about summary refs log tree commit
diff options
context:
space:
mode:
authorEric Wong <e@yhbt.net>2020-04-18 03:38:53 +0000
committerEric Wong <e@yhbt.net>2020-04-19 08:51:29 +0000
commit0c586dc64b3b6642a894e125d09df446667a4079 (patch)
tree38d6b349b64e6395e1910aaae2ef450b06d96a13
parent92de3139920992bfad32ef927153f27addfdc72c (diff)
downloadpublic-inbox-0c586dc64b3b6642a894e125d09df446667a4079.tar.gz
It's unnecessary overhead for anything which does Email::MIME
parsing.  It was never done for v2 indexing, even though v1->v2
conversions did NOT remove those From_ lines.  There was never a
need to remote From_ lines the v1 SearchIdx paths, either.

Hitting a /$INBOX_URL/$MSGID/T/ endpoint with an 18 message
thread reveals a ~0.5% speed improvement.  This will become
more apparent when we have a faster MIME parser.
-rw-r--r--lib/PublicInbox/Inbox.pm8
-rw-r--r--lib/PublicInbox/Mbox.pm7
-rw-r--r--lib/PublicInbox/NNTP.pm2
-rw-r--r--lib/PublicInbox/SearchIdx.pm2
-rw-r--r--t/psgi_v2.t28
5 files changed, 27 insertions, 20 deletions
diff --git a/lib/PublicInbox/Inbox.pm b/lib/PublicInbox/Inbox.pm
index 4bd82989..186eb420 100644
--- a/lib/PublicInbox/Inbox.pm
+++ b/lib/PublicInbox/Inbox.pm
@@ -311,9 +311,7 @@ sub nntp_usable {
 # for v1 users w/o SQLite only
 sub msg_by_path ($$;$) {
         my ($self, $path, $ref) = @_;
-        my $str = git($self)->cat_file('HEAD:'.$path, $ref);
-        $$str =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s if $str;
-        $str;
+        git($self)->cat_file('HEAD:'.$path, $ref);
 }
 
 sub msg_by_smsg ($$;$) {
@@ -324,9 +322,7 @@ sub msg_by_smsg ($$;$) {
         return unless defined $smsg;
         defined(my $blob = $smsg->{blob}) or return;
 
-        my $str = git($self)->cat_file($blob, $ref);
-        $$str =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s if $str;
-        $str;
+        git($self)->cat_file($blob, $ref);
 }
 
 sub smsg_mime {
diff --git a/lib/PublicInbox/Mbox.pm b/lib/PublicInbox/Mbox.pm
index 16de1a72..9995140c 100644
--- a/lib/PublicInbox/Mbox.pm
+++ b/lib/PublicInbox/Mbox.pm
@@ -106,8 +106,11 @@ sub msg_hdr ($$;$) {
                 'List-Post', "<mailto:$ibx->{-primary_address}>",
         );
         my $crlf = $header_obj->crlf;
-        my $buf = 'From mboxrd@z Thu Jan  1 00:00:00 1970' . $crlf .
-                        $header_obj->as_string;
+        my $buf = $header_obj->as_string;
+        # fixup old bug from import (pre-a0c07cba0e5d8b6a)
+        $buf =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s;
+        $buf = "From mboxrd\@z Thu Jan  1 00:00:00 1970" . $crlf . $buf;
+
         for (my $i = 0; $i < @append; $i += 2) {
                 my $k = $append[$i];
                 my $v = $append[$i + 1];
diff --git a/lib/PublicInbox/NNTP.pm b/lib/PublicInbox/NNTP.pm
index d1f75f6f..c79f198b 100644
--- a/lib/PublicInbox/NNTP.pm
+++ b/lib/PublicInbox/NNTP.pm
@@ -506,6 +506,8 @@ sub set_art {
 sub msg_hdr_write ($$$) {
         my ($self, $hdr, $body_follows) = @_;
         $hdr = $hdr->as_string;
+        # fixup old bug from import (pre-a0c07cba0e5d8b6a)
+        $hdr =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s;
         utf8::encode($hdr);
         $hdr =~ s/(?<!\r)\n/\r\n/sg; # Alpine barfs without this
 
diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm
index d1290dc2..579b85e3 100644
--- a/lib/PublicInbox/SearchIdx.pm
+++ b/lib/PublicInbox/SearchIdx.pm
@@ -553,8 +553,6 @@ sub do_cat_mail {
         my ($git, $blob, $sizeref) = @_;
         my $str = $git->cat_file($blob, $sizeref) or
                 die "BUG: $blob not found in $git->{git_dir}";
-        # fixup bugs from import:
-        $$str =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s;
         PublicInbox::MIME->new($str);
 }
 
diff --git a/t/psgi_v2.t b/t/psgi_v2.t
index c4f80869..57017de1 100644
--- a/t/psgi_v2.t
+++ b/t/psgi_v2.t
@@ -26,16 +26,16 @@ my $new_mid;
 my $im = PublicInbox::V2Writable->new($ibx, 1);
 $im->{parallel} = 0;
 
-my $mime = PublicInbox::MIME->create(
-        header => [
-                From => 'a@example.com',
-                To => 'test@example.com',
-                Subject => 'this is a subject',
-                'Message-ID' => '<a-mid@b>',
-                Date => 'Fri, 02 Oct 1993 00:00:00 +0000',
-        ],
-        body => "hello world\n",
-);
+my $mime = PublicInbox::MIME->new(<<'EOF');
+From oldbug-pre-a0c07cba0e5d8b6a Fri Oct  2 00:00:00 1993
+From: a@example.com
+To: test@example.com
+Subject: this is a subject
+Message-ID: <a-mid@b>
+Date: Fri, 02 Oct 1993 00:00:00 +0000
+
+hello world
+EOF
 ok($im->add($mime), 'added one message');
 $mime->body_set("hello world!\n");
 
@@ -48,6 +48,10 @@ my $mids = mids($mime->header_obj);
 $new_mid = $mids->[1];
 $im->done;
 
+my $msg = $ibx->msg_by_mid('a-mid@b');
+like($$msg, qr/\AFrom oldbug/s,
+        '"From_" line stored to test old bug workaround');
+
 my $cfgpfx = "publicinbox.v2test";
 my $cfg = <<EOF;
 $cfgpfx.address=$ibx->{-primary_address}
@@ -63,6 +67,7 @@ test_psgi(sub { $www->call(@_) }, sub {
                 'got v2 description missing message');
         $res = $cb->(GET('/v2test/a-mid@b/raw'));
         $raw = $res->content;
+        unlike($raw, qr/^From oldbug/sm, 'buggy "From_" line omitted');
         like($raw, qr/^hello world$/m, 'got first message');
         like($raw, qr/^hello world!$/m, 'got second message');
         @from_ = ($raw =~ m/^From /mg);
@@ -123,6 +128,7 @@ test_psgi(sub { $www->call(@_) }, sub {
                 my $out;
                 my $in = $res->content;
                 my $status = IO::Uncompress::Gunzip::gunzip(\$in => \$out);
+                unlike($out, qr/^From oldbug/sm, 'buggy "From_" line omitted');
                 like($out, qr/^hello world$/m, 'got first in t.mbox.gz');
                 like($out, qr/^hello world!$/m, 'got second in t.mbox.gz');
                 like($out, qr/^hello ghosts$/m, 'got third in t.mbox.gz');
@@ -133,6 +139,7 @@ test_psgi(sub { $www->call(@_) }, sub {
                 $res = $cb->(POST('/v2test/?q=m:a-mid@b&x=m'));
                 $in = $res->content;
                 $status = IO::Uncompress::Gunzip::gunzip(\$in => \$out);
+                unlike($out, qr/^From oldbug/sm, 'buggy "From_" line omitted');
                 like($out, qr/^hello world$/m, 'got first in mbox POST');
                 like($out, qr/^hello world!$/m, 'got second in mbox POST');
                 like($out, qr/^hello ghosts$/m, 'got third in mbox POST');
@@ -143,6 +150,7 @@ test_psgi(sub { $www->call(@_) }, sub {
                 $res = $cb->(GET('/v2test/all.mbox.gz'));
                 $in = $res->content;
                 $status = IO::Uncompress::Gunzip::gunzip(\$in => \$out);
+                unlike($out, qr/^From oldbug/sm, 'buggy "From_" line omitted');
                 like($out, qr/^hello world$/m, 'got first in all.mbox');
                 like($out, qr/^hello world!$/m, 'got second in all.mbox');
                 like($out, qr/^hello ghosts$/m, 'got third in all.mbox');