require 'thread'
require 'fileutils'
require 'ferret'
-#require_gem 'ferret', ">= 0.10.13"
module Redwood
-class IndexError < StandardError
- attr_reader :source
-
- def initialize source, s
- super s
- @source = source
- end
-end
-
class Index
include Singleton
end
def save
+ Redwood::log "saving index and sources..."
FileUtils.mkdir_p @dir unless File.exists? @dir
save_sources
save_index
@sources[source.id] = source
end
- def source_for name; @sources.values.find { |s| s.is_source_for? name }; end
+ def source_for uri; @sources.values.find { |s| s.is_source_for? uri }; end
def usual_sources; @sources.values.find_all { |s| s.usual? }; end
def sources; @sources.values; end
end
end
- ## update the message by deleting and re-adding
- def update_message m, source=nil, source_info=nil
- docid, entry = load_entry_for_id m.id
- if entry
- source ||= entry[:source_id].to_i
- source_info ||= entry[:source_info].to_i
- end
+ ## Syncs the message to the index: deleting if it's already there,
+ ## and adding either way. Index state will be determined by m.labels.
+ ##
+ ## docid and entry can be specified if they're already known.
+ def sync_message m, docid=nil, entry=nil
+ docid, entry = load_entry_for_id m.id unless docid && entry
+
+ raise "no source info for message #{m.id}" unless m.source && m.source_info
+ raise "trying deleting non-corresponding entry #{docid}" if docid && @index[docid][:message_id] != m.id
+
+ source_id =
+ if m.source.is_a? Integer
+ raise "Debugging: integer source set"
+ m.source
+ else
+ m.source.id or raise "unregistered source #{m.source} (id #{m.source.id.inspect})"
+ end
- ## this happens sometimes. i'm not sure why. ferret bug?
- raise "no entry and no source info for message #{m.id}: source #{source.inspect}, info #{source_info.inspect}, entry #{entry.inspect}, query #{Ferret::Search::TermQuery.new(:message_id, m.id)}, results #{@index.search(Ferret::Search::TermQuery.new(:message_id, m.id)).inspect}" unless source && source_info
+ to = (m.to + m.cc + m.bcc).map { |x| x.email }.join(" ")
+ d = {
+ :message_id => m.id,
+ :source_id => source_id,
+ :source_info => m.source_info,
+ :date => m.date.to_indexable_s,
+ :body => m.content,
+ :snippet => m.snippet,
+ :label => m.labels.join(" "),
+ :from => m.from ? m.from.email : "",
+ :to => (m.to + m.cc + m.bcc).map { |x| x.email }.join(" "),
+ :subject => wrap_subj(Message.normalize_subj(m.subj)),
+ :refs => (m.refs + m.replytos).uniq.join(" "),
+ }
- raise "deleting non-corresponding entry #{docid}" unless @index[docid][:message_id] == m.id
- @index.delete docid
- add_message m
+ @index.delete docid if docid
+ @index.add_document d
+
+ docid, entry = load_entry_for_id m.id
+ ## this hasn't been triggered in a long time. TODO: decide whether it's still a problem.
+ raise "just added message #{m.id} but couldn't find it in a search" unless docid
+ true
end
def save_index fn=File.join(@dir, "ferret")
- # don't have to do anything apparently
+ # don't have to do anything, apparently
end
def contains_id? id
def size; @index.size; end
## you should probably not call this on a block that doesn't break
- ## rather quickly because the results will probably be, as we say
- ## in scotland, frikkin' huuuge.
+ ## rather quickly because the results can be very large.
EACH_BY_DATE_NUM = 100
def each_id_by_date opts={}
return if @index.size == 0 # otherwise ferret barfs ###TODO: remove this once my ferret patch is accepted
end
## yield all messages in the thread containing 'm' by repeatedly
- ## querying the index. yields pairs of message ids and
+ ## querying the index. uields pairs of message ids and
## message-building lambdas, so that building an unwanted message
## can be skipped in the block if desired.
+ ##
+ ## stops loading any thread if a message with a :killed flag is found.
+
SAME_SUBJECT_DATE_LIMIT = 7
def each_message_in_thread_for m, opts={}
messages = {}
searched = {}
num_queries = 0
- ## temporarily disabling subject searching because it's a
- ## significant slowdown.
- ##
- ## TODO: make this configurable, i guess
- if false
+ ## todo: make subject querying configurable
+ if true # do subject queries
date_min = m.date - (SAME_SUBJECT_DATE_LIMIT * 12 * 3600)
date_max = m.date + (SAME_SUBJECT_DATE_LIMIT * 12 * 3600)
q.add_query sq, :must
q.add_query Ferret::Search::RangeQuery.new(:date, :>= => date_min.to_indexable_s, :<= => date_max.to_indexable_s), :must
+ q = build_query :qobj => q
+
pending = @index.search(q).hits.map { |hit| @index[hit.doc][:message_id] }
Redwood::log "found #{pending.size} results for subject query #{q}"
else
q.add_query Ferret::Search::TermQuery.new(:message_id, id), :should
q.add_query Ferret::Search::TermQuery.new(:refs, id), :should
+ q = build_query :qobj => q, :load_killed => true
+
num_queries += 1
@index.search_each(q, :limit => :all) do |docid, score|
break if opts[:limit] && messages.size >= opts[:limit]
+ break if @index[docid][:label].split(/\s+/).include? "killed" unless opts[:load_killed]
mid = @index[docid][:message_id]
unless messages.member? mid
messages[mid] ||= lambda { build_message docid }
end
end
end
- Redwood::log "ran #{num_queries} queries to build thread of #{messages.size} messages for #{m.id}"
+ Redwood::log "ran #{num_queries} queries to build thread of #{messages.size} messages for #{m.id}" if num_queries > 0
messages.each { |mid, builder| yield mid, builder }
end
def wrap_subj subj; "__START_SUBJECT__ #{subj} __END_SUBJECT__"; end
def unwrap_subj subj; subj =~ /__START_SUBJECT__ (.*?) __END_SUBJECT__/ && $1; end
- def add_message m
- return false if contains? m
-
- source_id =
- if m.source.is_a? Integer
- m.source
- else
- m.source.id or raise "unregistered source #{m.source} (id #{m.source.id.inspect})"
- end
-
- to = (m.to + m.cc + m.bcc).map { |x| x.email }.join(" ")
- d = {
- :message_id => m.id,
- :source_id => source_id,
- :source_info => m.source_info,
- :date => m.date.to_indexable_s,
- :body => m.content,
- :snippet => m.snippet,
- :label => m.labels.join(" "),
- :from => m.from ? m.from.email : "",
- :to => (m.to + m.cc + m.bcc).map { |x| x.email }.join(" "),
- :subject => wrap_subj(Message.normalize_subj(m.subj)),
- :refs => (m.refs + m.replytos).uniq.join(" "),
- }
-
- @index.add_document d
-
- ## TODO: figure out why this is sometimes triggered
- #docid, entry = load_entry_for_id m.id
- #raise "just added message #{m.id} but couldn't find it in a search" unless docid
- true
- end
-
def drop_entry docno; @index.delete docno; end
def load_entry_for_id mid
contacts.keys.compact
end
+ def load_sources fn=Redwood::SOURCE_FN
+ source_array = (Redwood::load_yaml_obj(fn) || []).map { |o| Recoverable.new o }
+ @sources = Hash[*(source_array).map { |s| [s.id, s] }.flatten]
+ @sources_dirty = false
+ end
+
protected
def parse_user_query_string str; @qparser.parse str; end
end
query.add_query Ferret::Search::TermQuery.new("label", "spam"), :must_not unless opts[:load_spam] || labels.include?(:spam)
+ query.add_query Ferret::Search::TermQuery.new("label", "deleted"), :must_not unless opts[:load_deleted] || labels.include?(:deleted)
query.add_query Ferret::Search::TermQuery.new("label", "killed"), :must_not unless opts[:load_killed] || labels.include?(:killed)
query
end
- def load_sources fn=Redwood::SOURCE_FN
- @sources = Hash[*(Redwood::load_yaml_obj(fn) || []).map { |s| [s.id, s] }.flatten]
- @sources_dirty = false
- end
-
def save_sources fn=Redwood::SOURCE_FN
if @sources_dirty || @sources.any? { |id, s| s.dirty? }
bakfn = fn + ".bak"
projects / sup / blobdiff