source file of the GNU LilyPond music typesetter
- (c) 1997--2000 Jan Nieuwenhuizen <janneke@gnu.org>
+ (c) 1997--2004 Jan Nieuwenhuizen <janneke@gnu.org>
*/
-#include <time.h>
-#include "debug.hh"
-#include "string.hh"
-#include "string-convert.hh"
+#include "performance.hh"
+
+#include <ctime>
+
+#include "audio-column.hh"
+#include "audio-staff.hh"
+#include "file-name.hh"
+#include "lily-version.hh"
#include "main.hh"
-#include "midi-def.hh"
#include "midi-item.hh"
#include "midi-stream.hh"
-#include "audio-column.hh"
-#include "audio-item.hh"
-#include "audio-staff.hh"
-#include "performance.hh"
#include "score.hh"
-#include "file-results.hh"
-#include "file-path.hh"
-#include "lily-version.hh"
+#include "string-convert.hh"
+#include "warn.hh"
#include "killing-cons.tcc"
Performance::Performance ()
{
- midi_l_ =0;
+ midi_ =0;
audio_elem_p_list_ = 0;
}
-Performance::~Performance()
+Performance::~Performance ()
{
delete audio_elem_p_list_;
}
void
Performance::output (Midi_stream& midi_stream)
{
- int tracks_i = audio_staff_l_arr_.size() + 1;
+ int tracks_i = audio_staffs_.size () + 1;
// ugh
int clocks_per_4_i = 384;
midi_stream << Midi_header (1, tracks_i, clocks_per_4_i);
output_header_track (midi_stream);
progress_indication ("\n");
- progress_indication (_ ("Track ... "));
- int channel = 1;
- for (int i =0; i < audio_staff_l_arr_.size (); i++)
+ progress_indication (_ ("Track...") + " ");
+ int channel = 0;
+ for (int i = 0; i < audio_staffs_.size (); i++)
{
- Audio_staff *s = audio_staff_l_arr_[i];
- if(verbose_global_b)
- progress_indication ("[" + to_str (i)) ;
+ Audio_staff *s = audio_staffs_[i];
+ if (verbose_global_b)
+ progress_indication ("[" + to_string (i)) ;
/*
- Aargh, let's hear it for the MIDI standard.
MIDI players tend to ignore instrument settings on
channel 10, the percussion channel by default.
*/
- if (channel == 9)
- channel++;
+ if (channel % 16 == 9)
+ channel++;
+
+
+ /*
+ Huh? Why does each staff also have a separate channel? We
+ should map channels to voices, not staves. --hwn.
+ */
+ if (s->channel_ < 0)
+ {
+ s->channel_ = channel % 16;
+ if (channel > 15)
+ warning ("MIDI channel wrapped around. Remapping modulo 16.");
+ }
+
s->output (midi_stream, channel++);
- if(verbose_global_b)
+ if (verbose_global_b)
progress_indication ("]");
}
}
+
void
Performance::output_header_track (Midi_stream& midi_stream)
{
Midi_track midi_track;
+ midi_track.channel_ = 9;
+
// perhaps multiple text events?
- String id_str;
- String str = String (_("Creator: "));
- if (no_timestamps_global_b)
- id_str = gnu_lilypond_str ();
- else
- id_str = gnu_lilypond_version_str();
- str += id_str;
- str += "\n";
+ String id_string;
+ String str = String (_ ("Creator: "));
+ id_string = String_convert::pad_to (gnu_lilypond_version_string (), 30);
+ str += id_string;
/*
This seems silly, but in fact the audio elements should
Midi_text creator (&creator_a);
midi_track.add (Moment (0), &creator);
- str = _("Generated automatically by: ");
- str += id_str;
- if (no_timestamps_global_b)
- str += ".\n";
- else
- {
- str += _(", at ");
- time_t t (time (0));
- str += ctime (&t);
- str = str.left_str (str.length_i() - 1);
- }
+ /* Better not translate this */
+ str = "Generated automatically by: ";
+ str += id_string;
+
Audio_text generate_a (Audio_text::TEXT, str);
Midi_text generate (&generate_a);
midi_track.add (Moment (0), &generate);
+
+ str = _ ("at ");
+ time_t t (time (0));
+ str += ctime (&t);
+ str = str.left_string (str.length () - 1);
+ str = String_convert::pad_to (str, 60);
+
+ Audio_text at_a (Audio_text::TEXT, str);
+ Midi_text at (&at_a);
+ midi_track.add (Moment (0), &at);
- str = _f ("from musical definition: %s", origin_str_);
+
+ // TODO:
+ // str = _f ("from musical definition: %s", origin_string_);
Audio_text from_a (Audio_text::TEXT, str);
Midi_text from (&from_a);
midi_track.add (Moment (0), &from);
Audio_text track_name_a (Audio_text::TRACK_NAME, "Track "
- + String_convert::i2dec_str (0, 0, '0'));
+ + String_convert::int2dec (0, 0, '0'));
Midi_text track_name (&track_name_a);
midi_track.add (Moment (0), &track_name);
// Some sequencers read track 0 last.
- // Audio_tempo tempo_a (midi_l_->get_tempo_i (Moment (1, 4)));
+ // Audio_tempo tempo_a (midi_->get_tempo (Moment (Rational (1, 4))));
// Midi_tempo tempo (&tempo_a);
// midi_track.add (Moment (0), &tempo);
{
if (Audio_staff*s=dynamic_cast<Audio_staff *> (p))
{
- audio_staff_l_arr_.push (s);
- }
- else if (Audio_column *c = dynamic_cast<Audio_column*>(p))
- {
- c->performance_l_ = this;
+ audio_staffs_.push (s);
}
audio_elem_p_list_ = new Killing_cons<Audio_element> (p, audio_elem_p_list_);
}
-
-void
-Performance::process()
+SCM
+Performance::process (String out)
{
- String out = outname_global;
if (out == "-")
out = "lelie.midi";
- int def = midi_l_->get_next_default_count ();
- if (def)
- {
- Path p = split_path (out);
- p.base += "-" + to_str (def);
- out = p.path ();
- }
+
+ /* Maybe a bit crude, but we had this before */
+ File_name file_name (out);
+ file_name.ext_ = "midi";
+ out = file_name.to_string ();
Midi_stream midi_stream (out);
- progress_indication ( _f ("MIDI output to %s...", out));
- target_str_global_array.push (out);
+ progress_indication (_f ("MIDI output to `%s'...", out));
output (midi_stream);
- progress_indication("\n");
+ progress_indication ("\n");
+ return SCM_UNDEFINED;
}
projects / lilypond.git / blobdiff