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