* [PATCH 2/9] v2writable: preserve timestamps from import
2020-03-20 8:18 5% ` [PATCH 0/9] preserve time and date of initial commit Eric Wong
@ 2020-03-20 8:18 7% ` Eric Wong
0 siblings, 0 replies; 2+ results
From: Eric Wong @ 2020-03-20 8:18 UTC (permalink / raw)
To: meta
While v2 indexing is triggered immediately after writing the
commit to the git repository, there may be a gap between when
PublicInbox::Import generates a timestamp and when
PublicInbox::SearchIdx sees the message. So follow the mirror
indexing behavior and take the to-be-indexed (time|date)stamps
directly from the git commit.
---
lib/PublicInbox/Import.pm | 12 ++++++++----
lib/PublicInbox/V2Writable.pm | 2 +-
2 files changed, 9 insertions(+), 5 deletions(-)
diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index 68dc0c7e..3853ff2b 100644
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -274,8 +274,8 @@ sub git_timestamp {
"$ts $zone";
}
-sub extract_cmt_info ($) {
- my ($mime) = @_;
+sub extract_cmt_info ($;$) {
+ my ($mime, $v2w) = @_;
my $sender = '';
my $from = $mime->header('From');
@@ -325,6 +325,10 @@ sub extract_cmt_info ($) {
utf8::encode($subject);
my $at = git_timestamp(my @at = msg_datestamp($hdr));
my $ct = git_timestamp(my @ct = msg_timestamp($hdr));
+ if ($v2w) { # set fallbacks in case message had no date
+ $v2w->{autime} = $at[0];
+ $v2w->{cotime} = $ct[0];
+ }
($name, $email, $at, $ct, $subject);
}
@@ -370,9 +374,9 @@ sub clean_tree_v2 ($$$) {
# returns undef on duplicate
# returns the :MARK of the most recent commit
sub add {
- my ($self, $mime, $check_cb) = @_; # mime = Email::MIME
+ my ($self, $mime, $check_cb, $v2w) = @_; # mime = Email::MIME
- my ($name, $email, $at, $ct, $subject) = extract_cmt_info($mime);
+ my ($name, $email, $at, $ct, $subject) = extract_cmt_info($mime, $v2w);
my $path_type = $self->{path_type};
my $path;
if ($path_type eq '2/38') {
diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm
index f1842843..d39a6f89 100644
--- a/lib/PublicInbox/V2Writable.pm
+++ b/lib/PublicInbox/V2Writable.pm
@@ -176,7 +176,7 @@ sub _add {
defined $num or return; # duplicate
defined $mid0 or die "BUG: $mid0 undefined\n";
my $im = $self->importer;
- my $cmt = $im->add($mime);
+ my $cmt = $im->add($mime, undef, $self); # sets $self->{(au|co)time}
$cmt = $im->get_mark($cmt);
$self->{last_commit}->[$self->{epoch_max}] = $cmt;
^ permalink raw reply related [relevance 7%]
* [PATCH 0/9] preserve time and date of initial commit
@ 2020-03-20 8:18 5% ` Eric Wong
2020-03-20 8:18 7% ` [PATCH 2/9] v2writable: preserve timestamps from import Eric Wong
0 siblings, 1 reply; 2+ results
From: Eric Wong @ 2020-03-20 8:18 UTC (permalink / raw)
To: meta
For messages lacking Date and/or Received headers, search
queries for "d:YYYYMMDD..YYYYMMDD" ranges can be unreliable in
mirrors, as can the $INBOX_URL/?t=$TIMESTAMP query which
only hits SQLite.
Yes, this ended up being a lot of work to deal with corner case
messages (probably most of which are spam), but there's also a
lot of internal cleanups which made the end result easier to
follow, I think...
The main fix is actually in 1/9, but it's gross.
Patch 2/9 fixes a small window where a race can happen and
cause searches to be off by a minute.
Patches 3-8 cleanup the mess left in 1 and 2,
Finally, patch 9 fixes the corner-case-of-corner-cases for
dealing with multi-MID messages which require a one-off queue to
store the git commit/author times instead of overloading msgmap.
Eric Wong (9):
index: use git commit times on missing Date/Received
v2writable: preserve timestamps from import if generated
rename PublicInbox::SearchMsg => PublicInbox::Smsg
smsg: to_doc_data: use existing fields
overidx: parse_references: less error-prone args
*idx: pass $smsg in more places instead of many args
v2: pass smsg in more places
*idx: pass smsg in even more places
v2: SDBM-based multi Message-ID queue
Documentation/mknews.perl | 4 +-
Documentation/technical/data_structures.txt | 4 +-
MANIFEST | 4 +-
lib/PublicInbox/ExtMsg.pm | 3 +-
lib/PublicInbox/Feed.pm | 4 +-
lib/PublicInbox/Import.pm | 19 +++--
lib/PublicInbox/Inbox.pm | 2 +-
lib/PublicInbox/Mbox.pm | 4 +-
lib/PublicInbox/MsgTime.pm | 12 +--
lib/PublicInbox/MultiMidQueue.pm | 57 +++++++++++++
lib/PublicInbox/NNTP.pm | 14 ++--
lib/PublicInbox/Over.pm | 8 +-
lib/PublicInbox/OverIdx.pm | 30 +++----
lib/PublicInbox/Search.pm | 8 +-
lib/PublicInbox/SearchIdx.pm | 68 ++++++++++-----
lib/PublicInbox/SearchIdxShard.pm | 26 ++++--
lib/PublicInbox/SearchView.pm | 8 +-
lib/PublicInbox/{SearchMsg.pm => Smsg.pm} | 19 +++--
lib/PublicInbox/SolverGit.pm | 2 +-
lib/PublicInbox/V2Writable.pm | 84 ++++++++++++-------
lib/PublicInbox/View.pm | 2 +-
t/import.t | 14 ++--
t/index-git-times.t | 93 +++++++++++++++++++++
t/multi-mid.t | 27 +++++-
t/search-thr-index.t | 17 +++-
t/thread-cycle.t | 2 +-
26 files changed, 386 insertions(+), 149 deletions(-)
create mode 100644 lib/PublicInbox/MultiMidQueue.pm
rename lib/PublicInbox/{SearchMsg.pm => Smsg.pm} (92%)
create mode 100644 t/index-git-times.t
^ permalink raw reply [relevance 5%]
Results 1-2 of 2 | reverse | options above
-- pct% links below jump to the message on this page, permalinks otherwise --
2020-03-05 5:13 [PATCH] index: use git commit times on missing Date/Received Eric Wong
2020-03-20 8:18 5% ` [PATCH 0/9] preserve time and date of initial commit Eric Wong
2020-03-20 8:18 7% ` [PATCH 2/9] v2writable: preserve timestamps from import Eric Wong
Code repositories for project(s) associated with this public 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).