-# Copyright (C) 2015, all contributors <meta@public-inbox.org>
+# Copyright (C) 2015 all contributors <meta@public-inbox.org>
# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
# based on notmuch, but with no concept of folders, files or flags
package PublicInbox::SearchIdx;
use strict;
use warnings;
use base qw(PublicInbox::Search);
-use PublicInbox::MID qw/mid_clean mid_compressed/;
+use PublicInbox::MID qw/mid_clean mid_compress/;
*xpfx = *PublicInbox::Search::xpfx;
use constant {
my $dir = $class->xdir($git_dir);
require Search::Xapian::WritableDatabase;
my $flag = Search::Xapian::DB_OPEN;
- if ($writable == 1) {
- require File::Path;
- File::Path::mkpath($dir);
- $flag = Search::Xapian::DB_CREATE_OR_OPEN;
- }
my $self = bless { git_dir => $git_dir }, $class;
- my $umask = _umask_for($self->_git_config_perm);
- my $old_umask = umask $umask;
- my $db = eval { Search::Xapian::WritableDatabase->new($dir, $flag) };
- my $err = $@;
- umask $old_umask;
- die $err if $err;
- $self->{xdb} = $db;
+ my $perm = $self->_git_config_perm;
+ my $umask = _umask_for($perm);
+ $self->{umask} = $umask;
+ $self->{xdb} = $self->with_umask(sub {
+ if ($writable == 1) {
+ require File::Path;
+ File::Path::mkpath($dir);
+ $flag = Search::Xapian::DB_CREATE_OR_OPEN;
+ }
+ Search::Xapian::WritableDatabase->new($dir, $flag);
+ });
$self;
}
my $db = $self->{xdb};
my $doc_id;
- my $mid_orig = mid_clean($mime->header_obj->header('Message-ID'));
- my $mid = mid_compressed($mid_orig);
+ my $mid = mid_clean($mime->header('Message-ID'));
my $was_ghost = 0;
my $ct_msg = $mime->header('Content-Type') || 'text/plain';
my $subj = $smsg->subject;
- if (length $subj) {
+ if ($subj ne '') {
$doc->add_term(xpfx('subject') . $subj);
my $path = $self->subject_path($subj);
- $doc->add_term(xpfx('path') . mid_compressed($path));
+ $doc->add_term(xpfx('path') . mid_compress($path));
}
- my $from = $smsg->from_name;
- my $date = $smsg->date;
my $ts = Search::Xapian::sortable_serialise($smsg->ts);
$doc->add_value(PublicInbox::Search::TS, $ts);
};
if ($@) {
- warn "failed to index message <$mid_orig>: $@\n";
+ warn "failed to index message <$mid>: $@\n";
return undef;
}
$doc_id;
# returns deleted doc_id on success, undef on missing
sub remove_message {
- my ($self, $mid_orig) = @_;
+ my ($self, $mid) = @_;
my $db = $self->{xdb};
my $doc_id;
- $mid_orig = mid_clean($mid_orig);
- my $mid = mid_compressed($mid_orig);
+ $mid = mid_clean($mid);
eval {
$doc_id = $self->find_unique_doc_id('mid', $mid);
};
if ($@) {
- warn "failed to remove message <$mid_orig>: $@\n";
+ warn "failed to remove message <$mid>: $@\n";
return undef;
}
$doc_id;
sub link_message_to_parents {
my ($self, $smsg) = @_;
my $doc = $smsg->{doc};
- my $mid = mid_compressed($smsg->mid);
+ my $mid = $smsg->mid;
my $mime = $smsg->mime;
- my $refs = $mime->header_obj->header('References');
+ my $refs = $mime->header('References');
my @refs = $refs ? ($refs =~ /<([^>]+)>/g) : ();
- my $irt = $mime->header_obj->header('In-Reply-To');
- if ($irt) {
- $irt = mid_compressed(mid_clean($irt));
-
- # maybe some crazies will try to make a circular reference:
- if ($irt eq $mid) {
- $irt = undef;
- } else {
- # last References should be $irt
- # we will de-dupe later
- push @refs, $irt;
- }
+ if (my $irt = $mime->header('In-Reply-To')) {
+ # last References should be $irt
+ # we will de-dupe later
+ push @refs, mid_clean($irt);
}
my $tid;
if (@refs) {
- my @crefs = map { mid_compressed($_) } @refs;
my %uniq = ($mid => 1);
+ my @orig_refs = @refs;
+ @refs = ();
# prevent circular references via References: here:
- @refs = ();
- foreach my $ref (@crefs) {
+ foreach my $ref (@orig_refs) {
next if $uniq{$ref};
$uniq{$ref} = 1;
push @refs, $ref;
}
}
if (@refs) {
- $doc->add_term(xpfx('inreplyto') . $irt) if defined $irt;
$smsg->{references_sorted} = '<'.join('><', @refs).'>';
- my $ref_pfx = xpfx('references');
-
# first ref *should* be the thread root,
# but we can never trust clients to do the right thing
my $ref = shift @refs;
- $doc->add_term($ref_pfx . $ref);
$tid = $self->_resolve_mid_to_tid($ref);
# the rest of the refs should point to this tid:
foreach $ref (@refs) {
- $doc->add_term($ref_pfx . $ref);
my $ptid = $self->_resolve_mid_to_tid($ref);
if ($tid ne $ptid) {
$self->merge_threads($tid, $ptid);
sub unindex_blob {
my ($self, $git, $blob) = @_;
my $mime = do_cat_mail($git, $blob) or return;
- my $mid = $mime->header_obj->header('Message-ID');
+ my $mid = $mime->header('Message-ID');
eval { $self->remove_message($mid) } if defined $mid;
warn "W: unindex_blob $blob: $@\n" if $@;
}
$@ ? undef : $mime;
}
-# indexes all unindexed messages
sub index_sync {
+ my ($self, $head) = @_;
+ $self->with_umask(sub { $self->_index_sync($head) });
+}
+
+# indexes all unindexed messages
+sub _index_sync {
my ($self, $head) = @_;
require PublicInbox::GitCatFile;
my $db = $self->{xdb};
my $git = PublicInbox::GitCatFile->new($self->{git_dir});
my $latest = $db->get_metadata('last_commit');
- my $range = length $latest ? "$latest..$head" : $head;
+ my $range = $latest eq '' ? $head : "$latest..$head";
$latest = undef;
# get indexed messages
sub create_ghost {
my ($self, $mid, $tid) = @_;
- $mid = mid_compressed($mid);
$tid = $self->next_thread_id unless defined $tid;
my $doc = Search::Xapian::Document->new;
sub _git_config_perm {
my $self = shift;
my $perm = scalar @_ ? $_[0] : _read_git_config_perm($self);
- return PERM_GROUP if (!defined($perm) || !length($perm));
+ return PERM_GROUP if (!defined($perm) || $perm eq '');
return PERM_UMASK if ($perm eq 'umask');
return PERM_GROUP if ($perm eq 'group');
if ($perm =~ /\A(?:all|world|everybody)\z/) {
(~$rv & 0777);
}
+sub with_umask {
+ my ($self, $cb) = @_;
+ my $old = umask $self->{umask};
+ my $rv = eval { $cb->() };
+ my $err = $@;
+ umask $old;
+ die $err if $@;
+ $rv;
+}
+
1;