parslib/tinyparser.c
Kevin Jerebica 12b3f750ac feat: implement parsing of headers
some todos left before I call it stable
2024-09-03 00:03:32 +02:00

272 lines
6.8 KiB
C

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "tinyparser.h"
#include "streecmp/streecmp.h"
extern char *headers;
enum errs err = err_init;
enum states state = state_init;
enum methods method = method_init;
char *uri = NULL;
char *ver = NULL;
int uri_len = 0;
int ver_len = 0;
struct nod *header_tree = NULL;
char **header_table = NULL;
int read_line(char **buffer, char **buffer_limit) {
int diff = 0;
if ((*buffer) >= (*buffer_limit)) {
return diff;
}
int ret = 0;
char *match = strstr(*buffer, "\r\n");
if (match) {
ret = match-(*buffer);
*buffer += 2;
} else {
ret = (*buffer_limit)-(*buffer);
}
*buffer += ret;
return ret;
}
int cmp_orig(char *ch, char *orig, char *limit) {
int diff = 0;
for (; *orig != '\0' && *ch != '\0' && ch < limit; orig++, ch++, diff++)
if (*orig != *ch)
return -1;
return diff;
}
int cmp_ascii(char *ch, char *limit) {
int diff = 0;
for (; *ch != '\0' && *ch != ' ' && ch < limit; ch++, diff++) {
if (*ch >= '0' && *ch <= '9') continue;
if (*ch >= 'A' && *ch <= 'Z') continue;
if (*ch >= 'a' && *ch <= 'z') continue;
if (*ch >= '#' && *ch <= '/') continue;
if (*ch == '!') continue;
if (*ch == ':') continue;
if (*ch == ';') continue;
if (*ch == '=') continue;
if (*ch == '?') continue;
if (*ch == '@') continue;
if (*ch == '[') continue;
if (*ch == ']') continue;
if (*ch == '_') continue;
if (*ch == '~') continue;
return -1;
}
return diff;
}
int parse_header(char *offset, int len) {
int cursor = 0;
int ret = 0;
char *header_limit = offset+len;
char *sep = strchr(offset, ':');
if (!sep) {
return -1;
}
if (sep > header_limit) {
return -1;
}
int htitle_len = sep-offset;
ret = streencmp(header_tree, offset, htitle_len);
if (ret == 0) {
return -1;
}
char *hvalue = sep+1;
header_table[ret] = hvalue;
return 0;
}
int parse_title(char *offset, int len) {
int cursor = 0;
while (state != state_title_ok && state != state_title_err) {
if (cursor >= len) {
state = state_title_err;
err = err_parse_title;
}
int ret = 0;
char *ch = offset+cursor;
char *lim = offset+len-1;
switch (state) {
case state_init:
state = state_title_method;
break;
case state_title_method:
if ((ret = cmp_orig(ch, "GET", lim))) {
cursor += ret;
state = state_title_after_method;
method = method_get;
break;
}
if ((ret = cmp_orig(ch, "POST", lim))) {
cursor += ret;
state = state_title_after_method;
method = method_post;
break;
}
if ((ret = cmp_orig(ch, "HEAD", lim))) {
cursor += ret;
state = state_title_after_method;
method = method_head;
break;
}
state = state_title_err;
err = err_parse_title_method;
break;
case state_title_after_method:
cursor++;
state = state_title_uri;
break;
case state_title_uri:
if ((ret = cmp_ascii(ch, lim)) <= 0) {
state = state_title_err;
err = err_parse_title_uri;
break;
}
if (!(uri = strndup(ch, ret))) {
state = state_title_err;
err = err_memory;
break;
}
uri_len = ret;
cursor += ret;
state = state_title_after_uri;
break;
case state_title_after_uri:
cursor++;
state = state_title_version;
break;
case state_title_version:
if ((ret = cmp_ascii(ch, lim)) <= 0) {
state = state_title_err;
err = err_parse_title_version;
break;
}
if (!(ver = strndup(ch, ret))) {
state = state_title_err;
err = err_memory;
break;
}
ver_len = ret;
cursor += ret;
state = state_title_ok;
break;
case state_title_ok:
for (;ch < lim;ch++) {
if (*ch == ' ') continue;
if (*ch == '\t') continue;
if (*ch == '\0') continue;
if (*ch == '\r') continue;
state = state_title_err;
err = err_parse_title;
break;
}
break;
case state_title_err:
fprintf(stderr, "failed\n");
return -1;
break;
default:
break;
}
}
return 0;
}
int parse_request(char *buffer) {
int ret;
char *buffer_limit = buffer+strlen(buffer);
char *title_offset = buffer;
int title_len = read_line(&buffer, &buffer_limit);
if ((ret = parse_title(title_offset, title_len)) < 0) {
fprintf(stderr, "Failed parsing title\n");
return -1;
}
for (int bound = 0; bound < MAX_BOUND; bound++) {
char *header_offset = buffer;
int header_len = read_line(&buffer, &buffer_limit);
char *header_limit = header_offset+header_len;
// IF END OF MESSAGE
if (!header_len) {
state = state_fin;
break;
}
if ((ret = parse_header(header_offset, header_len)) < 0) {
fprintf(stderr, "Failed parsing header\n");
return -1;
}
}
free(uri);
free(ver);
return 0;
}
int main(void) {
int ret;
char *str = strdup(TEST_ONE);
header_table = (char **) calloc(header_count, sizeof(char *));
if (!header_table) {
fprintf(stderr, "Not enough dynamic memory\n");
return -1;
}
header_tree = allocnod();
if (!header_tree) {
fprintf(stderr, "Not enough dynamic memory\n");
return -1;
}
ret = gentree(header_tree, headers, NULL);
if (gentree < 0) {
fprintf(stderr, "Failed generating the header comparison tree\n");
return -1;
}
ret = parse_request(str);
if (ret < 0) {
fprintf(stderr, "Failed parsing request\n");
return -1;
}
fprintf(stderr, "Finished parsing request\n");
frenod(header_tree);
free(header_table);
free(str);
return 0;
}