From: Sascha Silbe <sascha-pgp@silbe.org>
To: sup-devel <sup-devel@rubyforge.org>
Subject: [sup-devel] [PATCH] fix handling of multiple label: terms in search
Date: Wed, 29 Sep 2010 16:16:02 +0200 [thread overview]
Message-ID: <1285769762-15823-1-git-send-email-sascha-pgp@silbe.org> (raw)
By default Xapian will join query terms with the same prefix with OR instead
of AND, so searching for multiple labels doesn't return the expected results.
By making use of a parameter to add_boolean_prefix (added in Xapian 1.2) we
can tell Xapian to use OR only for the search terms that are guaranteed to be
unique.
Signed-off-by: Sascha Silbe <sascha-pgp@silbe.org>
---
lib/sup/index.rb | 74 +++++++++++++++++++++++++++---------------------------
1 files changed, 37 insertions(+), 37 deletions(-)
Tested on Debian Squeeze with Ruby 1.8.7.302 and Xapian 1.2.3.
diff --git a/lib/sup/index.rb b/lib/sup/index.rb
index 9273f18..a72bec6 100644
--- a/lib/sup/index.rb
+++ b/lib/sup/index.rb
@@ -419,8 +419,8 @@ EOS
qp.stemming_strategy = Xapian::QueryParser::STEM_SOME
qp.default_op = Xapian::Query::OP_AND
qp.add_valuerangeprocessor(Xapian::NumberValueRangeProcessor.new(DATE_VALUENO, 'date:', true))
- NORMAL_PREFIX.each { |k,vs| vs.each { |v| qp.add_prefix k, v } }
- BOOLEAN_PREFIX.each { |k,vs| vs.each { |v| qp.add_boolean_prefix k, v } }
+ NORMAL_PREFIX.each { |k,info| info[:prefix].each { |v| qp.add_prefix k, v } }
+ BOOLEAN_PREFIX.each { |k,info| info[:prefix].each { |v| qp.add_boolean_prefix k, v, info[:exclusive] } }
begin
xapian_query = qp.parse_query(subs, Xapian::QueryParser::FLAG_PHRASE|Xapian::QueryParser::FLAG_BOOLEAN|Xapian::QueryParser::FLAG_LOVEHATE|Xapian::QueryParser::FLAG_WILDCARD)
@@ -471,31 +471,31 @@ EOS
# Stemmed
NORMAL_PREFIX = {
- 'subject' => 'S',
- 'body' => 'B',
- 'from_name' => 'FN',
- 'to_name' => 'TN',
- 'name' => %w(FN TN),
- 'attachment' => 'A',
- 'email_text' => 'E',
- '' => %w(S B FN TN A E),
+ 'subject' => {:prefix => 'S', :exclusive => false},
+ 'body' => {:prefix => 'B', :exclusive => false},
+ 'from_name' => {:prefix => 'FN', :exclusive => false},
+ 'to_name' => {:prefix => 'TN', :exclusive => false},
+ 'name' => {:prefix => %w(FN TN), :exclusive => false},
+ 'attachment' => {:prefix => 'A', :exclusive => false},
+ 'email_text' => {:prefix => 'E', :exclusive => false},
+ '' => {:prefix => %w(S B FN TN A E), :exclusive => false},
}
# Unstemmed
BOOLEAN_PREFIX = {
- 'type' => 'K',
- 'from_email' => 'FE',
- 'to_email' => 'TE',
- 'email' => %w(FE TE),
- 'date' => 'D',
- 'label' => 'L',
- 'source_id' => 'I',
- 'attachment_extension' => 'O',
- 'msgid' => 'Q',
- 'id' => 'Q',
- 'thread' => 'H',
- 'ref' => 'R',
- 'location' => 'J',
+ 'type' => {:prefix => 'K', :exclusive => true},
+ 'from_email' => {:prefix => 'FE', :exclusive => false},
+ 'to_email' => {:prefix => 'TE', :exclusive => false},
+ 'email' => {:prefix => %w(FE TE), :exclusive => false},
+ 'date' => {:prefix => 'D', :exclusive => true},
+ 'label' => {:prefix => 'L', :exclusive => false},
+ 'source_id' => {:prefix => 'I', :exclusive => true},
+ 'attachment_extension' => {:prefix => 'O', :exclusive => false},
+ 'msgid' => {:prefix => 'Q', :exclusive => true},
+ 'id' => {:prefix => 'Q', :exclusive => true},
+ 'thread' => {:prefix => 'H', :exclusive => false},
+ 'ref' => {:prefix => 'R', :exclusive => false},
+ 'location' => {:prefix => 'J', :exclusive => false},
}
PREFIX = NORMAL_PREFIX.merge BOOLEAN_PREFIX
@@ -661,8 +661,8 @@ EOS
# Person names are indexed with several prefixes
person_termer = lambda do |d|
lambda do |p|
- doc.index_text p.name, PREFIX["#{d}_name"] if p.name
- doc.index_text p.email, PREFIX['email_text']
+ doc.index_text p.name, PREFIX["#{d}_name"][:prefix] if p.name
+ doc.index_text p.email, PREFIX['email_text'][:prefix]
doc.add_term mkterm(:email, d, p.email)
end
end
@@ -673,9 +673,9 @@ EOS
# Full text search content
subject_text = m.indexable_subject
body_text = m.indexable_body
- doc.index_text subject_text, PREFIX['subject']
- doc.index_text body_text, PREFIX['body']
- m.attachments.each { |a| doc.index_text a, PREFIX['attachment'] }
+ doc.index_text subject_text, PREFIX['subject'][:prefix]
+ doc.index_text body_text, PREFIX['body'][:prefix]
+ m.attachments.each { |a| doc.index_text a, PREFIX['attachment'][:prefix] }
# Miscellaneous terms
doc.add_term mkterm(:date, m.date) if m.date
@@ -753,25 +753,25 @@ EOS
def mkterm type, *args
case type
when :label
- PREFIX['label'] + args[0].to_s.downcase
+ PREFIX['label'][:prefix] + args[0].to_s.downcase
when :type
- PREFIX['type'] + args[0].to_s.downcase
+ PREFIX['type'][:prefix] + args[0].to_s.downcase
when :date
- PREFIX['date'] + args[0].getutc.strftime("%Y%m%d%H%M%S")
+ PREFIX['date'][:prefix] + args[0].getutc.strftime("%Y%m%d%H%M%S")
when :email
case args[0]
- when :from then PREFIX['from_email']
- when :to then PREFIX['to_email']
+ when :from then PREFIX['from_email'][:prefix]
+ when :to then PREFIX['to_email'][:prefix]
else raise "Invalid email term type #{args[0]}"
end + args[1].to_s.downcase
when :source_id
- PREFIX['source_id'] + args[0].to_s.downcase
+ PREFIX['source_id'][:prefix] + args[0].to_s.downcase
when :location
- PREFIX['location'] + [args[0]].pack('n') + args[1].to_s
+ PREFIX['location'][:prefix] + [args[0]].pack('n') + args[1].to_s
when :attachment_extension
- PREFIX['attachment_extension'] + args[0].to_s.downcase
+ PREFIX['attachment_extension'][:prefix] + args[0].to_s.downcase
when :msgid, :ref, :thread
- PREFIX[type.to_s] + args[0][0...(MAX_TERM_LENGTH-1)]
+ PREFIX[type.to_s][:prefix] + args[0][0...(MAX_TERM_LENGTH-1)]
else
raise "Invalid term type #{type}"
end
--
1.7.1
_______________________________________________
Sup-devel mailing list
Sup-devel@rubyforge.org
http://rubyforge.org/mailman/listinfo/sup-devel
next reply other threads:[~2010-09-29 14:26 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-09-29 14:16 Sascha Silbe [this message]
2010-09-29 15:27 ` Edward Z. Yang
2010-10-08 4:30 ` Rich Lane
2011-01-17 6:20 ` Rich Lane
2011-01-18 18:11 ` Sascha Silbe
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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1285769762-15823-1-git-send-email-sascha-pgp@silbe.org \
--to=sascha-pgp@silbe.org \
--cc=sascha-ml-reply-to-2010-3@silbe.org \
--cc=sup-devel@rubyforge.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox