C: Question about Beej's networking guide... Is there an assumption here?

I was just going through the Networking Guide by Beej and am curious about this part of the code (specifically marked with "From here" and "To here"):

// main loop
    for(;;) {
        read_fds = master; // copy it
        if (select(fdmax+1, &read_fds, NULL, NULL, NULL) == -1) {
            perror("select");
            exit(4);
        }

        // run through the existing connections looking for data to read
        for(i = 0; i <= fdmax; i++) {
            if (FD_ISSET(i, &read_fds)) { // we got one!!
                if (i == listener) {
                    // handle new connections
                    addrlen = sizeof remoteaddr;
                    newfd = accept(listener,
                        (struct sockaddr *)&remoteaddr,
                        &addrlen);

                    if (newfd == -1) {
                        perror("accept");
                    } else {
                        FD_SET(newfd, &master); // add to master set
                        if (newfd > fdmax) {    // keep track of the max
                            fdmax = newfd;
                        }
                        printf("selectserver: new connection from %s on "
                            "socket %d\n",
                            inet_ntop(remoteaddr.ss_family,
                                get_in_addr((struct sockaddr*)&remoteaddr),
                                remoteIP, INET6_ADDRSTRLEN),
                            newfd);
                    }
                } else {
                    // handle data from a client
                    //----------------- FROM HERE --------------------------
                    if ((nbytes = recv(i, buf, sizeof buf, 0)) <= 0) {
                        // got error or connection closed by client
                        if (nbytes == 0) {
                            // connection closed
                            printf("selectserver: socket %d hung up\n", i);
                        } else {
                            perror("recv");
                        }
                        close(i); // bye!
                        FD_CLR(i, &master); // remove from master set
                    //----------------- TO HERE ----------------------------
  开发者_如何学编程                  } else {
                        // we got some data from a client
                        for(j = 0; j <= fdmax; j++) {
                            // send to everyone!
                            if (FD_ISSET(j, &master)) {
                                // except the listener and ourselves
                                if (j != listener && j != i) {
                                    if (send(j, buf, nbytes, 0) == -1) {
                                        perror("send");
                                    }
                                }
                            }
                        }
                    }
                } // END handle data from client
            } // END got new incoming connection
        } // END looping through file descriptors
    } // END for(;;)--and you thought it would never end!

    return 0;

Now I know that read doesn't always read "everything" that is to be read on a socket and that it sometimes can return only part of it. In that case, wouldn't this code be incorrect? I mean, after one read, the connection is being closed... Instead, aren't we supposed to have some other mechanism in place? If so, what is the right approach here?

The socket is only going to get closed there if there was an error from recv(), otherwise it'll deal with the data that was read even if it isnt all read. It will then read more out when it loops through again. Pretty sure this is what you're asking?

Yes you would keep reading until you got all the data you expected, obviosuly you need someway of knowing how much to expect - which is why http puts the document size first

Your only calling close when recv() has returned a negative value which means that recv had some sort of error. Notice that the block where you do the close has a comment stating // got error or connection closed by client).

When you actually get some data (the else branch starting with // we got some data from a client), the connection is not being closed.

You are right that you can't assume the data arrives all at one time. Your mistake is in following how the code is working.