feat: implement parsing of headers

some todos left before I call it stable
This commit is contained in:
Kevin J. 2024-09-03 00:03:32 +02:00
parent f428190016
commit 12b3f750ac
8 changed files with 248 additions and 34 deletions

1
.gitignore vendored
View File

@ -1,2 +1,3 @@
*.o
compile_commands.json
tinyparser

3
.gitmodules vendored Normal file
View File

@ -0,0 +1,3 @@
[submodule "streecmp"]
path = streecmp
url = http://git.0xdeadbeer.xyz/0xdeadbeer/streecmp

View File

@ -1,10 +1,11 @@
CC=gcc
CFLAGS=-g3
CFILES=tinyparser.c
CFILES=tinyparser.c headers.c
CFILES_STREECMP=streecmp/streecmp.c
OUTPUT=tinyparser
all:
$(CC) $(CFLAGS) $(CFILES) -o $(OUTPUT)
$(CC) $(CFLAGS) $(CFILES) $(CFILES_STREECMP) -o $(OUTPUT)
run:
./$(OUTPUT)

108
headers.c Normal file
View File

@ -0,0 +1,108 @@
char *headers = "A-IM\n"
"Accept\n"
"Accept-Charset\n"
"Accept-Datetime\n"
"Accept-Encoding\n"
"Accept-Language\n"
"Access-Control-Request-Method\n"
"Access-Control-Request-Headers\n"
"Authorization\n"
"Cache-Control\n"
"Connection\n"
"Content-Encoding\n"
"Content-Length\n"
"Content-MD5\n"
"Content-Type\n"
"Cookie\n"
"Date\n"
"Expect\n"
"Forwarded\n"
"From\n"
"Host\n"
"HTTP2-Settings\n"
"If-Match\n"
"If-Modified-Since\n"
"If-None-Match\n"
"If-Range\n"
"If-Unmodified-Since\n"
"Max-Forwards\n"
"Origin\n"
"Pragma\n"
"Prefer\n"
"Proxy-Authorization\n"
"Range\n"
"Referer\n"
"TE\n"
"Trailer\n"
"Transfer-Encoding\n"
"User-Agent\n"
"Upgrade\n"
"Via\n"
"Warning\n"
"Upgrade-Insecure-Requests\n"
"X-Requested-With\n"
"DNT\n"
"X-Forwarded-For\n"
"X-Forwarded-Host\n"
"X-Forwarded-Proto\n"
"Front-End-Https\n"
"X-Http-Method-Override\n"
"X-ATT-DeviceID\n"
"X-Wap-Profile\n"
"Proxy-Connection\n"
"X-UIDH\n"
"X-Csrf-Token\n"
"X-Request-ID\n"
"X-Correlation-ID\n"
"Correlation-ID\n"
"Save-Data\n"
"Sec-GPC\n"
"Accept-CH\n"
"Access-Control-Allow-Origin\n"
"Access-Control-Allow-Credentials\n"
"Access-Control-Expose-Headers\n"
"Access-Control-Max-Age\n"
"Access-Control-Allow-Methods\n"
"Access-Control-Allow-Headers\n"
"Accept-Patch\n"
"Accept-Ranges\n"
"Age\n"
"Allow\n"
"Alt-Svc\n"
"Content-Disposition\n"
"Content-Language\n"
"Content-Location\n"
"Content-Range\n"
"Delta-Base\n"
"ETag\n"
"Expires\n"
"IM\n"
"Last-Modified\n"
"Link\n"
"Location\n"
"P3P\n"
"Preference-Applied\n"
"Proxy-Authenticate\n"
"Public-Key-Pins\n"
"Retry-After\n"
"Server\n"
"Set-Cookie\n"
"Strict-Transport-Security\n"
"Tk\n"
"Vary\n"
"WWW-Authenticate\n"
"X-Frame-Options\n"
"Content-Security-Policy\n"
"Expect-CT\n"
"NEL\n"
"Permissions-Policy\n"
"Refresh\n"
"Report-To\n"
"Status\n"
"Timing-Allow-Origin\n"
"X-Content-Duration\n"
"X-Content-Type-Options\n"
"X-Powered-By\n"
"X-Redirect-By\n"
"X-UA-Compatible\n"
"X-XSS-Protection";

1
streecmp Submodule

@ -0,0 +1 @@
Subproject commit 42ed96353d186078e34e4c541e7c92d433cda138

Binary file not shown.

View File

@ -2,6 +2,9 @@
#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;
@ -11,17 +14,25 @@ char *ver = NULL;
int uri_len = 0;
int ver_len = 0;
int read_line(char **cursor, int cursor_size) {
int diff = 0;
char *match = strchr(*cursor, '\n');
struct nod *header_tree = NULL;
char **header_table = NULL;
if (!match)
match = (*cursor)+cursor_size-1;
diff = match - (*cursor);
*cursor += diff + 1;
return diff;
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) {
@ -57,18 +68,43 @@ int cmp_ascii(char *ch, char *limit) {
return diff;
}
int parse_title(char *title_offset, int title_len) {
int title_cursor = 0;
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 (title_cursor >= title_len) {
if (cursor >= len) {
state = state_title_err;
err = err_parse_title;
}
int ret = 0;
char *ch = title_offset+title_cursor;
char *lim = title_offset+title_len-1;
char *ch = offset+cursor;
char *lim = offset+len-1;
switch (state) {
case state_init:
@ -76,19 +112,19 @@ int parse_title(char *title_offset, int title_len) {
break;
case state_title_method:
if ((ret = cmp_orig(ch, "GET", lim))) {
title_cursor += ret;
cursor += ret;
state = state_title_after_method;
method = method_get;
break;
}
if ((ret = cmp_orig(ch, "POST", lim))) {
title_cursor += ret;
cursor += ret;
state = state_title_after_method;
method = method_post;
break;
}
if ((ret = cmp_orig(ch, "HEAD", lim))) {
title_cursor += ret;
cursor += ret;
state = state_title_after_method;
method = method_head;
break;
@ -98,7 +134,7 @@ int parse_title(char *title_offset, int title_len) {
err = err_parse_title_method;
break;
case state_title_after_method:
title_cursor++;
cursor++;
state = state_title_uri;
break;
case state_title_uri:
@ -115,11 +151,11 @@ int parse_title(char *title_offset, int title_len) {
}
uri_len = ret;
title_cursor += ret;
cursor += ret;
state = state_title_after_uri;
break;
case state_title_after_uri:
title_cursor++;
cursor++;
state = state_title_version;
break;
case state_title_version:
@ -136,7 +172,7 @@ int parse_title(char *title_offset, int title_len) {
}
ver_len = ret;
title_cursor += ret;
cursor += ret;
state = state_title_ok;
break;
case state_title_ok:
@ -144,6 +180,7 @@ int parse_title(char *title_offset, int title_len) {
if (*ch == ' ') continue;
if (*ch == '\t') continue;
if (*ch == '\0') continue;
if (*ch == '\r') continue;
state = state_title_err;
err = err_parse_title;
@ -164,16 +201,35 @@ int parse_title(char *title_offset, int title_len) {
int parse_request(char *buffer) {
int ret;
int buffer_size = strlen(buffer);
char *buffer_limit = buffer+strlen(buffer);
char *title_offset = buffer;
int title_len = read_line(&buffer, buffer_size);
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;
}
@ -181,6 +237,24 @@ 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");
@ -189,6 +263,8 @@ int main(void) {
fprintf(stderr, "Finished parsing request\n");
frenod(header_tree);
free(header_table);
free(str);
return 0;

View File

@ -1,12 +1,19 @@
#ifndef PARSER_H
#define PARSER_H
#define CRLF "\r\n"
#define MAX_BOUND 1024
#define LINE_BUF 1024
#define TEST_ONE "GET / HTTP/1.1\r\n"\
"Host: archive.0xdeadbeer.xyz\r\n"\
"User-Agent: curl/8.9.1\r\n"\
"Accept: */*\r\n"
"Accept: */*\r\n"\
"\r\n"\
"{\"key\": \"kefjoiawejfojgorgjbosejrgo\"}"\
// ENUMS
// SRC:https://datatracker.ietf.org/doc/html/rfc2616
enum errs {
err_init = 0,
err_generic,
@ -14,9 +21,11 @@ enum errs {
err_parse_title,
err_parse_title_method,
err_parse_title_uri,
err_parse_title_version
err_parse_title_version,
err_parse_header_not_found
};
// SRC:https://datatracker.ietf.org/doc/html/rfc2616
enum states {
state_init = 0,
state_title_method,
@ -25,9 +34,23 @@ enum states {
state_title_after_uri,
state_title_version,
state_title_ok,
state_title_err
state_title_err,
state_header_init,
state_header_alloc,
state_header_parse,
state_header_not_found,
state_header_ok,
state_header_err,
state_body_init,
state_body_ok,
state_body_err,
state_fin,
};
// SRC:https://developer.mozilla.org/en-US/docs/Web/HTTP/Methods
enum methods {
method_init = 0,
method_get,
@ -35,8 +58,9 @@ enum methods {
method_head
};
// SRC:https://en.wikipedia.org/wiki/List_of_HTTP_header_fields
enum headers {
header_a_im = 0,
header_a_im = 1,
header_accept,
header_accept_charset,
header_accept_datetime,
@ -77,7 +101,6 @@ enum headers {
header_upgrade,
header_via,
header_warning,
header_upgrade_insecure_requests,
header_x_requested_with,
header_dnt,
@ -96,7 +119,6 @@ enum headers {
header_correlation_id,
header_save_data,
header_sec_gpc,
header_accept_ch,
header_access_control_allow_origin,
header_access_control_allow_credentials,
@ -145,7 +167,9 @@ enum headers {
header_x_powered_by,
header_x_redirect_by,
header_x_ua_compatible,
header_x_xss_protection
header_x_xss_protection,
header_count
};
#endif