UNPKG

@sqhead/mux.js

Version:

A collection of lightweight utilities for inspecting and manipulating video container formats.

1,600 lines (1,404 loc) 161 kB
/** * Copyright 2015 Brightcove * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */ (function(f){if(typeof exports==="object"&&typeof module!=="undefined"){module.exports=f()}else if(typeof define==="function"&&define.amd){define([],f)}else{var g;if(typeof window!=="undefined"){g=window}else if(typeof global!=="undefined"){g=global}else if(typeof self!=="undefined"){g=self}else{g=this}g.muxjs = f()}})(function(){var define,module,exports;return (function e(t,n,r){function s(o,u){if(!n[o]){if(!t[o]){var a=typeof require=="function"&&require;if(!u&&a)return a(o,!0);if(i)return i(o,!0);var f=new Error("Cannot find module '"+o+"'");throw f.code="MODULE_NOT_FOUND",f}var l=n[o]={exports:{}};t[o][0].call(l.exports,function(e){var n=t[o][1][e];return s(n?n:e)},l,l.exports,e,t,n,r)}return n[o].exports}var i=typeof require=="function"&&require;for(var o=0;o<r.length;o++)s(r[o]);return s})({1:[function(require,module,exports){ /** * mux.js * * Copyright (c) 2016 Brightcove * All rights reserved. * * A stream-based aac to mp4 converter. This utility can be used to * deliver mp4s to a SourceBuffer on platforms that support native * Media Source Extensions. */ 'use strict'; var Stream = require('../utils/stream.js'); // Constants var AacStream; /** * Splits an incoming stream of binary data into ADTS and ID3 Frames. */ AacStream = function() { var everything = new Uint8Array(), timeStamp = 0; AacStream.prototype.init.call(this); this.setTimestamp = function(timestamp) { timeStamp = timestamp; }; this.parseId3TagSize = function(header, byteIndex) { var returnSize = (header[byteIndex + 6] << 21) | (header[byteIndex + 7] << 14) | (header[byteIndex + 8] << 7) | (header[byteIndex + 9]), flags = header[byteIndex + 5], footerPresent = (flags & 16) >> 4; if (footerPresent) { return returnSize + 20; } return returnSize + 10; }; this.parseAdtsSize = function(header, byteIndex) { var lowThree = (header[byteIndex + 5] & 0xE0) >> 5, middle = header[byteIndex + 4] << 3, highTwo = header[byteIndex + 3] & 0x3 << 11; return (highTwo | middle) | lowThree; }; this.push = function(bytes) { var frameSize = 0, byteIndex = 0, bytesLeft, chunk, packet, tempLength; // If there are bytes remaining from the last segment, prepend them to the // bytes that were pushed in if (everything.length) { tempLength = everything.length; everything = new Uint8Array(bytes.byteLength + tempLength); everything.set(everything.subarray(0, tempLength)); everything.set(bytes, tempLength); } else { everything = bytes; } while (everything.length - byteIndex >= 3) { if ((everything[byteIndex] === 'I'.charCodeAt(0)) && (everything[byteIndex + 1] === 'D'.charCodeAt(0)) && (everything[byteIndex + 2] === '3'.charCodeAt(0))) { // Exit early because we don't have enough to parse // the ID3 tag header if (everything.length - byteIndex < 10) { break; } // check framesize frameSize = this.parseId3TagSize(everything, byteIndex); // Exit early if we don't have enough in the buffer // to emit a full packet if (frameSize > everything.length) { break; } chunk = { type: 'timed-metadata', data: everything.subarray(byteIndex, byteIndex + frameSize) }; this.trigger('data', chunk); byteIndex += frameSize; continue; } else if ((everything[byteIndex] & 0xff === 0xff) && ((everything[byteIndex + 1] & 0xf0) === 0xf0)) { // Exit early because we don't have enough to parse // the ADTS frame header if (everything.length - byteIndex < 7) { break; } frameSize = this.parseAdtsSize(everything, byteIndex); // Exit early if we don't have enough in the buffer // to emit a full packet if (frameSize > everything.length) { break; } packet = { type: 'audio', data: everything.subarray(byteIndex, byteIndex + frameSize), pts: timeStamp, dts: timeStamp }; this.trigger('data', packet); byteIndex += frameSize; continue; } byteIndex++; } bytesLeft = everything.length - byteIndex; if (bytesLeft > 0) { everything = everything.subarray(byteIndex); } else { everything = new Uint8Array(); } }; }; AacStream.prototype = new Stream(); module.exports = AacStream; },{"../utils/stream.js":17}],2:[function(require,module,exports){ 'use strict'; var Stream = require('../utils/stream.js'); var AdtsStream; var ADTS_SAMPLING_FREQUENCIES = [ 96000, 88200, 64000, 48000, 44100, 32000, 24000, 22050, 16000, 12000, 11025, 8000, 7350 ]; /* * Accepts a ElementaryStream and emits data events with parsed * AAC Audio Frames of the individual packets. Input audio in ADTS * format is unpacked and re-emitted as AAC frames. * * @see http://wiki.multimedia.cx/index.php?title=ADTS * @see http://wiki.multimedia.cx/?title=Understanding_AAC */ AdtsStream = function() { var buffer; AdtsStream.prototype.init.call(this); this.push = function(packet) { var i = 0, frameNum = 0, frameLength, protectionSkipBytes, frameEnd, oldBuffer, sampleCount, adtsFrameDuration; if (packet.type !== 'audio') { // ignore non-audio data return; } // Prepend any data in the buffer to the input data so that we can parse // aac frames the cross a PES packet boundary if (buffer) { oldBuffer = buffer; buffer = new Uint8Array(oldBuffer.byteLength + packet.data.byteLength); buffer.set(oldBuffer); buffer.set(packet.data, oldBuffer.byteLength); } else { buffer = packet.data; } // unpack any ADTS frames which have been fully received // for details on the ADTS header, see http://wiki.multimedia.cx/index.php?title=ADTS while (i + 5 < buffer.length) { // Loook for the start of an ADTS header.. if (buffer[i] !== 0xFF || (buffer[i + 1] & 0xF6) !== 0xF0) { // If a valid header was not found, jump one forward and attempt to // find a valid ADTS header starting at the next byte i++; continue; } // The protection skip bit tells us if we have 2 bytes of CRC data at the // end of the ADTS header protectionSkipBytes = (~buffer[i + 1] & 0x01) * 2; // Frame length is a 13 bit integer starting 16 bits from the // end of the sync sequence frameLength = ((buffer[i + 3] & 0x03) << 11) | (buffer[i + 4] << 3) | ((buffer[i + 5] & 0xe0) >> 5); sampleCount = ((buffer[i + 6] & 0x03) + 1) * 1024; adtsFrameDuration = (sampleCount * 90000) / ADTS_SAMPLING_FREQUENCIES[(buffer[i + 2] & 0x3c) >>> 2]; frameEnd = i + frameLength; // If we don't have enough data to actually finish this ADTS frame, return // and wait for more data if (buffer.byteLength < frameEnd) { return; } // Otherwise, deliver the complete AAC frame this.trigger('data', { pts: packet.pts + (frameNum * adtsFrameDuration), dts: packet.dts + (frameNum * adtsFrameDuration), sampleCount: sampleCount, audioobjecttype: ((buffer[i + 2] >>> 6) & 0x03) + 1, channelcount: ((buffer[i + 2] & 1) << 2) | ((buffer[i + 3] & 0xc0) >>> 6), samplerate: ADTS_SAMPLING_FREQUENCIES[(buffer[i + 2] & 0x3c) >>> 2], samplingfrequencyindex: (buffer[i + 2] & 0x3c) >>> 2, // assume ISO/IEC 14496-12 AudioSampleEntry default of 16 samplesize: 16, data: buffer.subarray(i + 7 + protectionSkipBytes, frameEnd) }); // If the buffer is empty, clear it and return if (buffer.byteLength === frameEnd) { buffer = undefined; return; } frameNum++; // Remove the finished frame from the buffer and start the process again buffer = buffer.subarray(frameEnd); } }; this.flush = function() { this.trigger('done'); }; }; AdtsStream.prototype = new Stream(); module.exports = AdtsStream; },{"../utils/stream.js":17}],3:[function(require,module,exports){ 'use strict'; var Stream = require('../utils/stream.js'); var ExpGolomb = require('../utils/exp-golomb.js'); var H264Stream, NalByteStream; var PROFILES_WITH_OPTIONAL_SPS_DATA; /** * Accepts a NAL unit byte stream and unpacks the embedded NAL units. */ NalByteStream = function() { var syncPoint = 0, i, buffer; NalByteStream.prototype.init.call(this); this.push = function(data) { var swapBuffer; if (!buffer) { buffer = data.data; } else { swapBuffer = new Uint8Array(buffer.byteLength + data.data.byteLength); swapBuffer.set(buffer); swapBuffer.set(data.data, buffer.byteLength); buffer = swapBuffer; } // Rec. ITU-T H.264, Annex B // scan for NAL unit boundaries // a match looks like this: // 0 0 1 .. NAL .. 0 0 1 // ^ sync point ^ i // or this: // 0 0 1 .. NAL .. 0 0 0 // ^ sync point ^ i // advance the sync point to a NAL start, if necessary for (; syncPoint < buffer.byteLength - 3; syncPoint++) { if (buffer[syncPoint + 2] === 1) { // the sync point is properly aligned i = syncPoint + 5; break; } } while (i < buffer.byteLength) { // look at the current byte to determine if we've hit the end of // a NAL unit boundary switch (buffer[i]) { case 0: // skip past non-sync sequences if (buffer[i - 1] !== 0) { i += 2; break; } else if (buffer[i - 2] !== 0) { i++; break; } // deliver the NAL unit if it isn't empty if (syncPoint + 3 !== i - 2) { this.trigger('data', buffer.subarray(syncPoint + 3, i - 2)); } // drop trailing zeroes do { i++; } while (buffer[i] !== 1 && i < buffer.length); syncPoint = i - 2; i += 3; break; case 1: // skip past non-sync sequences if (buffer[i - 1] !== 0 || buffer[i - 2] !== 0) { i += 3; break; } // deliver the NAL unit this.trigger('data', buffer.subarray(syncPoint + 3, i - 2)); syncPoint = i - 2; i += 3; break; default: // the current byte isn't a one or zero, so it cannot be part // of a sync sequence i += 3; break; } } // filter out the NAL units that were delivered buffer = buffer.subarray(syncPoint); i -= syncPoint; syncPoint = 0; }; this.flush = function() { // deliver the last buffered NAL unit if (buffer && buffer.byteLength > 3) { this.trigger('data', buffer.subarray(syncPoint + 3)); } // reset the stream state buffer = null; syncPoint = 0; this.trigger('done'); }; }; NalByteStream.prototype = new Stream(); // values of profile_idc that indicate additional fields are included in the SPS // see Recommendation ITU-T H.264 (4/2013), // 7.3.2.1.1 Sequence parameter set data syntax PROFILES_WITH_OPTIONAL_SPS_DATA = { 100: true, 110: true, 122: true, 244: true, 44: true, 83: true, 86: true, 118: true, 128: true, 138: true, 139: true, 134: true }; /** * Accepts input from a ElementaryStream and produces H.264 NAL unit data * events. */ H264Stream = function() { var nalByteStream = new NalByteStream(), self, trackId, currentPts, currentDts, discardEmulationPreventionBytes, readSequenceParameterSet, skipScalingList; H264Stream.prototype.init.call(this); self = this; this.push = function(packet) { if (packet.type !== 'video') { return; } trackId = packet.trackId; currentPts = packet.pts; currentDts = packet.dts; nalByteStream.push(packet); }; nalByteStream.on('data', function(data) { var event = { trackId: trackId, pts: currentPts, dts: currentDts, data: data }; switch (data[0] & 0x1f) { case 0x05: event.nalUnitType = 'slice_layer_without_partitioning_rbsp_idr'; break; case 0x06: event.nalUnitType = 'sei_rbsp'; event.escapedRBSP = discardEmulationPreventionBytes(data.subarray(1)); break; case 0x07: event.nalUnitType = 'seq_parameter_set_rbsp'; event.escapedRBSP = discardEmulationPreventionBytes(data.subarray(1)); event.config = readSequenceParameterSet(event.escapedRBSP); break; case 0x08: event.nalUnitType = 'pic_parameter_set_rbsp'; break; case 0x09: event.nalUnitType = 'access_unit_delimiter_rbsp'; break; default: break; } self.trigger('data', event); }); nalByteStream.on('done', function() { self.trigger('done'); }); this.flush = function() { nalByteStream.flush(); }; /** * Advance the ExpGolomb decoder past a scaling list. The scaling * list is optionally transmitted as part of a sequence parameter * set and is not relevant to transmuxing. * @param count {number} the number of entries in this scaling list * @param expGolombDecoder {object} an ExpGolomb pointed to the * start of a scaling list * @see Recommendation ITU-T H.264, Section 7.3.2.1.1.1 */ skipScalingList = function(count, expGolombDecoder) { var lastScale = 8, nextScale = 8, j, deltaScale; for (j = 0; j < count; j++) { if (nextScale !== 0) { deltaScale = expGolombDecoder.readExpGolomb(); nextScale = (lastScale + deltaScale + 256) % 256; } lastScale = (nextScale === 0) ? lastScale : nextScale; } }; /** * Expunge any "Emulation Prevention" bytes from a "Raw Byte * Sequence Payload" * @param data {Uint8Array} the bytes of a RBSP from a NAL * unit * @return {Uint8Array} the RBSP without any Emulation * Prevention Bytes */ discardEmulationPreventionBytes = function(data) { var length = data.byteLength, emulationPreventionBytesPositions = [], i = 1, newLength, newData; // Find all `Emulation Prevention Bytes` while (i < length - 2) { if (data[i] === 0 && data[i + 1] === 0 && data[i + 2] === 0x03) { emulationPreventionBytesPositions.push(i + 2); i += 2; } else { i++; } } // If no Emulation Prevention Bytes were found just return the original // array if (emulationPreventionBytesPositions.length === 0) { return data; } // Create a new array to hold the NAL unit data newLength = length - emulationPreventionBytesPositions.length; newData = new Uint8Array(newLength); var sourceIndex = 0; for (i = 0; i < newLength; sourceIndex++, i++) { if (sourceIndex === emulationPreventionBytesPositions[0]) { // Skip this byte sourceIndex++; // Remove this position index emulationPreventionBytesPositions.shift(); } newData[i] = data[sourceIndex]; } return newData; }; /** * Read a sequence parameter set and return some interesting video * properties. A sequence parameter set is the H264 metadata that * describes the properties of upcoming video frames. * @param data {Uint8Array} the bytes of a sequence parameter set * @return {object} an object with configuration parsed from the * sequence parameter set, including the dimensions of the * associated video frames. */ readSequenceParameterSet = function(data) { var frameCropLeftOffset = 0, frameCropRightOffset = 0, frameCropTopOffset = 0, frameCropBottomOffset = 0, sarScale = 1, expGolombDecoder, profileIdc, levelIdc, profileCompatibility, chromaFormatIdc, picOrderCntType, numRefFramesInPicOrderCntCycle, picWidthInMbsMinus1, picHeightInMapUnitsMinus1, frameMbsOnlyFlag, scalingListCount, sarRatio, aspectRatioIdc, i; expGolombDecoder = new ExpGolomb(data); profileIdc = expGolombDecoder.readUnsignedByte(); // profile_idc profileCompatibility = expGolombDecoder.readUnsignedByte(); // constraint_set[0-5]_flag levelIdc = expGolombDecoder.readUnsignedByte(); // level_idc u(8) expGolombDecoder.skipUnsignedExpGolomb(); // seq_parameter_set_id // some profiles have more optional data we don't need if (PROFILES_WITH_OPTIONAL_SPS_DATA[profileIdc]) { chromaFormatIdc = expGolombDecoder.readUnsignedExpGolomb(); if (chromaFormatIdc === 3) { expGolombDecoder.skipBits(1); // separate_colour_plane_flag } expGolombDecoder.skipUnsignedExpGolomb(); // bit_depth_luma_minus8 expGolombDecoder.skipUnsignedExpGolomb(); // bit_depth_chroma_minus8 expGolombDecoder.skipBits(1); // qpprime_y_zero_transform_bypass_flag if (expGolombDecoder.readBoolean()) { // seq_scaling_matrix_present_flag scalingListCount = (chromaFormatIdc !== 3) ? 8 : 12; for (i = 0; i < scalingListCount; i++) { if (expGolombDecoder.readBoolean()) { // seq_scaling_list_present_flag[ i ] if (i < 6) { skipScalingList(16, expGolombDecoder); } else { skipScalingList(64, expGolombDecoder); } } } } } expGolombDecoder.skipUnsignedExpGolomb(); // log2_max_frame_num_minus4 picOrderCntType = expGolombDecoder.readUnsignedExpGolomb(); if (picOrderCntType === 0) { expGolombDecoder.readUnsignedExpGolomb(); // log2_max_pic_order_cnt_lsb_minus4 } else if (picOrderCntType === 1) { expGolombDecoder.skipBits(1); // delta_pic_order_always_zero_flag expGolombDecoder.skipExpGolomb(); // offset_for_non_ref_pic expGolombDecoder.skipExpGolomb(); // offset_for_top_to_bottom_field numRefFramesInPicOrderCntCycle = expGolombDecoder.readUnsignedExpGolomb(); for (i = 0; i < numRefFramesInPicOrderCntCycle; i++) { expGolombDecoder.skipExpGolomb(); // offset_for_ref_frame[ i ] } } expGolombDecoder.skipUnsignedExpGolomb(); // max_num_ref_frames expGolombDecoder.skipBits(1); // gaps_in_frame_num_value_allowed_flag picWidthInMbsMinus1 = expGolombDecoder.readUnsignedExpGolomb(); picHeightInMapUnitsMinus1 = expGolombDecoder.readUnsignedExpGolomb(); frameMbsOnlyFlag = expGolombDecoder.readBits(1); if (frameMbsOnlyFlag === 0) { expGolombDecoder.skipBits(1); // mb_adaptive_frame_field_flag } expGolombDecoder.skipBits(1); // direct_8x8_inference_flag if (expGolombDecoder.readBoolean()) { // frame_cropping_flag frameCropLeftOffset = expGolombDecoder.readUnsignedExpGolomb(); frameCropRightOffset = expGolombDecoder.readUnsignedExpGolomb(); frameCropTopOffset = expGolombDecoder.readUnsignedExpGolomb(); frameCropBottomOffset = expGolombDecoder.readUnsignedExpGolomb(); } if (expGolombDecoder.readBoolean()) { // vui_parameters_present_flag if (expGolombDecoder.readBoolean()) { // aspect_ratio_info_present_flag aspectRatioIdc = expGolombDecoder.readUnsignedByte(); switch (aspectRatioIdc) { case 1: sarRatio = [1, 1]; break; case 2: sarRatio = [12, 11]; break; case 3: sarRatio = [10, 11]; break; case 4: sarRatio = [16, 11]; break; case 5: sarRatio = [40, 33]; break; case 6: sarRatio = [24, 11]; break; case 7: sarRatio = [20, 11]; break; case 8: sarRatio = [32, 11]; break; case 9: sarRatio = [80, 33]; break; case 10: sarRatio = [18, 11]; break; case 11: sarRatio = [15, 11]; break; case 12: sarRatio = [64, 33]; break; case 13: sarRatio = [160, 99]; break; case 14: sarRatio = [4, 3]; break; case 15: sarRatio = [3, 2]; break; case 16: sarRatio = [2, 1]; break; case 255: { sarRatio = [expGolombDecoder.readUnsignedByte() << 8 | expGolombDecoder.readUnsignedByte(), expGolombDecoder.readUnsignedByte() << 8 | expGolombDecoder.readUnsignedByte() ]; break; } } if (sarRatio) { sarScale = sarRatio[0] / sarRatio[1]; } } } return { profileIdc: profileIdc, levelIdc: levelIdc, profileCompatibility: profileCompatibility, width: Math.ceil((((picWidthInMbsMinus1 + 1) * 16) - frameCropLeftOffset * 2 - frameCropRightOffset * 2) * sarScale), height: ((2 - frameMbsOnlyFlag) * (picHeightInMapUnitsMinus1 + 1) * 16) - (frameCropTopOffset * 2) - (frameCropBottomOffset * 2) }; }; }; H264Stream.prototype = new Stream(); module.exports = { H264Stream: H264Stream, NalByteStream: NalByteStream }; },{"../utils/exp-golomb.js":16,"../utils/stream.js":17}],4:[function(require,module,exports){ var highPrefix = [33, 16, 5, 32, 164, 27]; var lowPrefix = [33, 65, 108, 84, 1, 2, 4, 8, 168, 2, 4, 8, 17, 191, 252]; var zeroFill = function(count) { var a = []; while (count--) { a.push(0); } return a; }; var makeTable = function(metaTable) { return Object.keys(metaTable).reduce(function(obj, key) { obj[key] = new Uint8Array(metaTable[key].reduce(function(arr, part) { return arr.concat(part); }, [])); return obj; }, {}); }; // Frames-of-silence to use for filling in missing AAC frames var coneOfSilence = { 96000: [highPrefix, [227, 64], zeroFill(154), [56]], 88200: [highPrefix, [231], zeroFill(170), [56]], 64000: [highPrefix, [248, 192], zeroFill(240), [56]], 48000: [highPrefix, [255, 192], zeroFill(268), [55, 148, 128], zeroFill(54), [112]], 44100: [highPrefix, [255, 192], zeroFill(268), [55, 163, 128], zeroFill(84), [112]], 32000: [highPrefix, [255, 192], zeroFill(268), [55, 234], zeroFill(226), [112]], 24000: [highPrefix, [255, 192], zeroFill(268), [55, 255, 128], zeroFill(268), [111, 112], zeroFill(126), [224]], 16000: [highPrefix, [255, 192], zeroFill(268), [55, 255, 128], zeroFill(268), [111, 255], zeroFill(269), [223, 108], zeroFill(195), [1, 192]], 12000: [lowPrefix, zeroFill(268), [3, 127, 248], zeroFill(268), [6, 255, 240], zeroFill(268), [13, 255, 224], zeroFill(268), [27, 253, 128], zeroFill(259), [56]], 11025: [lowPrefix, zeroFill(268), [3, 127, 248], zeroFill(268), [6, 255, 240], zeroFill(268), [13, 255, 224], zeroFill(268), [27, 255, 192], zeroFill(268), [55, 175, 128], zeroFill(108), [112]], 8000: [lowPrefix, zeroFill(268), [3, 121, 16], zeroFill(47), [7]] }; module.exports = makeTable(coneOfSilence); },{}],5:[function(require,module,exports){ /** * mux.js * * Copyright (c) 2015 Brightcove * All rights reserved. * * Reads in-band caption information from a video elementary * stream. Captions must follow the CEA-708 standard for injection * into an MPEG-2 transport streams. * @see https://en.wikipedia.org/wiki/CEA-708 * @see https://www.gpo.gov/fdsys/pkg/CFR-2007-title47-vol1/pdf/CFR-2007-title47-vol1-sec15-119.pdf */ 'use strict'; // ----------------- // Link To Transport // ----------------- // Supplemental enhancement information (SEI) NAL units have a // payload type field to indicate how they are to be // interpreted. CEAS-708 caption content is always transmitted with // payload type 0x04. var USER_DATA_REGISTERED_ITU_T_T35 = 4, RBSP_TRAILING_BITS = 128, Stream = require('../utils/stream'); /** * Parse a supplemental enhancement information (SEI) NAL unit. * Stops parsing once a message of type ITU T T35 has been found. * * @param bytes {Uint8Array} the bytes of a SEI NAL unit * @return {object} the parsed SEI payload * @see Rec. ITU-T H.264, 7.3.2.3.1 */ var parseSei = function(bytes) { var i = 0, result = { payloadType: -1, payloadSize: 0 }, payloadType = 0, payloadSize = 0; // go through the sei_rbsp parsing each each individual sei_message while (i < bytes.byteLength) { // stop once we have hit the end of the sei_rbsp if (bytes[i] === RBSP_TRAILING_BITS) { break; } // Parse payload type while (bytes[i] === 0xFF) { payloadType += 255; i++; } payloadType += bytes[i++]; // Parse payload size while (bytes[i] === 0xFF) { payloadSize += 255; i++; } payloadSize += bytes[i++]; // this sei_message is a 608/708 caption so save it and break // there can only ever be one caption message in a frame's sei if (!result.payload && payloadType === USER_DATA_REGISTERED_ITU_T_T35) { result.payloadType = payloadType; result.payloadSize = payloadSize; result.payload = bytes.subarray(i, i + payloadSize); break; } // skip the payload and parse the next message i += payloadSize; payloadType = 0; payloadSize = 0; } return result; }; // see ANSI/SCTE 128-1 (2013), section 8.1 var parseUserData = function(sei) { // itu_t_t35_contry_code must be 181 (United States) for // captions if (sei.payload[0] !== 181) { return null; } // itu_t_t35_provider_code should be 49 (ATSC) for captions if (((sei.payload[1] << 8) | sei.payload[2]) !== 49) { return null; } // the user_identifier should be "GA94" to indicate ATSC1 data if (String.fromCharCode(sei.payload[3], sei.payload[4], sei.payload[5], sei.payload[6]) !== 'GA94') { return null; } // finally, user_data_type_code should be 0x03 for caption data if (sei.payload[7] !== 0x03) { return null; } // return the user_data_type_structure and strip the trailing // marker bits return sei.payload.subarray(8, sei.payload.length - 1); }; // see CEA-708-D, section 4.4 var parseCaptionPackets = function(pts, userData) { var results = [], i, count, offset, data; // if this is just filler, return immediately if (!(userData[0] & 0x40)) { return results; } // parse out the cc_data_1 and cc_data_2 fields count = userData[0] & 0x1f; for (i = 0; i < count; i++) { offset = i * 3; data = { type: userData[offset + 2] & 0x03, pts: pts }; // capture cc data when cc_valid is 1 if (userData[offset + 2] & 0x04) { data.ccData = (userData[offset + 3] << 8) | userData[offset + 4]; results.push(data); } } return results; }; var CaptionStream = function() { CaptionStream.prototype.init.call(this); this.captionPackets_ = []; this.ccStreams_ = [ new Cea608Stream(0, 0), // eslint-disable-line no-use-before-define new Cea608Stream(0, 1), // eslint-disable-line no-use-before-define new Cea608Stream(1, 0), // eslint-disable-line no-use-before-define new Cea608Stream(1, 1) // eslint-disable-line no-use-before-define ]; this.reset(); // forward data and done events from CCs to this CaptionStream this.ccStreams_.forEach(function(cc) { cc.on('data', this.trigger.bind(this, 'data')); cc.on('done', this.trigger.bind(this, 'done')); }, this); }; CaptionStream.prototype = new Stream(); CaptionStream.prototype.push = function(event) { var sei, userData; // only examine SEI NALs if (event.nalUnitType !== 'sei_rbsp') { return; } // parse the sei sei = parseSei(event.escapedRBSP); // ignore everything but user_data_registered_itu_t_t35 if (sei.payloadType !== USER_DATA_REGISTERED_ITU_T_T35) { return; } // parse out the user data payload userData = parseUserData(sei); // ignore unrecognized userData if (!userData) { return; } // Sometimes, the same segment # will be downloaded twice. To stop the // caption data from being processed twice, we track the latest dts we've // received and ignore everything with a dts before that. However, since // data for a specific dts can be split across packets on either side of // a segment boundary, we need to make sure we *don't* ignore the packets // from the *next* segment that have dts === this.latestDts_. By constantly // tracking the number of packets received with dts === this.latestDts_, we // know how many should be ignored once we start receiving duplicates. if (event.dts < this.latestDts_) { // We've started getting older data, so set the flag. this.ignoreNextEqualDts_ = true; return; } else if ((event.dts === this.latestDts_) && (this.ignoreNextEqualDts_)) { this.numSameDts_--; if (!this.numSameDts_) { // We've received the last duplicate packet, time to start processing again this.ignoreNextEqualDts_ = false; } return; } // parse out CC data packets and save them for later this.captionPackets_ = this.captionPackets_.concat(parseCaptionPackets(event.pts, userData)); if (this.latestDts_ !== event.dts) { this.numSameDts_ = 0; } this.numSameDts_++; this.latestDts_ = event.dts; }; CaptionStream.prototype.flush = function() { // make sure we actually parsed captions before proceeding if (!this.captionPackets_.length) { this.ccStreams_.forEach(function(cc) { cc.flush(); }, this); return; } // In Chrome, the Array#sort function is not stable so add a // presortIndex that we can use to ensure we get a stable-sort this.captionPackets_.forEach(function(elem, idx) { elem.presortIndex = idx; }); // sort caption byte-pairs based on their PTS values this.captionPackets_.sort(function(a, b) { if (a.pts === b.pts) { return a.presortIndex - b.presortIndex; } return a.pts - b.pts; }); this.captionPackets_.forEach(function(packet) { if (packet.type < 2) { // Dispatch packet to the right Cea608Stream this.dispatchCea608Packet(packet); } // this is where an 'else' would go for a dispatching packets // to a theoretical Cea708Stream that handles SERVICEn data }, this); this.captionPackets_.length = 0; this.ccStreams_.forEach(function(cc) { cc.flush(); }, this); return; }; CaptionStream.prototype.reset = function() { this.latestDts_ = null; this.ignoreNextEqualDts_ = false; this.numSameDts_ = 0; this.activeCea608Channel_ = [null, null]; this.ccStreams_.forEach(function(ccStream) { ccStream.reset(); }); }; CaptionStream.prototype.dispatchCea608Packet = function(packet) { // NOTE: packet.type is the CEA608 field if (this.setsChannel1Active(packet)) { this.activeCea608Channel_[packet.type] = 0; } else if (this.setsChannel2Active(packet)) { this.activeCea608Channel_[packet.type] = 1; } if (this.activeCea608Channel_[packet.type] === null) { // If we haven't received anything to set the active channel, discard the // data; we don't want jumbled captions return; } this.ccStreams_[(packet.type << 1) + this.activeCea608Channel_[packet.type]].push(packet); }; CaptionStream.prototype.setsChannel1Active = function(packet) { return ((packet.ccData & 0x7800) === 0x1000); }; CaptionStream.prototype.setsChannel2Active = function(packet) { return ((packet.ccData & 0x7800) === 0x1800); }; // ---------------------- // Session to Application // ---------------------- var CHARACTER_TRANSLATION = { 0x2a: 0xe1, // á 0x5c: 0xe9, // é 0x5e: 0xed, // í 0x5f: 0xf3, // ó 0x60: 0xfa, // ú 0x7b: 0xe7, // ç 0x7c: 0xf7, // ÷ 0x7d: 0xd1, // Ñ 0x7e: 0xf1, // ñ 0x7f: 0x2588, // █ 0x0130: 0xae, // ® 0x0131: 0xb0, // ° 0x0132: 0xbd, // ½ 0x0133: 0xbf, // ¿ 0x0134: 0x2122, // ™ 0x0135: 0xa2, // ¢ 0x0136: 0xa3, // £ 0x0137: 0x266a, // ♪ 0x0138: 0xe0, // à 0x0139: 0xa0, // 0x013a: 0xe8, // è 0x013b: 0xe2, // â 0x013c: 0xea, // ê 0x013d: 0xee, // î 0x013e: 0xf4, // ô 0x013f: 0xfb, // û 0x0220: 0xc1, // Á 0x0221: 0xc9, // É 0x0222: 0xd3, // Ó 0x0223: 0xda, // Ú 0x0224: 0xdc, // Ü 0x0225: 0xfc, // ü 0x0226: 0x2018, // ‘ 0x0227: 0xa1, // ¡ 0x0228: 0x2a, // * 0x0229: 0x27, // ' 0x022a: 0x2014, // — 0x022b: 0xa9, // © 0x022c: 0x2120, // ℠ 0x022d: 0x2022, // • 0x022e: 0x201c, // “ 0x022f: 0x201d, // ” 0x0230: 0xc0, // À 0x0231: 0xc2, // Â 0x0232: 0xc7, // Ç 0x0233: 0xc8, // È 0x0234: 0xca, // Ê 0x0235: 0xcb, // Ë 0x0236: 0xeb, // ë 0x0237: 0xce, // Î 0x0238: 0xcf, // Ï 0x0239: 0xef, // ï 0x023a: 0xd4, // Ô 0x023b: 0xd9, // Ù 0x023c: 0xf9, // ù 0x023d: 0xdb, // Û 0x023e: 0xab, // « 0x023f: 0xbb, // » 0x0320: 0xc3, // Ã 0x0321: 0xe3, // ã 0x0322: 0xcd, // Í 0x0323: 0xcc, // Ì 0x0324: 0xec, // ì 0x0325: 0xd2, // Ò 0x0326: 0xf2, // ò 0x0327: 0xd5, // Õ 0x0328: 0xf5, // õ 0x0329: 0x7b, // { 0x032a: 0x7d, // } 0x032b: 0x5c, // \ 0x032c: 0x5e, // ^ 0x032d: 0x5f, // _ 0x032e: 0x7c, // | 0x032f: 0x7e, // ~ 0x0330: 0xc4, // Ä 0x0331: 0xe4, // ä 0x0332: 0xd6, // Ö 0x0333: 0xf6, // ö 0x0334: 0xdf, // ß 0x0335: 0xa5, // ¥ 0x0336: 0xa4, // ¤ 0x0337: 0x2502, // │ 0x0338: 0xc5, // Å 0x0339: 0xe5, // å 0x033a: 0xd8, // Ø 0x033b: 0xf8, // ø 0x033c: 0x250c, // ┌ 0x033d: 0x2510, // ┐ 0x033e: 0x2514, // └ 0x033f: 0x2518 // ┘ }; var getCharFromCode = function(code) { if (code === null) { return ''; } code = CHARACTER_TRANSLATION[code] || code; return String.fromCharCode(code); }; // the index of the last row in a CEA-608 display buffer var BOTTOM_ROW = 14; // This array is used for mapping PACs -> row #, since there's no way of // getting it through bit logic. var ROWS = [0x1100, 0x1120, 0x1200, 0x1220, 0x1500, 0x1520, 0x1600, 0x1620, 0x1700, 0x1720, 0x1000, 0x1300, 0x1320, 0x1400, 0x1420]; // CEA-608 captions are rendered onto a 34x15 matrix of character // cells. The "bottom" row is the last element in the outer array. var createDisplayBuffer = function() { var result = [], i = BOTTOM_ROW + 1; while (i--) { result.push(''); } return result; }; var Cea608Stream = function(field, dataChannel) { Cea608Stream.prototype.init.call(this); this.field_ = field || 0; this.dataChannel_ = dataChannel || 0; this.name_ = 'CC' + (((this.field_ << 1) | this.dataChannel_) + 1); this.setConstants(); this.reset(); this.push = function(packet) { var data, swap, char0, char1, text; // remove the parity bits data = packet.ccData & 0x7f7f; // ignore duplicate control codes; the spec demands they're sent twice if (data === this.lastControlCode_) { this.lastControlCode_ = null; return; } // Store control codes if ((data & 0xf000) === 0x1000) { this.lastControlCode_ = data; } else if (data !== this.PADDING_) { this.lastControlCode_ = null; } char0 = data >>> 8; char1 = data & 0xff; if (data === this.PADDING_) { return; } else if (data === this.RESUME_CAPTION_LOADING_) { this.mode_ = 'popOn'; } else if (data === this.END_OF_CAPTION_) { this.clearFormatting(packet.pts); // if a caption was being displayed, it's gone now this.flushDisplayed(packet.pts); // flip memory swap = this.displayed_; this.displayed_ = this.nonDisplayed_; this.nonDisplayed_ = swap; // start measuring the time to display the caption this.startPts_ = packet.pts; } else if (data === this.ROLL_UP_2_ROWS_) { this.topRow_ = BOTTOM_ROW - 1; this.mode_ = 'rollUp'; } else if (data === this.ROLL_UP_3_ROWS_) { this.topRow_ = BOTTOM_ROW - 2; this.mode_ = 'rollUp'; } else if (data === this.ROLL_UP_4_ROWS_) { this.topRow_ = BOTTOM_ROW - 3; this.mode_ = 'rollUp'; } else if (data === this.CARRIAGE_RETURN_) { this.clearFormatting(packet.pts); this.flushDisplayed(packet.pts); this.shiftRowsUp_(); this.startPts_ = packet.pts; } else if (data === this.BACKSPACE_) { if (this.mode_ === 'popOn') { this.nonDisplayed_[BOTTOM_ROW] = this.nonDisplayed_[BOTTOM_ROW].slice(0, -1); } else { this.displayed_[BOTTOM_ROW] = this.displayed_[BOTTOM_ROW].slice(0, -1); } } else if (data === this.ERASE_DISPLAYED_MEMORY_) { this.flushDisplayed(packet.pts); this.displayed_ = createDisplayBuffer(); } else if (data === this.ERASE_NON_DISPLAYED_MEMORY_) { this.nonDisplayed_ = createDisplayBuffer(); } else if (data === this.RESUME_DIRECT_CAPTIONING_) { this.mode_ = 'paintOn'; // Append special characters to caption text } else if (this.isSpecialCharacter(char0, char1)) { // Bitmask char0 so that we can apply character transformations // regardless of field and data channel. // Then byte-shift to the left and OR with char1 so we can pass the // entire character code to `getCharFromCode`. char0 = (char0 & 0x03) << 8; text = getCharFromCode(char0 | char1); this[this.mode_](packet.pts, text); this.column_++; // Append extended characters to caption text } else if (this.isExtCharacter(char0, char1)) { // Extended characters always follow their "non-extended" equivalents. // IE if a "è" is desired, you'll always receive "eè"; non-compliant // decoders are supposed to drop the "è", while compliant decoders // backspace the "e" and insert "è". // Delete the previous character if (this.mode_ === 'popOn') { this.nonDisplayed_[this.row_] = this.nonDisplayed_[this.row_].slice(0, -1); } else { this.displayed_[BOTTOM_ROW] = this.displayed_[BOTTOM_ROW].slice(0, -1); } // Bitmask char0 so that we can apply character transformations // regardless of field and data channel. // Then byte-shift to the left and OR with char1 so we can pass the // entire character code to `getCharFromCode`. char0 = (char0 & 0x03) << 8; text = getCharFromCode(char0 | char1); this[this.mode_](packet.pts, text); this.column_++; // Process mid-row codes } else if (this.isMidRowCode(char0, char1)) { // Attributes are not additive, so clear all formatting this.clearFormatting(packet.pts); // According to the standard, mid-row codes // should be replaced with spaces, so add one now this[this.mode_](packet.pts, ' '); this.column_++; if ((char1 & 0xe) === 0xe) { this.addFormatting(packet.pts, ['i']); } if ((char1 & 0x1) === 0x1) { this.addFormatting(packet.pts, ['u']); } // Detect offset control codes and adjust cursor } else if (this.isOffsetControlCode(char0, char1)) { // Cursor position is set by indent PAC (see below) in 4-column // increments, with an additional offset code of 1-3 to reach any // of the 32 columns specified by CEA-608. So all we need to do // here is increment the column cursor by the given offset. this.column_ += (char1 & 0x03); // Detect PACs (Preamble Address Codes) } else if (this.isPAC(char0, char1)) { // There's no logic for PAC -> row mapping, so we have to just // find the row code in an array and use its index :( var row = ROWS.indexOf(data & 0x1f20); if (row !== this.row_) { // formatting is only persistent for current row this.clearFormatting(packet.pts); this.row_ = row; } // All PACs can apply underline, so detect and apply // (All odd-numbered second bytes set underline) if ((char1 & 0x1) && (this.formatting_.indexOf('u') === -1)) { this.addFormatting(packet.pts, ['u']); } if ((data & 0x10) === 0x10) { // We've got an indent level code. Each successive even number // increments the column cursor by 4, so we can get the desired // column position by bit-shifting to the right (to get n/2) // and multiplying by 4. this.column_ = ((data & 0xe) >> 1) * 4; } if (this.isColorPAC(char1)) { // it's a color code, though we only support white, which // can be either normal or italicized. white italics can be // either 0x4e or 0x6e depending on the row, so we just // bitwise-and with 0xe to see if italics should be turned on if ((char1 & 0xe) === 0xe) { this.addFormatting(packet.pts, ['i']); } } // We have a normal character in char0, and possibly one in char1 } else if (this.isNormalChar(char0)) { if (char1 === 0x00) { char1 = null; } text = getCharFromCode(char0); text += getCharFromCode(char1); this[this.mode_](packet.pts, text); this.column_ += text.length; } // finish data processing }; }; Cea608Stream.prototype = new Stream(); // Trigger a cue point that captures the current state of the // display buffer Cea608Stream.prototype.flushDisplayed = function(pts) { var content = this.displayed_ // remove spaces from the start and end of the string .map(function(row) { return row.trim(); }) // combine all text rows to display in one cue .join('\n') // and remove blank rows from the start and end, but not the middle .replace(/^\n+|\n+$/g, ''); if (content.length) { this.trigger('data', { startPts: this.startPts_, endPts: pts, text: content, stream: this.name_ }); } }; /** * Zero out the data, used for startup and on seek */ Cea608Stream.prototype.reset = function() { this.mode_ = 'popOn'; // When in roll-up mode, the index of the last row that will // actually display captions. If a caption is shifted to a row // with a lower index than this, it is cleared from the display // buffer this.topRow_ = 0; this.startPts_ = 0; this.displayed_ = createDisplayBuffer(); this.nonDisplayed_ = createDisplayBuffer(); this.lastControlCode_ = null; // Track row and column for proper line-breaking and spacing this.column_ = 0; this.row_ = BOTTOM_ROW; // This variable holds currently-applied formatting this.formatting_ = []; }; /** * Sets up control code and related constants for this instance */ Cea608Stream.prototype.setConstants = function() { // The following attributes have these uses: // ext_ : char0 for mid-row codes, and the base for extended // chars (ext_+0, ext_+1, and ext_+2 are char0s for // extended codes) // control_: char0 for control codes, except byte-shifted to the // left so that we can do this.control_ | CONTROL_CODE // offset_: char0 for tab offset codes // // It's also worth noting that control codes, and _only_ control codes, // differ between field 1 and field2. Field 2 control codes are always // their field 1 value plus 1. That's why there's the "| field" on the // control value. if (this.dataChannel_ === 0) { this.BASE_ = 0x10; this.EXT_ = 0x11; this.CONTROL_ = (0x14 | this.field_) << 8; this.OFFSET_ = 0x17; } else if (this.dataChannel_ === 1) { this.BASE_ = 0x18; this.EXT_ = 0x19; this.CONTROL_ = (0x1c | this.field_) << 8; this.OFFSET_ = 0x1f; } // Constants for the LSByte command codes recognized by Cea608Stream. This // list is not exhaustive. For a more comprehensive listing and semantics see // http://www.gpo.gov/fdsys/pkg/CFR-2010-title47-vol1/pdf/CFR-2010-title47-vol1-sec15-119.pdf // Padding this.PADDING_ = 0x0000; // Pop-on Mode this.RESUME_CAPTION_LOADING_ = this.CONTROL_ | 0x20; this.END_OF_CAPTION_ = this.CONTROL_ | 0x2f; // Roll-up Mode this.ROLL_UP_2_ROWS_ = this.CONTROL_ | 0x25; this.ROLL_UP_3_ROWS_ = this.CONTROL_ | 0x26; this.ROLL_UP_4_ROWS_ = this.CONTROL_ | 0x27; this.CARRIAGE_RETURN_ = this.CONTROL_ | 0x2d; // paint-on mode (not supported) this.RESUME_DIRECT_CAPTIONING_ = this.CONTROL_ | 0x29; // Erasure this.BACKSPACE_ = this.CONTROL_ | 0x21; this.ERASE_DISPLAYED_MEMORY_ = this.CONTROL_ | 0x2c; this.ERASE_NON_DISPLAYED_MEMORY_ = this.CONTROL_ | 0x2e; }; /** * Detects if the 2-byte packet data is a special character * * Special characters have a second byte in the range 0x30 to 0x3f, * with the first byte being 0x11 (for data channel 1) or 0x19 (for * data channel 2). * * @param {Integer} char0 The first byte * @param {Integer} char1 The second byte * @return {Boolean} Whether the 2 bytes are an special character */ Cea608Stream.prototype.isSpecialCharacter = function(char0, char1) { return (char0 === this.EXT_ && char1 >= 0x30 && char1 <= 0x3f); }; /** * Detects if the 2-byte packet data is an extended character * * Extended characters have a second byte in the range 0x20 to 0x3f, * with the first byte being 0x12 or 0x13 (for data channel 1) or * 0x1a or 0x1b (for data channel 2). * * @param {Integer} char0 The first byte * @param {Integer} char1 The second byte * @return {Boolean} Whether the 2 bytes are an extended character */ Cea608Stream.prototype.isExtCharacter = function(char0, char1) { return ((char0 === (this.EXT_ + 1) || char0 === (this.EXT_ + 2)) && (char1 >= 0x20 && char1 <= 0x3f)); }; /** * Detects if the 2-byte packet is a mid-row code * * Mid-row codes have a second byte in the range 0x20 to 0x2f, with * the first byte being 0x11 (for data channel 1) or 0x19 (for data * channel 2). * * @param {Integer} char0 The first byte * @param {Integer} char1 The second byte * @return {Boolean} Whether the 2 bytes are a mid-row code */ Cea608Stream.prototype.isMidRowCode = function(char0, char1) { return (char0 === this.EXT_ && (char1 >= 0x20 && char1 <= 0x2f)); }; /** * Detects if the 2-byte packet is an offset control code * * Offset control codes have a second byte in the range 0x21 to 0x23, * with the first byte being 0x17 (for data channel 1) or 0x1f (for * data channel 2). * * @param {Integer} char0 The first byte * @param {Integer} char1 The second byte * @return {Boolean} Whether the 2 bytes are an offset control code */ Cea608Stream.prototype.isOffsetControlCode = function(char0, char1) { return (char0 === this.OFFSET_ && (char1 >= 0x21 && char1 <= 0x23)); }; /** * Detects if the 2-byte packet is a Preamble Address Code * * PACs have a first byte in the range 0x10 to 0x17 (for data channel 1) * or 0x18 to 0x1f (for data channel 2), with the second byte in the * range 0x40 to 0x7f. * * @param {Integer} char0 The first byte * @param {Integer} char1 The second byte * @return {Boolean} Whether the 2 bytes are a PAC */ Cea608Stream.prototype.isPAC = function(char0, char1) { return (char0 >= this.BASE_ && char0 < (this.BASE_ + 8) && (char1 >= 0x40 && char1 <= 0x7f)); }; /** * Detects if a packet's second byte is in the range of a PAC color code * * PAC color codes have the second byte be in the range 0x40 to 0x4f, or * 0x60 to 0x6f. * * @param {Integer} char1 The second byte * @return {Boolean} Whether the byte is a color PAC */ Cea608Stream.prototype.isColorPAC = function(char1) { return ((char1 >= 0x40 && char1 <= 0x4f) || (char1 >= 0x60 && char1 <= 0x7f)); }; /** * Detects if a single byte is in the range of a normal character * * Normal text bytes are in the range 0x20 to 0x7f. * * @param {Integer} char The byte * @return {Boolean} Whether the byte is a normal character */ Cea608Stream.prototype.isNormalChar = function(char) { return (char >= 0x20 && char <= 0x7f); }; // Adds the opening HTML tag for the passed character to the caption text, // and keeps track of it for later closing Cea608Stream.prototype.addFormatting = function(pts, format) { this.formatting_ = this.formatting_.concat(format); var text = format.reduce(function(text, format) { return text + '<' + format + '>'; }, ''); this[this.mode_](pts, text); }; // Adds HTML closing tags for current formatting to caption text and // clears remembered formatting Cea608Stream.prototype.clearFormatting = function(pts) { if (!this.formatting_.length) { return; } var text = this.formatting_.reverse().reduce(function(text, format) { return text + '</' + format + '>'; }, ''); this.formatting_ = []; this[this.mode_](pts, text); }; // Mode Implementations Cea608Stream.prototype.popOn = function(pts, text) { var baseRow = this.nonDisplayed_[this.row_]; // buffer characters baseRow += text; this.nonDisplayed_[this.row_] = baseRow; }; Cea608Stream.prototype.rollUp = function(pts, text) { var baseRow = this.displayed_[BOTTOM_ROW]; baseRow += text; this.displayed_[BOTTOM_ROW] = baseRow; }; Cea608Stream.prototype.shiftRowsUp_ = function() { var i; // clear out inactive rows for (i = 0; i < this.topRow_; i++) {