Markus Armbruster | 2dbc4eb | 2016-06-28 13:07:36 +0200 | [diff] [blame] | 1 | #!/usr/bin/perl -w |
| 2 | # |
| 3 | # Clean up include guards in headers |
| 4 | # |
| 5 | # Copyright (C) 2016 Red Hat, Inc. |
| 6 | # |
| 7 | # Authors: |
| 8 | # Markus Armbruster <armbru@redhat.com> |
| 9 | # |
| 10 | # This work is licensed under the terms of the GNU GPL, version 2 or |
| 11 | # (at your option) any later version. See the COPYING file in the |
| 12 | # top-level directory. |
| 13 | # |
| 14 | # Usage: scripts/clean-header-guards.pl [OPTION]... [FILE]... |
| 15 | # -c CC Use a compiler other than cc |
| 16 | # -n Suppress actual cleanup |
| 17 | # -v Show which files are cleaned up, and which are skipped |
| 18 | # |
| 19 | # Does the following: |
| 20 | # - Header files without a recognizable header guard are skipped. |
| 21 | # - Clean up any untidy header guards in-place. Warn if the cleanup |
| 22 | # renames guard symbols, and explain how to find occurences of these |
| 23 | # symbols that may have to be updated manually. |
| 24 | # - Warn about duplicate header guard symbols. To make full use of |
| 25 | # this warning, you should clean up *all* headers in one run. |
| 26 | # - Warn when preprocessing a header with its guard symbol defined |
| 27 | # produces anything but whitespace. The preprocessor is run like |
| 28 | # "cc -E -DGUARD_H -c -P -", and fed the test program on stdin. |
| 29 | |
| 30 | use strict; |
| 31 | use Getopt::Std; |
| 32 | |
| 33 | # Stuff we don't want to clean because we import it into our tree: |
| 34 | my $exclude = qr,^(disas/libvixl/|include/standard-headers/ |
| 35 | |linux-headers/|pc-bios/|tests/tcg/|tests/multiboot/),x; |
| 36 | # Stuff that is expected to fail the preprocessing test: |
| 37 | my $exclude_cpp = qr,^include/libdecnumber/decNumberLocal.h,; |
| 38 | |
| 39 | my %guarded = (); |
| 40 | my %old_guard = (); |
| 41 | |
| 42 | our $opt_c = "cc"; |
| 43 | our $opt_n = 0; |
| 44 | our $opt_v = 0; |
| 45 | getopts("c:nv"); |
| 46 | |
| 47 | sub skipping { |
| 48 | my ($fname, $msg, $line1, $line2) = @_; |
| 49 | |
| 50 | return if !$opt_v or $fname =~ $exclude; |
| 51 | print "$fname skipped: $msg\n"; |
| 52 | print " $line1" if defined $line1; |
| 53 | print " $line2" if defined $line2; |
| 54 | } |
| 55 | |
| 56 | sub gripe { |
| 57 | my ($fname, $msg) = @_; |
| 58 | return if $fname =~ $exclude; |
| 59 | print STDERR "$fname: warning: $msg\n"; |
| 60 | } |
| 61 | |
| 62 | sub slurp { |
| 63 | my ($fname) = @_; |
| 64 | local $/; # slurp |
| 65 | open(my $in, "<", $fname) |
| 66 | or die "can't open $fname for reading: $!"; |
| 67 | return <$in>; |
| 68 | } |
| 69 | |
| 70 | sub unslurp { |
| 71 | my ($fname, $contents) = @_; |
| 72 | open (my $out, ">", $fname) |
| 73 | or die "can't open $fname for writing: $!"; |
| 74 | print $out $contents |
| 75 | or die "error writing $fname: $!"; |
| 76 | close $out |
| 77 | or die "error writing $fname: $!"; |
| 78 | } |
| 79 | |
| 80 | sub fname2guard { |
| 81 | my ($fname) = @_; |
| 82 | $fname =~ tr/a-z/A-Z/; |
| 83 | $fname =~ tr/A-Z0-9/_/cs; |
| 84 | return $fname; |
| 85 | } |
| 86 | |
| 87 | sub preprocess { |
| 88 | my ($fname, $guard) = @_; |
| 89 | |
| 90 | open(my $pipe, "-|", "$opt_c -E -D$guard -c -P - <$fname") |
| 91 | or die "can't run $opt_c: $!"; |
| 92 | while (<$pipe>) { |
| 93 | if ($_ =~ /\S/) { |
| 94 | gripe($fname, "not blank after preprocessing"); |
| 95 | last; |
| 96 | } |
| 97 | } |
| 98 | close $pipe |
| 99 | or gripe($fname, "preprocessing failed ($opt_c exit status $?)"); |
| 100 | } |
| 101 | |
| 102 | for my $fname (@ARGV) { |
| 103 | my $text = slurp($fname); |
| 104 | |
| 105 | $text =~ m,\A(\s*\n|\s*//\N*\n|\s*/\*.*?\*/\s*\n)*|,msg; |
| 106 | my $pre = $&; |
| 107 | unless ($text =~ /\G(.*\n)/g) { |
| 108 | $text =~ /\G.*/; |
| 109 | skipping($fname, "no recognizable header guard", "$&\n"); |
| 110 | next; |
| 111 | } |
| 112 | my $line1 = $1; |
| 113 | unless ($text =~ /\G(.*\n)/g) { |
| 114 | $text =~ /\G.*/; |
| 115 | skipping($fname, "no recognizable header guard", "$&\n"); |
| 116 | next; |
| 117 | } |
| 118 | my $line2 = $1; |
| 119 | my $body = substr($text, pos($text)); |
| 120 | |
| 121 | unless ($line1 =~ /^\s*\#\s*(if\s*\!\s*defined(\s*\()?|ifndef)\s* |
| 122 | ([A-Za-z0-9_]+)/x) { |
| 123 | skipping($fname, "no recognizable header guard", $line1, $line2); |
| 124 | next; |
| 125 | } |
| 126 | my $guard = $3; |
| 127 | unless ($line2 =~ /^\s*\#\s*define\s+([A-Za-z0-9_]+)/) { |
| 128 | skipping($fname, "no recognizable header guard", $line1, $line2); |
| 129 | next; |
| 130 | } |
| 131 | my $guard2 = $1; |
| 132 | unless ($guard2 eq $guard) { |
| 133 | skipping($fname, "mismatched header guard ($guard vs. $guard2) ", |
| 134 | $line1, $line2); |
| 135 | next; |
| 136 | } |
| 137 | |
| 138 | unless ($body =~ m,\A((.*\n)*) |
| 139 | (\s*\#\s*endif\s*(/\*\s*.*\s*\*/\s*)?\n?) |
| 140 | (\n|\s)*\Z,x) { |
| 141 | skipping($fname, "can't find end of header guard"); |
| 142 | next; |
| 143 | } |
| 144 | $body = $1; |
| 145 | my $line3 = $3; |
| 146 | my $endif_comment = $4; |
| 147 | |
| 148 | my $oldg = $guard; |
| 149 | |
| 150 | unless ($fname =~ $exclude) { |
| 151 | my @issues = (); |
| 152 | $guard =~ tr/a-z/A-Z/ |
| 153 | and push @issues, "contains lowercase letters"; |
| 154 | $guard =~ s/^_+// |
| 155 | and push @issues, "is a reserved identifier"; |
| 156 | $guard =~ s/(_H)?_*$/_H/ |
| 157 | and $& ne "_H" and push @issues, "doesn't end with _H"; |
| 158 | unless ($guard =~ /^[A-Z][A-Z0-9_]*_H/) { |
| 159 | skipping($fname, "can't clean up odd guard symbol $oldg\n", |
| 160 | $line1, $line2); |
| 161 | next; |
| 162 | } |
| 163 | |
| 164 | my $exp = fname2guard($fname =~ s,.*/,,r); |
| 165 | unless ($guard =~ /\Q$exp\E\Z/) { |
| 166 | $guard = fname2guard($fname =~ s,^include/,,r); |
| 167 | push @issues, "doesn't match the file name"; |
| 168 | } |
| 169 | if (@issues and $opt_v) { |
| 170 | print "$fname guard $oldg needs cleanup:\n ", |
| 171 | join(", ", @issues), "\n"; |
| 172 | } |
| 173 | } |
| 174 | |
| 175 | $old_guard{$guard} = $oldg |
| 176 | if $guard ne $oldg; |
| 177 | |
| 178 | if (exists $guarded{$guard}) { |
| 179 | gripe($fname, "guard $guard also used by $guarded{$guard}"); |
| 180 | } else { |
| 181 | $guarded{$guard} = $fname; |
| 182 | } |
| 183 | |
| 184 | unless ($fname =~ $exclude) { |
| 185 | my $newl1 = "#ifndef $guard\n"; |
| 186 | my $newl2 = "#define $guard\n"; |
| 187 | my $newl3 = "#endif\n"; |
| 188 | $newl3 =~ s,\Z, /* $guard */, if defined $endif_comment; |
| 189 | if ($line1 ne $newl1 or $line2 ne $newl2 or $line3 ne $newl3) { |
| 190 | $pre =~ s/\n*\Z/\n\n/ if $pre =~ /\N/; |
| 191 | $body =~ s/\A\n*/\n/; |
| 192 | if ($opt_n) { |
| 193 | print "$fname would be cleaned up\n" if $opt_v; |
| 194 | } else { |
| 195 | unslurp($fname, "$pre$newl1$newl2$body$newl3"); |
| 196 | print "$fname cleaned up\n" if $opt_v; |
| 197 | } |
| 198 | } |
| 199 | } |
| 200 | |
| 201 | preprocess($fname, $opt_n ? $oldg : $guard) |
| 202 | unless $fname =~ $exclude or $fname =~ $exclude_cpp; |
| 203 | } |
| 204 | |
| 205 | if (%old_guard) { |
| 206 | print STDERR "warning: guard symbol renaming may break things\n"; |
| 207 | for my $guard (sort keys %old_guard) { |
| 208 | print STDERR " $old_guard{$guard} -> $guard\n"; |
| 209 | } |
| 210 | print STDERR "To find uses that may have to be updated try:\n"; |
| 211 | print STDERR " git grep -Ew '", join("|", sort values %old_guard), |
| 212 | "'\n"; |
| 213 | } |