curl/tests/symbol-scan.pl

#!/usr/bin/env perl
#***************************************************************************
#                                  _   _ ____  _
#  Project                     ___| | | |  _ \| |
#                             / __| | | | |_) | |
#                            | (__| |_| |  _ <| |___
#                             \___|\___/|_| \_\_____|
#
# Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al.
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution. The terms
# are also available at https://curl.se/docs/copyright.html.
#
# You may opt to use, copy, modify, merge, publish, distribute and/or sell
# copies of the Software, and permit persons to whom the Software is
# furnished to do so, under the terms of the COPYING file.
#
# This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
# KIND, either express or implied.
#
# SPDX-License-Identifier: curl
#
###########################################################################
#
# This script grew out of help from Przemyslaw Iskra and Balint Szilakszi
# a late evening in the #curl IRC channel.
#

use strict;
use warnings;
use vars qw($Cpreprocessor);

#
# configurehelp perl module is generated by configure script
#
my $rc = eval {
    require configurehelp;
    configurehelp->import(qw(
        $Cpreprocessor
    ));
    1;
};
# Set default values if configure has not generated a configurehelp.pm file.
# This is the case with cmake.
if (!$rc) {
    $Cpreprocessor = 'cpp';
}

# we may get the dir root pointed out
my $root=$ARGV[0] || ".";

# need an include directory when building out-of-tree
my $i = ($ARGV[1]) ? "-I$ARGV[1] " : '';

my $verbose=0;
my $summary=0;
my $misses=0;

my @manrefs;
my @syms;
my %doc;
my %rem;

# scanenum runs the preprocessor on curl.h so it will process all enums
# included by it, which *should* be all headers
sub scanenum {
    my ($file) = @_;
    open my $h_in, "-|", "$Cpreprocessor $i$file" || die "Cannot preprocess $file";
    while ( <$h_in> ) {
        if ( /enum\s+(\S+\s+)?{/ .. /}/ ) {
            s/^\s+//;
            next unless /^CURL/;
            chomp;
            s/[,\s].*//;
            push @syms, $_;
        }
    }
    close $h_in || die "Error preprocessing $file";
}

sub scanheader {
    my ($f)=@_;
    open my $h, "<", "$f";
    while(<$h>) {
        if (/^#define ((LIB|)CURL[A-Za-z0-9_]*)/) {
            push @syms, $1;
        }
    }
    close $h;
}

sub scanallheaders {
    my $d = "$root/include/curl";
    opendir(my $dh, $d) ||
        die "Can't opendir: $!";
    my @headers = grep { /.h\z/ } readdir($dh);
    closedir $dh;
    foreach my $h (@headers) {
        scanenum("$d/$h");
        scanheader("$d/$h");
    }
}

sub checkmanpage {
    my ($m) = @_;

    open(my $mh, "<", "$m");
    my $line = 1;
    while(<$mh>) {
        # strip off formatting
        $_ =~ s/\\f[BPRI]//;
        # detect global-looking 'CURL[BLABLA]_*' symbols
        while(s/\W(CURL(AUTH|E|H|MOPT|OPT|SHOPT|UE|M|SSH|SSLBACKEND|HEADER|FORM|FTP|PIPE|MIMEOPT|GSSAPI|ALTSVC|PROTO|PROXY|UPART|USESSL|_READFUNC|_WRITEFUNC|_CSELECT|_FORMADD|_IPRESOLVE|_REDIR|_RTSPREQ|_TIMECOND|_VERSION)_[a-zA-Z0-9_]+)//) {
            my $s = $1;
            # skip two "special" ones
            if($s !~ /^(CURLE_OBSOLETE|CURLOPT_TEMPLATE)/) {
                push @manrefs, "$1:$m:$line";
            }
        }
        $line++;
    }
    close($mh);
}

sub scanman3dir {
    my ($d) = @_;
    opendir(my $dh, $d) ||
        die "Can't opendir: $!";
    my @mans = grep { /.3\z/ } readdir($dh);
    closedir $dh;
    for my $m (@mans) {
        checkmanpage("$d/$m");
    }
}


scanallheaders();
scanman3dir("$root/docs/libcurl");
scanman3dir("$root/docs/libcurl/opts");

open my $s, "<", "$root/docs/libcurl/symbols-in-versions";
while(<$s>) {
    if(/(^[^ \n]+) +(.*)/) {
        my ($sym, $rest)=($1, $2);
        if($doc{$sym}) {
            print "Detected duplicate symbol: $sym\n";
            $misses++;
            next;
        }
        $doc{$sym}=$sym;
        my @a=split(/ +/, $rest);
        if($a[2]) {
            # this symbol is documented to have been present the last time
            # in this release
            $rem{$sym}=$a[2];
        }
    }
}
close $s;

my $ignored=0;
for my $e (sort @syms) {
    # OBSOLETE - names that are just placeholders for a position where we
    # previously had a name, that is now removed. The OBSOLETE names should
    # never be used for anything.
    #
    # CURL_EXTERN - is a define used for libcurl functions that are external,
    # public. No app or other code should ever use it.
    #
    # CURLINC_ - defines for header dual-include prevention, ignore those.
    #
    # CURL_TEMP_ - are defined and *undefined* again within the file
    #
    # *_LAST and *_LASTENTRY are just prefix for the placeholders used for the
    # last entry in many enum series.
    #

    if($e =~ /(OBSOLETE|^CURL_EXTERN|^CURLINC_|_LAST\z|_LASTENTRY\z|^CURL_TEMP_)/) {
        $ignored++;
        next;
    }
    if($doc{$e}) {
        if($verbose) {
            print $e."\n";
        }
        $doc{$e}="used";
        next;
    }
    else {
        print $e."\n";
        $misses++;
    }
}

#
# now scan through all symbols that were present in the symbols-in-versions
# but not in the headers
#
# If the symbols were marked 'removed' in symbols-in-versions we don't output
# anything about it since that is perfectly fine.
#

my $anyremoved;

for my $e (sort keys %doc) {
    if(($doc{$e} ne "used") && !$rem{$e}) {

        if(!$anyremoved++) {
            print "Missing symbols mentioned in symbols-in-versions\n";
            print "Add them to a header, or mark them as removed.\n";
        }

        print "$e\n";
        $misses++;
    }
}

my %warned;
for my $r (@manrefs) {
    if($r =~ /^([^:]+):(.*)/) {
        my ($sym, $file)=($1, $2);
        if(!$doc{$sym} && !$warned{$sym, $file}) {
            print "$file: $sym is not a public symbol\n";
            $warned{$sym, $file} = 1;
        }
    }
}

if($summary) {
    print "Summary:\n";
    printf "%d symbols in headers (out of which %d are ignored)\n", scalar(@syms),
    $ignored;
    printf "%d symbols in headers are interesting\n",
    scalar(@syms)- $ignored;
    printf "%d symbols are listed in symbols-in-versions\n (out of which %d are listed as removed)\n", scalar(keys %doc), scalar(keys %rem);
    printf "%d symbols in symbols-in-versions should match the ones in headers\n", scalar(keys %doc) - scalar(keys %rem);
}

if($misses) {
    exit 0; # there are stuff to attend to!
}
else {
    print "OK\n";
}