Infinite hangup when using CGI sys.stdin.read()

[MRAB]

pacsciadmin at gmail.com wrote:
> I maintain a CGI framework named PyTin, and while previously the
> developer was only allowed to access GET and POST by cgi.FieldStorage.
> I'm trying to add code to my request gatherer that looks like:
> 
> 1. request.rawpost = StringIO() # Just normal StringIO
> 2. request.rawpost.write(sys.stdin.read())
> 3. request.rawpost.seek(0)
> 4. request.fields = cgi.FieldStorage(fp=request.rawpost)
> 5. request.rawpost.seek(0)
> 
> Obviously, my actual code doesn't have the line numbers.
> 
> I'm testing it out on the CGIHTTPRequestHandler with BaseHTTPServer,
> and whenever my test script reaches line 2, it hangs up. It doesn't
> spit out a traceback, or stop in less than a minute, whenever I use
> sys.stdin.read(). Can anyone explain what is going on?
> 
sys.stdin.read() will read until EOF. If stdin is connected to a file 
when the entire file will be read. If stdin is connected to the console 
then it will return when it reaches the end of the input (will that ever 
happen?) or the EOF marker, such as ctrl-Z in Windows or ctrl-D (I 
think) in *nix.