checksrc: add COPYRIGHT year check

The check for updated copyright year is currently not enforced on any
file but only on files edited and/or committed locally. This is due to
the amount of files which aren't updated with their correct copyright
year at the time of their respective commit.

lib/checksrc.pl

@@ -314,7 +314,7 @@ sub scanfile {
     open(R, "<$file") || die "failed to open $file";
 
     my $incomment=0;
-    my $copyright=0;
+    my @copyright=();
     checksrc_clear(); # for file based ignores
     accept_violations();
 
@@ -330,9 +330,16 @@ sub scanfile {
             checksrc($cmd, $line, $file, $l)
         }
 
-        # check for a copyright statement
-        if(!$copyright && ($l =~ /copyright .* \d\d\d\d/i)) {
-            $copyright=1;
+        # check for a copyright statement and save the years
+        if($l =~ /\* +copyright .* \d\d\d\d/i) {
+            while($l =~ /([\d]{4})/g) {
+                push @copyright, {
+                  year => $1,
+                  line => $line,
+                  col => index($l, $1),
+                  code => $l
+                };
+            }
         }
 
         # detect long lines
@@ -650,9 +657,41 @@ sub scanfile {
         $prevl = $ol;
     }
 
-    if(!$copyright) {
+    # The check for updated copyrightyear is overly complicated in order to
+    # not punish current hacking for past sins. The copyright years are right
+    # now a bit behind, so enforcing copyright year checking on all files
+    # would cause hundreds of errors. Instead we only look at files which are
+    # tracked in the Git repo and edited in the workdir, or committed locally
+    # on the branch without being in upstream master.
+    #
+    # The simple and naive test is to simply check for the current year, but
+    # updating the year even without an edit is against project policy (and it
+    # would fail every file on January 1st).
+    #
+    # A rather more interesting, and correct, check would be to not test only
+    # locally committed files but inspect all files wrt the year of their last
+    # commit. Removing the `git rev-list origin/master..HEAD` condition below
+    # will enfore copyright year checks against the year the file was last
+    # committed (and thus edited to some degree).
+    my $commityear = undef;
+    @copyright = sort {$$b{year} cmp $$a{year}} @copyright;
+    if(!scalar(@copyright)) {
         checkwarn("COPYRIGHT", 1, 0, $file, "", "Missing copyright statement", 1);
     }
+    elsif(`git status -s -- $file` =~ /^ [MARCU]/) {
+        $commityear = (localtime(time))[5] + 1900;
+    }
+    elsif (`git rev-list --count origin/master..HEAD -- $file` !~ /^0/) {
+        my $grl = `git rev-list --max-count=1 --timestamp HEAD -- $file`;
+        $commityear = (localtime((split(/ /, $grl))[0]))[5] + 1900;
+    }
+
+    if(defined($commityear) && $copyright[0]{year}!= $commityear) {
+        checkwarn("COPYRIGHT", $copyright[0]{line}, $copyright[0]{col},
+                  $file, $copyright[0]{code},
+                  "Copyright year out of date, should be $commityear, is $copyright[0]{year}", 1);
+    }
+
     if($incomment) {
         checkwarn("OPENCOMMENT", 1, 0, $file, "", "Missing closing comment", 1);
     }

lib/md4.c

@@ -1,4 +1,5 @@
 /*
+ * !checksrc! disable COPYRIGHT
  * This is an OpenSSL-compatible implementation of the RSA Data Security, Inc.
  * MD4 Message-Digest Algorithm (RFC 1320).
  *