util/find-doc-nits

   1 #! /usr/bin/env perl
   2 # Copyright 2002-2016 The OpenSSL Project Authors. All Rights Reserved.
   3 #
   4 # Licensed under the OpenSSL license (the "License").  You may not use
   5 # this file except in compliance with the License.  You can obtain a copy
   6 # in the file LICENSE in the source distribution or at
   7 # https://www.openssl.org/source/license.html
   8
   9
  10 require 5.10.0;
  11 use warnings;
  12 use strict;
  13 use Pod::Checker;
  14 use File::Find;
  15 use File::Basename;
  16 use File::Spec::Functions;
  17 use Getopt::Std;
  18 use lib catdir(dirname($0), "perl");
  19 use OpenSSL::Util::Pod;
  20
  21 # Options.
  22 our($opt_s);
  23 our($opt_u);
  24 our($opt_h);
  25 our($opt_n);
  26 our($opt_l);
  27
  28 sub help()
  29 {
  30     print <<EOF;
  31 Find small errors (nits) in documentation.  Options:
  32     -l Print bogus links
  33     -n Print nits in POD pages
  34     -s Also print missing sections in POD pages (implies -n)
  35     -u List undocumented functions
  36     -h Print this help message
  37 EOF
  38     exit;
  39 }
  40
  41 my $temp = '/tmp/docnits.txt';
  42 my $OUT;
  43
  44 my %mandatory_sections =
  45     ( '*'    => [ 'NAME', 'DESCRIPTION', 'COPYRIGHT' ],
  46       1      => [ 'SYNOPSIS', 'OPTIONS' ],
  47       3      => [ 'SYNOPSIS', 'RETURN VALUES' ],
  48       5      => [ ],
  49       7      => [ ] );
  50
  51 # Cross-check functions in the NAME and SYNOPSIS section.
  52 sub name_synopsis()
  53 {
  54     my $id = shift;
  55     my $filename = shift;
  56     my $contents = shift;
  57
  58     # Get NAME section and all words in it.
  59     return unless $contents =~ /=head1 NAME(.*)=head1 SYNOPSIS/ms;
  60     my $tmp = $1;
  61     $tmp =~ tr/\n/ /;
  62     $tmp =~ s/-.*//g;
  63     $tmp =~ s/,//g;
  64
  65     my $dirname = dirname($filename);
  66     my $simplename = basename($filename);
  67     $simplename =~ s/.pod$//;
  68     my $foundfilename = 0;
  69     my %foundfilenames = ();
  70     my %names;
  71     foreach my $n ( split ' ', $tmp ) {
  72         $names{$n} = 1;
  73         $foundfilename++ if $n eq $simplename;
  74         $foundfilenames{$n} = 1
  75             if -f "$dirname/$n.pod" && $n ne $simplename;
  76     }
  77     print "$id the following exist as other .pod files:\n",
  78         join(" ", sort keys %foundfilenames), "\n"
  79         if %foundfilenames;
  80     print "$id $simplename (filename) missing from NAME section\n",
  81         unless $foundfilename;
  82
  83     # Find all functions in SYNOPSIS
  84     return unless $contents =~ /=head1 SYNOPSIS(.*)=head1 DESCRIPTION/ms;
  85     my $syn = $1;
  86     foreach my $line ( split /\n+/, $syn ) {
  87         my $sym;
  88         $line =~ s/STACK_OF\([^)]+\)/int/g;
  89         $line =~ s/__declspec\([^)]+\)//;
  90         if ( $line =~ /env (\S*)=/ ) {
  91             # environment variable env NAME=...
  92             $sym = $1;
  93         } elsif ( $line =~ /typedef.*\(\*(\S+)\)\(.*/ ) {
  94             # a callback function: typedef ... (*NAME)(...
  95             $sym = $1;
  96         } elsif ( $line =~ /typedef.* (\S+);/ ) {
  97             # a simple typedef: typedef ... NAME;
  98             $sym = $1;
  99         } elsif ( $line =~ /#define ([A-Za-z0-9_]+)/ ) {
 100             $sym = $1;
 101         } elsif ( $line =~ /([A-Za-z0-9_]+)\(/ ) {
 102             $sym = $1;
 103         }
 104         else {
 105             next;
 106         }
 107         print "$id $sym missing from NAME section\n"
 108             unless defined $names{$sym};
 109         $names{$sym} = 2;
 110
 111         # Do some sanity checks on the prototype.
 112         print "$id prototype missing spaces around commas: $line\n"
 113             if ( $line =~ /[a-z0-9],[^ ]/ );
 114     }
 115
 116     foreach my $n ( keys %names ) {
 117         next if $names{$n} == 2;
 118         print "$id $n missing from SYNOPSIS\n";
 119     }
 120 }
 121
 122 sub check()
 123 {
 124     my $filename = shift;
 125     my $dirname = basename(dirname($filename));
 126
 127     my $contents = '';
 128     {
 129         local $/ = undef;
 130         open POD, $filename or die "Couldn't open $filename, $!";
 131         $contents = <POD>;
 132         close POD;
 133     }
 134
 135     my $id = "${filename}:1:";
 136
 137     &name_synopsis($id, $filename, $contents)
 138         unless $contents =~ /=for comment generic/
 139             or $filename =~ m@man[157]/@;
 140
 141     print "$id doesn't start with =pod\n"
 142         if $contents !~ /^=pod/;
 143     print "$id doesn't end with =cut\n"
 144         if $contents !~ /=cut\n$/;
 145     print "$id more than one cut line.\n"
 146         if $contents =~ /=cut.*=cut/ms;
 147     print "$id missing copyright\n"
 148         if $contents !~ /Copyright .* The OpenSSL Project Authors/;
 149     print "$id copyright not last\n"
 150         if $contents =~ /head1 COPYRIGHT.*=head/ms;
 151     print "$id head2 in All uppercase\n"
 152         if $contents =~ /head2\s+[A-Z ]+\n/;
 153     print "$id extra space after head\n"
 154         if $contents =~ /=head\d\s\s+/;
 155     print "$id period in NAME section\n"
 156         if $contents =~ /=head1 NAME.*\.\n.*=head1 SYNOPSIS/ms;
 157     print "$id POD markup in NAME section\n"
 158         if $contents =~ /=head1 NAME.*[<>].*=head1 SYNOPSIS/ms;
 159
 160     # Look for multiple consecutive openssl #include lines.
 161     # Consecutive because of files like md5.pod. Sometimes it's okay
 162     # or necessary, as in ssl/SSL_set1_host.pod
 163     if ( $contents !~ /=for comment multiple includes/ ) {
 164         if ( $contents =~ /=head1 SYNOPSIS(.*)=head1 DESCRIPTION/ms ) {
 165             my $count = 0;
 166             foreach my $line ( split /\n+/, $1 ) {
 167                 if ( $line =~ m@include <openssl/@ ) {
 168                     if ( ++$count == 2 ) {
 169                         print "$id has multiple includes\n";
 170                     }
 171                 } else {
 172                     $count = 0;
 173                 }
 174             }
 175         }
 176     }
 177
 178     return unless $opt_s;
 179
 180     # Find what section this page is in.  If run from "." assume
 181     # section 3.
 182     my $section = 3;
 183     $section = $1 if $dirname =~ /man([1-9])/;
 184
 185     foreach ((@{$mandatory_sections{'*'}}, @{$mandatory_sections{$section}})) {
 186         print "$id: missing $_ head1 section\n"
 187             if $contents !~ /^=head1\s+${_}\s*$/m;
 188     }
 189
 190     open my $OUT, '>', $temp
 191         or die "Can't open $temp, $!";
 192     podchecker($filename, $OUT);
 193     close $OUT;
 194     open $OUT, '<', $temp
 195         or die "Can't read $temp, $!";
 196     while ( <$OUT> ) {
 197         next if /\(section\) in.*deprecated/;
 198         print;
 199     }
 200     close $OUT;
 201     unlink $temp || warn "Can't remove $temp, $!";
 202 }
 203
 204 my %dups;
 205
 206 sub parsenum()
 207 {
 208     my $file = shift;
 209     my @apis;
 210
 211     open my $IN, '<', $file
 212         or die "Can't open $file, $!, stopped";
 213
 214     while ( <$IN> ) {
 215         next if /\bNOEXIST\b/;
 216         next if /\bEXPORT_VAR_AS_FUNC\b/;
 217         push @apis, $1 if /([^\s]+).\s/;
 218     }
 219
 220     close $IN;
 221
 222     print "# Found ", scalar(@apis), " in $file\n";
 223     return sort @apis;
 224 }
 225
 226 sub getdocced()
 227 {
 228     my $dir = shift;
 229     my %return;
 230
 231     foreach my $pod ( glob("$dir/*.pod") ) {
 232         my %podinfo = extract_pod_info($pod);
 233         foreach my $n ( @{$podinfo{names}} ) {
 234             $return{$n} = $pod;
 235             print "# Duplicate $n in $pod and $dups{$n}\n"
 236                 if defined $dups{$n} && $dups{$n} ne $pod;
 237             $dups{$n} = $pod;
 238         }
 239     }
 240
 241     return %return;
 242 }
 243
 244 my %docced;
 245
 246 sub printem()
 247 {
 248     my $libname = shift;
 249     my $numfile = shift;
 250     my $count = 0;
 251
 252     foreach my $func ( &parsenum($numfile) ) {
 253         next if $docced{$func};
 254
 255         # Skip ASN1 utilities
 256         next if $func =~ /^ASN1_/;
 257
 258         print "$libname:$func\n";
 259         $count++;
 260     }
 261     print "# Found $count missing from $numfile\n\n";
 262 }
 263
 264
 265 # Collection of links in each POD file.
 266 # filename => [ "foo(1)", "bar(3)", ... ]
 267 my %link_collection = ();
 268 # Collection of names in each POD file.
 269 # "name(s)" => filename
 270 my %name_collection = ();
 271
 272 sub collectnames {
 273     my $filename = shift;
 274     $filename =~ m|man(\d)/|;
 275     my $section = $1;
 276     my $simplename = basename($filename, ".pod");
 277     my $id = "${filename}:1:";
 278
 279     my $contents = '';
 280     {
 281         local $/ = undef;
 282         open POD, $filename or die "Couldn't open $filename, $!";
 283         $contents = <POD>;
 284         close POD;
 285     }
 286
 287     $contents =~ /=head1 NAME([^=]*)=head1 /ms;
 288     my $tmp = $1;
 289     unless (defined $tmp) {
 290         print "$id weird name section\n";
 291         return;
 292     }
 293     $tmp =~ tr/\n/ /;
 294     $tmp =~ s/-.*//g;
 295
 296     my @names = map { s/\s+//g; $_ } split(/,/, $tmp);
 297     unless (grep { $simplename eq $_ } @names) {
 298         print "$id missing $simplename\n";
 299         push @names, $simplename;
 300     }
 301     foreach my $name (@names) {
 302         next if $name eq "";
 303         my $name_sec = "$name($section)";
 304         if (! exists $name_collection{$name_sec}) {
 305             $name_collection{$name_sec} = $filename;
 306         } else { #elsif ($filename ne $name_collection{$name_sec}) {
 307             print "$id $name_sec also in $name_collection{$name_sec}\n";
 308         }
 309     }
 310
 311     my @foreign_names =
 312         map { map { s/\s+//g; $_ } split(/,/, $_) }
 313         $contents =~ /=for\s+comment\s+foreign\s+manuals:\s*(.*)\n\n/;
 314     foreach (@foreign_names) {
 315         $name_collection{$_} = undef; # It still exists!
 316     }
 317
 318     my @links = $contents =~ /L<
 319                               # if the link is of the form L<something|name(s)>,
 320                               # then remove 'something'.  Note that 'something'
 321                               # may contain POD codes as well...
 322                               (?:(?:[^\|]|<[^>]*>)*\|)?
 323                               # we're only interested in referenses that have
 324                               # a one digit section number
 325                               ([^\/>\(]+\(\d\))
 326                              /gx;
 327     $link_collection{$filename} = [ @links ];
 328 }
 329
 330 sub checklinks {
 331     foreach my $filename (sort keys %link_collection) {
 332         foreach my $link (@{$link_collection{$filename}}) {
 333             print "${filename}:1: reference to non-existing $link\n"
 334                 unless exists $name_collection{$link};
 335         }
 336     }
 337 }
 338
 339 getopts('lnshu');
 340
 341 &help() if ( $opt_h );
 342
 343 die "Need one of -l -n -s or -u flags.\n"
 344     unless $opt_l or $opt_n or $opt_s or $opt_u;
 345
 346 if ( $opt_n or $opt_s ) {
 347     foreach (@ARGV ? @ARGV : glob('doc/*/*.pod')) {
 348         &check($_);
 349     }
 350 }
 351
 352 if ( $opt_l ) {
 353     foreach (@ARGV ? @ARGV : glob('doc/*/*.pod')) {
 354         collectnames($_);
 355     }
 356     checklinks();
 357 }
 358
 359 if ( $opt_u ) {
 360     my %temp = &getdocced('doc/man3');
 361     foreach ( keys %temp ) {
 362         $docced{$_} = $temp{$_};
 363     }
 364     &printem('crypto', 'util/libcrypto.num');
 365     &printem('ssl', 'util/libssl.num');
 366 }
 367
 368 exit;