Socket Programming — recv() cannot get all data - c

I am learning socket programming in C language, and this is an incomprehensible problem I encountered during my study.
Today I am trying to send a HTTP request to my test server which host an Apache example website, then receive the response from test server. Here is a part of my receive code.
unsigned long recv_size = 0;
unsigned long response_size = 4096;
int ret = 0;
char *recv_buff = (char *)malloc(response_size);
while (1)
{
// ret = recv(socket, recv_buff, response_size, MSG_WAITALL); // cannot get all data
ret = read(socket, recv_buff, response_size); // same effect as the above
recv_size += ret;
if (ret < 0)
error(strerror(errno));
else if (ret == 0)
break; // all data recved
}
The normal result of my test with burpsuit is this.
But what I received with the C language program was incomplete data.
I searched the reason for one night, but I still did not find a solution for my problem. Whether it is to set the buff to a super large size or any other method, the complete data cannot be accepted at all.
The traffic monitored from wireshark is ok, but my program still cannot receive the complete data. What is the problem?
If you know why, please let me know. THX. (o゜▽゜)o☆
UPDATE
The while loop will execute twice, and first time the value of ret is 3343, and second time is 0, so the loop will stop here.

You can get a short read on a socket.
But, your code to handle that has a few issues.
You're allocating a buffer of size response_size. You are always reading that amount instead of reducing the amount read by the amount you've already read on a prior loop iteration.
This can cause you to read past the end of the buffer causing UB (undefined behavior).
Your termination condition is if (ret == 0). This can fail if another packet arrives "early". You'll never see a ret of 0, because the partial data from the next packet will make it non-zero
Here's the corrected code:
#if 0
unsigned long recv_size = 0;
#endif
unsigned long response_size = 4096;
int ret = 0;
char *recv_buff = (char *) malloc(response_size);
#if 1
unsigned long remaining_size = response_size;
unsigned long offset = 0;
#endif
for (; remaining_size > 0; remaining_size -= ret, offset += ret) {
ret = read(socket, &recv_buff[offset], remaining_size);
if (ret < 0)
error(strerror(errno));
}
UPDATE:
The above code corrects some of the issues. But, for a variable length source [such as http], we don't know how much to read at the outset.
So, we have to parse the headers and look for the "Content-Length" field. This will tell us how much to read.
So, we'd like to have line oriented input for the headers. Or, manage our own buffer
Assuming we can parse that value, we have to wait for the empty line to denote the start of the payload. And, then we can loop on that exact amount.
Here's some code that attempts the header parsing and saving of the payload. I've coded it, but not compiled it. So, you can take it as pseudo code:
unsigned long recv_size = 0;
unsigned long response_size = 4096;
char *recv_buff = malloc(response_size + 1);
// line oriented header buffer
char *endl = NULL;
unsigned long linelen;
char linebuf[1000];
int ret = 0;
// read headers
while (1) {
// fill up a chunk of data
while (recv_size < response_size) {
recv_buff[recv_size] = 0;
// do we have a line end?
endl = strstr(recv_buff,"\r\n");
if (endl != NULL)
break;
ret = read(socket, &recv_buff[recv_size], response_size - recv_size);
if (ret < 0)
error(strerror(errno));
if (ret == 0)
break;
recv_size += ret;
}
// error -- no line end but short read
if (endl == NULL)
error(strerror(errno));
// copy header to work buffer
linelen = endl - recv_buff;
memcpy(linebuf,recv_buff,linelen);
linebuf[linelen] = 0;
// remove header from receive buffer
linelen += 2;
recv_size -= linelen;
if (recv_size > 0)
memcpy(recv_buff,&recv_buff[linelen],recv_size);
// stop on end of headers (back to back "\r\n")
if ((recv_size >= 2) && (recv_buff[0] == '\r') && (recv_buff[1] == '\n')) {
memcpy(recv_buff,&recv_buff[2],recv_size - 2);
recv_size -= 2;
break;
}
// parse line work buffer for keywords ... (e.g.)
content_length = ...;
}
// save payload to file
while (content_length > 0) {
// write out prior payload amount
if (recv_size > 0) {
write(file_fd,recv_buff,recv_size);
content_length -= recv_size;
recv_size = 0;
continue;
}
recv_size = read(socket,recv_buff,response_size);
if (recv_size < 0)
error(strerror(errno));
if (recv_size == 0)
break;
}
UPDATE #2:
Yeah, it hard to make the pseudo code run, and the returned values are all garbled
Okay, here is a soup-to-nuts working version that I've tested against my own http server.
I had to create my own routines for the parts you didn't post (e.g. connect, etc.).
At the core, there might have been a minor tweak to the buffer slide code [it was sliding by an extra 2 bytes in one place], but, otherwise it was pretty close to my previous version
// htprcv/htprcv.c -- HTTP receiver
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <fcntl.h>
#include <errno.h>
#include <error.h>
#include <netdb.h>
#include <arpa/inet.h>
#include <netinet/in.h>
#include <sys/socket.h>
typedef unsigned char byte;
#define HTPSLIDE(_rmlen) \
recv_size = htpslide(recv_buff,recv_size,_rmlen)
#define _dbgprt(_fmt...) \
fprintf(stderr,_fmt)
#if DEBUG || _USE_ZPRT_
#define dbgprt(_lvl,_fmt...) \
do { \
if (dbgok(_lvl)) \
_dbgprt(_fmt); \
} while (0)
#define dbgexec(_lvl,_expr) \
do { \
if (dbgok(_lvl)) \
_expr; \
} while (0)
#else
#define dbgprt(_lvl,_fmt...) \
do { \
} while (0)
#define dbgexec(_lvl,_expr) \
do { \
} while (0)
#endif
#define dbgok(_lvl) \
opt_d[(byte) #_lvl[0]]
byte opt_d[256];
char *opt_o;
#define HEXMAX 16
// htpconn -- do connect to server
int
htpconn(const char *hostname,unsigned short portno)
{
struct addrinfo hints, *res;
struct hostent *hostent;
int ret;
char portstr[20];
int sockfd;
/* Prepare hint (socket address input). */
hostent = gethostbyname(hostname);
if (hostent == NULL)
error(1,errno,"htpconn: gethostbyname -- %s\n",hostname);
memset(&hints, 0, sizeof hints);
hints.ai_family = AF_INET; // ipv4
hints.ai_socktype = SOCK_STREAM; // tcp
hints.ai_flags = AI_PASSIVE; // fill in my IP for me
sprintf(portstr, "%d", portno);
getaddrinfo(NULL, portstr, &hints, &res);
sockfd = socket(res->ai_family, res->ai_socktype, res->ai_protocol);
if (sockfd < 0)
error(1,errno,"htpconn: socket\n");
// do the actual connection
ret = connect(sockfd, res->ai_addr, res->ai_addrlen);
if (ret < 0)
error(1,errno,"htprcv: read header\n");
return sockfd;
}
// htpslide -- slide buffer (strip out processed data)
size_t
htpslide(char *recv_buff,size_t recv_size,int slidelen)
{
size_t new_size;
if (slidelen > recv_size)
slidelen = recv_size;
new_size = recv_size - slidelen;
dbgprt(S,"htpslide: slidelen=%d recv_size=%zu new_size=%zu\n",
slidelen,recv_size,new_size);
memcpy(&recv_buff[0],&recv_buff[slidelen],new_size);
return new_size;
}
// _htphex -- dump a line in hex
void
_htphex(unsigned long off,const void *vp,size_t xlen)
{
const byte *bp = vp;
int idx;
int chr;
char hexbuf[200];
char alfbuf[200];
char *hexptr = hexbuf;
char *alfptr = alfbuf;
for (idx = 0; idx < HEXMAX; ++idx) {
chr = bp[idx];
if ((idx % 4) == 0)
*hexptr++ = ' ';
if (idx < xlen) {
hexptr += sprintf(hexptr,"%2.2X",chr);
if ((chr < 0x20) || (chr > 0x7E))
chr = '.';
}
else {
hexptr += sprintf(hexptr," ");
chr = ' ';
}
*alfptr++ = chr;
}
*hexptr = 0;
*alfptr = 0;
_dbgprt(" %8.8lX: %s *%s*\n",off,hexbuf,alfbuf);
}
// htphex -- dump a buffer in hex
void
htphex(const char *buf,size_t buflen,const char *reason)
{
size_t off = 0;
size_t xlen;
if (reason != NULL)
_dbgprt("htphex: DUMP buf=%p buflen=%zu (from %s)\n",
buf,buflen,reason);
for (; buflen > 0; buflen -= xlen, buf += xlen, off += xlen) {
xlen = buflen;
if (xlen > HEXMAX)
xlen = HEXMAX;
_htphex(off,buf,xlen);
}
}
// htpsym -- get symbol/value
int
htpsym(char *linebuf,char *sym,char *val)
{
char *cp;
int match;
dbgprt(H,"htpsym: PARAM linebuf='%s'\n",linebuf);
// FORMAT:
// foo-bar: baz
do {
match = 0;
cp = strchr(linebuf,':');
if (cp == NULL)
break;
*cp++ = 0;
strcpy(sym,linebuf);
for (; (*cp == ' ') || (*cp == '\t'); ++cp);
strcpy(val,cp);
match = 1;
dbgprt(H,"htpsym: SYMBOL sym='%s' val='%s'\n",sym,val);
} while (0);
return match;
}
// htprcv -- receive server response
void
htprcv(int sockfd,int fdout)
{
size_t recv_size = 0;
size_t response_size = 4096;
char *recv_buff = malloc(response_size + 1);
// line oriented header buffer
char *endl = NULL;
size_t linelen;
char linebuf[1000];
ssize_t ret = 0;
off_t content_length = 0;
// read headers
while (1) {
// fill up a chunk of data
while (recv_size < response_size) {
recv_buff[recv_size] = 0;
// do we have a line end?
endl = strstr(recv_buff,"\r\n");
if (endl != NULL)
break;
// read a chunk of data
ret = read(sockfd,&recv_buff[recv_size],response_size - recv_size);
if (ret < 0)
error(1,errno,"htprcv: read header\n");
if (ret == 0)
break;
recv_size += ret;
dbgprt(R,"htprcv: READ ret=%zd\n",ret);
dbgexec(R,htphex(recv_buff,recv_size,"htprcv/READ"));
}
// error -- no line end but short read
if (endl == NULL)
error(1,0,"htprcv: no endl\n");
// copy header to work buffer
linelen = endl - recv_buff;
memcpy(linebuf,recv_buff,linelen);
linebuf[linelen] = 0;
// remove header from receive buffer
linelen += 2;
HTPSLIDE(linelen);
// stop on end of headers (back to back "\r\n")
if ((recv_size >= 2) &&
(recv_buff[0] == '\r') && (recv_buff[1] == '\n')) {
HTPSLIDE(2);
break;
}
// parse line work buffer for keywords ...
char sym[100];
char val[1000];
if (! htpsym(linebuf,sym,val))
continue;
if (strcasecmp(sym,"Content-Length") == 0) {
content_length = atoi(val);
continue;
}
}
// save payload to file
while (content_length > 0) {
// write out prior payload amount
if (recv_size > 0) {
dbgexec(W,htphex(recv_buff,recv_size,"htprcv/WRITE"));
ret = write(fdout,recv_buff,recv_size);
if (ret < 0)
error(1,errno,"htprcv: write body\n");
content_length -= recv_size;
recv_size = 0;
continue;
}
// read in new chunk of payload
ret = read(sockfd,recv_buff,response_size);
if (ret < 0)
error(1,errno,"htprcv: read body\n");
if (ret == 0)
break;
recv_size = ret;
}
free(recv_buff);
}
// htpget -- do initial dialog
void
htpget(int sockfd,const char *hostname,const char *file)
{
char *bp;
char buf[1024];
ssize_t resid;
ssize_t xlen;
size_t off;
bp = buf;
if (file == NULL)
file = "/";
bp += sprintf(bp,"GET %s HTTP/1.1\r\n",file);
if (hostname == NULL)
hostname = "localhost";
bp += sprintf(bp,"Host: %s\r\n",hostname);
if (0) {
bp += sprintf(bp,"User-Agent: %s\r\n","curl/7.61.1");
}
else {
bp += sprintf(bp,"User-Agent: %s\r\n","htprcv");
}
bp += sprintf(bp,"Accept: */*\r\n");
bp += sprintf(bp,"\r\n");
resid = bp - buf;
off = 0;
for (; resid > 0; resid -= xlen, off += xlen) {
xlen = write(sockfd,buf,resid);
if (xlen < 0)
error(1,errno,"htpget: write error\n");
}
}
// main -- main program
int
main(int argc,char **argv)
{
char *cp;
char *portstr;
unsigned short portno;
int sockfd;
int filefd;
char url[1000];
--argc;
++argv;
//setlinebuf(stdout);
setlinebuf(stderr);
for (; argc > 0; --argc, ++argv) {
cp = *argv;
if (*cp != '-')
break;
cp += 2;
switch(cp[-1]) {
case 'd': // debug options
if (*cp == 0)
cp = "SHRW";
for (; *cp != 0; ++cp)
opt_d[(byte) *cp] = 1;
break;
case 'o': // output file
opt_o = cp;
break;
}
}
// get the remote host:port
do {
if (argc <= 0) {
strcpy(url,"localhost:80");
break;
}
strcpy(url,*argv++);
--argc;
} while (0);
// get remote port number
portstr = strchr(url,':');
if (portstr != NULL)
*portstr++ = 0;
else
portstr = "80";
portno = atoi(portstr);
// open the output file (or send to stdout)
do {
if (opt_o == NULL) {
filefd = 1;
break;
}
filefd = open(opt_o,O_WRONLY | O_CREAT,0644);
if (filefd < 0)
filefd = 1;
} while (0);
// establish connection
sockfd = htpconn(url,portno);
// send the file request
htpget(sockfd,NULL,"/");
// receive the server response
htprcv(sockfd,filefd);
close(sockfd);
return 0;
}

Related

Is there any way to know the amount of bytes send from the client to the server and process the recv() in networks

I am trying to build a chat application between the server and the client. My doubt is for sending information from the client or from the server I was able to handle the partial send with the help of the loop, but I am unable to find out the length of the send data bytes from the client to the server or from the server to the client, thereby having problem in creating the memory for the received bytes and printing.
My chat function code for the client:
int chat_function(int sockfd)
{
char ch;
char *buf;
char *newp;
int ret_send = 0;
int ret_recv = 0;
int buf_size = 0;
while(1) {
printf("From client, enter the message : ");
buf = (char *)malloc(sizeof(char));
if (buf == NULL)
return -1;
while ((ch = getchar()) != '\n') {
buf[buf_size++] = ch;
newp = (char *)realloc(buf, (buf_size + 1) * sizeof(char));
if ( newp == NULL) {
free(buf);
return -1;
}
buf = newp;
}
buf[buf_size] = '\0';
ret_send = send_all(sockfd, buf, buf_size);
if (ret_send == -1)
error(1, errno, "error in send() function call\n");
memset(buf, 0, buf_size);
ret_recv = recv_all(sockfd, buf, buf_size);
if (ret_recv == -1) {
error(1, errno, "error in recv() function call\n");
} else if (ret_recv == -2) {
printf("Oops the server has closed the connection\n");
free(buf);
break;
}
printf("From Server : %s", buf);
if ((strncmp(buf, "exit", 4)) == 0) {
printf("Client Exit...\n");
free(buf);
break;
}
free(buf);
}
}
For handling partial send:
int send_all(int sockfd, char *buf, int buf_size)
{
int bytes_left = 0;
size_t send_bytes = 0;
bytes_left = buf_size
while (1) {
send_bytes = send(fd, buf, bytes_left, 0);
if (send_bytes == -1)
return -1;
buf = buf + send_bytes;
bytes_left = bytes_left - send_bytes;
if (bytes_left == 0)
break;
}
return 0;
}
TCP is a stream protocol, meaning there are no message boundaries: it is just a full-duplex (meaning data flows in both directions at the same time, as if there were two separate lanes) more or less continuous stream of data.
UDP is a datagram protocol, and does have message boundaries. There is an ioctl (FIONREAD/SIOCINQ) that provides the length of the next datagram, but because it involves a syscall, doing that for every message you receive is going to be slow and inefficient. Instead, you normally use a buffer large enough to hold the largest acceptable message, and copy it if/when necessary. However, UDP also has no reliability guarantees, and often UDP datagrams are completely lost without any trace or discernible reason; that's just what happens.
For a chat client-server connection, you'll want to use TCP.
Since the underlying connection is just a stream of data, you need to design a protocol for the communications, so that the stream can be split into messages, with each message processed separately.
The simplest case would be to use the nul character, \0, as a message separator.
The "send" function would then look something like this:
/* Returns 0 if message successfully sent,
nonzero errno code otherwise. */
int send_message(int descriptor, const char *message)
{
/* If message is NULL, we cannot use strlen(); use zero for that. */
const size_t message_len = (message) ? strlen(message) : 0;
/* Temporary variables for the sending part. */
const char *ptr = message;
const char *const end = message + message_len + 1; /* Include '\0' at end */
ssize_t bytes;
/* Check valid descriptor and message length. */
if (descriptor == -1 || message_len < 1)
return errno = EINVAL;
/* Write loop for sending the entire message. */
while (ptr < end) {
bytes = write(descriptor, ptr, (size_t)(end - ptr));
if (bytes > 0) {
ptr += bytes;
} else
if (bytes != -1) {
/* This should never happen. */
return errno = EIO;
} else
if (errno != EINTR) {
/* We do not consider EINTR an actual error; others we do. */
return errno;
}
}
return 0;
}
The above send_message() function writes the specified string, including the string terminating nul character \0, to the specified descriptor.
On the read end, we need a buffer large enough to hold at least one full message. Instead of always waiting for incoming data, we need to check if the buffer already contains a full message, and if it does, return that. Also, you do not necessarily want to always wait for an incoming message, because that would mean you cannot send two messages in a row.
So, here's my suggestion:
static int incoming_desc = -1;
static char *incoming_data = NULL;
static size_t incoming_size = 0;
static char *incoming_next = NULL; /* First received but not handled */
static char *incoming_ends = NULL; /* Last received but not handled */
#define INCOMING_CHUNK 4096
/* Receive a new message into dynamically allocated buffer,
and return the length. Returns 0 when no message, with errno set.
Waits at most ms milliseconds for a new message to arrive.
errno == EAGAIN: no message, timeout elapsed.
errno == ECONNABORTED: other end closed the connection.
*/
size_t get_message(char **message, size_t *size, long ms)
{
struct timeval timeout;
/* Make sure the parameters are sane. */
if (!message || !size || ms < 0) {
errno = EINVAL;
return 0;
}
/* For this function to work like getline() and getdelim() do,
we need to treat *message as NULL if *size == 0. */
if (!*size)
*message = NULL;
timeout.tv_sec = ms / 1000;
timeout.tv_usec = (ms % 1000) * 1000;
/* Timeout loop. */
while (1) {
fd_set readfds;
ssize_t bytes;
size_t used;
int result;
/* Is there a pending complete message in the buffer? */
if (incoming_ends > incoming_next) {
char *endmark = memchr(incoming_next, '\0', (size_t)(incoming_ends - incoming_next));
if (endmark) {
const size_t len = (size_t)(endmark - incoming_next) + 1;
/* Reallocate the message buffer, if necessary. */
if (len > *size) {
char *temp = realloc(*message, len);
if (!temp) {
errno = ENOMEM;
return 0;
}
*message = temp;
*size = len;
}
/* Copy message, */
memcpy(*message, incoming_next, len);
/* and remove it from the buffer. */
incoming_next += len;
/* In case the other end sent just the separator, clear errno. */
errno = 0;
/* We return the length sans the separator. */
return len - 1;
}
}
/* Do we have time left to check for input? */
if (timeout.tv_sec <= 0 && timeout.tv_usec <= 0)
break; /* Nope. */
/* Is incoming_desc one we can select() for? */
if (incoming_desc < 0 || incoming_desc >= FD_SETSIZE)
break; /* Nope. */
FD_ZERO(&readfds);
FD_SET(incoming_desc, &readfds);
result = select(incoming_desc + 1, &readfds, NULL, NULL, &timeout);
if (result < 1)
break; /* Nothing interesting happened (we ignore error here). */
if (!FD_ISSET(incoming_fd, &readfds))
break;
/* Number of bytes used in the buffer right now. */
used = (size_t)(incoming_ends - incoming_data);
/* Do we have at least INCOMING_CHUNK bytes available? */
if (used + INCOMING_CHUNK >= incoming_size) {
/* Nope. Repack the incoming buffer first. */
if (incoming_next > incoming_data) {
const size_t len = (size_t)(incoming_ends - incoming_next);
if (len > 0)
memmove(incoming_data, incoming_next, len);
incoming_next = incoming_data;
incoming_ends = incoming_data + len;
}
/* Recalculate the number of bytes we have free now. Enough? */
used = (size_t)(incoming_ends - incoming_data);
if (used + INCOMING_CHUNK > incoming_size) {
/* Grow incoming buffer. */
const size_t newsize = used + INCOMING_CHUNK;
char *temp = realloc(incoming_data, newsize);
if (!temp) {
errno = ENOMEM;
return 0;
}
incoming_next = temp + (size_t)(incoming_next - incoming_data);
incoming_ends = temp + used;
incoming_data = temp;
incoming_size = newsize;
}
}
/* Read more data into the buffer; up to a full buffer. */
bytes = read(incoming_fd, incoming_ends, incoming_size - used);
if (bytes > 0) {
incoming_ends += bytes;
} else
if (bytes == 0) {
/* Other end closed the connection. We may have a partial message
in the buffer, and should handle that too, but for now, we
just error out. */
errno = ECONNABORTED;
return 0;
} else
if (bytes != -1) {
/* Should never happen. */
errno = EIO;
return 0;
} else
if (errno == EINTR || errno == EAGAIN || errno == EWOULDBLOCK) {
/* No data yet, interrupted by signal delivery, etc. */
continue;
} else {
/* errno is set to indicate which error happened. */
return 0;
}
}
/* Timeout. */
errno = EAGAIN;
return 0;
}
Note that get_message() works like getline(): you do e.g.
char *msg = NULL;
size_t size = 0;
size_t len;
len = get_message(&msg, &size, 100); /* 100 ms = 0.1 seconds */
if (len) {
/* msg contains a full message of len characters */
} else
if (errno == ECONNABORTED) {
/* Other end closed the connection */
} else
if (errno != EAGAIN) {
fprintf(stderr, "Error receiving data: %s.\n", strerror(errno));
}
Then, you can reuse the same dynamically allocated buffer by just calling e.g.
len = get_message(&msg, &size, 100); /* 100 ms = 0.1 seconds */
again.
There is no such mechanism built into TCP or UDP. You need to implement your own protocol on top of it. One of the possible solutions is:
If the content delivered is static.
If the sending end knows the size of the data that is being delivered prior, your client and server can agree on specific terms. For example, the first four bytes sent by the server is the size of the remaining message represented in network byte order.
Server code
uint32_t n_size = htonl(size); // Convert the data size into network byte order.
write(sockfd, &n_size, sizeof(n_size)); // Send to the client.
Client code
uint32_t n_size;
int n_read = 0;
for ( ; ; ) {
int rd_status = read(sockfd, (void*) &n_size + n_read, sizeof(n_size) - n_read);
if (rd_status <= 0)
goto handle_this_case;
n_read = n_read + rd_status;
if (n_read == sizeof(n_size))
break;
}
uint32_t size = ntohl(n_size);
If the content delivered is generated on the fly.
In this case, even the server is not aware of the size of the message. You need to build your functions for handling this case. Below I have shown a bare minimal implementation:
Client-Side:
struct data_unit
{
void* data;
int size;
};
struct data_storage
{
struct data_unit unit;
struct data_storage* next;
};
void append_data(struct data_storage* storage, struct data_unit* unit);
struct data_unit* dump_data(struct data_storage* storage);
int main()
{
struct data_storage storage;
struct data_unit unit;
unit.data = malloc(MAX_SIZE);
for ( ; ; ) {
int rd_status = read(sockfd, unit.data, MAX_SIZE);
if (rd_status < 0)
goto handle_this_case;
else if (rd_status == 0)
break;
unit.size = rd_status;
append_data(&storage, &unit);
}
struct data_unit* t_data = dump_data(&storage);
}

Receiving multiple TCP segments

I have an assignment in which a TCP client sends data to the TCP server in the form of:
IP_address\0port\0message\n
Now, the server (IP address 10.0.2.15) receives the packet fine when I send some data through a terminal like this:
printf "127.0.0.1\0004444\000Some message\n" | nc -N 10.0.2.15 3333
However, the second part of the assignment is to read a packet that comes in multiple segments:
(printf "127.0.0.1"; sleep 0.3; printf "\0004444\000"; sleep 0.3; \
printf "It works"; sleep 0.5; printf "\n") | nc -N 10.0.2.15 3333
How should I implement the read function on the server so that, if possible, all the segments are stored into a buffer?
The number of bytes recv() returns can be as few as 1 byte up to as many bytes as requested. TCP is a byte stream, it has no concept of messages, that has to be handled in the application code instead.
The receiver must know how many bytes to expect, and then keep reading in a loop until it has read that many bytes, however many reads it takes.
However, in this situation, the receiver does not know the exact length of the message, because the sender is not sending the message length before sending the message itself, so the only option available is for the receiver to read from the socket byte-by-byte until it encounters the terminating \n.
For example:
int readLine(int socket, char **line)
{
int r, len = 0, cap = 256;
char b;
*line = NULL;
char *outline = (char*) malloc(cap);
if (!outline) return -2;
do
{
r = recv(socket, &b, 1, 0);
if (r <= 0)
{
free(outline);
return r;
}
if (b == '\n')
break;
if (len == cap)
{
cap += 256;
char *newline = (char*) realloc(outline, cap);
if (!newline)
{
free(outline);
return -2;
}
outline = newline;
}
outline[len] = b;
++len;
}
while (true);
if ((len > 0) && (line[len-1] == '\r'))
--len;
if (len == cap)
{
char *newline = (char*) realloc(outline, cap + 1);
if (!newline)
{
free(outline);
return -2;
}
outline = newline;
}
outline[len] = '\0';
*line = outline;
return 1;
}
char *line;
int r;
do
{
r = readLine(cliSock, &line);
if (r <= 0)
{
if (r == 0)
printf("client disconnected\n");
else if (r == -2)
printf("memory error\n");
else
printf("read error\n");
break;
}
// process line as needed...
free(line);
}
while (true);
Alternatively, you can use an intermediate buffer to help you cache data between reads and get data out of the socket more efficiently:
char *buffer;
int buflen, bufcap;
int readLine(int socket, char **line)
{
char *ptr;
int r, idx = 0;
*line = NULL;
do
{
ptr = memchr(buffer + idx, '\n', buflen - idx);
if (ptr)
{
int total = ((ptr + 1) - buffer);
int len = (total - 1);
if ((len > 0) && (buffer[len-1] == '\r'))
--len;
*line = (char*) malloc(len + 1);
if (*line == NULL)
return -2;
memcpy(*line, buffer, len);
(*line)[len] = '\0';
if (total < buflen)
memmove(buffer, buffer + total, buflen - total);
buflen -= total;
break;
}
if (buflen == bufcap)
{
int newcap = bufcap + 256;
char *newbuffer = (char*) realloc(buffer, newcap);
if (!newbuffer)
return -2;
buffer = newbuffer;
bufcap = newcap;
}
r = recv(socket, buffer + buflen, bufcap - buflen, 0);
if (r <= 0)
return r;
buflen += r;
}
while (true);
return 1;
}
buflen = 0;
bufcap = 256;
buffer = (char*) malloc(bufcap);
if (buffer)
{
char *line;
int r;
do
{
r = readLine(cliSock, &line);
if (r <= 0)
{
if (r == 0)
printf("client disconnected\n");
else if (r == -2)
printf("memory error\n");
else
printf("read error\n");
break;
}
// process line as needed...
free(line);
}
while (true);
free(buffer);
}

recv() on socket by dynamically allocating space

I'm trying to get the source code of my website using c, I'm able to connect and everything but when I implement the recv() code, it only receives the last few bytes of the source code. I'd like to dynamically allocate space for the buffer to receive more using the C functions malloc and realloc.
This is the code I have so far:
char *buffer = NULL;
unsigned int i = 0;
unsigned long LEN = 200;
unsigned long cur_size = 0;
buffer = (char*)malloc(sizeof(char)*LEN);
do
{
if( status >= LEN )
{
cur_size += status;
buffer = (char*)realloc(buffer, cur_size);
}
status = recv(cSocket, buffer, LEN, 0);
if( status == 0 )
{
printf("Bye\n");
}
else if( status > 0 )
{
printf("%d\n", status);
}
else
{
printf("socket error=%d\n", WSAGetLastError());
break;
}
}while( status > 0 );
printf("%s\n", buffer);
It still doesn't print the whole source code. How should I go about this?
Pseudocode:
buffer = 'len chars';
loop:
if( status >= buffer ) buffer = 'resize to status chars';
status = recv(sock, buffer, len, 0);
end loop
As you resize the buffer in advance this needs to be reflected by its size. Which currently is not the case.
To fix this you could, for example, initialise cur_size with LEN by changing
unsigned long cur_size = 0;
to
unsigned long cur_size = LEN;
Assuming the fix above, you want to append to the buffer and not overwrite it with every call to recv().
To do so change this line
status = recv(cSocket, buffer, LEN, 0);
to be
status = recv(cSocket, buffer + cur_size - LEN, LEN, 0);
A more straight forward approach would be to not track the size of the buffer, but the number of bytes received and just always increase the buffer by a constant size.
Also the two calls to allocate memory can be replaced by one:
char *buffer = NULL;
unsigned long LEN = 200;
unsigned long bytes_received = 0;
unsigned long cur_size = 0;
int status = 0;
do
{
if (bytes_received >= cur_size)
{
char * tmp;
cur_size += LEN;
tmp = realloc(buffer, cur_size);
if (NULL == tmp)
{
fprintf(stderr, "realloc error=%d\n", WSAGetLastError());
break;
}
buffer = tmp;
}
status = recv(cSocket, buffer + bytes_received, LEN, 0);
if (status == 0)
{
printf("Bye\n");
}
else if (status > 0)
{
bytes_received += status;
printf("%d\n", status);
}
else /* < 0 */
{
fprintf(stderr, "socket error=%d\n", WSAGetLastError());
}
} while (status > 0);
printf("%s\n", buffer);
Well, after a bit of research, I came across this website and finally found what I was looking for.
Binary tides
Although it uses linux's fcntl, the windows equivalent is ioctlsocket which is used to set the socket's non-blocking mode.
To see the exact function, visit the website. I modified the version and set my socket to blocking mode.
int total_recv(SOCKET s)
{
int size_recv = 0, total_size = 0, block = 00;
char chunk[BUFLEN];
ioctlsocket(s, FIONBIO, (unsigned long*)&block); // set mode to block
// not necessary but clarification of function, mode is block by
// default
while( 1 )
{
memset(chunk, 0, BUFLEN);
if( ( size_recv = recv(s, chunk, BUFLEN, 0) ) == SOCKET_ERROR )
{
printf("Error receiving\n");
}
else if( size_recv == 0 )
{
break;
}
else
{
total_size += size_recv;
// i used file since console wouldn't show full source code
FILE *fp = NULL;
fp = fopen("source.txt", "a");
fprintf(fp, chunk);
fclose(fp);
}
}
return total_size;
}

Read line by line from a socket buffer

I want to write a function that read line by line from a socket buffer obtained from third parameter from read() function from unistd.h header.
I have wrote this:
int sgetline(int fd, char ** out)
{
int buf_size = 128;
int bytesloaded = 0;
char buf[2];
char * buffer = malloc(buf_size);
char * newbuf;
int size = 0;
assert(NULL != buffer);
while( read(fd, buf, 1) > 0 )
{
strcat(buffer, buf);
buf[1] = '\0';
bytesloaded += strlen(buf);
size = size + buf_size;
if(buf[0] == '\n')
{
*out = buffer;
return bytesloaded;
}
if(bytesloaded >= size)
{
size = size + buf_size;
newbuf = realloc(buffer, size);
if(NULL != newbuf)
{
buffer = newbuf;
}
else
{
printf("sgetline() allocation failed!\n");
exit(1);
}
}
}
*out = buffer;
return bytesloaded;
}
but I have some problems with this function, for example, if the input is something like:
HTTP/1.1 301 Moved Permanently\r\n
Cache-Control:no-cache\r\n
Content-Length:0\r\n
Location\r\nhttp://bing.com/\r\n
\r\n\r\n
and I do
int sockfd = socket( ... );
//....
char* tbuf;
while(sgetline(sockfd, &tbuf) > 0)
{
if(strcmp(tbuf,"\r\n\r\n") == 0)
{
printf("End of Headers detected.\n");
}
}
the above C application does not output "End of Header detected.". Why is this, and how can I fix this?
It's not OK to read one byte at a time, because you are making too many system calls - better is to use a buffer, read a chunk and check if you got \n. After getting a line, the rest of the bytes read remains in the buffer, so you cannot mix read/recv with read_line. Another version of read n bytes using this kind of buffer can be write...
My version to read a line, and a little example to use it.
#include <stdio.h>
#include <errno.h>
#include <stdlib.h>
#include <sys/socket.h>
#include <unistd.h>
#include <netinet/in.h>
#include <netdb.h>
#include <arpa/inet.h>
#include <string.h>
#define CBSIZE 2048
typedef struct cbuf {
char buf[CBSIZE];
int fd;
unsigned int rpos, wpos;
} cbuf_t;
int read_line(cbuf_t *cbuf, char *dst, unsigned int size)
{
unsigned int i = 0;
ssize_t n;
while (i < size) {
if (cbuf->rpos == cbuf->wpos) {
size_t wpos = cbuf->wpos % CBSIZE;
//if ((n = read(cbuf->fd, cbuf->buf + wpos, (CBSIZE - wpos))) < 0) {
if((n = recv(cbuf->fd, cbuf->buf + wpos, (CBSIZE - wpos), 0)) < 0) {
if (errno == EINTR)
continue;
return -1;
} else if (n == 0)
return 0;
cbuf->wpos += n;
}
dst[i++] = cbuf->buf[cbuf->rpos++ % CBSIZE];
if (dst[i - 1] == '\n')
break;
}
if(i == size) {
fprintf(stderr, "line too large: %d %d\n", i, size);
return -1;
}
dst[i] = 0;
return i;
}
int main()
{
cbuf_t *cbuf;
char buf[512];
struct sockaddr_in saddr;
struct hostent *h;
char *ip;
char host[] = "www.google.com";
if(!(h = gethostbyname(host))) {
perror("gethostbyname");
return NULL;
}
ip = inet_ntoa(*(struct in_addr*)h->h_addr);
cbuf = calloc(1, sizeof(*cbuf));
fprintf(stdout, "Connecting to ip: %s\n", ip);
if((cbuf->fd = socket(AF_INET, SOCK_STREAM, 0)) < 0) {
perror("socket");
return 1;
}
memset(&saddr, 0, sizeof(saddr));
saddr.sin_family = AF_INET;
saddr.sin_port = htons(80);
inet_aton(ip, &saddr.sin_addr);
if(connect(cbuf->fd, (struct sockaddr*)&saddr, sizeof(saddr)) < 0) {
perror("connect");
return 1;
}
snprintf(buf, sizeof(buf), "GET / HTTP/1.1\r\nHost: %s\r\nConnection: close\r\n\r\n", host);
write(cbuf->fd, buf, strlen(buf));
while(read_line(cbuf, buf, sizeof(buf)) > 0) {
// if it's an empty \r\n on a line, header ends //
if(buf[0]=='\r' && buf[1] == '\n') {
printf("------------------------\n");
}
printf("[%s]", buf);
}
close(cbuf->fd);
free(cbuf);
return 0;
}
Try this implementation instead:
int sgetline(int fd, char ** out)
{
int buf_size = 0;
int in_buf = 0;
int ret;
char ch;
char * buffer = NULL;
char * new_buffer;
do
{
// read a single byte
ret = read(fd, &ch, 1);
if (ret < 1)
{
// error or disconnect
free(buffer);
return -1;
}
// has end of line been reached?
if (ch == '\n')
break; // yes
// is more memory needed?
if ((buf_size == 0) || (in_buf == buf_size))
{
buf_size += 128;
new_buffer = realloc(buffer, buf_size);
if (!new_buffer)
{
free(buffer);
return -1;
}
buffer = new_buffer;
}
buffer[in_buf] = ch;
++in_buf;
}
while (true);
// if the line was terminated by "\r\n", ignore the
// "\r". the "\n" is not in the buffer
if ((in_buf > 0) && (buffer[in_buf-1] == '\r'))
--in_buf;
// is more memory needed?
if ((buf_size == 0) || (in_buf == buf_size))
{
++buf_size;
new_buffer = realloc(buffer, buf_size);
if (!new_buffer)
{
free(buffer);
return -1;
}
buffer = new_buffer;
}
// add a null terminator
buffer[in_buf] = '\0';
*out = buffer; // complete line
return in_buf; // number of chars in the line, not counting the line break and null terminator
}
int sockfd = socket( ... );
//....
char* tbuf;
int ret;
// keep reading until end of headers is detected.
// headers are terminated by a 0-length line
do
{
// read a single line
ret = sgetline(sockfd, &tbuf);
if (ret < 0)
break; // error/disconnect
// is it a 0-length line?
if (ret == 0)
{
printf("End of Headers detected.\n");
free(tbuf);
break;
}
// tbuf contains a header line, use as needed...
free(tbuf);
}
while (true);
You are making things more difficult for yourself than they need to be. You really don't need to do strcats to get the single character you read on each read added at the current position.
But your bug is that the routine returns as soon as it sees a \n, so the string it returns can never contain anything following the first \n.

Writing memory to socket in chunks in C

I'm attempting to write memory contents to a socket in chunks. I can write files that are smaller than my buffer, but anything else and I'm in deep water.
/* allocate memory for file contents */
char fileContents = malloc(sizeof(char)*filesize);
/* read a file into memory */
read(fileDescriptor, fileContents , filesize);
int chunksWritten;
/* Write the memory to socket? */
if (filesize > MAX_BLOCK_SIZE){
while (chunksWritten < filesize){
// what goes here?
}
} else {
chunksWritten = writen(sd, fileContents, filesize); // this works for files < MAX_BLOCK_SIZE
}
writen here writes to my socket:
int writen(int fd, char *buf, int nbytes) {
short data_size = nbytes;
int n, nw;
if (nbytes > MAX_BLOCK_SIZE)
return (-3);
data_size = htons(data_size);
if (write(fd, (char *) & data_size, 1) != 1) return (-1);
if (write(fd, (char *) (&data_size) + 1, 1) != 1) return (-1);
/* send nbytes */
for (n = 0; n < nbytes; n += nw) {
if ((nw = write(fd, buf + n, nbytes - n)) <= 0)
return (nw);
}
return (n);
}
This seems like it should be quite easy, but I'm struggling to find any good examples.
/* outside the loop */
chunksWritten = 0;
int smaller;
int r;
int sizeRemaining = filesize;
//char *fileChunk = malloc(sizeof(char)*MAX_BLOCK_SIZE+1);
//memcpy(fileChunk, fileContents, sizeof(char)*MAX_BLOCK_SIZE);
//r = writen(sd, fileChunk, MAX_BLOCK_SIZE);
r = writen(sd, fileContents, MAX_BLOCK_SIZE);
if(r==-1) {
/* deal with error in a manner that fits the rest of your program */
}
chunksWritten = chunksWritten + r;
sizeRemaining = sizeRemaining - MAX_BLOCK_SIZE;
while(sizeRemaining > 0){
if(sizeRemaining > MAX_BLOCK_SIZE){
smaller = MAX_BLOCK_SIZE;
} else {
smaller = sizeRemaining;
}
//memcpy(fileChunk, fileContents+sizeof(char)*chunksWritten, sizeof(char)*smaller);
//r = writen(sd, fileChunk, MAX_BLOCK_SIZE);
r = writen(sd, fileContents[filesize - sizeRemaining], smaller);
if(r==-1) {
/* deal with error in a manner that fits the rest of your program */
}
sizeRemaining = sizeRemaining - MAX_BLOCK_SIZE;
}
/*
Reminder: clean-up fileChunk & fileContents if you don't need them later on
*/
You certainly can rework the loop to count up instead of down. I can think better counting down.
Edit: made a few changes based on comments.
Not sure why you want this, but seems like you want something like:
#define MIN(x, y) ((x) < (y) ? (x) : (y))
while (chunksWritten < filesize) {
int writtenThisPass = writen(fd,
fileContents + chunksWritten,
MIN(filesize - chunksWritten, MAX_BLOCK_SIZE));
if (writtenThisPass <= 0)
{
// TODO: handle the error
}
else
{
chunksWritten += writtenThisPass;
}
}

Resources