stream-proxy

#!/usr/bin/python3
#
# stream-proxy - a CGI script to act as a proxy for a streaming service

import os
import sys
import socket
import select

debug = False

# comment out the second line to turn logging off
logfile = None
logfile = "/home/tbird/work/labcontrol/lc-data/stream-proxy.log"

def do_log(data):
    if logfile:
        with open(logfile, "ab") as fd:
            fd.write(data)
            fd.flush()

# specify time in seconds to wait for streamer
# note: 0 = socket is non-blocking
# non-zero = socket is in timeout mode (nonblocking, with a timeout)
# None = socket is in blocking mode
proxy_connection_timeout = "10"

# max amount of data to read at a time from the streaming server
max_data = 8192

SERVER_CONF_FILENAME = "/etc/lcserver.conf"

# read config file variables as strings into the global name space
# hacky, but it's quick and it works
lines = open(SERVER_CONF_FILENAME, "r").readlines()
for line in lines:
    if not line or line.startswith("#"):
        continue
    try:
        name, value = line.split("=", 1)
        globals()[name] = value.strip()
    except:
        continue

PATH_INFO=os.environ.get("PATH_INFO", "")
# use the last element of the path as the stream identifier
# This means that http://../stream-proxy/one/one is equivalent
# to http://../stream-proxy/one/ - but I don't care
stream_id=PATH_INFO.split("/")[-1]

# read stream-proxy config file
proxy_data_file = base_dir + "/stream-proxies.dat"
try:
    fd = open(base_dir+"/stream-proxies.dat", "r")
except:
    err_msg = "Error: proxy data file is missing"
    print("Content-type: text\html\n")
    print(err_msg)
    sys.exit(0)

stream_url = None
for line in fd.readlines():
    if not line or line.startswith("#"):
        continue
    try:
        id_name, value = line.split("=", 1)
        if id_name == stream_id:
            stream_url = value.strip()
    except:
        continue

if not stream_url:
    err_msg = "Error: Could not find proxy for stream_id '%s'" % stream_id
    print("Content-type: text\html\n")
    print(err_msg)
    sys.exit(0)

if stream_id == "debug":
    debug = True
    stream_url = "http://www.google.com/"

# parse the proxied_stream
method, rest = stream_url.split("://", 1)
addr, path = rest.split("/", 1)
path = "/" + path
if ":" in addr:
    server, port = addr.split(":", 1)
    port = int(port)
else:
    server = addr
    port = 80

# create the request
request = b"GET %s HTTP/1.1\r\n" % path.encode("utf-8")
# test 2
request += b"""Host: %s\r
Connection: keep-alive\r
User-Agent: STREAM-PROXY/1.0\r
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avig,image/webp,image/apng,*/*;q=0.8\r
Accept-Language: en-US,en;q=0.9\r
\r
\r
""" % server.encode("utf-8")

# attempt at a correct-syntax header
request += b"""Host: %s\r
User-Agent: STREAM-PROXY/1.0\r
Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avig,image/webp,image/apng,*/*;q=0.8\r
Accept-Language: en-US,en;q=0.9\r
\r
\r
"""

request = b"GET %s HTTP/1.1\r\nHost: %s\r\n\r\n\r\n" % (path.encode("utf-8"), server.encode("utf-8"))

do_log(b"==================================================\n")
do_log(b"request='%s'\n" % request)

if debug:
    print("Content-type: text\html\n")
    print("Hello world - from stream-proxy\n<p>\n");
    print("PATH_INFO=%s<BR>" % PATH_INFO)
    print("stream_id=%s<BR>" % stream_id)
    if not proxied_stream:
        print("Could not find stream_url for stream_id %s" % stream_id)
    else:
        print("stream_url=%s<BR>" % stream_url)
else:
    # read the data from the streamer, and return it unmodified
    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)

    # set duration of connection timeout
    #s.settimeout(float(proxy_connection_timeout))
    # use blocking mode, for now
    s.settimeout(None)
    s.connect((server, port))
    s.sendall(request)
    s.shutdown(socket.SHUT_WR)

    while True:
        try:
            data = s.recv(max_data)
        except TimeoutError:
            break

# partial code to handle non-blocking recvs
#        except socket.error, e:
#            err = e.args[0]
#            if err == errno.EAGAIN or err == errno.EWOULDBLOCK:
#                sleep(1)
#                do_log("got not data this time")
#            else:
#                break

        #do_log(b"## received data of length %d\n" % len(data))

        # check for and strip HTTP headers
        if data.startswith(b"HTTP"):
            in_headers = True
            #sys.stdout.buffer.write(b"Content-type: text/html\n\n")
            #sys.stdout.buffer.write(b"## got the following data:")
            #do_log(b"## data:\n")
            #do_log(data)

            #sys.stdout.buffer.write(b"<BR> ## outputting the following response:")

            lines = data.split(b"\n")
            for line in lines:
                #do_log(b"line=%s\n" % line)
                if line.startswith(b"HTTP"):
                    continue
                if line.startswith(b"Connection:"):
                    # should handle this?
                    continue
                if line.startswith(b"Content-Length:"):
                    # should handle this?
                    continue
                if line.startswith(b"Content-Type:"):
                    do_log(b"## found '%s'\n" % line)
                    content_type = line
                    continue
                # FIXTHIS - what about cookies?
                if in_headers and not line.strip():
                    do_log(b"## found blank link in headers\n")
                    in_headers = False
                    sys.stdout.buffer.write(content_type + b"\n\n")
                    continue
                if not in_headers:
                    sys.stdout.buffer.write(line+b"\n")
                else:
                    do_log(b"still in header\n")

            # done with this block, recv some more
            continue

        do_log(data)

        if len(data) == 0:
            break

        sys.stdout.buffer.write(data)

sys.stdout.buffer.flush()
sys.exit(0)