fetch.c revision 63340 
1/*-
2 * Copyright (c) 1998 Dag-Erling Co�dan 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 * $FreeBSD: head/lib/libfetch/fetch.c 63340 2000-07-17 21:25:00Z des $
29 */
30
31#include <sys/param.h>
32#include <sys/errno.h>
33
34#include <ctype.h>
35#include <stdio.h>
36#include <stdlib.h>
37#include <string.h>
38
39#include "fetch.h"
40#include "common.h"
41
42
43int	 fetchLastErrCode;
44char	 fetchLastErrString[MAXERRSTRING];
45int	 fetchTimeout;
46int	 fetchRestartCalls = 1;
47
48
49/*** Local data **************************************************************/
50
51/*
52 * Error messages for parser errors
53 */
54#define URL_MALFORMED		1
55#define URL_BAD_SCHEME		2
56#define URL_BAD_PORT		3
57static struct fetcherr _url_errlist[] = {
58    { URL_MALFORMED,	FETCH_URL,	"Malformed URL" },
59    { URL_BAD_SCHEME,	FETCH_URL,	"Invalid URL scheme" },
60    { URL_BAD_PORT,	FETCH_URL,	"Invalid server port" },
61    { -1,		FETCH_UNKNOWN,	"Unknown parser error" }
62};
63
64
65/*** Public API **************************************************************/
66
67/*
68 * Select the appropriate protocol for the URL scheme, and return a
69 * read-only stream connected to the document referenced by the URL.
70 * Also fill out the struct url_stat.
71 */
72FILE *
73fetchXGet(struct url *URL, struct url_stat *us, char *flags)
74{
75    int direct;
76
77    direct = (flags && strchr(flags, 'd'));
78    if (strcasecmp(URL->scheme, "file") == 0)
79	return fetchXGetFile(URL, us, flags);
80    else if (strcasecmp(URL->scheme, "http") == 0)
81	return fetchXGetHTTP(URL, us, flags);
82    else if (strcasecmp(URL->scheme, "ftp") == 0) {
83	if (!direct &&
84	    getenv("FTP_PROXY") == NULL && getenv("HTTP_PROXY") != NULL)
85	    return fetchXGetHTTP(URL, us, flags);
86	return fetchXGetFTP(URL, us, flags);
87    } else {
88	_url_seterr(URL_BAD_SCHEME);
89	return NULL;
90    }
91}
92
93/*
94 * Select the appropriate protocol for the URL scheme, and return a
95 * read-only stream connected to the document referenced by the URL.
96 */
97FILE *
98fetchGet(struct url *URL, char *flags)
99{
100    return fetchXGet(URL, NULL, flags);
101}
102
103/*
104 * Select the appropriate protocol for the URL scheme, and return a
105 * write-only stream connected to the document referenced by the URL.
106 */
107FILE *
108fetchPut(struct url *URL, char *flags)
109{
110    int direct;
111
112    direct = (flags && strchr(flags, 'd'));
113    if (strcasecmp(URL->scheme, "file") == 0)
114	return fetchPutFile(URL, flags);
115    else if (strcasecmp(URL->scheme, "http") == 0)
116	return fetchPutHTTP(URL, flags);
117    else if (strcasecmp(URL->scheme, "ftp") == 0) {
118	if (!direct &&
119	    getenv("FTP_PROXY") == NULL && getenv("HTTP_PROXY") != NULL)
120	    return fetchPutHTTP(URL, flags);
121	return fetchPutFTP(URL, flags);
122    } else {
123	_url_seterr(URL_BAD_SCHEME);
124	return NULL;
125    }
126}
127
128/*
129 * Select the appropriate protocol for the URL scheme, and return the
130 * size of the document referenced by the URL if it exists.
131 */
132int
133fetchStat(struct url *URL, struct url_stat *us, char *flags)
134{
135    int direct;
136
137    direct = (flags && strchr(flags, 'd'));
138    if (strcasecmp(URL->scheme, "file") == 0)
139	return fetchStatFile(URL, us, flags);
140    else if (strcasecmp(URL->scheme, "http") == 0)
141	return fetchStatHTTP(URL, us, flags);
142    else if (strcasecmp(URL->scheme, "ftp") == 0) {
143	if (!direct &&
144	    getenv("FTP_PROXY") == NULL && getenv("HTTP_PROXY") != NULL)
145	    return fetchStatHTTP(URL, us, flags);
146	return fetchStatFTP(URL, us, flags);
147    } else {
148	_url_seterr(URL_BAD_SCHEME);
149	return -1;
150    }
151}
152
153/*
154 * Select the appropriate protocol for the URL scheme, and return a
155 * list of files in the directory pointed to by the URL.
156 */
157struct url_ent *
158fetchList(struct url *URL, char *flags)
159{
160    int direct;
161
162    direct = (flags && strchr(flags, 'd'));
163    if (strcasecmp(URL->scheme, "file") == 0)
164	return fetchListFile(URL, flags);
165    else if (strcasecmp(URL->scheme, "http") == 0)
166	return fetchListHTTP(URL, flags);
167    else if (strcasecmp(URL->scheme, "ftp") == 0) {
168	if (!direct &&
169	    getenv("FTP_PROXY") == NULL && getenv("HTTP_PROXY") != NULL)
170	    return fetchListHTTP(URL, flags);
171	return fetchListFTP(URL, flags);
172    } else {
173	_url_seterr(URL_BAD_SCHEME);
174	return NULL;
175    }
176}
177
178/*
179 * Attempt to parse the given URL; if successful, call fetchXGet().
180 */
181FILE *
182fetchXGetURL(char *URL, struct url_stat *us, char *flags)
183{
184    struct url *u;
185    FILE *f;
186
187    if ((u = fetchParseURL(URL)) == NULL)
188	return NULL;
189
190    f = fetchXGet(u, us, flags);
191
192    fetchFreeURL(u);
193    return f;
194}
195
196/*
197 * Attempt to parse the given URL; if successful, call fetchGet().
198 */
199FILE *
200fetchGetURL(char *URL, char *flags)
201{
202    return fetchXGetURL(URL, NULL, flags);
203}
204
205/*
206 * Attempt to parse the given URL; if successful, call fetchPut().
207 */
208FILE *
209fetchPutURL(char *URL, char *flags)
210{
211    struct url *u;
212    FILE *f;
213
214    if ((u = fetchParseURL(URL)) == NULL)
215	return NULL;
216
217    f = fetchPut(u, flags);
218
219    fetchFreeURL(u);
220    return f;
221}
222
223/*
224 * Attempt to parse the given URL; if successful, call fetchStat().
225 */
226int
227fetchStatURL(char *URL, struct url_stat *us, char *flags)
228{
229    struct url *u;
230    int s;
231
232    if ((u = fetchParseURL(URL)) == NULL)
233	return -1;
234
235    s = fetchStat(u, us, flags);
236
237    fetchFreeURL(u);
238    return s;
239}
240
241/*
242 * Attempt to parse the given URL; if successful, call fetchList().
243 */
244struct url_ent *
245fetchListURL(char *URL, char *flags)
246{
247    struct url *u;
248    struct url_ent *ue;
249
250    if ((u = fetchParseURL(URL)) == NULL)
251	return NULL;
252
253    ue = fetchList(u, flags);
254
255    fetchFreeURL(u);
256    return ue;
257}
258
259/*
260 * Make a URL
261 */
262struct url *
263fetchMakeURL(char *scheme, char *host, int port, char *doc,
264    char *user, char *pwd)
265{
266    struct url *u;
267
268    if (!scheme || (!host && !doc)) {
269	_url_seterr(URL_MALFORMED);
270	return NULL;
271    }
272
273    if (port < 0 || port > 65535) {
274	_url_seterr(URL_BAD_PORT);
275	return NULL;
276    }
277
278    /* allocate struct url */
279    if ((u = calloc(1, sizeof *u)) == NULL) {
280	_fetch_syserr();
281	return NULL;
282    }
283
284    if ((u->doc = strdup(doc ? doc : "/")) == NULL) {
285	_fetch_syserr();
286	free(u);
287	return NULL;
288    }
289
290#define seturl(x) snprintf(u->x, sizeof u->x, "%s", x)
291    seturl(scheme);
292    seturl(host);
293    seturl(user);
294    seturl(pwd);
295#undef seturl
296    u->port = port;
297
298    return u;
299}
300
301/*
302 * Split an URL into components. URL syntax is:
303 * method:[//[user[:pwd]@]host[:port]]/[document]
304 * This almost, but not quite, RFC1738 URL syntax.
305 */
306struct url *
307fetchParseURL(char *URL)
308{
309    char *p, *q;
310    struct url *u;
311    int i;
312
313    /* allocate struct url */
314    if ((u = calloc(1, sizeof *u)) == NULL) {
315	_fetch_syserr();
316	return NULL;
317    }
318
319    /* scheme name */
320    for (i = 0; *URL && (*URL != ':'); URL++)
321	if (i < URL_SCHEMELEN)
322	    u->scheme[i++] = *URL;
323    if (!URL[0] || (URL[1] != '/')) {
324	_url_seterr(URL_BAD_SCHEME);
325	goto ouch;
326    }
327    else URL++;
328    if (URL[1] != '/') {
329	p = URL;
330	goto nohost;
331    }
332    else URL += 2;
333
334    p = strpbrk(URL, "/@");
335    if (p && *p == '@') {
336	/* username */
337	for (q = URL, i = 0; (*q != ':') && (*q != '@'); q++)
338	    if (i < URL_USERLEN)
339		u->user[i++] = *q;
340
341	/* password */
342	if (*q == ':')
343	    for (q++, i = 0; (*q != ':') && (*q != '@'); q++)
344		if (i < URL_PWDLEN)
345		    u->pwd[i++] = *q;
346
347	p++;
348    } else p = URL;
349
350    /* hostname */
351#ifdef INET6
352    if (*p == '[' && (q = strchr(p + 1, ']')) != NULL &&
353	(*++q == '\0' || *q == '/' || *q == ':')) {
354	if ((i = q - p - 2) > MAXHOSTNAMELEN)
355	    i = MAXHOSTNAMELEN;
356	strncpy(u->host, ++p, i);
357	p = q;
358    } else
359#endif
360	for (i = 0; *p && (*p != '/') && (*p != ':'); p++)
361	    if (i < MAXHOSTNAMELEN)
362		u->host[i++] = *p;
363
364    /* port */
365    if (*p == ':') {
366	for (q = ++p; *q && (*q != '/'); q++)
367	    if (isdigit(*q))
368		u->port = u->port * 10 + (*q - '0');
369	    else {
370		/* invalid port */
371		_url_seterr(URL_BAD_PORT);
372		goto ouch;
373	    }
374	while (*p && (*p != '/'))
375	    p++;
376    }
377
378nohost:
379    /* document */
380    if (!*p)
381	p = "/";
382
383    if ((u->doc = strdup(p)) == NULL) {
384	_fetch_syserr();
385	goto ouch;
386    }
387
388    DEBUG(fprintf(stderr,
389		  "scheme:   [\033[1m%s\033[m]\n"
390		  "user:     [\033[1m%s\033[m]\n"
391		  "password: [\033[1m%s\033[m]\n"
392		  "host:     [\033[1m%s\033[m]\n"
393		  "port:     [\033[1m%d\033[m]\n"
394		  "document: [\033[1m%s\033[m]\n",
395		  u->scheme, u->user, u->pwd,
396		  u->host, u->port, u->doc));
397
398    return u;
399
400ouch:
401    free(u);
402    return NULL;
403}
404
405/*
406 * Free a URL
407 */
408void
409fetchFreeURL(struct url *u)
410{
411    free(u->doc);
412    free(u);
413}
414