sensible-cinema 0.28.1 → 0.34.0

data/history_and_related_works_list.txt

@@ -1,32 +1,31 @@
-
 == Overly Verbose History ==
 
-Sensible Cinema was preceded/inspired by EDL's of editors in the 90's, cleanflicks (both analog and digital incantations, late 90's), 
-then mplayer began to support EDL, zoomplayer added a scene-cut editor, 
+Sensible Cinema was preceded by EDL's [2] of video editors in the 90's, 
+cleanflicks (both analog and digital/DVD incantations, late 90's), 
+then more recently mplayer began to support EDL, zoomplayer added a scene-cut editor some years ago, 
 http://code.google.com/p/movie-content-editor came along, and also many
-other editors have historically supported EDL for content creation and/or playback (see wikipedia's "edit decision list"),
-such as mythtv/XBMC's EDL support (as suggested by the phantom edit [1], oreillynet.com), 
-VLC also added playlists with optional start and stop positions per entry.  Ffmpeg, mencoder were high quality video editing open source tools. 
-Once I discovered all of these previously existing programs, I knew the right elements were ready for 
-something like sensible cinema.
+other editors have supported EDL for content creation and/or playback,
+such as mythtv/XBMC's EDL support (as suggested by the phantom edit [1], from oreillynet.com), 
+VLC also added playlists with optional start and stop positions per entry.  
+Ffmpeg, mencoder were high quality video editing open source video tools. 
+Once I discovered all of these previously existing programs, I knew the right pieces were ready.
 
 See also the "related" section, and "motivation" section.  See also the change_log_with_feature_list.txt file, and several
-blog entries I've posted historically.
+blog entries I've posted to various blogs.
 
 Originally the project started as a prototype to "watch hulu (online) and netflix instant movies
-edited" (it uses/used screen snapshots (of specific locations where the digits are expected, based on windows position) 
+edited" (it uses screen snapshots (of specific locations where the digits are expected, based on window positions) 
 and then tracks, using OCR of the on-screen (convert to timestamp), to monitor it for specific times relative 
-to an EDL, and react appropriately.
-However, this functionality, though working and available, is disabled currently for potential patent
-issues.  Some older released versions had it enabled, but not the later releases.
-Then I realized I could basically have the same effect by instructing a video player (VLC) to "play from second x to y"
+to an EDL, and reacts appropriately.
+Then I realized I could basically have the same effect by instructing any video player (VLC) to "play from second x to y"
 ok "now play from z to q muted" etc. (so kind of a stop and go video player, kind of like controlling the remote for you).
 It basically was implementing Edit Decision Lists (EDL's) using playlists and
 VLC.  
-I just passed VLC a playlist, it did all the work for me, without any necessary real-time tracking.  This seemed patent friendly,
-so I went with this.  There are a few released versions with this feature enabled.  
+I just passed VLC a playlist, it did all the work for me, without any necessary tracking on my part.  
+This seemed to be a different approach than some other programs out there take, so I 
+explored this route.  There were even a few released versions with this feature.  
 They avoid real-time tracking by specifying "play from here to here"
-not "skip from here to here" as some other editing players do.
+not "skip from here to here" as some other players do.
 
 Then it occurred to me that one could grab the several clean segments
 off a movie file, and piece them together to create an edited clean version ( a la cleanflix ).  This also avoids 
@@ -34,7 +33,8 @@ realtime tracking, so I also included this ability.
 The playlist option is disabled in later releases, as it was deemed too user-non-friendly, and because VLC timing
 is "different" than standard DVD players, so I decided to just use standardized timing, though it worked fine
 for VLC playbacks.
-I replaced it with support for mplayer's EDL format, which works better, and has consistent timing, though is still a bit buggy.
+I replaced it with support for mplayer's EDL format, which works better, and has consistent timing, 
+though is still a bit buggy, please report any issues you may find on it.
 
 You can see a full list of other features that have been published and demonstrated
 in the accompanying file change_log_with_feature_list.txt.  Each feature mentioned works 
@@ -44,33 +44,34 @@ in some released version of sensible-cinema though some features have been remov
 Who knows what the future will hold...
 
 [1] http://www.oreillynet.com/sysadmin/blog/2005/06/make_your_own_phantom_edit_wit.html
+[2] http://en.wikipedia.org/wiki/Edit_decision_list
 
 == Related ==
 
 The concepts I use aren't novel.  Some other previously existing related type examples:
 
 Edited "airplane edit" style movies.
-Commercial TV (ABC, NBC) show their own (edited) movies.
+Commercial TV (ABC, NBC) show their own (edited for TV) movies, for instance Schindler's List was slightly edited when shown on television.
 http://en.wikipedia.org/wiki/Linear_video_editing
 http://en.wikipedia.org/wiki/Edit_decision_list#Systems_supporting_EDL_playback.2C_not_just_EDL_cutting lists a lot of players that support EDL DVD playback in some form or another.
-http://imdb.com tends to have reasonably good lists of what occurs in movies (find a movie, click on "parent's guide" on the left).  Sometimes it even lists the time signatures for events (ex: "Labyrinth" http://www.imdb.com/title/tt0091369/parentalguide) which you could use to translate into a sensible-cinema compatible list.
+http://imdb.com tends to have reasonably good lists of what occurs in movies (find a movie, click on "parent's guide" on the left).  Sometimes it even lists the time signatures for events (ex: "Labyrinth" http://www.imdb.com/title/tt0091369/parentalguide) which you could use to translate into a sensible-cinema compatible list.  I've even released a version of SC that has working time edits pulled from a wiki like that.
 Normal DVD's versus "directors cut" versions of movies (one is at times preferable, content wise)
 http://en.wikipedia.org/wiki/Edit_decision_list (it's a linear editing tool, after all--so see the vast list of those)
 Clean Flicks (used to) sell edited movies
 http://college.swankmp.com/insti/edited.asp prisoners apparently have access to a few "edited" versions of movies (not many though, but there are some listed there)
 http://www.videoredo.com/en/index.htm lets you "cut out" offensive content (commercials et al) and burn DVD's from mpeg's/h264 streams, and has its own EDL syntax http://wiki.xbmc.org/?title=EDL_%28commercial_skipping%29_and_SceneMarker_support
 http://www.microsoft.com/windows/windowsmedia/howto/articles/device_control.aspx Windows Media Encoder supports EDL's by timestamp
-http://cutlist.at shared "commercial skip" cutlists
+http://cutlist.at has community shared cutlists (in this instance, for commercials)
 http://www.mplayerhq.hu/DOCS/HTML/en/edl.html
-Time programmable VCR's (at the least, 'skip the first few minutes')
+Time programmable VCR's (at the least, it can 'skip the first few minutes' of a movie or what not, and, if you're handy with the pause and rewind buttons on you remote, edit offensive content)
 http://www.amazon.com/Make-Edited-Movies-ebook/dp/B004EYT3BU
 http://www.sharewareconnection.com/fast-compress-content-and-anti-profanity.htm replaces profanity in web pages with euphemized content.
 MovieMask apparently had capability of "overlaying" DVD playback, like adding new clothing to people, and other features.
 http://www.oreillynet.com/sysadmin/blog/2005/06/make_your_own_phantom_edit_wit.html (mplayer has had this ability for awhile, with no complaints).
 http://fanedit.org/49 fanedits take the original and remix it through some means for more enjoyable playbacks.
 http://clearplay.com a commercial "filtering" DVD player See the "motivation" section for a description.
-  It does offer a few features that sensible cinema supports but disables by default, for patent reasons, 
-  like automatic (user flexible) content decisions.
+  It does offer a few features that sensible cinema supports but does not provide access to, 
+  like (user flexible) content decisions.
 http://en.wikipedia.org/wiki/RiffTrax has an interesting idea of "overlaying" audio over the original video, and synchronizing between the two (their re-player does, anyway).
 http://forum.bsplayer.com/feature-requests-feedback-suggestions/7157-chapter-playlist-scene-cut-3.html
 http://www.inmatrix.com zoom player's "scene cut editor"
@@ -78,27 +79,30 @@ edited released movies by the producer, like http://en.wikipedia.org/wiki/Seven_
 http://wiki.xbmc.org/?title=EDL_(commercial_skipping)_and_SceneMarker_support (XBMC's scene cut supporter--also contains links to some other editors' support for EDL's)
 http://dvdshrink.info "Re-author" mode: to make "movie-only" backups, compilations, combine "flippers",.... DvdShrink has the ability to "crop or cut parts of a title" etc.
 The VCR with its record button, coupled with the stop+rewind button and a list of edits (if even in your head).
-http://kb2.adobe.com/cps/191/tn_19151.html Adobe Director MX 2004 can "script" DVD playback.
+http://kb2.adobe.com/cps/191/tn_19151.html Adobe Director MX 2004 can "script" DVD playback apparently.
 DVD's themselves allow for "scripted" playback of certain content, allowing controls for jumping from location to location, switching audio/video tracks, etc.
-http://code.google.com/p/movie-content-editor (player written in Python, controls VLC based on subtitles and user-input, using edit decision lists in realtime)
+DVD's with different titles based on content.
+http://code.google.com/p/movie-content-editor (player written in Python, controls VLC based on subtitles and user-input'ed EDL's, using edit decision lists in realtime)
 http://forums.gbpvr.com/showthread.php?10606-Automated-Cutting-and-Transcoding-Guide-Using-Mencoder/page45
 http://www.imdb.com/swiki/special?ParentalGuideHelp (search for "scene description")
 http://www.mythtv.org/wiki/Removing_Commercials
 showanalyzer http://www.dragonglobal.biz/showanalyzer.html and comskip analyze videos to look for "segments to cut out" automatically (commercials, in this case, but the principle still applies) 
 Old-school: using the remote control with the pause, mute, stop, fast forward, and play buttons, along with previous knowledge of "questionable" scene locations
-The scissors and old VHS tapes (Clean Flicks' origin, if I remember correctly).  http://www.familysafemedia.com/home_movie_editor.html (was a helper kit for cutting VHS tape)
+The scissors and old VHS tapes (Clean Flicks' original method, if I remember correctly).  http://www.familysafemedia.com/home_movie_editor.html (was a helper kit for cutting VHS tape)
 Windows Movie Maker.  Allows for users to cut and copy scenes of movies.  Really any editor can do this.
 http://forum.videolan.org/viewtopic.php?f=7&t=56998 VLC Media Player using EDL's via playlists
-Corel WinDVD has "Quick Clip" settings for capturing from DVD.
+Corel WinDVD has "Quick Clip" settings for capturing movie from DVD's.
 Microsoft's built-in magnifier basically copies DVD's screen output if mouse is placed over playing DVD.
 http://superuser.com/questions/85278/how-can-i-cut-scenes-out-of-a-dvd
 moviemask once had several advanced features: http://theeditedmovieencyclopedia.blogspot.com/2008/06/cleaning-up-movies-part-2.html
 http://blog.josephhall.com/2008/09/command-line-dvd-authoring-part-1.html
 AviSynth can script dvd playback http://forum.doom9.org/archive/index.php/t-109558.html
-ps3 media server has edl support: https://github.com/chocolateboy/PMS-EDL maybe would work with DVD's dunno
-Anydvd HD's "magic file replacement" for DVD's/blu-ray's
-RiffTrax allows for an audio "overlay" to a DVD playback, have their own player to coordinate the two.
-"fan subs" are edited subtitles, displayed overlaid on the original track/video
+ps3 media server has edl support: https://github.com/chocolateboy/PMS-EDL
+Anydvd HD's "magic file replacement" for DVD's/blu-ray's.  Google for it.
+RiffTrax allows for an audio "overlay" to a DVD playback, have their own player to coordinate the extra audio track with the playing video.
+http://www.youtube.com/watch?v=0vumR5Hcz7s is an Emacs "EDL editor" which uses mplayer and Media Lovin' Toolkit framework.
+"fan subs" are edited subtitles, displayed overlaid on the original track/video, or so I'm told.
 Avatar DVD was released with an audio track that was "euphemized" of the profanity.
+kids-in-mind.com lists basically all the bad things that happen in movies (very detailedly) as also does common sense media, though without any timestamps unfortunately.
 
-BYU's Varsity theater used to edit movies, and show them.
+BYU's Varsity theater used to edit movies, and show them on the big screen.

data/legal_draft

@@ -0,0 +1,23 @@
+The design, not the idea.
+
+OCR can work on internet or standalone players.
+time elapsed "since the timestamp changed" on screen infers the more precise millisecond 
+you are at, for being able to skip, mute, overlay, apply overlaid transparent image/video more precisely.
+You can grab the specific coordinates by precise location, or by iterating over windows, or by relative location to screen coordinates/window coordinates.
+You can use the url to infer the title desired.
+You can use user preferences to infer how to use each setting within the "EDL" by saving the user preference, and comparing it 
+against each entry in the EDL when you parse it.
+
+It can mute by clicking a mute button, by accessing an API.
+It can work by monitoring, using an API, the current playback time.
+
+It can react to "seek" commands to coordinate position within EDL's.
+
+Another invention: it can apply editing "commands" to a player that is playing back blu-ray discs (hardware or software), as originally described in other clearplay
+patents, like US6889383 and those issued to them afterward.
+
+You can get the specific digits by "absolute digit location" or by "OCR on the entire timestamp"
+
+Could also playback using avisynth.
+
+Can create edited files.

data/lib/add_any_bundled_gems_to_load_path.rb

@@ -17,9 +17,10 @@ This file is part of Sensible Cinema.
 =end
 
 def add_any_bundled_gems_to_load_path
-  raise 'no vendor dir?' unless File.directory? 'vendor'
-  if File.directory? 'vendor/cache'
-    Dir['vendor/**/lib'].sort.reverse.each{|lib_dir| # prefers newer versioned copies of gems in case I have duplicates
+  vendor_dir = File.dirname(__FILE__) +  '/../vendor'
+  raise 'no vendor dir?' unless File.directory?(vendor_dir)
+  if File.directory? vendor_dir + '/cache'
+    Dir[vendor_dir + '/**/lib'].sort.reverse.each{|lib_dir| # prefers newer versioned copies of gems in case I have duplicates
       $: << lib_dir
     }
   else

data/lib/check_installed_mac.rb

@@ -2,8 +2,11 @@ require 'rubygems'
 require 'os'
 success = true
 
-# MPlayer OSX Extended instead of smplayer... ?
+ENV['PATH'] = '/opt/rdp_project_local/bin/:' + ENV['PATH'] # macports' bin in first
+
 module CheckInstalledMac
+
+ # should output an error message...
  def self.check_for_installed name
   if name == 'mencoder'
     output = `mencoder --fail 2>&1`
@@ -16,17 +19,14 @@ module CheckInstalledMac
   end
 
   if name == 'mplayer'
-    for path in ["/Applications/MPlayerX.app/Contents/Resources/binaries/m32", File.expand_path('~') + "/Downloads/MPlayerX.app/Contents/Resources/binaries/m32/"]
-      if File.exist? path + "/mplayer"
-        ENV['PATH'] = path + ':' + ENV['PATH']
-        return true
-      end
+    unless File.exist?('/opt/rdp_project_local/bin/mplayer')
+      puts "please install mplayer edl, please install mplayer-edl, see website http://rogerdpack.t28.net/sensible-cinema/content_editor.html"
+      return false
     end
-    puts 'lacking mplayer! please install MPlayerX from the App Store first'
-    return false
+    return true
   end
 
-  # check for the others as generics
+  # check for the others generically
 
   command = {"gocr" => "gocr --help", "convert" => "convert --help", "ffmpeg" => "ffmpeg -version"}[name]
 
@@ -34,7 +34,7 @@ module CheckInstalledMac
 
   unless system(command + " 1> " + OS.dev_null + " 2>" + OS.dev_null)
      name = 'ImageMagick' if name == 'convert' # special case...
-     puts 'lacking dependency! Please install ' + name + ' by installing macports and running in terminal: $ sudo port install ' + name
+     puts 'lacking dependency! Please install ' + name + ' by installing the contrib pkg from the website'
      false
   else
     true

data/lib/count_down_timer_jruby_swing.rb

@@ -1,5 +1,5 @@
 require 'rubygems'
-require 'sane'
+require 'sane' # require_relative
 require_relative 'jruby-swing-helpers/swing_helpers'
 
 include SwingHelpers
@@ -7,9 +7,9 @@ include SwingHelpers
 class MainWindow < JFrame
 
   def show_blocking_message_dialog(message, title = message.split("\n")[0], style= JOptionPane::INFORMATION_MESSAGE)
-# I think I'm already on top...
- setVisible(true);
- toFront()
+    # I think I'm already on top...
+    setVisible(true);
+    toFront()
     JOptionPane.showMessageDialog(nil, message, title, style)
     true
   end

data/lib/edl_parser.rb

@@ -17,8 +17,34 @@ This file is part of Sensible Cinema.
 =end
 
 require 'sane'
+require_relative 'convert_thirty_fps'
+
 class EdlParser
 
+  # this one is 1:01:02.0 => 36692.0
+  # its reverse is this: translate_time_to_human_readable
+  def self.translate_string_to_seconds s
+    # might actually already be a float, or int, depending on the yaml
+    # int for 8 => 9 and also for 1:09 => 1:10
+    if s.is_a? Numeric
+      return s.to_f # easy out.
+    end
+    
+    s = s.strip
+    # s is like 1:01:02.0
+    total = 0.0
+    seconds = nil
+    seconds = s.split(":")[-1]
+    raise 'does not look like a timestamp? ' + seconds.inspect unless seconds =~ /^\d+(|[,.]\d+)$/
+    seconds.gsub!(',', '.')
+    total += seconds.to_f
+    minutes = s.split(":")[-2] || "0"
+    total += 60 * minutes.to_i
+    hours = s.split(":")[-3] || "0"
+    total += 60* 60 * hours.to_i
+    total
+  end
+
   EDL_DIR = File.expand_path(__DIR__  + "/../zamples/edit_decision_lists/dvds")
   
   if File::ALT_SEPARATOR
@@ -35,7 +61,7 @@ class EdlParser
       output.merge! new_input
     end
     
-    require_relative 'gui/sensible-cinema-dependencies' # for download method...
+    require_relative 'gui/dependencies' # for download methods...
     
     if expand
     
@@ -52,65 +78,78 @@ class EdlParser
   end
   
   def self.parse_imdb output, imdb_id
-        require_relative 'convert_thirty_fps'
-        url = "http://www.imdb.com/title/#{imdb_id}/parentalguide"
-        all = SensibleSwing::MainWindow.download_to_string(url)
-        
-        header, violence_word, violence_section, profanity_word, profanity_section, alcohol_word, alcohol_section, frightening_word, frightening_section = 
-        sections = all.split(/<span>(Violence|Profanity|Alcohol|Frightening)/)
-        header = sections.shift
-        all ={}
-        while(!sections.empty?) # my klugey to_hash method
-          word_type = sections.shift
-          settings = sections.shift
-          assert word_type.in? ['Violence', 'Profanity', 'Alcohol', 'Frightening']
-          all[word_type] = settings
-        end
-        # blank_outs or mutes for each...
-        # TODO make the -> optional
-        split_into_timestamps = /([\d:]+(?:\.\d+|))\W*-&gt;\W*([\d:]+(?:\.\d+|))([^\d\n]+)/
-        for type, settings in all
-          settings.scan(split_into_timestamps) do |begin_ts, end_ts, description|
-            puts "parsing from wiki imdb  entry violence: #{begin_ts} #{end_ts} #{description} #{type}"
-            start_seconds = translate_string_to_seconds begin_ts
-            end_seconds = translate_string_to_seconds end_ts
-            # convert from 30 to 29.97 fps ... we presume ...
-            start_seconds = ConvertThirtyFps.from_twenty_nine_nine_seven start_seconds
-            start_seconds = ("%.02f" % start_seconds).to_f # round
-            start_seconds = translate_time_to_human_readable start_seconds, true
-            end_seconds = ConvertThirtyFps.from_twenty_nine_nine_seven end_seconds
-            end_seconds = ("%.02f" % end_seconds).to_f # round
-            end_seconds = translate_time_to_human_readable end_seconds, true
-            p end_seconds
-            if type == 'Profanity'
-              output['mutes'] << [start_seconds, end_seconds]
-            else
-              output['blank_outs'] << [start_seconds, end_seconds]
-            end
-          end
+    require_relative 'convert_thirty_fps'
+    url = "http://www.imdb.com/title/#{imdb_id}/parentalguide"
+    all = SensibleSwing::MainWindow.download_to_string(url)
+    
+    header, violence_word, violence_section, profanity_word, profanity_section, alcohol_word, alcohol_section, frightening_word, frightening_section = 
+    sections = all.split(/<span>(Violence|Profanity|Alcohol|Frightening)/)
+    header = sections.shift
+    all ={}
+    while(!sections.empty?) # my klugey to_hash method
+      word_type = sections.shift
+      settings = sections.shift
+      assert word_type.in? ['Violence', 'Profanity', 'Alcohol', 'Frightening']
+      all[word_type] = settings
+    end
+    # blank_outs or mutes for each...
+    # TODO make the -> optional
+    split_into_timestamps = /([\d:]+(?:\.\d+|))\W*-&gt;\W*([\d:]+(?:\.\d+|))([^\d\n]+)/
+    for type, settings in all
+      settings.scan(split_into_timestamps) do |begin_ts, end_ts, description|
+        puts "parsing from wiki imdb  entry violence: #{begin_ts} #{end_ts} #{description} #{type}"
+        start_seconds = translate_string_to_seconds begin_ts
+        end_seconds = translate_string_to_seconds end_ts
+        # convert from 30 to 29.97 fps ... we presume ...
+        start_seconds = ConvertThirtyFps.from_twenty_nine_nine_seven start_seconds
+        start_seconds = ("%.02f" % start_seconds).to_f # round
+        start_seconds = translate_time_to_human_readable start_seconds, true
+        end_seconds = ConvertThirtyFps.from_twenty_nine_nine_seven end_seconds
+        end_seconds = ("%.02f" % end_seconds).to_f # round
+        end_seconds = translate_time_to_human_readable end_seconds, true
+        p end_seconds
+        if type == 'Profanity'
+          output['mutes'] << [start_seconds, end_seconds]
+        else
+          output['blank_outs'] << [start_seconds, end_seconds]
         end
+      end
+    end
   end
   
   private
 
   def self.download full_url, to_here
-    require 'open-uri'
-    writeOut = open(to_here, "wb")
-    writeOut.write(open(full_url).read)
-    writeOut.close
+    ::SwingHelpers::MainWindow.download full_url, to_here
   end
   
   # better eye-ball these before letting people run them, eh? TODO
   # but I couldn't think of any other way to parse the files tho
-  def self.parse_string string, filename, ok_categories_array = []
+  def self.parse_string string, filename = nil, ok_categories_array = []
     string = '{' + string + "\n}"
-    if filename
-     raw = eval(string, binding, filename)
-    else
-     raw = eval string
-    end
+	  begin
+      if filename
+       raw = eval(string, binding, filename, 0)
+      else
+       raw = eval string
+      end
+    rescue Exception => e
+      string.strip.lines.to_a[0..-3].each_with_index{|l, idx| # last line doesn't need a comma check
+	    orig_line = l
+	    l = l.split('#')[0]
+		  l = l.strip
+		  unless l.empty?
+		    # todo strip off # comments at the end of lines too...
+		    end_char = l[-1..-1]
+          if !end_char.in? [']', '[', '{'] # these are probably all ok...
+            puts "warning: line #{idx} might be bad: (maybe needs comma after?) " + l unless end_char == ',' 
+          end
+        end
+      }
+	    raise SyntaxError.new(e.to_s) # to_s as a workaround for jruby #6353
+	  end
     
-    raise SyntaxError.new("maybe missing quotation marks?" + string) if raw.keys.contain?(nil)
+    raise SyntaxError.new("maybe missing quotation marks somewhere?" + string) if raw.keys.contain?(nil)
     
     # mutes and blank_outs need to be special parsed into arrays...
     mutes = raw["mutes"] || []
@@ -120,7 +159,7 @@ class EdlParser
     raw
   end
   
-  # converts "blanks" => ["00:00:00", "00", "reason", "01", "01", "02", "02"] into sane arrays, also filters based on category, though disabled for production
+  # converts "blanks" => ["00:00:00", "00", "reason", "01", "01", "02", "02"] into sane arrays, also filters based on category, though not used in production
   def self.convert_to_timestamp_arrays array, ok_categories_array
     out = []
     while(single_element = extract_entry!(array))
@@ -151,10 +190,9 @@ class EdlParser
     out
   end
   
-  #TimeStamp = /(^\d+:\d\d[\d:\.]*$|\d+)/ # this one also allows for 4444 [?] and also weirdness like "don't kill the nice butterfly 2!" ...
-  TimeStamp = /(^\d+:\d\d[\d:\.]*|\d+\.\d+)$/ # allow 00:00:00 00:00:00.0 1222.4
-  # disallow 1905 too but in the code
-  # starts with a digit, has at least one colon followed by two digits,then some combo of digits and colons and periods...
+  # TimeStamp = /(^\d+:\d\d[\d:\.]*$|\d+)/ # this one also allows for 4444 [?] and also weirdness like "don't kill the nice butterfly 2!" ...
+  TimeStamp = /^\s*(\d+:\d\d[\d:\.]*|\d+\.\d+)\s*$/ # allow 00:00:00 00:00:00.0 1222.4 " 2:04 "
+  # disallow's 1905 too but elsewhere in the code
   
   def self.extract_entry! from_this
     return nil if from_this.length == 0
@@ -162,7 +200,7 @@ class EdlParser
     out = from_this.shift(2)
     out.each{|d|
       unless d =~ TimeStamp
-        raise SyntaxError.new('non timestamp? ' + d) 
+        raise SyntaxError.new('non timestamp? ' + d.inspect) 
       end
     }
     while(from_this[0] && from_this[0] !~ TimeStamp)
@@ -172,99 +210,45 @@ class EdlParser
     out
   end
   
-  def self.get_secs timestamp_string_begin, timestamp_string_end, add_begin, add_end, splits
-    answers = []
-    unless timestamp_string_begin
-      raise 'non begin' 
-    end
-    unless timestamp_string_end
-      raise 'non end' 
-    end
-    for type, offset, multiplier in [[timestamp_string_begin, add_begin, -1], [timestamp_string_end, add_end, 1]]
-      original_secs = translate_string_to_seconds(type) + offset
-      # now if splits is 900 and we'are at 909, then we're just 9
-      closest_split_idx = splits.reverse.index{|t| t < original_secs}
-      if closest_split_idx
-        closest_split = splits.reverse[closest_split_idx]
-        # add some extra seconds onto these if they're "past" a split, too
-        original_secs = original_secs - closest_split + multiplier * (splits.length - closest_split_idx)
-        original_secs = [0, original_secs].max # no negatives allowed :)
-      end
-      answers << original_secs
-    end
-    answers
-  end
-  
   public 
   
-  # called later, from external
+  # called later, from external files
   # divides up mutes and blanks so that they don't overlap, preferring blanks over mutes
   # returns it like [[start,end,type], [s,e,t]...] type like either :blank and :mute
-  # [[70.0, 73.0, :blank], [378.0, 379.1, :mute]]
-  def self.convert_incoming_to_split_sectors incoming, add_this_to_all_ends = 0, subtract_this_from_beginnings = 0, splits = []
+  # [ [70.0, 73.0, :blank], [378.0, 379.1, :mute], ... ]
+  def self.convert_incoming_to_split_sectors incoming, add_this_to_all_ends = 0, subtract_this_from_beginnings = 0, splits = [], subtract_this_from_ends = 0
     raise if subtract_this_from_beginnings < 0
     raise if add_this_to_all_ends < 0
-    if splits != []
-      # allow it to do all the double checks we later skip, just in case :)
-      self.convert_incoming_to_split_sectors incoming
-    end
+    raise if subtract_this_from_ends < 0
+    add_this_to_all_ends -= subtract_this_from_ends # now we allow negative :)
+	#raise if splits.size > 0 # for now
     mutes = incoming["mutes"] || {}
     blanks = incoming["blank_outs"] || {}
-    mutes = mutes.map{|k, v| get_secs(k, v, -subtract_this_from_beginnings, add_this_to_all_ends, splits) + [:mute]}
-    blanks = blanks.map{|k, v| get_secs(k, v, -subtract_this_from_beginnings, add_this_to_all_ends, splits) + [:blank]}
-    combined = (mutes+blanks).sort
+    mutes = mutes.map{|k, v| [k,v,:mute]}
+    blanks = blanks.map{|k, v| [k,v,:blank]}
+    combined = (mutes+blanks).sort.map{|s,e,type|  [translate_string_to_seconds(s),  translate_string_to_seconds(e), type]}
     
-    # detect overlap...
+    # detect any weirdness...
     previous = nil
-    combined.each_with_index{|current, idx|
-      s,e,t = current
-      if e < s
-       raise SyntaxError.new("detected an end before a start: #{e} < #{s}") if e < s unless splits.length > 0
+    combined.map!{|current|
+      s,e,type = current
+      if e < s || !s || !e || !type
+       raise SyntaxError.new("detected an end before a start or other weirdness: #{e} < #{s}")
       end
       if previous
-        ps, pe, pt = previous
-        if (s < pe)
-          raise SyntaxError.new("detected an overlap #{[s,e,t].join(' ')} #{previous.join(' ')}") unless splits.length > 0
-          # our start might be within the previous' in which case its their start, with (greater of our, their ending)
-          preferred_end = [e,pe].max
-          preferred_type = [t,pt].detect{|t| t == :blank} || :mute # prefer blank to mute
-          combined[idx-1] = [ps, preferred_end, preferred_type]
-          combined[idx] = nil # allow it to be culled later
+        ps, previous_end, pt = previous
+        if (s < previous_end)
+          raise SyntaxError.new("detected an overlap #{current.join(' ')} #{previous.join(' ')}")
         end
         
       end
       previous = current
+	  # do the math later to allow for ones that barely hit into each other 1.0 2.0, 2.0 3.0
+	  [s-subtract_this_from_beginnings, e+add_this_to_all_ends,type]
     }
-    combined.compact
+    combined
   end
   
-  # this one is 1:01:02.0 => 36692.0
-  # its reverse is: translate_time_to_human_readable
-  def self.translate_string_to_seconds s
-    # might actually already be a float, or int, depending on the yaml
-    # int for 8 => 9 and also for 1:09 => 1:10
-    if s.is_a? Numeric
-      return s.to_f
-    end
-    
-    # s is like 1:01:02.0
-    total = 0.0
-    seconds = nil
-    begin
-      seconds = s.split(":")[-1]
-    rescue Exception => e
-     p 'failed!', s
-     raise e
-    end
-    raise unless seconds =~ /^\d+(|[,.]\d+)$/
-    seconds.gsub!(',', '.')
-    total += seconds.to_f
-    minutes = s.split(":")[-2] || "0"
-    total += 60 * minutes.to_i
-    hours = s.split(":")[-3] || "0"
-    total += 60* 60 * hours.to_i
-    total
-  end
   
   # its reverse: translate_string_to_seconds
   def self.translate_time_to_human_readable seconds, force_hour_stamp = false
@@ -282,11 +266,12 @@ class EdlParser
     out << ":"
     
     # avoid an ugly .0 at the end
-    if seconds == seconds.to_i
-      out << "%02d" % seconds
-    else
+#    if seconds == seconds.to_i
+#      out << "%02d" % seconds
+#    else
       out << "%05.2f" % seconds # man that printf syntax is tricky...
-    end
+#    end
+    
   end
   
   def self.all_edl_files_parsed use_all_not_just_dvds
@@ -298,36 +283,61 @@ class EdlParser
           [filename, parsed]
         rescue SyntaxError => e
           # ignore poorly formed edit lists for the auto choose phase...
-          p 'warning, unable to parse a file:' + filename + " " + e.to_s
+          puts 'warning, unable to parse a file:' + filename + " " + e.to_s
           nil
         end
      }.compact
   end
   
   # returns single matching filename
-  def self.find_single_edit_list_matching use_all = false
-    matching = all_edl_files_parsed(use_all).map{|filename, parsed|
+  # requires a block
+  def self.find_single_edit_list_matching use_all_not_just_dvd_dir = false, return_first_if_there_are_several = false
+    matching = all_edl_files_parsed(use_all_not_just_dvd_dir).map{|filename, parsed|
       yield(parsed) ? filename : nil
     }.compact
     if matching.length == 1
       file = matching[0]
-      p "selecting the one only matching EDL: #{file}"
       file
     elsif matching.length > 1
-      p "found multiple matches for media? #{matching.inspect}"
-      nil
+          p "found multiple matches for media? #{matching.inspect}"
+	  if return_first_if_there_are_several
+	    matching[0]
+	  else
+            nil
+	  end
     else
       nil
     end
   end
   
-  def self.single_edit_list_matches_dvd dvd_id
+  def self.single_edit_list_matches_dvd dvd_id, return_first_if_there_are_several = false
     return nil unless dvd_id
-    find_single_edit_list_matching {|parsed|
+    find_single_edit_list_matching(false, return_first_if_there_are_several)  {|parsed|
       parsed["disk_unique_id"] == dvd_id
     }
   end
-
+  
+  def self.convert_to_dvd_nav_times combined, start_type, start_mpeg_time, dvd_nav_packet_offsets, time_multiplier
+    start_dvdnav_time = dvd_nav_packet_offsets[1] - dvd_nav_packet_offsets[0]
+    raise unless start_type == 'dvd_start_offset' # for now :P
+    out = []
+    add_this_to_all_of_them = start_dvdnav_time - start_mpeg_time
+    #[[70.0, 73.0, :blank], [378.0, 379.1, :mute]]
+    for start, endy, type in combined
+     if time_multiplier == 30
+       # ok
+     elsif time_multiplier == 29.97
+       start = ConvertThirtyFps.from_twenty_nine_nine_seven start
+       endy  = ConvertThirtyFps.from_twenty_nine_nine_seven endy
+    else
+      raise time_multiplier
+    end
+    start += start_dvdnav_time
+    endy += start_dvdnav_time
+    out << [start, endy, type]
+   end
+   out
+  end
   
 end
 
@@ -345,6 +355,9 @@ if $0 == __FILE__
   p 'syntax: filename'
   require 'rubygems'
   require 'sane'
-  p EdlParser.parse_file(*ARGV)
+  parsed = EdlParser.parse_file(*ARGV)
   p 'parsed well'
+  p parsed
+  require 'yaml'
+  print YAML.dump parsed
 end