No ext_urls

[public-inbox.git] / lib / PublicInbox / InboxWritable.pm

diff --git a/lib/PublicInbox/InboxWritable.pm b/lib/PublicInbox/InboxWritable.pm
index c3acc4f923ad03367da0faa74a0a7cd16f39785f..17dfbe18500a6dfa00e87f50209baec4126f61c3 100644 (file)
--- a/lib/PublicInbox/InboxWritable.pm
+++ b/lib/PublicInbox/InboxWritable.pm
@@ -47,7 +47,7 @@ sub _init_v1 {
                require PublicInbox::Msgmap;
                my $sidx = PublicInbox::SearchIdx->new($self, 1); # just create
                $sidx->begin_txn_lazy;
-               my $mm = PublicInbox::Msgmap->new($self->{inboxdir}, 1);
+               my $mm = PublicInbox::Msgmap->new_file($self, 1);
                if (defined $skip_artnum) {
                        $mm->{dbh}->begin_work;
                        $mm->skip_artnum($skip_artnum);
@@ -131,13 +131,9 @@ sub eml_from_path ($) {
        }
 }
 
-sub _each_maildir_fn {
-       my ($fn, $im, $self) = @_;
-       if ($fn =~ /:2,([A-Za-z]*)\z/) {
-               my $fl = $1;
-               return if $fl =~ /[DT]/; # no Drafts or Trash for public
-       }
-       my $eml = eml_from_path($fn) or return;
+sub _each_maildir_eml {
+       my ($fn, $kw, $eml, $im, $self) = @_;
+       return if grep(/\Adraft\z/, @$kw);
        if ($self && (my $filter = $self->filter($im))) {
                my $ret = $filter->scrub($eml) or return;
                return if $ret == REJECT();
@@ -146,6 +142,7 @@ sub _each_maildir_fn {
        $im->add($eml);
 }
 
+# XXX does anybody use this?
 sub import_maildir {
        my ($self, $dir) = @_;
        foreach my $sub (qw(cur new tmp)) {
@@ -153,54 +150,29 @@ sub import_maildir {
        }
        my $im = $self->importer(1);
        my @self = $self->filter($im) ? ($self) : ();
-       PublicInbox::MdirReader::maildir_each_file(\&_each_maildir_fn,
-                                               $im, @self);
+       require PublicInbox::MdirReader;
+       PublicInbox::MdirReader->new->maildir_each_eml($dir,
+                                       \&_each_maildir_eml, $im, @self);
        $im->done;
 }
 
-# asctime: From example@example.com Fri Jun 23 02:56:55 2000
-my $from_strict = qr/^From \S+ +\S+ \S+ +\S+ [^:]+:[^:]+:[^:]+ [^:]+/;
-
-sub mb_add ($$$$) {
-       my ($im, $variant, $filter, $msg) = @_;
-       $$msg =~ s/(\r?\n)+\z/$1/s;
-       if ($variant eq 'mboxrd') {
-               $$msg =~ s/^>(>*From )/$1/gms;
-       } elsif ($variant eq 'mboxo') {
-               $$msg =~ s/^>From /From /gms;
-       }
-       my $mime = PublicInbox::Eml->new($msg);
+sub _mbox_eml_cb { # MboxReader->mbox* callback
+       my ($eml, $im, $filter) = @_;
        if ($filter) {
-               my $ret = $filter->scrub($mime) or return;
+               my $ret = $filter->scrub($eml) or return;
                return if $ret == REJECT();
-               $mime = $ret;
+               $eml = $ret;
        }
-       $im->add($mime)
+       $im->add($eml);
 }
 
 sub import_mbox {
        my ($self, $fh, $variant) = @_;
-       if ($variant !~ /\A(?:mboxrd|mboxo)\z/) {
-               die "variant must be 'mboxrd' or 'mboxo'\n";
-       }
+       require PublicInbox::MboxReader;
+       my $cb = PublicInbox::MboxReader->reads($variant) or
+               die "$variant not supported\n";
        my $im = $self->importer(1);
-       my $prev = undef;
-       my $msg = '';
-       my $filter = $self->filter;
-       while (defined(my $l = <$fh>)) {
-               if ($l =~ /$from_strict/o) {
-                       if (!defined($prev) || $prev =~ /^\r?$/) {
-                               mb_add($im, $variant, $filter, \$msg) if $msg;
-                               $msg = '';
-                               $prev = $l;
-                               next;
-                       }
-                       warn "W[$.] $l\n";
-               }
-               $prev = $l;
-               $msg .= $l;
-       }
-       mb_add($im, $variant, $filter, \$msg) if $msg;
+       $cb->(undef, $fh, \&_mbox_eml_cb, $im, $self->filter);
        $im->done;
 }