[cplay-dev] Re: better mpc code

[Ulf Betlehem]

Han Boetes <address@hidden> writes:

| Hi Ulf,
| 
| After some hours of hacking on the wrong problem I came up with
| the following code for mppdec:
| 
| class FrameOffsetPlayerMpp(Player):
|     re_progress = re.compile(".*\s(\d+):(\d+).*\s(\d+):(\d+)")
|     def parse_buf(self):
|         match = self.re_progress.search(self.buf)
|         if match:
|             m1, s1, m2, s2 = map(string.atoi, match.groups())
|             head = m1*60+s1
|           tail = (m2*60+s2) - head
|             self.set_position(head, head+tail , [head, tail])
| 
| FrameOffsetPlayerMpp("mppdec --gain 2 --start %d %s /dev/audio", "\.mp[cp+]$")
| 
| 
| Seems somewhat simpler to me.
| 
| /dev/audio is more portable than /dev/esd and --gain 2 is
| track-replaygain instead of album-replaygain
| 
| Hard choice of course. The only way to decide between album and
| track RG is to detect wether you play concurrent tracks.
| 
| I couldn't find the email-address of Artur [Ar't] Polaczynski,
| otherwise I would have Cc'ed him.
| 
| 
| 
| # Han

Hi,

Good work. Vanilla cplay-1.50 really should have mpc support.

Optimally, the command-line for mppdec should not contain a device at
all as I think it has / should have a sane default set at compile-
time. I tested mppdec without /dev/audio and it works for me.

It has been a while and I find the current player-framework a little
confusing myself now. In a perfect world there should be one generic
TimeOffsetPlayer and one FrameOffsetPlayer with the ability parse most
formats. Hmm, or perhaps it would be a good idea to go back to
something like the "one class per player" framework used in cplay a
long time ago. For example, I really hate passing "38.28" to the
constructor of some FrameOffsetPlayer's. At one time I just became
obsessed with making the cplay source shorter.

-- 
 ulf