[RFC] player: "cue prev" reliably hits previous cue breakpoint

[Eric Wong <e@80x24.org>]

Seeking around a track should not be done relative to
a previous seek, but should rather be analogous to the
"tl prev" command to skip around the tracklist.

I'm not sure what drugs I was on when I originally wrote
this original version, but I'm fairly certain this is a
bugfix and not intentional behavior.
---
  Thoughts?  I'm _pretty_ sure this is the right thing to do;
  will think about it for a bit...

 lib/dtas/player/client_handler.rb | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/lib/dtas/player/client_handler.rb b/lib/dtas/player/client_handler.rb
index 7b33967..8a15974 100644
--- a/lib/dtas/player/client_handler.rb
+++ b/lib/dtas/player/client_handler.rb
@@ -745,9 +745,9 @@ def __bp_prev_next(io, msg, cur, bp)
       return io.emit("INVALID TYPE")
     end
     fmt = cur.format
+    ds = __current_decoded_samples
     case msg[0]
     when "next"
-      ds = __current_decoded_samples
       bp.each do |ci|
         next if ci.offset_samples(fmt) < ds
         seek_internal(cur, ci.offset)
@@ -756,11 +756,15 @@ def __bp_prev_next(io, msg, cur, bp)
       # go to the next (real) track if not found
       __current_drop
     when "prev"
-      os = cur.offset_samples # where we currently started
+      prev = nil
       bp.reverse_each do |ci|
-        next if ci.offset_samples(fmt) >= os
-        seek_internal(cur, ci.offset)
-        return io.emit("OK")
+        if prev.nil?
+          next if ci.offset_samples(fmt) >= ds
+          prev = ci
+        else
+          seek_internal(cur, ci.offset)
+          return io.emit("OK")
+        end
       end
       # offset may be nil/zero if we couldn't find a previous breakpoint
       seek_internal(cur, '0')
-- 
EW