Zero-copy FTP/HTTP Daemon compatible with all POSIX systems
| 1 | /* |
| 2 | MIT License |
| 3 | |
| 4 | Copyright (c) 2026 Seregon |
| 5 | |
| 6 | Permission is hereby granted, free of charge, to any person obtaining a copy |
| 7 | of this software and associated documentation files (the "Software"), to deal |
| 8 | in the Software without restriction, including without limitation the rights |
| 9 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 10 | copies of the Software, and to permit persons to whom the Software is |
| 11 | furnished to do so, subject to the following conditions: |
| 12 | |
| 13 | The above copyright notice and this permission notice shall be included in all |
| 14 | copies or substantial portions of the Software. |
| 15 | |
| 16 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 17 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 18 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 19 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 20 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 21 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE |
| 22 | SOFTWARE. |
| 23 | */ |
| 24 | |
| 25 | /** |
| 26 | * @file http_parser.c |
| 27 | * @brief Minimal HTTP/1.1 request parser (no strtok, no strncpy) |
| 28 | * |
| 29 | * Rewrites the original parser to avoid: |
| 30 | * - strtok() : hidden global state, not reentrant |
| 31 | * - strncpy(): no guaranteed NUL termination |
| 32 | * |
| 33 | * Uses bounded pointer arithmetic with explicit length checks. |
| 34 | * The input buffer is mutated in-place (NUL inserted at delimiters) |
| 35 | * so header name/value pointers remain valid for the buffer's lifetime. |
| 36 | */ |
| 37 | |
| 38 | #include "http_parser.h" |
| 39 | #include <ctype.h> |
| 40 | #include <stdio.h> |
| 41 | #include <string.h> |
| 42 | |
| 43 | /*===========================================================================* |
| 44 | * HELPERS |
| 45 | * |
| 46 | * find_char() — bounded strchr (stays within [p, end)) |
| 47 | * find_crlf() — bounded strstr for "\r\n" |
| 48 | *===========================================================================*/ |
| 49 | |
| 50 | /** |
| 51 | * @brief Find first occurrence of c in [p, end) |
| 52 | * @return Pointer to c, or NULL if not found |
| 53 | */ |
| 54 | static char *find_char(char *p, const char *end, char c) { |
| 55 | while (p < end) { |
| 56 | if (*p == c) { |
| 57 | return p; |
| 58 | } |
| 59 | p++; |
| 60 | } |
| 61 | return NULL; |
| 62 | } |
| 63 | |
| 64 | /** |
| 65 | * @brief Find "\r\n" in [p, end) |
| 66 | * @return Pointer to the '\r', or NULL if not found |
| 67 | */ |
| 68 | static char *find_crlf(char *p, const char *end) { |
| 69 | while ((p + 1) < end) { |
| 70 | if (p[0] == '\r' && p[1] == '\n') { |
| 71 | return p; |
| 72 | } |
| 73 | p++; |
| 74 | } |
| 75 | return NULL; |
| 76 | } |
| 77 | |
| 78 | /*===========================================================================* |
| 79 | * PARSE REQUEST LINE |
| 80 | * |
| 81 | * "GET /index.html HTTP/1.1\r\n" |
| 82 | * ^method ^uri ^version |
| 83 | * |
| 84 | * Splits on spaces with bounded search, no strtok. |
| 85 | *===========================================================================*/ |
| 86 | |
| 87 | /** |
| 88 | * @brief Parse the request line: METHOD SP URI SP HTTP/x.y |
| 89 | * |
| 90 | * @param[in,out] line Start of request line (NUL-terminated by caller) |
| 91 | * @param[in] line_len Length of the line (excluding NUL) |
| 92 | * @param[out] request Populated with method, uri, version |
| 93 | * |
| 94 | * @return 0 on success, -1 on malformed request |
| 95 | */ |
| 96 | static int parse_request_line(char *line, size_t line_len, |
| 97 | http_request_t *request) { |
| 98 | const char *end = line + line_len; |
| 99 | |
| 100 | /*-- METHOD --*/ |
| 101 | char *sp1 = find_char(line, end, ' '); |
| 102 | if (sp1 == NULL) { |
| 103 | return -1; |
| 104 | } |
| 105 | *sp1 = '\0'; |
| 106 | |
| 107 | if (strcmp(line, "GET") == 0) { |
| 108 | request->method = HTTP_METHOD_GET; |
| 109 | } else if (strcmp(line, "POST") == 0) { |
| 110 | request->method = HTTP_METHOD_POST; |
| 111 | } else if (strcmp(line, "HEAD") == 0) { |
| 112 | request->method = HTTP_METHOD_HEAD; |
| 113 | } else { |
| 114 | request->method = HTTP_METHOD_UNKNOWN; |
| 115 | } |
| 116 | |
| 117 | /*-- URI --*/ |
| 118 | char *uri_start = sp1 + 1; |
| 119 | if (uri_start >= end) { |
| 120 | return -1; |
| 121 | } |
| 122 | char *sp2 = find_char(uri_start, end, ' '); |
| 123 | if (sp2 == NULL) { |
| 124 | return -1; |
| 125 | } |
| 126 | *sp2 = '\0'; |
| 127 | |
| 128 | /* Bounded copy into fixed-size uri field */ |
| 129 | size_t uri_len = (size_t)(sp2 - uri_start); |
| 130 | if (uri_len >= HTTP_URI_MAX_LENGTH) { |
| 131 | uri_len = HTTP_URI_MAX_LENGTH - 1U; |
| 132 | } |
| 133 | memcpy(request->uri, uri_start, uri_len); |
| 134 | request->uri[uri_len] = '\0'; |
| 135 | |
| 136 | /*-- VERSION "HTTP/x.y" --*/ |
| 137 | char *ver_start = sp2 + 1; |
| 138 | if (ver_start >= end) { |
| 139 | return -1; |
| 140 | } |
| 141 | if (sscanf(ver_start, "HTTP/%d.%d", &request->version_major, |
| 142 | &request->version_minor) != 2) { |
| 143 | return -1; |
| 144 | } |
| 145 | |
| 146 | return 0; |
| 147 | } |
| 148 | |
| 149 | /*===========================================================================* |
| 150 | * PARSE HEADERS |
| 151 | * |
| 152 | * "Content-Type: application/json\r\n" |
| 153 | * ^name ^colon ^value |
| 154 | * |
| 155 | * Each header line is NUL-terminated at the \r\n boundary. |
| 156 | * Colon is replaced with NUL to split name/value in-place. |
| 157 | * Leading whitespace on value is skipped. |
| 158 | *===========================================================================*/ |
| 159 | |
| 160 | /** |
| 161 | * @brief Parse a single header line into request->headers[] |
| 162 | * |
| 163 | * @param[in,out] line Header line, already NUL-terminated |
| 164 | * @param[out] request Target request struct |
| 165 | */ |
| 166 | static void parse_header_line(char *line, http_request_t *request) { |
| 167 | if (request->num_headers >= HTTP_HEADER_MAX_COUNT) { |
| 168 | return; /* silently drop excess headers */ |
| 169 | } |
| 170 | |
| 171 | char *colon = strchr(line, ':'); |
| 172 | if (colon == NULL) { |
| 173 | return; /* malformed header, skip */ |
| 174 | } |
| 175 | |
| 176 | *colon = '\0'; |
| 177 | char *value = colon + 1; |
| 178 | |
| 179 | /* Skip leading whitespace (OWS per RFC 7230 §3.2.6) */ |
| 180 | while (*value == ' ' || *value == '\t') { |
| 181 | value++; |
| 182 | } |
| 183 | |
| 184 | request->headers[request->num_headers].name = line; |
| 185 | request->headers[request->num_headers].value = value; |
| 186 | request->num_headers++; |
| 187 | } |
| 188 | |
| 189 | /*===========================================================================* |
| 190 | * PUBLIC API |
| 191 | *===========================================================================*/ |
| 192 | |
| 193 | /** |
| 194 | * @brief Parse a complete HTTP/1.1 request from a mutable buffer |
| 195 | * |
| 196 | * The buffer is mutated in-place: NUL bytes are inserted at |
| 197 | * line boundaries and colon separators so that request->uri, |
| 198 | * header name/value, and body pointers reference the buffer |
| 199 | * directly with zero-copy semantics. |
| 200 | * |
| 201 | * @param[in,out] buffer Raw HTTP request data (mutable) |
| 202 | * @param[in] length Number of valid bytes in buffer |
| 203 | * @param[out] request Parsed result |
| 204 | * |
| 205 | * @return 0 on success, negative on parse error |
| 206 | * -1: NULL input or malformed request line |
| 207 | * -2: missing CRLF (incomplete request) |
| 208 | */ |
| 209 | int http_parse_request(char *buffer, size_t length, http_request_t *request) { |
| 210 | if ((buffer == NULL) || (request == NULL) || (length == 0U)) { |
| 211 | return -1; |
| 212 | } |
| 213 | |
| 214 | memset(request, 0, sizeof(*request)); |
| 215 | |
| 216 | const char *buf_end = buffer + length; |
| 217 | |
| 218 | /*-- Request line --*/ |
| 219 | char *crlf = find_crlf(buffer, buf_end); |
| 220 | if (crlf == NULL) { |
| 221 | return -2; |
| 222 | } |
| 223 | *crlf = '\0'; /* NUL-terminate request line */ |
| 224 | |
| 225 | size_t line_len = (size_t)(crlf - buffer); |
| 226 | if (parse_request_line(buffer, line_len, request) != 0) { |
| 227 | return -1; |
| 228 | } |
| 229 | |
| 230 | /*-- Headers --*/ |
| 231 | char *line = crlf + 2; /* skip past \r\n */ |
| 232 | while (line < buf_end) { |
| 233 | crlf = find_crlf(line, buf_end); |
| 234 | if (crlf == NULL) { |
| 235 | break; /* truncated headers, stop */ |
| 236 | } |
| 237 | *crlf = '\0'; |
| 238 | |
| 239 | /* Empty line = end of headers, body follows */ |
| 240 | if (line[0] == '\0') { |
| 241 | char *body_start = crlf + 2; |
| 242 | if (body_start < buf_end) { |
| 243 | request->body = body_start; |
| 244 | request->body_length = (size_t)(buf_end - body_start); |
| 245 | } |
| 246 | break; |
| 247 | } |
| 248 | |
| 249 | parse_header_line(line, request); |
| 250 | line = crlf + 2; |
| 251 | } |
| 252 | |
| 253 | return 0; |
| 254 | } |
| 255 | |
| 256 | /** |
| 257 | * @brief Look up a header value by name (case-insensitive) |
| 258 | * |
| 259 | * @param request Parsed request |
| 260 | * @param name Header name to search for |
| 261 | * |
| 262 | * @return Header value string, or NULL if not found |
| 263 | */ |
| 264 | const char *http_get_header(const http_request_t *request, const char *name) { |
| 265 | if ((request == NULL) || (name == NULL)) { |
| 266 | return NULL; |
| 267 | } |
| 268 | |
| 269 | for (size_t i = 0; i < request->num_headers; i++) { |
| 270 | if (strcasecmp(request->headers[i].name, name) == 0) { |
| 271 | return request->headers[i].value; |
| 272 | } |
| 273 | } |
| 274 | |
| 275 | return NULL; |
| 276 | } |
| 277 |