9 def to_s; sprintf '%.2f', self; end
11 infinite? ? "unknown" : super
18 sprintf "%d:%02d:%02d", i / 3600, (i / 60) % 60, i % 60
28 opts = Trollop::options do
29 version "sup-sync (sup #{Redwood::VERSION})"
31 Synchronizes the Sup index with one or more message sources by adding
32 messages, deleting messages, or changing message state in the index as
35 "Message state" means read/unread, archived/inbox, starred/unstarred,
36 and all user-defined labels on each message.
38 "Default source state" refers to any state that a source itself has
39 keeps about a message. Sup-sync uses this information when adding a
40 new message to the index. The source state is typically limited to
41 read/unread, archived/inbox status and a single label based on the
42 source name. Messages using the default source state are placed in
43 the inbox (i.e. not archived) and unstarred.
46 sup-sync [options] <source>*
48 where <source>* is zero or more source URIs. If no sources are given,
49 sync from all usual sources. Supported source URI schemes can be seen
50 by running "sup-add --help".
52 Options controlling WHICH messages sup-sync operates on:
54 opt :new, "Operate on new messages only. Don't scan over the entire source. (Default.)", :short => :none
55 opt :changed, "Scan over the entire source for messages that have been deleted, altered, or moved from another source. (In the case of mbox sources, this includes all messages AFTER an altered message.)"
56 opt :restored, "Operate only on those messages included in a dump file as specified by --restore which have changed state."
57 opt :all, "Operate on all messages in the source, regardless of newness or changedness."
58 opt :start_at, "For --changed and --all, start at a particular offset.", :type => :int
62 Options controlling HOW message state is altered:
64 opt :asis, "If the message is already in the index, preserve its state. Otherwise, use default source state. (Default.)", :short => :none
65 opt :restore, "Restore message state from a dump file created with sup-dump. If a message is not in this dumpfile, act as --asis.", :type => String, :short => :none
66 opt :discard, "Discard any message state in the index and use the default source state. Dangerous!", :short => :none
67 opt :archive, "When using the default source state, mark messages as archived.", :short => "-x"
68 opt :read, "When using the default source state, mark messages as read."
69 opt :extra_labels, "When using the default source state, also apply these user-defined labels. Should be a comma-separated list.", :type => String, :short => :none
75 opt :verbose, "Print message ids as they're processed."
76 opt :optimize, "As the final operation, optimize the index."
77 opt :all_sources, "Scan over all sources.", :short => :none
78 opt :dry_run, "Don't actually modify the index. Probably only useful with --verbose.", :short => "-n"
79 opt :version, "Show version information", :short => :none
81 conflicts :changed, :all, :new, :restored
82 conflicts :asis, :restore, :discard
84 Trollop::die :restored, "requires --restore" if opts[:restored] unless opts[:restore]
86 Trollop::die :start_at, "must be non-negative" if opts[:start_at] < 0
87 Trollop::die :start_at, "requires either --changed or --all" unless opts[:changed] || opts[:all]
90 target = [:new, :changed, :all, :restored].find { |x| opts[x] } || :new
91 op = [:asis, :restore, :discard].find { |x| opts[x] } || :asis
94 index = Redwood::Index.new
99 $stderr.puts "Loading state dump from #{opts[:restore]}..."
100 IO.foreach opts[:restore] do |l|
101 l =~ /^(\S+) \((.*?)\)$/ or raise "Can't read dump line: #{l.inspect}"
103 dump[mid] = labels.symbolistize
105 $stderr.puts "Read #{dump.size} entries from dump file."
116 sources = ARGV.map do |uri|
117 index.source_for uri or Trollop::die "Unknown source: #{uri}. Did you add it with sup-add first?"
120 sources = index.usual_sources if sources.empty?
121 sources = index.sources if opts[:all_sources]
123 unless target == :new
125 sources.each { |s| s.seek_to! opts[:start_at] }
127 sources.each { |s| s.reset! }
131 sources.each do |source|
132 $stderr.puts "Scanning #{source}..."
133 num_added = num_updated = num_scanned = num_restored = 0
134 last_info_time = start_time = Time.now
136 Redwood::PollManager.add_messages_from source, :force_overwrite => true do |m, offset, entry|
140 ## skip if we're operating only on changed messages, the message
141 ## is in the index, and it's unchanged from what the source is
143 next if target == :changed && entry && entry[:source_id].to_i == source.id && entry[:source_info].to_i == offset
145 ## get the state currently in the index
146 index_state = entry[:label].symbolistize if entry
148 ## skip if we're operating on restored messages, and this one
150 next if target == :restored && (!restored_state[m.id] || (index_state && restored_state[m.id].sort_by { |s| s.to_s } == index_state.sort_by { |s| s.to_s }))
152 ## m.labels is the default source labels. tweak these according
153 ## to default source state modification flags.
154 m.labels -= [:inbox] if opts[:archive]
155 m.labels -= [:unread] if opts[:read]
156 m.labels += opts[:extra_labels].strip.split(/\s*,\s*/).map { |x| x.intern } if opts[:extra_labels]
158 ## assign message labels based on the operation we're performing
161 m.labels = ((m.labels - [:unread, :inbox]) + index_state).uniq if index_state
163 ## if the entry exists on disk
164 if restored_state[m.id]
165 m.labels = restored_state[m.id]
168 m.labels = index_state
171 ## nothin! use default source labels
174 if Time.now - last_info_time > 60
175 last_info_time = Time.now
176 elapsed = last_info_time - start_time
177 pctdone = source.respond_to?(:pct_done) ? source.pct_done : 100.0 * (source.cur_offset.to_f - source.start_offset).to_f / (source.end_offset - source.start_offset).to_f
178 remaining = (100.0 - pctdone) * (elapsed.to_f / pctdone)
179 $stderr.puts "## #{num_scanned} (#{pctdone}%) read; #{elapsed.to_time_s} elapsed; #{remaining.to_time_s} remaining"
183 puts "Adding message #{source}##{offset} with state {#{m.labels * ', '}}" if opts[:verbose]
186 puts "Updating message #{source}##{offset}, source #{entry[:source_id]} => #{source.id}, offset #{entry[:source_info]} => #{offset}, state {#{index_state * ', '}} => {#{m.labels * ', '}}" if opts[:verbose]
190 opts[:dry_run] ? nil : m
192 $stderr.puts "Scanned #{num_scanned}, added #{num_added}, updated #{num_updated} messages from #{source}."
193 $stderr.puts "Restored state on #{num_restored} (#{100.0 * num_restored / num_scanned}%) messages." if num_restored > 0
196 ## delete any messages in the index that claim they're from one of
197 ## these sources, but that we didn't see.
199 ## kinda crappy code here, because we delve directly into the Ferret
202 ## TODO: move this to Index, i suppose.
205 if target == :all || target == :changed
206 $stderr.puts "Deleting missing messages from the index..."
207 num_del, num_scanned = 0, 0
208 sources.each do |source|
209 raise "no source id for #{source}" unless source.id
210 q = "+source_id:#{source.id}"
211 q += " +source_info: >= #{opts[:start_at]}" if opts[:start_at]
212 index.index.search_each(q, :limit => :all) do |docid, score|
214 mid = index.index[docid][:message_id]
216 puts "Deleting #{mid}" if opts[:verbose]
217 index.index.delete docid unless opts[:dry_run]
222 $stderr.puts "Deleted #{num_del} / #{num_scanned} messages"
228 $stderr.puts "Optimizing index..."
229 optt = time { index.index.optimize unless opts[:dry_run] }
230 $stderr.puts "Optimized index of size #{index.size} in #{optt}s."
232 rescue Redwood::FatalSourceError => e
233 $stderr.puts "Sorry, I couldn't communicate with a source: #{e.message}"
234 rescue Exception => e
235 File.open("sup-exception-log.txt", "w") { |f| f.puts e.backtrace }