user/dev discussion of public-inbox itself
 help / color / mirror / code / Atom feed
From: Eric Wong <e@yhbt.net>
To: meta@public-inbox.org
Subject: [PATCH 65/82] index: account for CRLF conversion when storing bytes
Date: Wed, 10 Jun 2020 07:05:02 +0000	[thread overview]
Message-ID: <20200610070519.18252-66-e@yhbt.net> (raw)
In-Reply-To: <20200610070519.18252-1-e@yhbt.net>

NNTP and IMAP both require CRLF conversions on the wire.
They're also the only components which care about
$smsg->{bytes}, so store the CRLF-adjusted value in over.sqlite3
and Xapian DBs..

This will allow us to optimize RFC822.SIZE fetch item in IMAP
without triggering size mismatch errors in some clients' default
configurations (e.g. Mail::IMAPClient), but not most others.

It could also fix hypothetical problems with NNTP clients that
report discrepancies between overview and article data.
---
 lib/PublicInbox/Import.pm         |  2 +-
 lib/PublicInbox/SearchIdx.pm      | 12 ++++++++++++
 lib/PublicInbox/SearchIdxShard.pm | 11 ++++++-----
 lib/PublicInbox/V2Writable.pm     | 10 ++++++----
 t/import.t                        |  5 +++--
 t/nntpd.t                         |  5 ++++-
 t/search.t                        |  8 ++++++++
 7 files changed, 40 insertions(+), 13 deletions(-)

diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index ab75aa00dc2..af35905be49 100644
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -400,7 +400,7 @@ sub add {
 	# v2: we need this for Xapian
 	if ($smsg) {
 		$smsg->{blob} = $self->get_mark(":$blob");
-		$smsg->{bytes} = $n;
+		$smsg->{raw_bytes} = $n;
 		$smsg->{-raw_email} = \$raw_email;
 	}
 	my $ref = $self->{ref};
diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm
index a790ac4076a..85821ea706a 100644
--- a/lib/PublicInbox/SearchIdx.pm
+++ b/lib/PublicInbox/SearchIdx.pm
@@ -549,11 +549,23 @@ sub unindex_mm {
 	$self->{mm}->mid_delete(mid_mime($mime));
 }
 
+# returns the number of bytes to add if given a non-CRLF arg
+sub crlf_adjust ($) {
+	if (index($_[0], "\r\n") < 0) {
+		# common case is LF-only, every \n needs an \r;
+		# so favor a cheap tr// over an expensive m//g
+		$_[0] =~ tr/\n/\n/;
+	} else { # count number of '\n' w/o '\r', expensive:
+		scalar(my @n = ($_[0] =~ m/(?<!\r)\n/g));
+	}
+}
+
 sub index_both { # git->cat_async callback
 	my ($bref, $oid, $type, $size, $sync) = @_;
 	my ($nr, $max) = @$sync{qw(nr max)};
 	++$$nr;
 	$$max -= $size;
+	$size += crlf_adjust($$bref);
 	my $smsg = bless { bytes => $size, blob => $oid }, 'PublicInbox::Smsg';
 	my $self = $sync->{sidx};
 	my $eml = PublicInbox::Eml->new($bref);
diff --git a/lib/PublicInbox/SearchIdxShard.pm b/lib/PublicInbox/SearchIdxShard.pm
index c1f52d8b884..f7ba293ff5b 100644
--- a/lib/PublicInbox/SearchIdxShard.pm
+++ b/lib/PublicInbox/SearchIdxShard.pm
@@ -71,11 +71,11 @@ sub shard_worker_loop ($$$$$) {
 		} else {
 			chomp $line;
 			# n.b. $mid may contain spaces(!)
-			my ($bytes, $num, $blob, $ds, $ts, $mid) =
-							split(/ /, $line, 6);
+			my ($to_read, $bytes, $num, $blob, $ds, $ts, $mid) =
+							split(/ /, $line, 7);
 			$self->begin_txn_lazy;
-			my $n = read($r, my $msg, $bytes) or die "read: $!\n";
-			$n == $bytes or die "short read: $n != $bytes\n";
+			my $n = read($r, my $msg, $to_read) or die "read: $!\n";
+			$n == $to_read or die "short read: $n != $to_read\n";
 			my $mime = PublicInbox::Eml->new(\$msg);
 			my $smsg = bless {
 				bytes => $bytes,
@@ -96,7 +96,8 @@ sub index_raw {
 	my ($self, $msgref, $mime, $smsg) = @_;
 	if (my $w = $self->{w}) {
 		# mid must be last, it can contain spaces (but not LF)
-		print $w join(' ', @$smsg{qw(bytes num blob ds ts mid)}),
+		print $w join(' ', @$smsg{qw(raw_bytes bytes
+						num blob ds ts mid)}),
 			"\n", $$msgref or die "failed to write shard $!\n";
 	} else {
 		$$msgref = undef;
diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm
index 79bee7f9f3d..91379431633 100644
--- a/lib/PublicInbox/V2Writable.pm
+++ b/lib/PublicInbox/V2Writable.pm
@@ -155,10 +155,12 @@ sub add {
 # indexes a message, returns true if checkpointing is needed
 sub do_idx ($$$$) {
 	my ($self, $msgref, $mime, $smsg) = @_;
+	$smsg->{bytes} = $smsg->{raw_bytes} +
+			PublicInbox::SearchIdx::crlf_adjust($$msgref);
 	$self->{over}->add_overview($mime, $smsg);
 	my $idx = idx_shard($self, $smsg->{num} % $self->{shards});
 	$idx->index_raw($msgref, $mime, $smsg);
-	my $n = $self->{transact_bytes} += $smsg->{bytes};
+	my $n = $self->{transact_bytes} += $smsg->{raw_bytes};
 	$n >= ($PublicInbox::SearchIdx::BATCH_BYTES * $self->{shards});
 }
 
@@ -568,7 +570,7 @@ W: $list
 	for my $smsg (@$need_reindex) {
 		my $new_smsg = bless {
 			blob => $blob,
-			bytes => $bytes,
+			raw_bytes => $bytes,
 			num => $smsg->{num},
 			mid => $smsg->{mid},
 		}, 'PublicInbox::Smsg';
@@ -962,7 +964,7 @@ sub reindex_oid_m ($$$$;$) {
 	}
 	$sync->{nr}++;
 	my $smsg = bless {
-		bytes => $len,
+		raw_bytes => $len,
 		num => $num,
 		blob => $oid,
 		mid => $mid0,
@@ -1054,7 +1056,7 @@ sub reindex_oid ($$$$) {
 		die "failed to delete <$mid0> for article #$num\n";
 	$sync->{nr}++;
 	my $smsg = bless {
-		bytes => $len,
+		raw_bytes => $len,
 		num => $num,
 		blob => $oid,
 		mid => $mid0,
diff --git a/t/import.t b/t/import.t
index f987b1141f7..abbc8229d0e 100644
--- a/t/import.t
+++ b/t/import.t
@@ -32,8 +32,9 @@ like($im->add($mime, undef, $smsg), qr/\A:[0-9]+\z/, 'added one message');
 
 if ($v2) {
 	like($smsg->{blob}, qr/\A[a-f0-9]{40}\z/, 'got last object_id');
-	is($mime->as_string, ${$smsg->{-raw_email}}, 'string matches');
-	is($smsg->{bytes}, length(${$smsg->{-raw_email}}), 'length matches');
+	my $raw_email = $smsg->{-raw_email};
+	is($mime->as_string, $$raw_email, 'string matches');
+	is($smsg->{raw_bytes}, length($$raw_email), 'length matches');
 	my @cmd = ('git', "--git-dir=$git->{git_dir}", qw(hash-object --stdin));
 	my $in = tempfile();
 	print $in $mime->as_string or die "write failed: $!";
diff --git a/t/nntpd.t b/t/nntpd.t
index eee67ea65bb..d2f31323115 100644
--- a/t/nntpd.t
+++ b/t/nntpd.t
@@ -73,7 +73,10 @@ EOF
 		my $list_id = $addr;
 		$list_id =~ s/@/./;
 		$mime->header_set('List-Id', "<$list_id>");
-		$len = length($mime->as_string);
+		my $str = $mime->as_string;
+		$str =~ s/(?<!\r)\n/\r\n/sg;
+		$len = length($str);
+		undef $str;
 		$im->add($mime);
 		$im->done;
 		if ($version == 1) {
diff --git a/t/search.t b/t/search.t
index d4ca28c794f..82caf9e41c3 100644
--- a/t/search.t
+++ b/t/search.t
@@ -59,6 +59,14 @@ sub oct_is ($$$) {
 	}
 }
 
+{
+	my $crlf_adjust = \&PublicInbox::SearchIdx::crlf_adjust;
+	is($crlf_adjust->("hi\r\nworld\r\n"), 0, 'no adjustment needed');
+	is($crlf_adjust->("hi\nworld\n"), 2, 'LF-only counts two CR');
+	is($crlf_adjust->("hi\r\nworld\n"), 1, 'CRLF/LF-mix 1 counts 1 CR');
+	is($crlf_adjust->("hi\nworld\r\n"), 1, 'CRLF/LF-mix 2 counts 1 CR');
+}
+
 $ibx->with_umask(sub {
 	my $root = PublicInbox::Eml->new(<<'EOF');
 Date: Fri, 02 Oct 1993 00:00:00 +0000

  parent reply	other threads:[~2020-06-10  7:07 UTC|newest]

Thread overview: 84+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-10  7:03 [PATCH 00/82] public-inbox-imapd: read-only IMAP server Eric Wong
2020-06-10  7:03 ` [PATCH 01/82] doc: add some IMAP standards Eric Wong
2020-06-10  7:03 ` [PATCH 02/82] nntpd: restrict allowed newsgroup names Eric Wong
2020-06-10  7:04 ` [PATCH 03/82] preliminary imap server implementation Eric Wong
2020-06-10  7:04 ` [PATCH 04/82] inboxidle: new class to detect inbox changes Eric Wong
2020-06-10  7:04 ` [PATCH 05/82] imap: support IDLE Eric Wong
2020-06-10  7:04 ` [PATCH 06/82] msgmap: split ->max into its own method Eric Wong
2020-06-10  7:04 ` [PATCH 07/82] imap: delay InboxIdle start, support refresh Eric Wong
2020-06-10  7:04 ` [PATCH 08/82] imap: implement STATUS command Eric Wong
2020-06-10  7:04 ` [PATCH 09/82] imap: use Text::ParseWords::parse_line to handle quoted words Eric Wong
2020-06-10  7:04 ` [PATCH 10/82] imap: support LIST command Eric Wong
2020-06-10  7:04 ` [PATCH 11/82] t/imapd: support FakeInotify and KQNotify Eric Wong
2020-06-10  7:04 ` [PATCH 12/82] imap: support fetch for BODYSTRUCTURE and BODY Eric Wong
2020-06-10  7:04 ` [PATCH 13/82] eml: each_part: single part $idx is 1 Eric Wong
2020-06-10  7:04 ` [PATCH 14/82] imap: allow fetch of partial of BODY[...] and headers Eric Wong
2020-06-10  7:04 ` [PATCH 15/82] imap: always include `resp-text' in responses Eric Wong
2020-06-10  7:04 ` [PATCH 16/82] imap: split out unit tests and benchmarks Eric Wong
2020-06-10  7:04 ` [PATCH 17/82] imap: fix multi-message partial header fetches Eric Wong
2020-06-10  7:04 ` [PATCH 18/82] imap: simplify partial fetch structure Eric Wong
2020-06-10  7:04 ` [PATCH 19/82] imap: support sequence number FETCH Eric Wong
2020-06-10  7:04 ` [PATCH 20/82] imap: do not include ".PEEK" in responses Eric Wong
2020-06-10  7:04 ` [PATCH 21/82] imap: support the CLOSE command Eric Wong
2020-06-10  7:04 ` [PATCH 22/82] imap: speed up HEADER.FIELDS[.NOT] range fetches Eric Wong
2020-06-10  7:04 ` [PATCH 23/82] git: async: flatten the inflight array Eric Wong
2020-06-10  7:04 ` [PATCH 24/82] git: do our own read buffering for cat-file Eric Wong
2020-06-10  7:04 ` [PATCH 25/82] imap: use git-cat-file asynchronously Eric Wong
2020-06-10  7:04 ` [PATCH 26/82] git: idle rbuf for async Eric Wong
2020-06-10  7:04 ` [PATCH 27/82] imap: support LSUB command Eric Wong
2020-06-10  7:04 ` [PATCH 28/82] imap: FETCH: support comma-delimited ranges Eric Wong
2020-06-10  7:04 ` [PATCH 29/82] add imapd compression test Eric Wong
2020-06-10  7:04 ` [PATCH 30/82] testcommon: tcp_(server|connect): BAIL_OUT on failure Eric Wong
2020-06-10  7:04 ` [PATCH 31/82] *deflate: drop invalid comment about rbuf Eric Wong
2020-06-10  7:04 ` [PATCH 32/82] imap: fix pipelining with async git Eric Wong
2020-06-10  7:04 ` [PATCH 33/82] git: cat_async: provide requested OID + "missing" on missing blobs Eric Wong
2020-06-10  7:04 ` [PATCH 34/82] git: move async_cat reference to PublicInbox::Git Eric Wong
2020-06-10  7:04 ` [PATCH 35/82] git: async: automatic retry on alternates change Eric Wong
2020-06-10  7:04 ` [PATCH 36/82] imapclient: wrapper for Mail::IMAPClient Eric Wong
2020-06-10  7:04 ` [PATCH 37/82] xt: add imapd-validate and imapd-mbsync-oimap Eric Wong
2020-06-10  7:04 ` [PATCH 38/82] imap: support out-of-bounds ranges Eric Wong
2020-06-10  7:04 ` [PATCH 39/82] xt/perf-imap-list: time refresh_inboxlist Eric Wong
2020-06-10  7:04 ` [PATCH 40/82] imap: case-insensitive mailbox name comparisons Eric Wong
2020-06-10  7:04 ` [PATCH 41/82] imap: break giant inboxes into sub-inboxes of 50K messages Eric Wong
2020-06-10  7:04 ` [PATCH 42/82] imap: start doing iterative config reloading Eric Wong
2020-06-10  7:04 ` [PATCH 43/82] imap: require ".$UID_MIN-$UID_END" suffix Eric Wong
2020-06-10  7:04 ` [PATCH 44/82] imapd: ensure LIST is sorted alphabetically, for now Eric Wong
2020-06-10  7:04 ` [PATCH 45/82] imap: omit $UID_END from mailbox name, use index Eric Wong
2020-06-10  7:04 ` [PATCH 46/82] t/config.t: always compare against git bool behavior Eric Wong
2020-06-10  7:04 ` [PATCH 47/82] xt/*: show some tunable parameters Eric Wong
2020-06-10  7:04 ` [PATCH 48/82] imap: STATUS and LIST are case-insensitive, too Eric Wong
2020-06-10  7:04 ` [PATCH 49/82] imap: EXAMINE/STATUS: return correct counts Eric Wong
2020-06-10  7:04 ` [PATCH 50/82] imap: avoid uninitialized warnings on incomplete commands Eric Wong
2020-06-10  7:04 ` [PATCH 51/82] imap: start parsing out queries for SQLite and Xapian Eric Wong
2020-06-10  7:04 ` [PATCH 52/82] imap: SEARCH: clamp results to the 50K UID range Eric Wong
2020-06-10  7:04 ` [PATCH 53/82] imap: allow UID range search on timestamps Eric Wong
2020-06-10  7:04 ` [PATCH 54/82] over: get_art: use dbh->prepare_cached Eric Wong
2020-06-10  7:04 ` [PATCH 55/82] search: index byte size of a message for IMAP search Eric Wong
2020-06-10  7:04 ` [PATCH 56/82] search: index UID for IMAP search, too Eric Wong
2020-06-10  7:04 ` [PATCH 57/82] imap: remove dummies from sequence number FETCH Eric Wong
2020-06-10  7:04 ` [PATCH 58/82] imap: compile UID FETCH to opcodes Eric Wong
2020-06-10  7:04 ` [PATCH 59/82] imap: UID FETCH: optimize for smsg-only case Eric Wong
2020-06-10  7:04 ` [PATCH 60/82] imap: UID FETCH: optimize (UID FLAGS) harder Eric Wong
2020-06-10  7:04 ` [PATCH 61/82] imap: IDLE: avoid extraneous wakeups, keep-alive Eric Wong
2020-06-10  7:04 ` [PATCH 62/82] imap: 30 minute auto-logout timer Eric Wong
2020-06-10  7:05 ` [PATCH 63/82] imap: split ->logged_in attribute into a separate class Eric Wong
2020-06-10  7:05 ` [PATCH 64/82] searchidx: v1 (re)-index uses git asynchronously Eric Wong
2020-06-10  7:05 ` Eric Wong [this message]
2020-06-10  7:05 ` [PATCH 66/82] imap: rely on smsg->{bytes} for RFC822.SIZE Eric Wong
2020-06-10  7:05 ` [PATCH 67/82] imap: UID FETCH requires at least one data item Eric Wong
2020-06-10  7:05 ` [PATCH 68/82] imap: LIST shows "INBOX" in all caps Eric Wong
2020-06-10  7:05 ` [PATCH 69/82] imap: support 8000 octet lines Eric Wong
2020-06-10  7:05 ` [PATCH 70/82] imap: reinstate some message sequence number support Eric Wong
2020-06-10  7:05 ` [PATCH 71/82] imap: cleanup ->{uid_base} usage Eric Wong
2020-06-10  7:05 ` [PATCH 72/82] imap: FETCH: more granular CRLF conversion Eric Wong
2020-06-10  7:05 ` [PATCH 73/82] imap: further speed up HEADER.FIELDS FETCH requests Eric Wong
2020-06-10  7:05 ` [PATCH 74/82] imap: FETCH: try to make fake MSNs sequentially Eric Wong
2020-06-10  7:05 ` [PATCH 75/82] imap: STATUS/EXAMINE: rely on SQLite overview Eric Wong
2020-06-10  7:05 ` [PATCH 76/82] imap: UID SEARCH: support multiple ranges Eric Wong
2020-06-10  7:05 ` [PATCH 77/82] imap: wire up Xapian, MSN SEARCH and multi sequence-sets Eric Wong
2020-06-10  7:05 ` [PATCH 78/82] imap: misc cleanups and notes Eric Wong
2020-06-10  7:05 ` [PATCH 79/82] imapd: don't bother sorting LIST output Eric Wong
2020-06-10  7:05 ` [PATCH 80/82] imap: remove non-UID SEARCH for now Eric Wong
2020-06-10  7:05 ` [PATCH 81/82] over: uid_range: remove LIMIT Eric Wong
2020-06-10  7:05 ` [PATCH 82/82] imap: FETCH: proper MSN => UID mapping for requests Eric Wong
2020-06-12 23:49 ` [PATCH 83/82] imap: introduce memory-efficient uo2m mapping Eric Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://public-inbox.org/README

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200610070519.18252-66-e@yhbt.net \
    --to=e@yhbt.net \
    --cc=meta@public-inbox.org \
    --subject='Re: [PATCH 65/82] index: account for CRLF conversion when storing bytes' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Code repositories for project(s) associated with this inbox:

	https://80x24.org/public-inbox.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).