FLV 分析脚本

一、需求

　　通过脚本，可以检查本地flv文件格式是否正确，可以打印每个Tag中的二进制内容

二、效果

　　可以看到VideoTag中开始处增加了一段SEI数据，并且可以看到部分字段，gameid、time、network

三、实施

　　1）为了不重复造轮子，利用python下面的flvlib库

　　　　可以通过 pip install flvlib来安装，需要注意的是这个库工作在python2下面的环境

　　2）写一段代码，可以打印出flv中所有的tag

import sys
import logging
 
from optparse import OptionParser
 
from flvlib import __versionstr__
from flvlib import tags
from flvlib import helpers
from flvlib.astypes import MalformedFLV
from array import array
import HexTest
 
 
class Packet:
 
    def __init__(self, f, offset, size):
        self.f = f
        self.offset = offset
        self.size = size
        self.data = None
        self.load()
 
 
    def load(self):
        cur_offset = f.tell()
        f.seek(self.offset, 0)
        self.data = f.read(self.size)
        f.seek(cur_offset, 0)
        self.prety_dump()
 
 
    def prety_dump(self):
        HexTest.dump_memory(array('B', self.data), self.offset)
 
 
 
 
 
 
 
 
if __name__ == '__main__':
    filename = "test4.flv"
    with open(filename, 'rb') as f:
        open(filename, 'rb')
 
        flv = tags.FLV(f)
 
        try:
            tag_generator = flv.iter_tags()
            print(flv.f.tell())
            for i, tag in enumerate(tag_generator):
                # Print the tag information
                print("#%05d %s" % (i + 1, tag))
                # Print the content of onMetaData tags
                if (isinstance(tag, tags.ScriptTag)
                    and tag.name == "onMetaData"):
                    helpers.pprint(tag.variable)
                print(str(type(tag)))
                if isinstance(tag, tags.VideoTag) and (tag.frame_type == 1):
                        p = Packet(flv.f, tag.offset, tag.size)
 
 
        except MalformedFLV as e:
            message = e[0] % e[1:]
            print("The file `%s' is not a valid FLV file: %s",
                  filename, message)
        except tags.EndOfFile:
            print("Unexpected end of file on file `%s'", filename)

　　　　main中的方法能够输出所有Tag，同时能够检查FLV文件是否正常

#00672 <AudioTag at offset 0x001DD15B, time 9842, size 163, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00673 <VideoTag at offset 0x001DD20D, time 9855, size 11404, H.264 (interframe), NAL unit>
<class 'flvlib.tags.VideoTag'>
#00674 <AudioTag at offset 0x001DFEA8, time 9863, size 162, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00675 <AudioTag at offset 0x001DFF59, time 9884, size 168, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00676 <VideoTag at offset 0x001E0010, time 9903, size 6748, H.264 (interframe), NAL unit>
<class 'flvlib.tags.VideoTag'>
#00677 <AudioTag at offset 0x001E1A7B, time 9906, size 166, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00678 <AudioTag at offset 0x001E1B30, time 9927, size 159, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00679 <AudioTag at offset 0x001E1BDE, time 9948, size 162, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00680 <VideoTag at offset 0x001E1C8F, time 9950, size 8690, H.264 (interframe), NAL unit>
<class 'flvlib.tags.VideoTag'>
#00681 <AudioTag at offset 0x001E3E90, time 9970, size 166, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00682 <AudioTag at offset 0x001E3F45, time 9991, size 164, AAC, raw>
<class 'flvlib.tags.AudioTag'>
#00683 <VideoTag at offset 0x001E3FF8, time 9998, size 20543, H.264 (keyframe), NAL unit>

　　同时将每个Tag对应的开始位置和大小确定之后，读入内存中转成字节数组，并且格式化输出

import sys
import logging
import struct
from array import array
 
 
def dump_newline():
    sys.stdout.write("\n")
 
def dump_with_space(str):
    sys.stdout.write(str + " ")
 
def dump_line_number(x):
    dump_with_space("%010xh:" % (x))
 
def dump_one_byte(x):
    dump_with_space("%02x" % (x))
 
def dump_memory(data, memory_start_offset):
    sys.stdout.flush()
    start_line = (memory_start_offset/16)*16
    for i in range(0, len(data), 16):
        dump_line_number(i + start_line)
        byte_arr = []
        for j in range(0, 16):
            index = i + j
            if index < len(data):
                if (index + start_line) < memory_start_offset:
                    dump_with_space("  ")
                else:
                    byte = data[index]
                    dump_one_byte(byte)
                    byte_arr.append(byte)
        line_str = "; " + ("".join(map(chr, byte_arr))).replace('\n','').replace('\t','').replace('\r','')
        dump_with_space(line_str)
        dump_newline()