Merge branch 'parser-user-query-fix'

[sup] / lib / sup / mbox / loader.rb

diff --git a/lib/sup/mbox/loader.rb b/lib/sup/mbox/loader.rb
index 20bab2d20fce88d61723faa39e75914fb6ca2b86..ebb2aed675e171397b99e877023e24611cf119c7 100644 (file)
--- a/lib/sup/mbox/loader.rb
+++ b/lib/sup/mbox/loader.rb
@@ -6,9 +6,10 @@ module MBox
 
 class Loader < Source
   yaml_properties :uri, :cur_offset, :usual, :archived, :id, :labels
+  attr_accessor :labels
 
   ## uri_or_fp is horrific. need to refactor.
-  def initialize uri_or_fp, start_offset=nil, usual=true, archived=false, id=nil, labels=[]
+  def initialize uri_or_fp, start_offset=0, usual=true, archived=false, id=nil, labels=[]
     @mutex = Mutex.new
     @labels = ((labels || []) - LabelManager::RESERVED_LABELS).uniq.freeze
 
@@ -37,7 +38,7 @@ class Loader < Source
     if File.dirname(path) =~ /\b(var|usr|spool)\b/
       []
     else
-      [File.basename(path).intern]
+      [File.basename(path).downcase.intern]
     end
   end
 
@@ -55,10 +56,10 @@ class Loader < Source
     @mutex.synchronize do
       @f.seek offset
       l = @f.gets
-      unless l =~ BREAK_RE
+      unless MBox::is_break_line? l
         raise OutOfSyncSourceError, "mismatch in mbox file offset #{offset.inspect}: #{l.inspect}." 
       end
-      header = MBox::read_header @f
+      header = parse_raw_email_header @f
     end
     header
   end
@@ -67,25 +68,36 @@ class Loader < Source
     @mutex.synchronize do
       @f.seek offset
       begin
-        RMail::Mailbox::MBoxReader.new(@f).each_message do |input|
-          m = RMail::Parser.read(input)
-          if m.body && m.body.is_a?(String)
-            m.body.gsub!(/^>From /, "From ")
-          end
-          return m
-        end
+        ## don't use RMail::Mailbox::MBoxReader because it doesn't properly ignore
+        ## "From" at the start of a message body line.
+        string = ""
+        l = @f.gets
+        string << l until @f.eof? || MBox::is_break_line?(l = @f.gets)
+        RMail::Parser.read string
       rescue RMail::Parser::Error => e
         raise FatalSourceError, "error parsing mbox file: #{e.message}"
       end
     end
   end
 
+  ## scan forward until we're at the valid start of a message
+  def correct_offset!
+    @mutex.synchronize do
+      @f.seek cur_offset
+      string = ""
+      until @f.eof? || (l = @f.gets) =~ BREAK_RE
+        string << l
+      end
+      self.cur_offset += string.length
+    end
+  end
+
   def raw_header offset
     ret = ""
     @mutex.synchronize do
       @f.seek offset
-      until @f.eof? || (l = @f.gets) =~ /^$/
-        ret += l
+      until @f.eof? || (l = @f.gets) =~ /^\r*$/
+        ret << l
       end
     end
     ret
@@ -93,7 +105,7 @@ class Loader < Source
 
   def raw_message offset
     ret = ""
-    each_raw_message_line(offset) { |l| ret += l }
+    each_raw_message_line(offset) { |l| ret << l }
     ret
   end
 
@@ -107,7 +119,7 @@ class Loader < Source
     @mutex.synchronize do
       @f.seek offset
       yield @f.gets
-      until @f.eof? || (l = @f.gets) =~ BREAK_RE
+      until @f.eof? || MBox::is_break_line?(l = @f.gets)
         yield l
       end
     end
@@ -128,7 +140,7 @@ class Loader < Source
         ## 2. at the beginning of an mbox separator (in all other
         ##    cases).
 
-        l = @f.gets or raise "next while at EOF"
+        l = @f.gets or return nil
         if l =~ /^\s*$/ # case 1
           returned_offset = @f.tell
           @f.gets # now we're at a BREAK_RE, so skip past it
@@ -138,7 +150,7 @@ class Loader < Source
         end
 
         while(line = @f.gets)
-          break if line =~ BREAK_RE
+          break if MBox::is_break_line? line
           next_offset = @f.tell
         end
       end
@@ -147,7 +159,7 @@ class Loader < Source
     end
 
     self.cur_offset = next_offset
-    [returned_offset, (@labels + [:unread]).uniq]
+    [returned_offset, (self.labels + [:unread]).uniq]
   end
 end