Migrate PSI parsing

Move the code that parses PMTs and PATs from segment parser into the re-organized transmuxing module. Add tests.

@@ -7,7 +7,187 @@
 (function(window, videojs, undefined) {
 'use strict';
 
-var Transmuxer = function() {
+var PacketStream, ParseStream, Transmuxer, MP2T_PACKET_LENGTH;
+
+MP2T_PACKET_LENGTH = 188; // bytes
+
+/**
+ * Splits an incoming stream of binary data into MP2T packets.
+ */
+PacketStream = function() {
+  var
+    buffer = new Uint8Array(MP2T_PACKET_LENGTH),
+    end = 0;
+
+  PacketStream.prototype.init.call(this);
+
+  /**
+   * Deliver new bytes to the stream.
+   */
+  this.push = function(bytes) {
+    var remaining, i;
+
+    // clear out any partial packets in the buffer
+    if (end > 0) {
+      remaining = MP2T_PACKET_LENGTH - end;
+      buffer.set(bytes.subarray(0, remaining), end);
+
+      // we still didn't write out a complete packet
+      if (bytes.byteLength < remaining) {
+        end += bytes.byteLength;
+        return;
+      }
+
+      bytes = bytes.subarray(remaining);
+      end = 0;
+      this.trigger('data', buffer);
+    }
+
+    // if less than a single packet is available, buffer it up for later
+    if (bytes.byteLength < MP2T_PACKET_LENGTH) {
+      buffer.set(bytes.subarray(i), end);
+      end += bytes.byteLength;
+      return;
+    }
+    // parse out all the completed packets
+    i = 0;
+    do {
+      this.trigger('data', bytes.subarray(i, i + MP2T_PACKET_LENGTH));
+      i += MP2T_PACKET_LENGTH;
+      remaining = bytes.byteLength - i;
+    } while (i < bytes.byteLength && remaining >= MP2T_PACKET_LENGTH);
+    // buffer any partial packets left over
+    if (remaining > 0) {
+      buffer.set(bytes.subarray(i));
+      end = remaining;
+    }
+  };
+};
+PacketStream.prototype = new videojs.Hls.Stream();
+
+/**
+ * Accepts an MP2T PacketStream and emits data events with parsed
+ * forms of the individual packets.
+ */
+ParseStream = function() {
+  var parsePsi, parsePat, parsePmt, self;
+  PacketStream.prototype.init.call(this);
+  self = this;
+
+  this.programMapTable = {};
+
+  parsePsi = function(payload, psi) {
+    var offset = 0;
+
+    // PSI packets may be split into multiple sections and those
+    // sections may be split into multiple packets. If a PSI
+    // section starts in this packet, the payload_unit_start_indicator
+    // will be true and the first byte of the payload will indicate
+    // the offset from the current position to the start of the
+    // section.
+    if (psi.payloadUnitStartIndicator) {
+      offset += payload[offset] + 1;
+    }
+
+    if (psi.type === 'pat') {
+      parsePat(payload.subarray(offset), psi);
+    } else {
+      parsePmt(payload.subarray(offset), psi);
+    }
+  };
+
+  parsePat = function(payload, pat) {
+    // skip the PSI header and parse the first PMT entry
+    self.pmtPid = (payload[10] & 0x1F) << 8 | payload[11];
+    pat.pmtPid = self.pmtPid;
+  };
+
+  /**
+   * Parse out the relevant fields of a Program Map Table (PMT).
+   * @param payload {Uint8Array} the PMT-specific portion of an MP2T
+   * packet. The first byte in this array should be the table_id
+   * field.
+   * @param pmt {object} the object that should be decorated with
+   * fields parsed from the PMT.
+   */
+  parsePmt = function(payload, pmt) {
+    var tableEnd, programInfoLength, offset;
+
+    // PMTs can be sent ahead of the time when they should actually
+    // take effect. We don't believe this should ever be the case
+    // for HLS but we'll ignore "forward" PMT declarations if we see
+    // them. Future PMT declarations have the current_next_indicator
+    // set to zero.
+    if (!(payload[5] & 0x01)) {
+      return;
+    }
+
+    // overwrite any existing program map table
+    self.programMapTable = {};
+
+    // the mapping table ends right before the 32-bit CRC
+    tableEnd = payload.byteLength - 4;
+    // to determine where the table starts, we have to figure out how
+    // long the program info descriptors are
+    programInfoLength = (payload[10] & 0x0f) << 8 | payload[11];
+
+    // advance the offset to the first entry in the mapping table
+    offset = 12 + programInfoLength;
+    while (offset < tableEnd) {
+      // add an entry that maps the elementary_pid to the stream_type
+      self.programMapTable[(payload[offset + 1] & 0x1F) << 8 | payload[offset + 2]] = payload[offset];
+
+      // move to the next table entry
+      // skip past the elementary stream descriptors, if present
+      offset += ((payload[offset + 3] & 0x0F) << 8 | payload[offset + 4]) + 5;
+    }
+  };
+
+  /**
+   * Deliver a new MP2T packet to the stream.
+   */
+  this.push = function(packet) {
+    var
+      result = {},
+      offset = 4,
+      stream;
+    // make sure packet is aligned on a sync byte
+    if (packet[0] !== 0x47) {
+      return this.trigger('error', 'mis-aligned packet');
+    }
+    result.payloadUnitStartIndicator = !!(packet[1] & 0x40);
+
+    // pid is a 13-bit field starting at the last bit of packet[1]
+    result.pid = packet[1] & 0x1f;
+    result.pid <<= 8;
+    result.pid |= packet[2];
+
+    // if an adaption field is present, its length is specified by
+    // the fifth byte of the PES header. The adaptation field is
+    // used to specify some forms of timing and control data that we
+    // do not currently use.
+    if (((packet[3] & 0x30) >>> 4) > 0x01) {
+      offset += packet[offset] + 1;
+    }
+
+    // parse the rest of the packet based on the type
+    if (result.pid === 0) {
+      result.type = 'pat';
+      parsePsi(packet.subarray(offset), result);
+    } else if (result.pid === this.pmtPid) {
+      result.type = 'pmt';
+      parsePsi(packet.subarray(offset), result);
+    } else {
+      result.stream = this.programMapTable[result.pid];
+      result.type = 'pes';
+    }
+
+    this.trigger('data', result);
+  };
+};
+ParseStream.prototype = new videojs.Hls.Stream();
+
+Transmuxer = function() {
   Transmuxer.prototype.init.call(this);
   this.push = function() {
     this.mp4 = new Uint8Array();
@@ -15,5 +195,13 @@ var Transmuxer = function() {
 };
 Transmuxer.prototype = new videojs.Hls.Stream();
 
-window.videojs.Hls.Transmuxer = Transmuxer;
+window.videojs.mp2t = {
+  PAT_PID: 0x0000,
+  MP2T_PACKET_LENGTH: MP2T_PACKET_LENGTH,
+  H264_STREAM_TYPE: 0x1b,
+  ADTS_STREAM_TYPE: 0x0f,
+  Transmuxer: Transmuxer,
+  PacketStream: PacketStream,
+  ParseStream: ParseStream
+};
 })(window, window.videojs);

@@ -21,9 +21,250 @@
   throws(block, [expected], [message])
 */
 var
-  Transmuxer = videojs.Hls.Transmuxer,
+  PacketStream = videojs.mp2t.PacketStream,
+  packetStream,
+  ParseStream = videojs.mp2t.ParseStream,
+  parseStream,
+  Transmuxer = videojs.mp2t.Transmuxer,
   transmuxer;
 
+module('MP2T Packet Stream', {
+  setup: function() {
+    packetStream = new PacketStream();
+  }
+});
+
+test('empty input does not error', function() {
+  packetStream.push(new Uint8Array([]));
+  ok(true, 'did not throw');
+});
+test('parses a generic packet', function() {
+  var datas = [];
+  packetStream.on('data', function(event) {
+    datas.push(event);
+  });
+  packetStream.push(new Uint8Array(188));
+
+  equal(1, datas.length, 'fired one event');
+  equal(datas[0].byteLength, 188, 'delivered the packet');
+});
+
+test('buffers partial packets', function() {
+  var datas = [];
+  packetStream.on('data', function(event) {
+    datas.push(event);
+  });
+  packetStream.push(new Uint8Array(187));
+
+  equal(0, datas.length, 'did not fire an event');
+
+  packetStream.push(new Uint8Array(189));
+  equal(2, datas.length, 'fired events');
+  equal(188, datas[0].byteLength, 'parsed the first packet');
+  equal(188, datas[1].byteLength, 'parsed the second packet');
+});
+
+test('parses multiple packets delivered at once', function() {
+  var datas = [];
+  packetStream.on('data', function(event) {
+    datas.push(event);
+  });
+
+  packetStream.push(new Uint8Array(188 * 3));
+  equal(3, datas.length, 'fired three events');
+  equal(188, datas[0].byteLength, 'parsed the first packet');
+  equal(188, datas[1].byteLength, 'parsed the second packet');
+  equal(188, datas[2].byteLength, 'parsed the third packet');
+});
+
+test('buffers extra after multiple packets', function() {
+  var datas = [];
+  packetStream.on('data', function(event) {
+    datas.push(event);
+  });
+
+  packetStream.push(new Uint8Array(188 * 2 + 10));
+  equal(2, datas.length, 'fired two events');
+  equal(188, datas[0].byteLength, 'parsed the first packet');
+  equal(188, datas[1].byteLength, 'parsed the second packet');
+
+  packetStream.push(new Uint8Array(178));
+  equal(3, datas.length, 'fired a final event');
+  equal(188, datas[2].length, 'parsed the finel packet');
+});
+
+module('MP2T Parse Stream', {
+  setup: function() {
+    parseStream = new ParseStream();
+  }
+});
+
+test('emits an error on an invalid packet', function() {
+  var errors = [];
+  parseStream.on('error', function(error) {
+    errors.push(error);
+  });
+  parseStream.push(new Uint8Array(188));
+
+  equal(1, errors.length, 'emitted an error');
+});
+
+test('parses generic packet properties', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0000 0001 tsc:01 afc:10 cc:11 padding: 00
+    0x40, 0x01, 0x6c
+  ]));
+  ok(packet.payloadUnitStartIndicator, 'parsed payload_unit_start_indicator');
+  ok(packet.pid, 'parsed PID');
+});
+
+test('parses a data packet with adaptation fields', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0000 0000 tsc:01 afc:10 cc:11 afl:00 0000 00 stuffing:00 0000 00 pscp:00 0001 padding:0000
+    0x40, 0x00, 0x6c, 0x00, 0x00, 0x10
+  ]));
+  strictEqual(packet.type, 'pat', 'parsed the packet type');
+});
+
+test('parses a PES packet', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+
+  // setup a program map table
+  parseStream.programMapTable = {
+    0x0010: videojs.mp2t.H264_STREAM_TYPE
+  };
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0000 0010 tsc:01 afc:01 cc:11 padding:00
+    0x40, 0x02, 0x5c
+  ]));
+  strictEqual(packet.type, 'pes', 'parsed a PES packet');
+});
+
+test('parses packets with variable length adaptation fields and a payload', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+
+  // setup a program map table
+  parseStream.programMapTable = {
+    0x0010: videojs.mp2t.H264_STREAM_TYPE
+  };
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0000 0010 tsc:01 afc:11 cc:11 afl:00 0000 11 stuffing:00 0000 0000 00 pscp:00 0001
+    0x40, 0x02, 0x7c, 0x0c, 0x00, 0x01
+  ]));
+  strictEqual(packet.type, 'pes', 'parsed a PES packet');
+});
+
+/*
+ Packet Header:
+ | sb | tei pusi tp pid:5 | pid | tsc afc cc |
+ with af:
+ | afl | ... | <data> |
+ without af:
+ | <data> |
+
+PAT:
+ | pf? | ... |
+ | tid | ssi '0' r sl:4 | sl | tsi:8 |
+ | tsi | r vn cni | sn | lsn |
+
+with program_number == '0':
+ | pn | pn | r np:5 | np |
+otherwise:
+ | pn | pn | r pmp:5 | pmp |
+*/
+
+test('parses the program map table pid from the program association table (PAT)', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0000 0000
+    0x40, 0x00,
+    // tsc:01 afc:01 cc:0000 pointer_field:0000 0000
+    0x50, 0x00,
+    // tid:0000 0000 ssi:0 0:0 r:00 sl:0000 0000 0000
+    0x00, 0x00, 0x00,
+    // tsi:0000 0000 0000 0000
+    0x00, 0x00,
+    // r:00 vn:00 000 cni:1 sn:0000 0000 lsn:0000 0000
+    0x01, 0x00, 0x00,
+    // pn:0000 0000 0000 0001
+    0x00, 0x01,
+    // r:000 pmp:0 0000 0010 0000
+    0x00, 0x10,
+    // crc32:0000 0000 0000 0000 0000 0000 0000 0000
+    0x00, 0x00, 0x00, 0x00
+  ]));
+  strictEqual(0x0010, parseStream.pmtPid, 'parsed PMT pid');
+});
+
+test('parse the elementary streams from a program map table', function() {
+  var packet;
+  parseStream.on('data', function(data) {
+    packet = data;
+  });
+  parseStream.pmtPid = 0x0010;
+
+  parseStream.push(new Uint8Array([
+    0x47, // sync byte
+    // tei:0 pusi:1 tp:0 pid:0 0000 0010 0000
+    0x40, 0x10,
+    // tsc:01 afc:01 cc:0000 pointer_field:0000 0000
+    0x50, 0x00,
+    // tid:0000 0000 ssi:0 0:0 r:00 sl:0000 0010 1111
+    0x00, 0x00, 0x2f,
+    // pn:0000 0000 0000 0001
+    0x00, 0x01,
+    // r:00 vn:00 000 cni:1 sn:0000 0000 lsn:0000 0000
+    0x01, 0x00, 0x00,
+    // r:000 ppid:0 0011 1111 1111
+    0x03, 0xff,
+    // r:0000 pil:0000 0000 0000
+    0x00, 0x00,
+    // h264
+    // st:0001 1010 r:000 epid:0 0000 0001 0001
+    0x1b, 0x00, 0x11,
+    // r:0000 esil:0000 0000 0000
+    0x00, 0x00,
+    // adts
+    // st:0000 1111 r:000 epid:0 0000 0001 0010
+    0x0f, 0x00, 0x12,
+    // r:0000 esil:0000 0000 0000
+    0x00, 0x00,
+    // crc
+    0x00, 0x00, 0x00, 0x00
+  ]));
+
+  ok(parseStream.programMapTable, 'parsed a program map');
+  strictEqual(0x1b, parseStream.programMapTable[0x11], 'associated h264 with pid 0x11');
+  strictEqual(0x0f, parseStream.programMapTable[0x12], 'associated adts with pid 0x12');
+});
+
 module('MP4 Transmuxer', {
   setup: function() {
     transmuxer = new Transmuxer();