av/parser/h264.go

/*
NAME
  h264.go

DESCRIPTION
  See Readme.md

AUTHOR
  Saxon Nelson-Milton <saxon@ausocean.org>

LICENSE
  h264.go is Copyright (C) 2017 the Australian Ocean Lab (AusOcean)

  It is free software: you can redistribute it and/or modify them
  under the terms of the GNU General Public License as published by the
  Free Software Foundation, either version 3 of the License, or (at your
  option) any later version.

  It is distributed in the hope that it will be useful, but WITHOUT
  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
 for more details.

  You should have received a copy of the GNU General Public License
  along with revid in gpl.txt.  If not, see http://www.gnu.org/licenses.
*/
package parser

import (
	"time"

	"bitbucket.org/ausocean/av/itut"
)

const (
	inputChanSize    = 100000
	outputBufferSize = 10000
)

// h264Parser provides properties and methods to allow for the parsing of a
// h264 stream - i.e. to allow extraction of the individual access units
type h264Parser struct {
	inputBuffer         []byte
	isParsing           bool
	parserOutputChanRef chan []byte
	userOutputChanRef   chan []byte
	inputChan           chan byte
	delay               uint
}

// NewH264Parser returns an instance of the h264Parser struct
func NewH264Parser() (p *h264Parser) {
	p = new(h264Parser)
	p.isParsing = true
	p.inputChan = make(chan byte, inputChanSize)
	p.delay = 0
	return
}

// Stop simply sets the isParsing flag to false to indicate to the parser that
// we don't want to interpret incoming data anymore - this will also make the
// parser jump out of the parse func
func (p *h264Parser) Stop() {
	p.isParsing = false
}

// Start starts the parse func as a goroutine so that incoming data is interpreted
func (p *h264Parser) Start() {
	p.isParsing = true
	go p.parse()
}

// SetDelay sets a delay inbetween each buffer output. Useful if we're parsing
// a file but want to replicate the speed of incoming video frames from a
// camera
func (p *h264Parser) SetDelay(delay uint) {
	p.delay = delay
}

// InputChan returns a handle to the input channel of the parser
func (p *h264Parser) InputChan() chan byte {
	return p.inputChan
}

// OutputChan returns a handle to the output chan of the parser
func (p *h264Parser) OutputChan() <-chan []byte {
	return p.userOutputChanRef
}

// SetOutputChan sets the parser output chan to the passed output chan. This is
// useful if we want the parser output to go directly to a generator of some sort
// for packetization.
func (p *h264Parser) SetOutputChan(o chan []byte) {
	p.parserOutputChanRef = o
	p.userOutputChanRef = o
}

// parse interprets an incoming h264 stream and extracts individual frames
// aka access units
func (p *h264Parser) parse() {
	outputBuffer := make([]byte, 0, outputBufferSize)
	searchingForEnd := false
	for p.isParsing {
		var aByte uint8
		if p.isParsing {
			aByte = <-p.inputChan
		} else {
			return
		}

		outputBuffer = append(outputBuffer, aByte)
		for i := 1; aByte == 0x00 && i != 4; i++ {
			if p.isParsing {
				aByte = <-p.inputChan
			} else {
				return
			}
			outputBuffer = append(outputBuffer, aByte)
			if (aByte == 0x01 && i == 2) || (aByte == 0x01 && i == 3) {
				if searchingForEnd {
					output := append(append(itut.StartCode1(), itut.AUD()...), outputBuffer[:len(outputBuffer)-(i+1)]...)
					time.Sleep(time.Duration(p.delay) * time.Millisecond)
					p.parserOutputChanRef <- output
					outputBuffer = outputBuffer[len(outputBuffer)-1-i:]
					searchingForEnd = false
				}
				if p.isParsing {
					aByte = <-p.inputChan
				} else {
					return
				}
				outputBuffer = append(outputBuffer, aByte)
				if nalType := aByte & 0x1F; nalType == 1 || nalType == 5 || nalType == 8 {
					searchingForEnd = true
				}
			}
		}
	}
}