remove Email::Address dependency

[public-inbox.git] / script / public-inbox-learn

diff --git a/script/public-inbox-learn b/script/public-inbox-learn
index 0c7b4199abeea2def7b452101ce83e4e24b9b7a1..c79f24790e32870478a31d6c5202fada392ca738 100755 (executable)
--- a/script/public-inbox-learn
+++ b/script/public-inbox-learn
@@ -8,8 +8,10 @@ my $usage = "$0 (spam|ham) < /path/to/message";
 use strict;
 use warnings;
 use PublicInbox::Config;
+use PublicInbox::Git;
+use PublicInbox::Import;
 use Email::MIME;
-use Email::Address;
+use PublicInbox::Address;
 use IPC::Run qw/run/;
 my $train = shift or die "usage: $usage\n";
 if ($train !~ /\A(?:ham|spam)\z/) {
@@ -17,26 +19,28 @@ if ($train !~ /\A(?:ham|spam)\z/) {
 }
 
 my $pi_config = PublicInbox::Config->new;
-my $mime = Email::MIME->new(eval { local $/; <> });
+my $mime = Email::MIME->new(eval {
+       local $/;
+       my $data = scalar <STDIN>;
+       $data =~ s/\AFrom [^\r\n]*\r?\n//s;
+       $data
+});
 
 # get all recipients
 my %dests;
 foreach my $h (qw(Cc To)) {
-       foreach my $recipient (Email::Address->parse($mime->header($h))) {
-               $dests{lc($recipient->address)} = 1;
+       my $val = $mime->header($h) or next;
+       foreach my $email (PublicInbox::Address::emails($val)) {
+               $dests{lc($email)} = 1;
        }
 }
 
-my ($name, $email, $date);
-
 if ($train eq "ham") {
        require PublicInbox::MDA;
        require PublicInbox::Filter;
        PublicInbox::Filter->run($mime);
-       ($name, $email, $date) = PublicInbox::MDA->author_info($mime);
 }
 
-my $in = $mime->as_string;
 my $err = 0;
 my @output = qw(> /dev/null > /dev/null);
 
@@ -45,37 +49,30 @@ foreach my $recipient (keys %dests) {
        my $dst = $pi_config->lookup($recipient) or next;
        my $git_dir = $dst->{mainrepo} or next;
        my ($out, $err) = ("", "");
-
+       my $git = PublicInbox::Git->new($git_dir);
        # We do not touch GIT_COMMITTER_* env here so we can track
        # who trained the message.
-       # We will not touch GIT_AUTHOR_* when learning spam messages, either
+       my $name = $ENV{GIT_COMMITTER_NAME} || $dst->{inbox};
+       my $email = $ENV{GIT_COMMITTER_EMAIL} || $recipient;
+       my $im = PublicInbox::Import->new($git, $name, $email);
+
        if ($train eq "spam") {
                # This needs to be idempotent, as my inotify trainer
                # may train for each cross-posted message, and this
                # script already learns for every list in
                # ~/.public-inbox/config
-               if (!run(["ssoma-rm", $git_dir], \$in, \$out, \$err)) {
-                       if ($err !~ /^git cat-file .+ failed: 32768$/) {
-                               $err = 1;
-                       }
-               }
+               $im->remove($mime);
        } else { # $train eq "ham"
                # no checking for spam here, we assume the message has
                # been reviewed by a human at this point:
                PublicInbox::MDA->set_list_headers($mime, $dst);
-               my $s  = $mime->as_string;
-
-               local $ENV{GIT_AUTHOR_NAME} = $name;
-               local $ENV{GIT_AUTHOR_EMAIL} = $email;
-               local $ENV{GIT_AUTHOR_DATE} = $date;
 
                # Ham messages are trained when they're marked into
                # a SEEN state, so this is idempotent:
-               run([PublicInbox::MDA->cmd, $git_dir], \$s, \$out, \$err);
-               if ($err !~ /CONFLICT/) {
-                       $err = 1;
-               }
+               $im->add($mime);
        }
+       $im->done;
+       my $in = $mime->as_string;
        if (!run([qw(spamc -L), $train], \$in, @output)) {
                $err = 1;
        }