| 1 |
/* |
| 2 |
* Copyright (c) 2020-2022 joshua stein <jcs@jcs.org> |
| 3 |
* |
| 4 |
* Permission to use, copy, modify, and distribute this software for any |
| 5 |
* purpose with or without fee is hereby granted, provided that the above |
| 6 |
* copyright notice and this permission notice appear in all copies. |
| 7 |
* |
| 8 |
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES |
| 9 |
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF |
| 10 |
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR |
| 11 |
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES |
| 12 |
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN |
| 13 |
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF |
| 14 |
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. |
| 15 |
*/ |
| 16 |
|
| 17 |
#include <stdio.h> |
| 18 |
#include <string.h> |
| 19 |
#include "dnr.h" |
| 20 |
#include "http.h" |
| 21 |
#include "util.h" |
| 22 |
#include "wikipedia.h" |
| 23 |
|
| 24 |
struct url * |
| 25 |
url_parse(const char *str) |
| 26 |
{ |
| 27 |
struct url *url = NULL; |
| 28 |
char *buf, *scheme, *host, *path; |
| 29 |
unsigned short port; |
| 30 |
short ret, pos; |
| 31 |
size_t len, schemelen, hostlen, pathlen; |
| 32 |
|
| 33 |
len = strlen(str); |
| 34 |
scheme = xmalloc(len + 1); |
| 35 |
if (scheme == NULL) { |
| 36 |
warn("http: Failed allocating %ld", len + 1); |
| 37 |
return NULL; |
| 38 |
} |
| 39 |
host = xmalloc(len + 1); |
| 40 |
if (host == NULL) { |
| 41 |
warn("http: Failed allocating %ld", len + 1); |
| 42 |
xfree(&scheme); |
| 43 |
return NULL; |
| 44 |
} |
| 45 |
path = xmalloc(len + 1); |
| 46 |
if (path == NULL) { |
| 47 |
warn("http: Failed allocating %ld", len + 1); |
| 48 |
xfree(&host); |
| 49 |
xfree(&scheme); |
| 50 |
return NULL; |
| 51 |
} |
| 52 |
|
| 53 |
/* scheme://host:port/path */ |
| 54 |
ret = sscanf(str, "%[^:]://%[^:]:%d%s%n", scheme, host, &port, path, |
| 55 |
&pos); |
| 56 |
if (ret == 4) { |
| 57 |
if (pos > len) |
| 58 |
panic("url_parse sscanf overflow"); |
| 59 |
goto consolidate; |
| 60 |
} |
| 61 |
|
| 62 |
/* scheme://host/path */ |
| 63 |
ret = sscanf(str, "%[^:]://%[^/]%s%n", scheme, host, path, &pos); |
| 64 |
if (ret == 3) { |
| 65 |
if (pos > len) |
| 66 |
panic("url_parse sscanf overflow"); |
| 67 |
if (strcmp(scheme, "http") == 0) |
| 68 |
port = 80; |
| 69 |
else if (strcmp(scheme, "https") == 0) |
| 70 |
port = 443; |
| 71 |
else |
| 72 |
goto cleanup; |
| 73 |
goto consolidate; |
| 74 |
} |
| 75 |
|
| 76 |
goto cleanup; |
| 77 |
|
| 78 |
consolidate: |
| 79 |
schemelen = strlen(scheme); |
| 80 |
hostlen = strlen(host); |
| 81 |
pathlen = strlen(path); |
| 82 |
|
| 83 |
/* |
| 84 |
* Put everything in a single chunk of memory so the caller can just |
| 85 |
* free(url) |
| 86 |
*/ |
| 87 |
len = sizeof(struct url) + schemelen + 1 + hostlen + 1 + pathlen + 1; |
| 88 |
url = xmalloc(len); |
| 89 |
if (url == NULL) { |
| 90 |
warn("http: Failed allocating %ld for URL", len); |
| 91 |
goto cleanup; |
| 92 |
} |
| 93 |
|
| 94 |
url->scheme = (char *)url + sizeof(struct url); |
| 95 |
len = strlcpy(url->scheme, scheme, schemelen + 1); |
| 96 |
|
| 97 |
url->host = url->scheme + len + 1; |
| 98 |
len = strlcpy(url->host, host, hostlen + 1); |
| 99 |
|
| 100 |
url->path = url->host + len + 1; |
| 101 |
len = strlcpy(url->path, path, pathlen + 1); |
| 102 |
|
| 103 |
url->port = port; |
| 104 |
|
| 105 |
cleanup: |
| 106 |
xfree(&scheme); |
| 107 |
xfree(&host); |
| 108 |
xfree(&path); |
| 109 |
|
| 110 |
return url; |
| 111 |
} |
| 112 |
|
| 113 |
char * |
| 114 |
url_encode(unsigned char *str) |
| 115 |
{ |
| 116 |
char *ret = NULL; |
| 117 |
size_t len, n; |
| 118 |
bool encode = false; |
| 119 |
char a, b; |
| 120 |
|
| 121 |
encode: |
| 122 |
for (n = 0, len = 0; str[n] != '\0'; n++) { |
| 123 |
if ((str[n] >= 'A' && str[n] <= 'Z') || |
| 124 |
(str[n] >= 'a' && str[n] <= 'z') || |
| 125 |
(str[n] >= '0' && str[n] <= '9') || |
| 126 |
(str[n] == '-' || str[n] == '_' || str[n] == '.' || |
| 127 |
str[n] == '~')) { |
| 128 |
if (ret) |
| 129 |
ret[len] = str[n]; |
| 130 |
len++; |
| 131 |
} else { |
| 132 |
if (ret) { |
| 133 |
sprintf(ret + len, "%%%02X", str[n]); |
| 134 |
} |
| 135 |
len += 3; |
| 136 |
} |
| 137 |
} |
| 138 |
|
| 139 |
if (ret) { |
| 140 |
ret[len] = '\0'; |
| 141 |
return ret; |
| 142 |
} |
| 143 |
|
| 144 |
ret = xmalloc(len + 1); |
| 145 |
if (ret == NULL) { |
| 146 |
warn("http: Failed allocating %ld", len + 1); |
| 147 |
return NULL; |
| 148 |
} |
| 149 |
len = 0; |
| 150 |
goto encode; |
| 151 |
} |
| 152 |
|
| 153 |
struct http_request * |
| 154 |
http_get(const char *surl) |
| 155 |
{ |
| 156 |
struct url *url; |
| 157 |
struct http_request *req; |
| 158 |
size_t len, alen; |
| 159 |
short err; |
| 160 |
char ip_s[16]; |
| 161 |
ip_addr local_ip; |
| 162 |
tcp_port local_port; |
| 163 |
|
| 164 |
url = url_parse(surl); |
| 165 |
if (url == NULL) |
| 166 |
return NULL; |
| 167 |
|
| 168 |
req = xmalloczero(sizeof(struct http_request)); |
| 169 |
if (req == NULL) { |
| 170 |
warn("http: Failed allocating http_request"); |
| 171 |
return NULL; |
| 172 |
} |
| 173 |
req->url = url; |
| 174 |
req->tcp_buf_size = (4 * 1500) + sizeof(req->chunk); |
| 175 |
req->tcp_buf = xmalloc(req->tcp_buf_size); |
| 176 |
if (req->tcp_buf == NULL) { |
| 177 |
warn("http: Failed allocating tcp_buf"); |
| 178 |
xfree(&req); |
| 179 |
return NULL; |
| 180 |
} |
| 181 |
|
| 182 |
err = _TCPCreate(&req->tcp_iopb, &req->tcp_stream, (Ptr)req->tcp_buf, |
| 183 |
req->tcp_buf_size, nil, nil, nil, false); |
| 184 |
if (err) { |
| 185 |
warn("TCPCreate failed: %d", err); |
| 186 |
goto error; |
| 187 |
} |
| 188 |
|
| 189 |
err = DNSResolveName(&req->url->host, &req->host_ip, NULL); |
| 190 |
if (err) { |
| 191 |
warn("Couldn't resolve host %s (%d)", req->url->host, err); |
| 192 |
goto error; |
| 193 |
} |
| 194 |
|
| 195 |
long2ip(req->host_ip, (char *)&ip_s); |
| 196 |
|
| 197 |
err = _TCPActiveOpen(&req->tcp_iopb, req->tcp_stream, req->host_ip, |
| 198 |
req->url->port, &local_ip, &local_port, nil, nil, false); |
| 199 |
if (err) { |
| 200 |
warn("Failed connecting to %s (%s) port %d: %d", |
| 201 |
req->url->host, ip_s, req->url->port, err); |
| 202 |
goto error; |
| 203 |
} |
| 204 |
|
| 205 |
alen = 256 + strlen(req->url->host) + strlen(req->url->path); |
| 206 |
req->message = xmalloc(alen); |
| 207 |
if (req->message == NULL) { |
| 208 |
warn("http: Failed allocating message"); |
| 209 |
goto error; |
| 210 |
} |
| 211 |
len = snprintf(req->message, alen, |
| 212 |
"GET %s HTTP/1.0\r\n" |
| 213 |
"Host: %s\r\n" |
| 214 |
"User-Agent: %s\r\n" |
| 215 |
"Accept: */*\r\n" |
| 216 |
"\r\n", req->url->path, req->url->host, PROGRAM_NAME); |
| 217 |
if (len > alen) |
| 218 |
panic("snprintf overflow"); |
| 219 |
|
| 220 |
memset(&req->tcp_wds, 0, sizeof(req->tcp_wds)); |
| 221 |
req->tcp_wds[0].ptr = req->message; |
| 222 |
req->tcp_wds[0].length = len; |
| 223 |
|
| 224 |
err = _TCPSend(&req->tcp_iopb, req->tcp_stream, req->tcp_wds, nil, nil, |
| 225 |
false); |
| 226 |
if (err) { |
| 227 |
warn("TCPSend to %s (%s) failed: %d", req->url->host, ip_s, err); |
| 228 |
goto error; |
| 229 |
} |
| 230 |
|
| 231 |
return req; |
| 232 |
|
| 233 |
error: |
| 234 |
http_req_free(&req); |
| 235 |
return NULL; |
| 236 |
} |
| 237 |
|
| 238 |
ssize_t |
| 239 |
http_req_read(struct http_request *req, char *data, size_t len) |
| 240 |
{ |
| 241 |
short err; |
| 242 |
unsigned short rlen; |
| 243 |
|
| 244 |
if (!req) |
| 245 |
return -1; |
| 246 |
|
| 247 |
err = _TCPStatus(&req->tcp_iopb, req->tcp_stream, &req->tcp_status_pb, |
| 248 |
nil, nil, false); |
| 249 |
if (err) |
| 250 |
return -1; |
| 251 |
|
| 252 |
if (req->tcp_status_pb.amtUnreadData == 0) |
| 253 |
return 0; |
| 254 |
|
| 255 |
rlen = MIN(req->tcp_status_pb.amtUnreadData, len); |
| 256 |
|
| 257 |
err = _TCPRcv(&req->tcp_iopb, req->tcp_stream, data, &rlen, nil, nil, |
| 258 |
false); |
| 259 |
if (err) |
| 260 |
return -1; |
| 261 |
|
| 262 |
return rlen; |
| 263 |
} |
| 264 |
|
| 265 |
bool |
| 266 |
http_req_skip_header(struct http_request *req) |
| 267 |
{ |
| 268 |
size_t len, n; |
| 269 |
bool last_nl = false; |
| 270 |
|
| 271 |
for (;;) { |
| 272 |
if (req->chunk_len == sizeof(req->chunk)) |
| 273 |
/* this sure is a long header... */ |
| 274 |
req->chunk_len = 0; |
| 275 |
|
| 276 |
len = http_req_read(req, req->chunk + req->chunk_len, |
| 277 |
sizeof(req->chunk) - req->chunk_len); |
| 278 |
if (len < 0) |
| 279 |
return false; |
| 280 |
if (len == 0) |
| 281 |
continue; |
| 282 |
req->chunk_len += len; |
| 283 |
|
| 284 |
for (n = 1; n < req->chunk_len; n++) { |
| 285 |
if (req->chunk[n - 1] != '\r' || req->chunk[n] != '\n') { |
| 286 |
last_nl = false; |
| 287 |
continue; |
| 288 |
} |
| 289 |
|
| 290 |
/* newline, shift chunk back */ |
| 291 |
if (strncmp(req->chunk, "Content-Length: ", 16) == 0) { |
| 292 |
if (sscanf(req->chunk, "Content-Length: %ld", &len) == 1) |
| 293 |
req->content_len = len; |
| 294 |
} |
| 295 |
req->chunk_len -= n + 1; |
| 296 |
memmove(req->chunk, req->chunk + n + 1, req->chunk_len); |
| 297 |
req->chunk_off = 0; |
| 298 |
|
| 299 |
if (last_nl) |
| 300 |
return true; |
| 301 |
|
| 302 |
last_nl = true; |
| 303 |
n = 0; /* start at 1 on next iteration */ |
| 304 |
} |
| 305 |
} |
| 306 |
|
| 307 |
return false; |
| 308 |
} |
| 309 |
|
| 310 |
short |
| 311 |
http_req_chunk_peek(void *cookie) |
| 312 |
{ |
| 313 |
struct http_request *req = (struct http_request *)cookie; |
| 314 |
|
| 315 |
if (req->chunk_len == 0 || (req->chunk_off + 1 > req->chunk_len)) { |
| 316 |
req->chunk_len = http_req_read(req, req->chunk, sizeof(req->chunk)); |
| 317 |
req->chunk_off = 0; |
| 318 |
} |
| 319 |
|
| 320 |
if (req->chunk_len == 0 || (req->chunk_off + 1 > req->chunk_len)) |
| 321 |
return EOF; |
| 322 |
|
| 323 |
return req->chunk[req->chunk_off]; |
| 324 |
} |
| 325 |
|
| 326 |
short |
| 327 |
http_req_chunk_read(void *cookie) |
| 328 |
{ |
| 329 |
struct http_request *req = (struct http_request *)cookie; |
| 330 |
short c; |
| 331 |
|
| 332 |
c = http_req_chunk_peek(req); |
| 333 |
if (c == EOF) |
| 334 |
return c; |
| 335 |
|
| 336 |
req->chunk_off++; |
| 337 |
|
| 338 |
return c; |
| 339 |
} |
| 340 |
|
| 341 |
void |
| 342 |
http_req_free(void *reqptr) |
| 343 |
{ |
| 344 |
unsigned long *addr = (unsigned long *)reqptr; |
| 345 |
void *ptr = (void *)*addr; |
| 346 |
struct http_request *req = (struct http_request *)ptr; |
| 347 |
|
| 348 |
if (req == NULL) |
| 349 |
return; |
| 350 |
|
| 351 |
_TCPRelease(&req->tcp_iopb, req->tcp_stream, nil, nil, false); |
| 352 |
|
| 353 |
if (req->message != NULL) |
| 354 |
xfree(&req->message); |
| 355 |
xfree(&req->tcp_buf); |
| 356 |
xfree(&req->url); |
| 357 |
xfree(&req); |
| 358 |
|
| 359 |
*addr = 0L; |
| 360 |
} |