openssl/util/find-doc-nits

#! /usr/bin/env perl
# Copyright 2002-2021 The OpenSSL Project Authors. All Rights Reserved.
#
# Licensed under the Apache License 2.0 (the "License").  You may not use
# this file except in compliance with the License.  You can obtain a copy
# in the file LICENSE in the source distribution or at
# https://www.openssl.org/source/license.html


require 5.10.0;
use warnings;
use strict;

use Carp qw(:DEFAULT cluck);
use Pod::Checker;
use File::Find;
use File::Basename;
use File::Spec::Functions;
use Getopt::Std;
use FindBin;
use lib "$FindBin::Bin/perl";

use OpenSSL::Util::Pod;

use lib '.';
use configdata;

# Set to 1 for debug output
my $debug = 0;

# Options.
our($opt_d);
our($opt_e);
our($opt_s);
our($opt_o);
our($opt_h);
our($opt_l);
our($opt_m);
our($opt_n);
our($opt_p);
our($opt_u);
our($opt_v);
our($opt_c);

# Print usage message and exit.
sub help {
    print <<EOF;
Find small errors (nits) in documentation.  Options:
    -c List undocumented commands, undocumented options and unimplemented options.
    -d Detailed list of undocumented (implies -u)
    -e Detailed list of new undocumented (implies -v)
    -h Print this help message
    -l Print bogus links
    -m Name(s) of manuals to focus on. Default: man1,man3,man5,man7
    -n Print nits in POD pages
    -o Causes -e/-v to count symbols added since 1.1.1 as new (implies -v)
    -u Count undocumented functions
    -v Count new undocumented functions
EOF
    exit;
}

getopts('cdehlm:nouv');

help() if $opt_h;
$opt_u = 1 if $opt_d;
$opt_v = 1 if $opt_o || $opt_e;
die "Cannot use both -u and -v"
    if $opt_u && $opt_v;
die "Cannot use both -d and -e"
    if $opt_d && $opt_e;

# We only need to check c, l, n, u and v.
# Options d, e, o imply one of the above.
die "Need one of -[cdehlnouv] flags.\n"
    unless $opt_c or $opt_l or $opt_n or $opt_u or $opt_v;


my $temp = '/tmp/docnits.txt';
my $OUT;
my $status = 0;

$opt_m = "man1,man3,man5,man7" unless $opt_m;
die "Argument of -m option may contain only man1, man3, man5, and/or man7"
    unless $opt_m =~ /^(man[1357][, ]?)*$/;
my @sections = ( split /[, ]/, $opt_m );

my %mandatory_sections = (
    '*' => [ 'NAME', 'DESCRIPTION', 'COPYRIGHT' ],
    1   => [ 'SYNOPSIS', 'OPTIONS' ],
    3   => [ 'SYNOPSIS', 'RETURN VALUES' ],
    5   => [ ],
    7   => [ ]
                         );

# Symbols that we ignored.
# They are reserved macros that we currently don't document
my $ignored = qr/(?| ^i2d_
                 |   ^d2i_
                 |   ^DEPRECATEDIN
                 |   ^OSSL_DEPRECATED
                 |   \Q_fnsig(3)\E$
                 |   ^IMPLEMENT_
                 |   ^_?DECLARE_
                 |   ^sk_
                 |   ^SKM_DEFINE_STACK_OF_INTERNAL
                 |   ^lh_
                 |   ^DEFINE_LHASH_OF_(INTERNAL|DEPRECATED)
                 )/x;

# A common regexp for C symbol names
my $C_symbol = qr/\b[[:alpha:]][_[:alnum:]]*\b/;

# Collect all POD files, both internal and public, and regardless of location
# We collect them in a hash table with each file being a key, so we can attach
# tags to them.  For example, internal docs will have the word "internal"
# attached to them.
my %files = ();
# We collect files names on the fly, on known tag basis
my %collected_tags = ();
# We cache results based on tags
my %collected_results = ();

# files OPTIONS
#
# Example:
#
#       files(TAGS => 'manual');
#       files(TAGS => [ 'manual', 'man1' ]);
#
# This function returns an array of files corresponding to a set of tags
# given with the options "TAGS".  The value of this option can be a single
# word, or an array of several words, which work as inclusive or exclusive
# selectors.  Inclusive selectors are used to add one more set of files to
# the returned array, while exclusive selectors limit the set of files added
# to the array.  The recognised tag values are:
#
# 'public_manual'       - inclusive selector, adds public manuals to the
#                         returned array of files.
# 'internal_manual'     - inclusive selector, adds internal manuals to the
#                         returned array of files.
# 'manual'              - inclusive selector, adds any manual to the returned
#                         array of files.  This is really a shorthand for
#                         'public_manual' and 'internal_manual' combined.
# 'public_header'       - inclusive selector, adds public headers to the
#                         returned array of files.
# 'header'              - inclusive selector, adds any header file to the
#                         returned array of files.  Since we currently only
#                         care about public headers, this is exactly
#                         equivalent to 'public_header', but is present for
#                         consistency.
#
# 'man1', 'man3', 'man5', 'man7'
#                       - exclusive selectors, only applicable together with
#                         any of the manual selectors.  If any of these are
#                         present, only the manuals from the given sections
#                         will be included.  If none of these are present,
#                         the manuals from all sections will be returned.
#
# All returned manual files come from configdata.pm.
# All returned header files come from looking inside
# "$config{sourcedir}/include/openssl"
#
sub files {
    my %opts = ( @_ );          # Make a copy of the arguments

    $opts{TAGS} = [ $opts{TAGS} ] if ref($opts{TAGS}) eq '';

    croak "No tags given, or not an array"
        unless exists $opts{TAGS} && ref($opts{TAGS}) eq 'ARRAY';

    my %tags = map { $_ => 1 } @{$opts{TAGS}};
    $tags{public_manual} = 1
        if $tags{manual} && ($tags{public} // !$tags{internal});
    $tags{internal_manual} = 1
        if $tags{manual} && ($tags{internal} // !$tags{public});
    $tags{public_header} = 1
        if $tags{header} && ($tags{public} // !$tags{internal});
    delete $tags{manual};
    delete $tags{header};
    delete $tags{public};
    delete $tags{internal};

    my $tags_as_key = join(':', sort keys %tags);

    cluck  "DEBUG[files]: This is how we got here!" if $debug;
    print STDERR "DEBUG[files]: tags: $tags_as_key\n" if $debug;

    my %tags_to_collect = ( map { $_ => 1 }
                            grep { !exists $collected_tags{$_} }
                            keys %tags );

    if ($tags_to_collect{public_manual}) {
        print STDERR "DEBUG[files]: collecting public manuals\n"
            if $debug;

        # The structure in configdata.pm is that $unified_info{mandocs}
        # contains lists of man files, and in turn, $unified_info{depends}
        # contains hash tables showing which POD file each of those man
        # files depend on.  We use that information to find the POD files,
        # and to attach the man section they belong to as tags
        foreach my $mansect ( @sections ) {
            foreach ( map { @{$unified_info{depends}->{$_}} }
                      @{$unified_info{mandocs}->{$mansect}} ) {
                $files{$_} = { $mansect => 1, public_manual => 1 };
            }
        }
        $collected_tags{public_manual} = 1;
    }

    if ($tags_to_collect{internal_manual}) {
        print STDERR "DEBUG[files]: collecting internal manuals\n"
            if $debug;

        # We don't have the internal docs in configdata.pm.  However, they
        # are all in the source tree, so they're easy to find.
        foreach my $mansect ( @sections ) {
            foreach ( glob(catfile($config{sourcedir},
                                   'doc', 'internal', $mansect, '*.pod')) ) {
                $files{$_} = { $mansect => 1, internal_manual => 1 };
            }
        }
        $collected_tags{internal_manual} = 1;
    }

    if ($tags_to_collect{public_header}) {
        print STDERR "DEBUG[files]: collecting public headers\n"
            if $debug;

        foreach ( glob(catfile($config{sourcedir},
                               'include', 'openssl', '*.h')) ) {
            $files{$_} = { public_header => 1 };
        }
    }

    my @result = @{$collected_results{$tags_as_key} // []};

    if (!@result) {
        # Produce a result based on caller tags
        foreach my $type ( ( 'public_manual', 'internal_manual' ) ) {
            next unless $tags{$type};

            # If caller asked for specific sections, we care about sections.
            # Otherwise, we give back all of them.
            my @selected_sections =
                grep { $tags{$_} } @sections;
            @selected_sections = @sections unless @selected_sections;

            foreach my $section ( ( @selected_sections ) ) {
                push @result,
                    ( sort { basename($a) cmp basename($b) }
                      grep { $files{$_}->{$type} && $files{$_}->{$section} }
                      keys %files );
            }
        }
        if ($tags{public_header}) {
            push @result,
                ( sort { basename($a) cmp basename($b) }
                  grep { $files{$_}->{public_header} }
                  keys %files );
        }

        if ($debug) {
            print STDERR "DEBUG[files]: result:\n";
            print STDERR "DEBUG[files]:     $_\n" foreach @result;
        }
        $collected_results{$tags_as_key} = [ @result ];
    }

    return @result;
}

# Print error message, set $status.
sub err {
    print join(" ", @_), "\n";
    $status = 1
}

# Cross-check functions in the NAME and SYNOPSIS section.
sub name_synopsis {
    my $id = shift;
    my $filename = shift;
    my $contents = shift;

    # Get NAME section and all words in it.
    return unless $contents =~ /=head1 NAME(.*)=head1 SYNOPSIS/ms;
    my $tmp = $1;
    $tmp =~ tr/\n/ /;
    err($id, "Trailing comma before - in NAME")
        if $tmp =~ /, *-/;
    $tmp =~ s/ -.*//g;
    err($id, "POD markup among the names in NAME")
        if $tmp =~ /[<>]/;
    $tmp =~ s/  */ /g;
    err($id, "Missing comma in NAME")
        if $tmp =~ /[^,] /;

    my $dirname = dirname($filename);
    my $section = basename($dirname);
    my $simplename = basename($filename, ".pod");
    my $foundfilename = 0;
    my %foundfilenames = ();
    my %names;
    foreach my $n ( split ',', $tmp ) {
        $n =~ s/^\s+//;
        $n =~ s/\s+$//;
        err($id, "The name '$n' contains white-space")
            if $n =~ /\s/;
        $names{$n} = 1;
        $foundfilename++ if $n eq $simplename;
        $foundfilenames{$n} = 1
            if ( ( grep { basename($_) eq "$n.pod" }
                   files(TAGS => [ 'manual', $section ]) )
                 && $n ne $simplename );
    }
    err($id, "The following exist as other .pod files:",
         sort keys %foundfilenames)
        if %foundfilenames;
    err($id, "$simplename (filename) missing from NAME section")
        unless $foundfilename;

    # Find all functions in SYNOPSIS
    return unless $contents =~ /=head1 SYNOPSIS(.*)=head1 DESCRIPTION/ms;
    my $syn = $1;
    my $ignore_until = undef;   # If defined, this is a regexp
    # Remove all non-code lines
    $syn =~ s/^(?:\s*?|\S.*?)$//msg;
    # Remove all comments
    $syn =~ s/\/\*.*?\*\///msg;
    while ( $syn ) {
        # "env" lines end at a newline.
        # Preprocessor lines start with a # and end at a newline.
        # Other lines end with a semicolon, and may cover more than
        # one physical line.
        if ( $syn !~ /^ \s*(env .*?|#.*?|.*?;)\s*$/ms ) {
            err($id, "Can't parse rest of synopsis:\n$syn\n(declarations not ending with a semicolon (;)?)");
            last;
        }
        my $line = $1;
        $syn = $';

        print STDERR "DEBUG[name_synopsis] \$line = '$line'\n" if $debug;

        # Special code to skip over documented structures
        if ( defined $ignore_until) {
            next if $line !~ /$ignore_until/;
            $ignore_until = undef;
            next;
        }
        if ( $line =~ /^\s*(?:typedef\s+)?struct(?:\s+\S+)\s*\{/ ) {
            $ignore_until = qr/\}.*?;/;
            next;
        }

        my $sym;
        my $is_prototype = 1;
        $line =~ s/LHASH_OF\([^)]+\)/int/g;
        $line =~ s/STACK_OF\([^)]+\)/int/g;
        $line =~ s/SPARSE_ARRAY_OF\([^)]+\)/int/g;
        $line =~ s/__declspec\([^)]+\)//;

        ## We don't prohibit that space, to allow typedefs looking like
        ## this:
        ##
        ## typedef int (fantastically_long_name_breaks_80char_limit)
        ##     (fantastically_long_name_breaks_80char_limit *something);
        ##
        #if ( $line =~ /typedef.*\(\*?\S+\)\s+\(/ ) {
        #    # a callback function with whitespace before the argument list:
        #    # typedef ... (*NAME) (...
        #    # typedef ... (NAME) (...
        #    err($id, "Function typedef has space before arg list: $line");
        #}

        if ( $line =~ /env (\S*)=/ ) {
            # environment variable env NAME=...
            $sym = $1;
        } elsif ( $line =~ /typedef.*\(\*?($C_symbol)\)\s*\(/ ) {
            # a callback function pointer: typedef ... (*NAME)(...
            # a callback function signature: typedef ... (NAME)(...
            $sym = $1;
        } elsif ( $line =~ /typedef.*($C_symbol)\s*\(/ ) {
            # a callback function signature: typedef ... NAME(...
            $sym = $1;
        } elsif ( $line =~ /typedef.*($C_symbol);/ ) {
            # a simple typedef: typedef ... NAME;
            $is_prototype = 0;
            $sym = $1;
        } elsif ( $line =~ /enum ($C_symbol) \{/ ) {
            # an enumeration: enum ... {
            $sym = $1;
        } elsif ( $line =~ /#\s*(?:define|undef) ($C_symbol)/ ) {
            $is_prototype = 0;
            $sym = $1;
        } elsif ( $line =~ /^[^\(]*?\(\*($C_symbol)\s*\(/ ) {
            # a function returning a function pointer: TYPE (*NAME(args))(args)
            $sym = $1;
        } elsif ( $line =~ /^[^\(]*?($C_symbol)\s*\(/ ) {
            # a simple function declaration
            $sym = $1;
        }
        else {
            next;
        }

        print STDERR "DEBUG[name_synopsis] \$sym = '$sym'\n" if $debug;

        err($id, "$sym missing from NAME section")
            unless defined $names{$sym};
        $names{$sym} = 2;

        # Do some sanity checks on the prototype.
        err($id, "Prototype missing spaces around commas: $line")
            if $is_prototype && $line =~ /[a-z0-9],[^\s]/;
    }

    foreach my $n ( keys %names ) {
        next if $names{$n} == 2;
        err($id, "$n missing from SYNOPSIS")
    }
}

# Check if SECTION ($3) is located before BEFORE ($4)
sub check_section_location {
    my $id = shift;
    my $contents = shift;
    my $section = shift;
    my $before = shift;

    return unless $contents =~ /=head1 $section/
        and $contents =~ /=head1 $before/;
    err($id, "$section should appear before $before section")
        if $contents =~ /=head1 $before.*=head1 $section/ms;
}

# Check if a =head1 is duplicated, or a =headX is duplicated within a
# =head1.  Treats =head2 =head3 as equivalent -- it doesn't reset the head3
# sets if it finds a =head2 -- but that is good enough for now. Also check
# for proper capitalization, trailing periods, etc.
sub check_head_style {
    my $id = shift;
    my $contents = shift;
    my %head1;
    my %subheads;

    foreach my $line ( split /\n+/, $contents ) {
        next unless $line =~ /^=head/;
        if ( $line =~ /head1/ ) {
            err($id, "Duplicate section $line")
                if defined $head1{$line};
            $head1{$line} = 1;
            %subheads = ();
        } else {
            err($id, "Duplicate subsection $line")
                if defined $subheads{$line};
            $subheads{$line} = 1;
        }
        err($id, "Period in =head")
            if $line =~ /\.[^\w]/ or $line =~ /\.$/;
        err($id, "not all uppercase in =head1")
            if $line =~ /head1.*[a-z]/;
        err($id, "All uppercase in subhead")
            if $line =~ /head[234][ A-Z0-9]+$/;
    }
}

# Because we have options and symbols with extra markup, we need
# to take that into account, so we need a regexp that extracts
# markup chunks, including recursive markup.
# please read up on /(?R)/ in perlre(1)
# (note: order is important, (?R) needs to come before .)
# (note: non-greedy is important, or something like 'B<foo> and B<bar>'
# will be captured as one item)
my $markup_re =
    qr/(                        # Capture group
           [BIL]<               # The start of what we recurse on
           (?:(?-1)|.)*?        # recurse the whole regexp (referring to
                                # the last opened capture group, i.e. the
                                # start of this regexp), or pick next
                                # character.  Do NOT be greedy!
           >                    # The end of what we recurse on
       )/x;                     # (the x allows this sort of split up regexp)

# Options must start with a dash, followed by a letter, possibly
# followed by letters, digits, dashes and underscores, and the last
# character must be a letter or a digit.
# We do also accept the single -? or -n, where n is a digit
my $option_re =
    qr/(?:
            \?                  # Single question mark
            |
            \d                  # Single digit
            |
            -                   # Single dash (--)
            |
            [[:alpha:]](?:[-_[:alnum:]]*?[[:alnum:]])?
       )/x;

# Helper function to check if a given $thing is properly marked up
# option.  It returns one of these values:
#     undef         if it's not an option
#     ""            if it's a malformed option
#     $unwrapped    the option with the outermost B<> wrapping removed.
sub normalise_option {
    my $id = shift;
    my $filename = shift;
    my $thing = shift;

    my $unwrapped = $thing;
    my $unmarked = $thing;

    # $unwrapped is the option with the outer B<> markup removed
    $unwrapped =~ s/^B<//;
    $unwrapped =~ s/>$//;
    # $unmarked is the option with *all* markup removed
    $unmarked =~ s/[BIL]<|>//msg;


    # If we found an option, check it, collect it
    if ( $unwrapped =~ /^\s*-/ ) {
        return $unwrapped       # return option with outer B<> removed
            if $unmarked =~ /^-${option_re}$/;
        return "";              # Malformed option
    }
    return undef;               # Something else
}

# Checks of command option (man1) formatting.  The man1 checks are
# restricted to the SYNOPSIS and OPTIONS sections, the rest is too
# free form, we simply cannot be too strict there.

sub option_check {
    my $id = shift;
    my $filename = shift;
    my $contents = shift;

    my $synopsis = ($contents =~ /=head1\s+SYNOPSIS(.*?)=head1/s, $1);

    # Some pages have more than one OPTIONS section, let's make sure
    # to get them all
    my $options = '';
    while ( $contents =~ /=head1\s+[A-Z ]*?OPTIONS$(.*?)(?==head1)/msg ) {
        $options .= $1;
    }

    # Look for options with no or incorrect markup
    while ( $synopsis =~
            /(?<![-<[:alnum:]])-(?:$markup_re|.)*(?![->[:alnum:]])/msg ) {
        err($id, "Malformed option [1] in SYNOPSIS: $&");
    }

    my @synopsis;
    while ( $synopsis =~ /$markup_re/msg ) {
        my $found = $&;
        push @synopsis, $found if $found =~ /^B<-/;
        print STDERR "$id:DEBUG[option_check] SYNOPSIS: found $found\n"
            if $debug;
        my $option_uw = normalise_option($id, $filename, $found);
        err($id, "Malformed option [2] in SYNOPSIS: $found")
            if defined $option_uw && $option_uw eq '';
    }

    # In OPTIONS, we look for =item paragraphs.
    # (?=^\s*$) detects an empty line.
    my @options;
    while ( $options =~ /=item\s+(.*?)(?=^\s*$)/msg ) {
        my $item = $&;

        while ( $item =~ /(\[\s*)?($markup_re)/msg ) {
            my $found = $2;
            print STDERR "$id:DEBUG[option_check] OPTIONS: found $&\n"
                if $debug;
            err($id, "Unexpected bracket in OPTIONS =item: $item")
                if ($1 // '') ne '' && $found =~ /^B<\s*-/;

            my $option_uw = normalise_option($id, $filename, $found);
            err($id, "Malformed option in OPTIONS: $found")
                if defined $option_uw && $option_uw eq '';
            if ($found =~ /^B<-/) {
                push @options, $found;
                err($id, "OPTIONS entry $found missing from SYNOPSIS")
                    unless (grep /^\Q$found\E$/, @synopsis)
                         || $id =~ /(openssl|-options)\.pod:1:$/;
            }
        }
    }
    foreach (@synopsis) {
        my $option = $_;
        err($id, "SYNOPSIS entry $option missing from OPTIONS")
            unless (grep /^\Q$option\E$/, @options);
    }
}

# Normal symbol form
my $symbol_re = qr/[[:alpha:]_][_[:alnum:]]*?/;

# Checks of function name (man3) formatting.  The man3 checks are
# easier than the man1 checks, we only check the names followed by (),
# and only the names that have POD markup.
sub functionname_check {
    my $id = shift;
    my $filename = shift;
    my $contents = shift;

    while ( $contents =~ /($markup_re)\(\)/msg ) {
        print STDERR "$id:DEBUG[functionname_check] SYNOPSIS: found $&\n"
            if $debug;

        my $symbol = $1;
        my $unmarked = $symbol;
        $unmarked =~ s/[BIL]<|>//msg;

        err($id, "Malformed symbol: $symbol")
            unless $symbol =~ /^B<.*?>$/ && $unmarked =~ /^${symbol_re}$/
    }

    # We can't do the kind of collecting coolness that option_check()
    # does, because there are too many things that can't be found in
    # name repositories like the NAME sections, such as symbol names
    # with a variable part (typically marked up as B<foo_I<TYPE>_bar>
}

# This is from http://man7.org/linux/man-pages/man7/man-pages.7.html
my %preferred_words = (
    '16bit'         => '16-bit',
    'a.k.a.'        => 'aka',
    'bitmask'       => 'bit mask',
    'builtin'       => 'built-in',
   #'epoch'         => 'Epoch', # handled specially, below
    'fall-back'     => 'fallback',
    'file name'     => 'filename',
    'file system'   => 'filesystem',
    'host name'     => 'hostname',
    'i-node'        => 'inode',
    'lower case'    => 'lowercase',
    'lower-case'    => 'lowercase',
    'manpage'       => 'man page',
    'non-blocking'  => 'nonblocking',
    'non-default'   => 'nondefault',
    'non-empty'     => 'nonempty',
    'non-negative'  => 'nonnegative',
    'non-zero'      => 'nonzero',
    'path name'     => 'pathname',
    'pre-allocated' => 'preallocated',
    'pseudo-terminal' => 'pseudoterminal',
    'real time'     => 'real-time',
    'realtime'      => 'real-time',
    'reserved port' => 'privileged port',
    'runtime'       => 'run time',
    'saved group ID'=> 'saved set-group-ID',
    'saved set-GID' => 'saved set-group-ID',
    'saved set-UID' => 'saved set-user-ID',
    'saved user ID' => 'saved set-user-ID',
    'set-GID'       => 'set-group-ID',
    'set-UID'       => 'set-user-ID',
    'setgid'        => 'set-group-ID',
    'setuid'        => 'set-user-ID',
    'sub-system'    => 'subsystem',
    'super block'   => 'superblock',
    'super-block'   => 'superblock',
    'super user'    => 'superuser',
    'super-user'    => 'superuser',
    'system port'   => 'privileged port',
    'time stamp'    => 'timestamp',
    'time zone'     => 'timezone',
    'upper case'    => 'uppercase',
    'upper-case'    => 'uppercase',
    'useable'       => 'usable',
    'user name'     => 'username',
    'userspace'     => 'user space',
    'zeroes'        => 'zeros'
);

# Search manpage for words that have a different preferred use.
sub wording {
    my $id = shift;
    my $contents = shift;

    foreach my $k ( keys %preferred_words ) {
        # Sigh, trademark
        next if $k eq 'file system'
            and $contents =~ /Microsoft Encrypted File System/;
        err($id, "Found '$k' should use '$preferred_words{$k}'")
            if $contents =~ /\b\Q$k\E\b/i;
    }
    err($id, "Found 'epoch' should use 'Epoch'")
        if $contents =~ /\bepoch\b/;
    if ( $id =~ m@man1/@ ) {
        err($id, "found 'tool' in NAME, should use 'command'")
            if $contents =~ /=head1 NAME.*\btool\b.*=head1 SYNOPSIS/s;
        err($id, "found 'utility' in NAME, should use 'command'")
            if $contents =~ /NAME.*\butility\b.*=head1 SYNOPSIS/s;

    }
}

# Perform all sorts of nit/error checks on a manpage
sub check {
    my %podinfo = @_;
    my $filename = $podinfo{filename};
    my $dirname = basename(dirname($filename));
    my $contents = $podinfo{contents};

    my $id = "${filename}:1:";
    check_head_style($id, $contents);

    # Check ordering of some sections in man3
    if ( $filename =~ m|man3/| ) {
        check_section_location($id, $contents, "RETURN VALUES", "EXAMPLES");
        check_section_location($id, $contents, "SEE ALSO", "HISTORY");
        check_section_location($id, $contents, "EXAMPLES", "SEE ALSO");
    }

    # Make sure every link has a section.
    while ( $contents =~ /$markup_re/msg ) {
        my $target = $1;
        next unless $target =~ /^L<(.*)>$/;     # Skip if not L<...>
        $target = $1;                           # Peal away L< and >
        $target =~ s/\/[^\/]*$//;               # Peal away possible anchor
        $target =~ s/.*\|//g;                   # Peal away possible link text
        next if $target eq '';                  # Skip if links within page, or
        next if $target =~ /::/;                #   links to a Perl module, or
        next if $target =~ /^https?:/;          #   is a URL link, or
        next if $target =~ /\([1357]\)$/;       #   it has a section
        err($id, "Section missing in $target")
    }
    # Check for proper links to commands.
    while ( $contents =~ /L<([^>]*)\(1\)(?:\/.*)?>/g ) {
        my $target = $1;
        next if $target =~ /openssl-?/;
        next if ( grep { basename($_) eq "$target.pod" }
                  files(TAGS => [ 'manual', 'man1' ]) );
        next if $target =~ /ps|apropos|sha1sum|procmail|perl/;
        err($id, "Bad command link L<$target(1)>") if grep /man1/, @sections;
    }
    # Check for proper in-man-3 API links.
    while ( $contents =~ /L<([^>]*)\(3\)(?:\/.*)?>/g ) {
        my $target = $1;
        err($id, "Bad L<$target>")
            unless $target =~ /^[_[:alpha:]][_[:alnum:]]*$/
    }

    unless ( $contents =~ /^=for openssl generic/ms ) {
        if ( $filename =~ m|man3/| ) {
            name_synopsis($id, $filename, $contents);
            functionname_check($id, $filename, $contents);
        } elsif ( $filename =~ m|man1/| ) {
            option_check($id, $filename, $contents)
        }
    }

    wording($id, $contents);

    err($id, "Doesn't start with =pod")
        if $contents !~ /^=pod/;
    err($id, "Doesn't end with =cut")
        if $contents !~ /=cut\n$/;
    err($id, "More than one cut line.")
        if $contents =~ /=cut.*=cut/ms;
    err($id, "EXAMPLE not EXAMPLES section.")
        if $contents =~ /=head1 EXAMPLE[^S]/;
    err($id, "WARNING not WARNINGS section.")
        if $contents =~ /=head1 WARNING[^S]/;
    err($id, "Missing copyright")
        if $contents !~ /Copyright .* The OpenSSL Project Authors/;
    err($id, "Copyright not last")
        if $contents =~ /head1 COPYRIGHT.*=head/ms;
    err($id, "head2 in All uppercase")
        if $contents =~ /head2\s+[A-Z ]+\n/;
    err($id, "Extra space after head")
        if $contents =~ /=head\d\s\s+/;
    err($id, "Period in NAME section")
        if $contents =~ /=head1 NAME.*\.\n.*=head1 SYNOPSIS/ms;
    err($id, "Duplicate $1 in L<>")
        if $contents =~ /L<([^>]*)\|([^>]*)>/ && $1 eq $2;
    err($id, "Bad =over $1")
        if $contents =~ /=over([^ ][^24])/;
    err($id, "Possible version style issue")
        if $contents =~ /OpenSSL version [019]/;

    if ( $contents !~ /=for openssl multiple includes/ ) {
        # Look for multiple consecutive openssl #include lines
        # (non-consecutive lines are okay; see man3/MD5.pod).
        if ( $contents =~ /=head1 SYNOPSIS(.*)=head1 DESCRIPTION/ms ) {
            my $count = 0;
            foreach my $line ( split /\n+/, $1 ) {
                if ( $line =~ m@include <openssl/@ ) {
                    err($id, "Has multiple includes")
                        if ++$count == 2;
                } else {
                    $count = 0;
                }
            }
        }
    }

    open my $OUT, '>', $temp
        or die "Can't open $temp, $!";
    err($id, "POD errors")
        if podchecker($filename, $OUT) != 0;
    close $OUT;
    open $OUT, '<', $temp
        or die "Can't read $temp, $!";
    while ( <$OUT> ) {
        next if /\(section\) in.*deprecated/;
        print;
    }
    close $OUT;
    unlink $temp || warn "Can't remove $temp, $!";

    # Find what section this page is in; assume 3.
    my $section = 3;
    $section = $1 if $dirname =~ /man([1-9])/;

    foreach ( (@{$mandatory_sections{'*'}}, @{$mandatory_sections{$section}}) ) {
        err($id, "Missing $_ head1 section")
            if $contents !~ /^=head1\s+${_}\s*$/m;
    }
}

# Information database ###############################################

# Map of links in each POD file; filename => [ "foo(1)", "bar(3)", ... ]
my %link_map = ();
# Map of names in each POD file or from "missing" files; possible values are:
# If found in a POD files, "name(s)" => filename
# If found in a "missing" file or external, "name(s)" => ''
my %name_map = ();

# State of man-page names.
# %state is affected by loading util/*.num and util/*.syms
# Values may be one of:
# 'crypto' : belongs in libcrypto (loaded from libcrypto.num)
# 'ssl' : belongs in libssl (loaded from libssl.num)
# 'other' : belongs in libcrypto or libssl (loaded from other.syms)
# 'internal' : Internal
# 'public' : Public (generic name or external documentation)
# Any of these values except 'public' may be prefixed with 'missing_'
# to indicate that they are known to be missing.
my %state;
# %missing is affected by loading util/missing*.txt.  Values may be one of:
# 'crypto' : belongs in libcrypto (loaded from libcrypto.num)
# 'ssl' : belongs in libssl (loaded from libssl.num)
# 'other' : belongs in libcrypto or libssl (loaded from other.syms)
# 'internal' : Internal
my %missing;

# Parse libcrypto.num, etc., and return sorted list of what's there.
sub loadnum ($;$) {
    my $file = shift;
    my $type = shift;
    my @symbols;

    open my $IN, '<', catfile($config{sourcedir}, $file)
        or die "Can't open $file, $!, stopped";

    while ( <$IN> ) {
        next if /^#/;
        next if /\bNOEXIST\b/;
        my @fields = split();
        die "Malformed line $. in $file: $_"
            if scalar @fields != 2 && scalar @fields != 4;
        $state{$fields[0].'(3)'} = $type // 'internal';
    }
    close $IN;
}

# Load file of symbol names that we know aren't documented.
sub loadmissing($;$)
{
    my $missingfile = shift;
    my $type = shift;

    open FH, catfile($config{sourcedir}, $missingfile)
        or die "Can't open $missingfile";
    while ( <FH> ) {
        chomp;
        next if /^#/;
        $missing{$_} = $type // 'internal';
    }
    close FH;
}

# Check that we have consistent public / internal documentation and declaration
sub checkstate () {
    # Collect all known names, no matter where they come from
    my %names = map { $_ => 1 } (keys %name_map, keys %state, keys %missing);

    # Check section 3, i.e. functions and macros
    foreach ( grep { $_ =~ /\(3\)$/ } sort keys %names ) {
        next if ( $name_map{$_} // '') eq '' || $_ =~ /$ignored/;

        # If a man-page isn't recorded public or if it's recorded missing
        # and internal, it's declared to be internal.
        my $declared_internal =
            ($state{$_} // 'internal') eq 'internal'
            || ($missing{$_} // '') eq 'internal';
        # If a man-page isn't recorded internal or if it's recorded missing
        # and not internal, it's declared to be public
        my $declared_public =
            ($state{$_} // 'internal') ne 'internal'
            || ($missing{$_} // 'internal') ne 'internal';

        err("$_ is supposedly public but is documented as internal")
            if ( $declared_public && $name_map{$_} =~ /\/internal\// );
        err("$_ is supposedly internal (maybe missing from other.syms) but is documented as public")
            if ( $declared_internal && $name_map{$_} !~ /\/internal\// );
    }
}

# Check for undocumented macros; ignore those in the "missing" file
# and do simple check for #define in our header files.
sub checkmacros {
    my $count = 0;
    my %seen;

    foreach my $f ( files(TAGS => 'public_header') ) {
        # Skip some internals we don't want to document yet.
        my $b = basename($f);
        next if $b eq 'asn1.h';
        next if $b eq 'asn1t.h';
        next if $b eq 'err.h';
        open(IN, $f)
            or die "Can't open $f, $!";
        while ( <IN> ) {
            next unless /^#\s*define\s*(\S+)\(/;
            my $macro = "$1(3)"; # We know they're all in section 3
            next if defined $name_map{$macro}
                || defined $missing{$macro}
                || defined $seen{$macro}
                || $macro =~ /$ignored/;

            err("$f:", "macro $macro undocumented")
                if $opt_d || $opt_e;
            $count++;
            $seen{$macro} = 1;
        }
        close(IN);
    }
    err("# $count macros undocumented (count is approximate)")
        if $count > 0;
}

# Find out what is undocumented (filtering out the known missing ones)
# and display them.
sub printem ($) {
    my $type = shift;
    my $count = 0;

    foreach my $func ( grep { $state{$_} eq $type } sort keys %state ) {
        next if defined $name_map{$func}
            || defined $missing{$func};

        err("$type:", "function $func undocumented")
            if $opt_d || $opt_e;
        $count++;
    }
    err("# $count lib$type names are not documented")
        if $count > 0;
}

# Collect all the names in a manpage.
sub collectnames {
    my %podinfo = @_;
    my $filename = $podinfo{filename};
    $filename =~ m|man(\d)/|;
    my $section = $1;
    my $simplename = basename($filename, ".pod");
    my $id = "${filename}:1:";
    my $is_generic = $podinfo{contents} =~ /^=for openssl generic/ms;

    unless ( grep { $simplename eq $_ } @{$podinfo{names}} ) {
        err($id, "$simplename not in NAME section");
        push @{$podinfo{names}}, $simplename;
    }
    foreach my $name ( @{$podinfo{names}} ) {
        next if $name eq "";
        err($id, "'$name' contains whitespace")
            if $name =~ /\s/;
        my $name_sec = "$name($section)";
        if ( !defined $name_map{$name_sec} ) {
            $name_map{$name_sec} = $filename;
            $state{$name_sec} //=
                ( $filename =~ /\/internal\// ? 'internal' : 'public' )
                if $is_generic;
        } elsif ( $filename eq $name_map{$name_sec} ) {
            err($id, "$name_sec duplicated in NAME section of",
                 $name_map{$name_sec});
        } elsif ( $name_map{$name_sec} ne '' ) {
            err($id, "$name_sec also in NAME section of",
                 $name_map{$name_sec});
        }
    }

    if ( $podinfo{contents} =~ /=for openssl foreign manual (.*)\n/ ) {
        foreach my $f ( split / /, $1 ) {
            $name_map{$f} = ''; # It still exists!
            $state{$f} = 'public'; # We assume!
        }
    }

    my @links = ();
    # Don't use this regexp directly on $podinfo{contents}, as it causes
    # a regexp recursion, which fails on really big PODs.  Instead, use
    # $markup_re to pick up general markup, and use this regexp to check
    # that the markup that was found is indeed a link.
    my $linkre = qr/L<
                    # if the link is of the form L<something|name(s)>,
                    # then remove 'something'.  Note that 'something'
                    # may contain POD codes as well...
                    (?:(?:[^\|]|<[^>]*>)*\|)?
                    # we're only interested in references that have
                    # a one digit section number
                    ([^\/>\(]+\(\d\))
                   /x;
    while ( $podinfo{contents} =~ /$markup_re/msg ) {
        my $x = $1;

        if ($x =~ $linkre) {
            push @links, $1;
        }
    }
    $link_map{$filename} = [ @links ];
}

# Look for L<> ("link") references that point to files that do not exist.
sub checklinks {
    foreach my $filename ( sort keys %link_map ) {
        foreach my $link ( @{$link_map{$filename}} ) {
            err("${filename}:1:", "reference to non-existing $link")
                unless defined $name_map{$link} || defined $missing{$link};
            err("${filename}:1:", "reference of internal $link in public documentation $filename")
                if ( ( ($state{$link} // '') eq 'internal'
                       || ($missing{$link} // '') eq 'internal' )
                     && $filename !~ /\/internal\// );
        }
    }
}

# Cipher/digests to skip if they show up as "not implemented"
# because they are, via the "-*" construct.
my %skips = (
    'aes128' => 1,
    'aes192' => 1,
    'aes256' => 1,
    'aria128' => 1,
    'aria192' => 1,
    'aria256' => 1,
    'camellia128' => 1,
    'camellia192' => 1,
    'camellia256' => 1,
    'des' => 1,
    'des3' => 1,
    'idea' => 1,
    'cipher' => 1,
    'digest' => 1,
);

my %genopts; # generic options parsed from apps/include/opt.h

# Check the flags of a command and see if everything is in the manpage
sub checkflags {
    my $cmd = shift;
    my $doc = shift;
    my @cmdopts;
    my %docopts;

    # Get the list of options in the command source file.
    my $active = 0;
    my $expect_helpstr = "";
    open CFH, "apps/$cmd.c"
        or die "Can't open apps/$cmd.c to list options for $cmd, $!";
    while ( <CFH> ) {
        chop;
        if ($active) {
            last if m/^\s*};/;
            if ($expect_helpstr ne "") {
                next if m/^\s*#\s*if/;
                err("$cmd does not implement help for -$expect_helpstr") unless m/^\s*"/;
                $expect_helpstr = "";
            }
            if (m/\{\s*"([^"]+)"\s*,\s*OPT_[A-Z0-9_]+\s*,\s*('[-\/:<>cEfFlMnNpsuU]'|0)(.*)$/
                    && !($cmd eq "s_client" && $1 eq "wdebug")) {
                push @cmdopts, $1;
                $expect_helpstr = $1;
                $expect_helpstr = "" if $3 =~ m/^\s*,\s*"/;
            } elsif (m/[\s,](OPT_[A-Z]+_OPTIONS?)\s*(,|$)/) {
                push @cmdopts, @{ $genopts{$1} };
            }
        } elsif (m/^const\s+OPTIONS\s*/) {
            $active = 1;
        }
    }
    close CFH;

    # Get the list of flags from the synopsis
    open CFH, "<$doc"
        or die "Can't open $doc, $!";
    while ( <CFH> ) {
        chop;
        last if /DESCRIPTION/;
        my $opt;
        if ( /\[B<-([^ >]+)/ ) {
            $opt = $1;
        } elsif ( /^B<-([^ >]+)/ ) {
            $opt = $1;
        } else {
            next;
        }
        $opt = $1 if $opt =~ /I<(.*)/;
        $docopts{$1} = 1;
    }
    close CFH;

    # See what's in the command not the manpage.
    my @undocced = sort grep { !defined $docopts{$_} } @cmdopts;
    foreach ( @undocced ) {
        err("$doc: undocumented $cmd option -$_");
    }

    # See what's in the command not the manpage.
    my @unimpl = sort grep { my $e = $_; !(grep /^\Q$e\E$/, @cmdopts) } keys %docopts;
    foreach ( @unimpl ) {
        next if $_ eq "-"; # Skip the -- end-of-flags marker
        next if defined $skips{$_};
        err("$doc: $cmd does not implement -$_");
    }
}

##
##  MAIN()
##  Do the work requested by the various getopt flags.
##  The flags are parsed in alphabetical order, just because we have
##  to have *some way* of listing them.
##

if ( $opt_c ) {
    my @commands = ();

    # Get the lists of generic options.
    my $active = "";
    open OFH, catdir($config{sourcedir}, "apps/include/opt.h")
        or die "Can't open apps/include/opt.h to list generic options, $!";
    while ( <OFH> ) {
        chop;
        push @{ $genopts{$active} }, $1 if $active ne "" && m/^\s+\{\s*"([^"]+)"\s*,\s*OPT_/;
        $active = $1 if m/^\s*#\s*define\s+(OPT_[A-Z]+_OPTIONS?)\s*\\\s*$/;
        $active = "" if m/^\s*$/;
    }
    close OFH;

    # Get list of commands.
    opendir(DIR, "apps");
    @commands = grep(/\.c$/, readdir(DIR));
    closedir(DIR);

    # See if each has a manpage.
    foreach my $cmd ( @commands ) {
        $cmd =~ s/\.c$//;
        next if $cmd eq 'progs' || $cmd eq 'vms_decc_init';
        my @doc = ( grep { basename($_) eq "openssl-$cmd.pod"
                           # For "tsget" and "CA.pl" pod pages
                           || basename($_) eq "$cmd.pod" }
                    files(TAGS => [ 'manual', 'man1' ]) );
        my $num = scalar @doc;
        if ($num > 1) {
            err("$num manuals for 'openssl $cmd': ".join(", ", @doc));
        } elsif ($num < 1) {
            err("no manual for 'openssl $cmd'");
        } else {
            checkflags($cmd, @doc);
        }
    }
}

# Populate %state
loadnum('util/libcrypto.num', 'crypto');
loadnum('util/libssl.num', 'ssl');
loadnum('util/other.syms', 'other');
loadnum('util/other-internal.syms');
if ( $opt_o ) {
    loadmissing('util/missingmacro111.txt', 'crypto');
    loadmissing('util/missingcrypto111.txt', 'crypto');
    loadmissing('util/missingssl111.txt', 'ssl');
} elsif ( !$opt_u ) {
    loadmissing('util/missingmacro.txt', 'crypto');
    loadmissing('util/missingcrypto.txt', 'crypto');
    loadmissing('util/missingssl.txt', 'ssl');
    loadmissing('util/missingcrypto-internal.txt');
    loadmissing('util/missingssl-internal.txt');
}

if ( $opt_n || $opt_l || $opt_u || $opt_v ) {
    my @files_to_read = ( $opt_n && @ARGV ) ? @ARGV : files(TAGS => 'manual');

    foreach (@files_to_read) {
        my %podinfo = extract_pod_info($_, { debug => $debug });

        collectnames(%podinfo)
            if ( $opt_l || $opt_u || $opt_v );

        check(%podinfo)
            if ( $opt_n );
    }
}

if ( $opt_l ) {
    checklinks();
}

if ( $opt_n ) {
    # If not given args, check that all man1 commands are named properly.
    if ( scalar @ARGV == 0 && grep /man1/, @sections ) {
        foreach ( files(TAGS => [ 'public_manual', 'man1' ]) ) {
            next if /openssl\.pod/
                || /CA\.pl/ || /tsget\.pod/; # these commands are special cases
            err("$_ doesn't start with openssl-") unless /openssl-/;
        }
    }
}

checkstate();

if ( $opt_u || $opt_v) {
    printem('crypto');
    printem('ssl');
    checkmacros();
}

exit $status;