[PATCH 1/2] av_ff_common: attempt to fix encodings for ffprobe/avprobe

[PATCH 1/2] av_ff_common: attempt to fix encodings for ffprobe/avprobe

[Eric Wong]

As with sox, these will be dealing with legacy encodings
and badly-encoding software until the end of days, so do
our best to fix them up.
---
 lib/dtas/source/av_ff_common.rb | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/lib/dtas/source/av_ff_common.rb b/lib/dtas/source/av_ff_common.rb
index 5299fdb..332a0bb 100644
--- a/lib/dtas/source/av_ff_common.rb
+++ b/lib/dtas/source/av_ff_common.rb
@@ -104,13 +104,14 @@ def av_ff_ok?
       prev_cmd = cmd
     end while incomplete.compact[0]
 
+    enc = Encoding.default_external # typically Encoding::UTF_8
     # old avprobe
     s.scan(%r{^\[FORMAT\]\n(.*?)\n\[/FORMAT\]\n}m) do |_|
       f = $1.dup
       f =~ /^duration=([\d\.]+)\s*$/nm and @duration = $1.to_f
       # TODO: multi-line/multi-value/repeated tags
       f.gsub!(/^TAG:([^=]+)=(.*)$/ni) { |_|
-        @comments[$1.upcase] = -($2)
+        @comments[-DTAS.try_enc($1.upcase, enc)] = $2
       }
     end
 
@@ -118,13 +119,17 @@ def av_ff_ok?
     s.scan(%r{^\[format\.tags\]\n(.*?)\n\n}m) do |_|
       f = $1.dup
       f.gsub!(/^([^=]+)=(.*)$/ni) { |_|
-        @comments[$1.upcase] = -$2
+        @comments[-DTAS.try_enc($1.upcase, enc)] = $2
       }
     end
     s.scan(%r{^\[format\]\n(.*?)\n\n}m) do |_|
       f = $1.dup
       f =~ /^duration=([\d\.]+)\s*$/nm and @duration = $1.to_f
     end
+    comments.each do |k,v|
+      v.chomp!
+      comments[k] = -DTAS.try_enc(v, enc)
+    end
 
     ! @astreams.compact.empty?
   end

[PATCH 2/2] av_ff_common: use mcache

[Eric Wong]

As with soxi(1), waiting for ffprobe(1) or avprobe(1) is still
an expensive operation despite Process.spawn being optimized to
use vfork(2).
---
 lib/dtas/source/av.rb           |  1 +
 lib/dtas/source/av_ff_common.rb | 19 ++++++++++++++++---
 lib/dtas/source/ff.rb           |  1 +
 3 files changed, 18 insertions(+), 3 deletions(-)

diff --git a/lib/dtas/source/av.rb b/lib/dtas/source/av.rb
index 39cad6c..823b29f 100644
--- a/lib/dtas/source/av.rb
+++ b/lib/dtas/source/av.rb
@@ -20,6 +20,7 @@ class DTAS::Source::Av # :nodoc:
 
   def initialize
     command_init(AV_DEFAULTS)
+    @mcache = nil
     @av_ff_probe = "avprobe"
   end
 
diff --git a/lib/dtas/source/av_ff_common.rb b/lib/dtas/source/av_ff_common.rb
index 332a0bb..0c84b2f 100644
--- a/lib/dtas/source/av_ff_common.rb
+++ b/lib/dtas/source/av_ff_common.rb
@@ -21,10 +21,23 @@ module DTAS::Source::AvFfCommon # :nodoc:
   attr_reader :format
   attr_reader :duration
 
+  CACHE_KEYS = [ :@duration, :@probe_harder, :@comments, :@astreams,
+                 :@format ].freeze
+
+  def mcache_lookup(infile)
+    (@mcache ||= DTAS::Mcache.new).lookup(infile) do |input, dst|
+      tmp = source_file_dup(infile, nil, nil)
+      tmp.av_ff_ok? or return nil
+      CACHE_KEYS.each { |k| dst[k] = tmp.instance_variable_get(k) }
+      dst
+    end
+  end
+
   def try(infile, offset = nil, trim = nil)
-    rv = source_file_dup(infile, offset, trim)
-    rv.av_ff_ok? or return
-    rv
+    ent = mcache_lookup(infile) or return
+    ret = source_file_dup(infile, offset, trim)
+    CACHE_KEYS.each { |k| ret.instance_variable_set(k, ent[k]) }
+    ret
   end
 
   def __parse_astream(cmd, stream)
diff --git a/lib/dtas/source/ff.rb b/lib/dtas/source/ff.rb
index 687cd18..491e580 100644
--- a/lib/dtas/source/ff.rb
+++ b/lib/dtas/source/ff.rb
@@ -21,6 +21,7 @@ class DTAS::Source::Ff  # :nodoc:
 
   def initialize
     command_init(FF_DEFAULTS)
+    @mcache = nil
     @av_ff_probe = "ffprobe"
   end