines; # Free memory
214 213
215 if ($adj) { 214 if ($adj) {
216 die unless 215 die unless
217 ($l =~ /^\@\@\s+\-([0- 216 ($l =~ /^\@\@\s+\-([0-9]+),([0-9]+)\s+\+([0-9]+),([0-9]+)\s\@\@(.*)$/);
218 my $mstart = $1; 217 my $mstart = $1;
219 my $mlin = $2; 218 my $mlin = $2;
220 my $pstart = $3; 219 my $pstart = $3;
221 my $plin = $4; 220 my $plin = $4;
222 my $tail = $5; # doesn't i 221 my $tail = $5; # doesn't include the final newline
223 222
224 $l = sprintf("@@ -%d,%d +% 223 $l = sprintf("@@ -%d,%d +%d,%d @@%s\n",
225 $mstart, $mli 224 $mstart, $mlin, $pstart, $plin-$adj,
226 $tail); 225 $tail);
227 } 226 }
228 unshift(@h, $l); 227 unshift(@h, $l);
229 228
230 # Transfer to the output array 229 # Transfer to the output array
231 foreach $l (@h) { 230 foreach $l (@h) {
232 $out_bytes += length($l); 231 $out_bytes += length($l);
233 push(@lines, $l); 232 push(@lines, $l);
234 } 233 }
235 234
236 $in_hunk = 0; 235 $in_hunk = 0;
237 } 236 }
238 } 237 }
239 } 238 }
240 239
241 if ($in_hunk) { 240 if ($in_hunk) {
242 print STDERR "$name: $f: malformed pat 241 print STDERR "$name: $f: malformed patch\n";
243 $err = 1; 242 $err = 1;
244 } 243 }
245 244
246 if (!$err) { 245 if (!$err) {
247 if ($in_bytes != $out_bytes) { 246 if ($in_bytes != $out_bytes) {
248 # Only write to the file if change 247 # Only write to the file if changed
249 seek(FILE, 0, 0); 248 seek(FILE, 0, 0);
250 print FILE @lines; 249 print FILE @lines;
251 250
252 if ( !defined($where = tell(FILE)) 251 if ( !defined($where = tell(FILE)) ||
253 !truncate(FILE, $where) ) { 252 !truncate(FILE, $where) ) {
254 die "$name: Failed to truncate 253 die "$name: Failed to truncate modified file: $f: $!\n";
255 } 254 }
256 } 255 }
257 } 256 }
258 257
259 close(FILE); 258 close(FILE);
260 } 259 }
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.