fetch.c revision 252375 
1/*-
2 * Copyright (c) 1998-2004 Dag-Erling Sm��rgrav
3 * All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 *    notice, this list of conditions and the following disclaimer
10 *    in this position and unchanged.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 *    notice, this list of conditions and the following disclaimer in the
13 *    documentation and/or other materials provided with the distribution.
14 * 3. The name of the author may not be used to endorse or promote products
15 *    derived from this software without specific prior written permission
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
22 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27 */
28
29#include <sys/cdefs.h>
30__FBSDID("$FreeBSD: head/lib/libfetch/fetch.c 252375 2013-06-29 15:51:27Z kientzle $");
31
32#include <sys/param.h>
33#include <sys/errno.h>
34
35#include <ctype.h>
36#include <stdio.h>
37#include <stdlib.h>
38#include <string.h>
39
40#include "fetch.h"
41#include "common.h"
42
43auth_t	 fetchAuthMethod;
44int	 fetchLastErrCode;
45char	 fetchLastErrString[MAXERRSTRING];
46int	 fetchTimeout;
47int	 fetchRestartCalls = 1;
48int	 fetchDebug;
49
50
51/*** Local data **************************************************************/
52
53/*
54 * Error messages for parser errors
55 */
56#define URL_MALFORMED		1
57#define URL_BAD_SCHEME		2
58#define URL_BAD_PORT		3
59static struct fetcherr url_errlist[] = {
60	{ URL_MALFORMED,	FETCH_URL,	"Malformed URL" },
61	{ URL_BAD_SCHEME,	FETCH_URL,	"Invalid URL scheme" },
62	{ URL_BAD_PORT,		FETCH_URL,	"Invalid server port" },
63	{ -1,			FETCH_UNKNOWN,	"Unknown parser error" }
64};
65
66
67/*** Public API **************************************************************/
68
69/*
70 * Select the appropriate protocol for the URL scheme, and return a
71 * read-only stream connected to the document referenced by the URL.
72 * Also fill out the struct url_stat.
73 */
74FILE *
75fetchXGet(struct url *URL, struct url_stat *us, const char *flags)
76{
77
78	if (us != NULL) {
79		us->size = -1;
80		us->atime = us->mtime = 0;
81	}
82	if (strcasecmp(URL->scheme, SCHEME_FILE) == 0)
83		return (fetchXGetFile(URL, us, flags));
84	else if (strcasecmp(URL->scheme, SCHEME_FTP) == 0)
85		return (fetchXGetFTP(URL, us, flags));
86	else if (strcasecmp(URL->scheme, SCHEME_HTTP) == 0)
87		return (fetchXGetHTTP(URL, us, flags));
88	else if (strcasecmp(URL->scheme, SCHEME_HTTPS) == 0)
89		return (fetchXGetHTTP(URL, us, flags));
90	url_seterr(URL_BAD_SCHEME);
91	return (NULL);
92}
93
94/*
95 * Select the appropriate protocol for the URL scheme, and return a
96 * read-only stream connected to the document referenced by the URL.
97 */
98FILE *
99fetchGet(struct url *URL, const char *flags)
100{
101	return (fetchXGet(URL, NULL, flags));
102}
103
104/*
105 * Select the appropriate protocol for the URL scheme, and return a
106 * write-only stream connected to the document referenced by the URL.
107 */
108FILE *
109fetchPut(struct url *URL, const char *flags)
110{
111
112	if (strcasecmp(URL->scheme, SCHEME_FILE) == 0)
113		return (fetchPutFile(URL, flags));
114	else if (strcasecmp(URL->scheme, SCHEME_FTP) == 0)
115		return (fetchPutFTP(URL, flags));
116	else if (strcasecmp(URL->scheme, SCHEME_HTTP) == 0)
117		return (fetchPutHTTP(URL, flags));
118	else if (strcasecmp(URL->scheme, SCHEME_HTTPS) == 0)
119		return (fetchPutHTTP(URL, flags));
120	url_seterr(URL_BAD_SCHEME);
121	return (NULL);
122}
123
124/*
125 * Select the appropriate protocol for the URL scheme, and return the
126 * size of the document referenced by the URL if it exists.
127 */
128int
129fetchStat(struct url *URL, struct url_stat *us, const char *flags)
130{
131
132	if (us != NULL) {
133		us->size = -1;
134		us->atime = us->mtime = 0;
135	}
136	if (strcasecmp(URL->scheme, SCHEME_FILE) == 0)
137		return (fetchStatFile(URL, us, flags));
138	else if (strcasecmp(URL->scheme, SCHEME_FTP) == 0)
139		return (fetchStatFTP(URL, us, flags));
140	else if (strcasecmp(URL->scheme, SCHEME_HTTP) == 0)
141		return (fetchStatHTTP(URL, us, flags));
142	else if (strcasecmp(URL->scheme, SCHEME_HTTPS) == 0)
143		return (fetchStatHTTP(URL, us, flags));
144	url_seterr(URL_BAD_SCHEME);
145	return (-1);
146}
147
148/*
149 * Select the appropriate protocol for the URL scheme, and return a
150 * list of files in the directory pointed to by the URL.
151 */
152struct url_ent *
153fetchList(struct url *URL, const char *flags)
154{
155
156	if (strcasecmp(URL->scheme, SCHEME_FILE) == 0)
157		return (fetchListFile(URL, flags));
158	else if (strcasecmp(URL->scheme, SCHEME_FTP) == 0)
159		return (fetchListFTP(URL, flags));
160	else if (strcasecmp(URL->scheme, SCHEME_HTTP) == 0)
161		return (fetchListHTTP(URL, flags));
162	else if (strcasecmp(URL->scheme, SCHEME_HTTPS) == 0)
163		return (fetchListHTTP(URL, flags));
164	url_seterr(URL_BAD_SCHEME);
165	return (NULL);
166}
167
168/*
169 * Attempt to parse the given URL; if successful, call fetchXGet().
170 */
171FILE *
172fetchXGetURL(const char *URL, struct url_stat *us, const char *flags)
173{
174	struct url *u;
175	FILE *f;
176
177	if ((u = fetchParseURL(URL)) == NULL)
178		return (NULL);
179
180	f = fetchXGet(u, us, flags);
181
182	fetchFreeURL(u);
183	return (f);
184}
185
186/*
187 * Attempt to parse the given URL; if successful, call fetchGet().
188 */
189FILE *
190fetchGetURL(const char *URL, const char *flags)
191{
192	return (fetchXGetURL(URL, NULL, flags));
193}
194
195/*
196 * Attempt to parse the given URL; if successful, call fetchPut().
197 */
198FILE *
199fetchPutURL(const char *URL, const char *flags)
200{
201	struct url *u;
202	FILE *f;
203
204	if ((u = fetchParseURL(URL)) == NULL)
205		return (NULL);
206
207	f = fetchPut(u, flags);
208
209	fetchFreeURL(u);
210	return (f);
211}
212
213/*
214 * Attempt to parse the given URL; if successful, call fetchStat().
215 */
216int
217fetchStatURL(const char *URL, struct url_stat *us, const char *flags)
218{
219	struct url *u;
220	int s;
221
222	if ((u = fetchParseURL(URL)) == NULL)
223		return (-1);
224
225	s = fetchStat(u, us, flags);
226
227	fetchFreeURL(u);
228	return (s);
229}
230
231/*
232 * Attempt to parse the given URL; if successful, call fetchList().
233 */
234struct url_ent *
235fetchListURL(const char *URL, const char *flags)
236{
237	struct url *u;
238	struct url_ent *ue;
239
240	if ((u = fetchParseURL(URL)) == NULL)
241		return (NULL);
242
243	ue = fetchList(u, flags);
244
245	fetchFreeURL(u);
246	return (ue);
247}
248
249/*
250 * Make a URL
251 */
252struct url *
253fetchMakeURL(const char *scheme, const char *host, int port, const char *doc,
254    const char *user, const char *pwd)
255{
256	struct url *u;
257
258	if (!scheme || (!host && !doc)) {
259		url_seterr(URL_MALFORMED);
260		return (NULL);
261	}
262
263	if (port < 0 || port > 65535) {
264		url_seterr(URL_BAD_PORT);
265		return (NULL);
266	}
267
268	/* allocate struct url */
269	if ((u = calloc(1, sizeof(*u))) == NULL) {
270		fetch_syserr();
271		return (NULL);
272	}
273
274	if ((u->doc = strdup(doc ? doc : "/")) == NULL) {
275		fetch_syserr();
276		free(u);
277		return (NULL);
278	}
279
280#define seturl(x) snprintf(u->x, sizeof(u->x), "%s", x)
281	seturl(scheme);
282	seturl(host);
283	seturl(user);
284	seturl(pwd);
285#undef seturl
286	u->port = port;
287
288	return (u);
289}
290
291/*
292 * Return value of the given hex digit.
293 */
294static int
295fetch_hexval(char ch)
296{
297
298	if (ch >= '0' && ch <= '9')
299		return (ch - '0');
300	else if (ch >= 'a' && ch <= 'f')
301		return (ch - 'a' + 10);
302	else if (ch >= 'A' && ch <= 'F')
303		return (ch - 'A' + 10);
304	return (-1);
305}
306
307/*
308 * Decode percent-encoded URL component from src into dst, stopping at end
309 * of string, or at @ or : separators.  Returns a pointer to the unhandled
310 * part of the input string (null terminator, @, or :).  No terminator is
311 * written to dst (it is the caller's responsibility).
312 */
313static const char *
314fetch_pctdecode(char *dst, const char *src, size_t dlen)
315{
316	int d1, d2;
317	char c;
318	const char *s;
319
320	for (s = src; *s != '\0' && *s != '@' && *s != ':'; s++) {
321		if (s[0] == '%' && (d1 = fetch_hexval(s[1])) >= 0 &&
322		    (d2 = fetch_hexval(s[2])) >= 0 && (d1 > 0 || d2 > 0)) {
323			c = d1 << 4 | d2;
324			s += 2;
325		} else {
326			c = *s;
327		}
328		if (dlen-- > 0)
329			*dst++ = c;
330	}
331	return (s);
332}
333
334/*
335 * Split an URL into components. URL syntax is:
336 * [method:/][/[user[:pwd]@]host[:port]/][document]
337 * This almost, but not quite, RFC1738 URL syntax.
338 */
339struct url *
340fetchParseURL(const char *URL)
341{
342	char *doc;
343	const char *p, *q;
344	struct url *u;
345	int i;
346
347	/* allocate struct url */
348	if ((u = calloc(1, sizeof(*u))) == NULL) {
349		fetch_syserr();
350		return (NULL);
351	}
352
353	/* scheme name */
354	if ((p = strstr(URL, ":/"))) {
355		snprintf(u->scheme, URL_SCHEMELEN+1,
356		    "%.*s", (int)(p - URL), URL);
357		URL = ++p;
358		/*
359		 * Only one slash: no host, leave slash as part of document
360		 * Two slashes: host follows, strip slashes
361		 */
362		if (URL[1] == '/')
363			URL = (p += 2);
364	} else {
365		p = URL;
366	}
367	if (!*URL || *URL == '/' || *URL == '.' ||
368	    (u->scheme[0] == '\0' &&
369		strchr(URL, '/') == NULL && strchr(URL, ':') == NULL))
370		goto nohost;
371
372	p = strpbrk(URL, "/@");
373	if (p && *p == '@') {
374		/* username */
375		q = fetch_pctdecode(u->user, URL, URL_USERLEN);
376
377		/* password */
378		if (*q == ':')
379			q = fetch_pctdecode(u->pwd, q + 1, URL_PWDLEN);
380
381		p++;
382	} else {
383		p = URL;
384	}
385
386	/* hostname */
387#ifdef INET6
388	if (*p == '[' && (q = strchr(p + 1, ']')) != NULL &&
389	    (*++q == '\0' || *q == '/' || *q == ':')) {
390		if ((i = q - p - 2) > MAXHOSTNAMELEN)
391			i = MAXHOSTNAMELEN;
392		strncpy(u->host, ++p, i);
393		p = q;
394	} else
395#endif
396		for (i = 0; *p && (*p != '/') && (*p != ':'); p++)
397			if (i < MAXHOSTNAMELEN)
398				u->host[i++] = *p;
399
400	/* port */
401	if (*p == ':') {
402		for (q = ++p; *q && (*q != '/'); q++)
403			if (isdigit((unsigned char)*q))
404				u->port = u->port * 10 + (*q - '0');
405			else {
406				/* invalid port */
407				url_seterr(URL_BAD_PORT);
408				goto ouch;
409			}
410		p = q;
411	}
412
413nohost:
414	/* document */
415	if (!*p)
416		p = "/";
417
418	if (strcasecmp(u->scheme, SCHEME_HTTP) == 0 ||
419	    strcasecmp(u->scheme, SCHEME_HTTPS) == 0) {
420		const char hexnums[] = "0123456789abcdef";
421
422		/* percent-escape whitespace. */
423		if ((doc = malloc(strlen(p) * 3 + 1)) == NULL) {
424			fetch_syserr();
425			goto ouch;
426		}
427		u->doc = doc;
428		while (*p != '\0') {
429			if (!isspace((unsigned char)*p)) {
430				*doc++ = *p++;
431			} else {
432				*doc++ = '%';
433				*doc++ = hexnums[((unsigned int)*p) >> 4];
434				*doc++ = hexnums[((unsigned int)*p) & 0xf];
435				p++;
436			}
437		}
438		*doc = '\0';
439	} else if ((u->doc = strdup(p)) == NULL) {
440		fetch_syserr();
441		goto ouch;
442	}
443
444	DEBUG(fprintf(stderr,
445		  "scheme:   [%s]\n"
446		  "user:     [%s]\n"
447		  "password: [%s]\n"
448		  "host:     [%s]\n"
449		  "port:     [%d]\n"
450		  "document: [%s]\n",
451		  u->scheme, u->user, u->pwd,
452		  u->host, u->port, u->doc));
453
454	return (u);
455
456ouch:
457	free(u);
458	return (NULL);
459}
460
461/*
462 * Free a URL
463 */
464void
465fetchFreeURL(struct url *u)
466{
467	free(u->doc);
468	free(u);
469}
470