xref: /freebsd/lib/libfetch/fetch.3 (revision ee41f1b1cf5e3d4f586cb85b46123b416275862c)
1.\" Copyright (c) 1998 Dag-Erling Co�dan Sm�rgrav
2.\" All rights reserved.
3.\"
4.\" Redistribution and use in source and binary forms, with or without
5.\" modification, are permitted provided that the following conditions
6.\" are met:
7.\" 1. Redistributions of source code must retain the above copyright
8.\"    notice, this list of conditions and the following disclaimer.
9.\" 2. Redistributions in binary form must reproduce the above copyright
10.\"    notice, this list of conditions and the following disclaimer in the
11.\"    documentation and/or other materials provided with the distribution.
12.\"
13.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
14.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
16.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
17.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
18.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
19.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
20.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
21.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
22.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
23.\" SUCH DAMAGE.
24.\"
25.\" $FreeBSD$
26.\"
27.Dd July 1, 1998
28.Dt FETCH 3
29.Os
30.Sh NAME
31.Nm fetchMakeURL ,
32.Nm fetchParseURL ,
33.Nm fetchFreeURL ,
34.Nm fetchXGetURL ,
35.Nm fetchGetURL ,
36.Nm fetchPutURL ,
37.Nm fetchStatURL ,
38.Nm fetchListURL ,
39.Nm fetchXGet ,
40.Nm fetchGet ,
41.Nm fetchPut ,
42.Nm fetchStat ,
43.Nm fetchList ,
44.Nm fetchXGetFile ,
45.Nm fetchGetFile ,
46.Nm fetchPutFile ,
47.Nm fetchStatFile ,
48.Nm fetchListFile ,
49.Nm fetchXGetHTTP ,
50.Nm fetchGetHTTP ,
51.Nm fetchPutHTTP ,
52.Nm fetchStatHTTP ,
53.Nm fetchListHTTP ,
54.Nm fetchXGetFTP ,
55.Nm fetchGetFTP ,
56.Nm fetchPutFTP ,
57.Nm fetchStatFTP ,
58.Nm fetchListFTP
59.Nd file transfer functions
60.Sh LIBRARY
61.Lb libfetch
62.Sh SYNOPSIS
63.Fd #include <sys/param.h>
64.Fd #include <stdio.h>
65.Fd #include <fetch.h>
66.Ft struct url *
67.Fn fetchMakeURL "char *scheme" "char *host" "int port" "char *doc" "char *user" "char *pwd"
68.Ft struct url *
69.Fn fetchParseURL "char *URL"
70.Ft void
71.Fn fetchFreeURL "struct url *URL"
72.Ft FILE *
73.Fn fetchXGetURL "char *URL" "struct url_stat *us" "char *flags"
74.Ft FILE *
75.Fn fetchGetURL "char *URL" "char *flags"
76.Ft FILE *
77.Fn fetchPutURL "char *URL" "char *flags"
78.Ft int
79.Fn fetchStatURL "char *URL" "struct url_stat *us" "char *flags"
80.Ft struct url_ent *
81.Fn fetchListURL "char *URL" "char *flags"
82.Ft FILE *
83.Fn fetchXGet "struct url *URL" "struct url_stat *us" "char *flags"
84.Ft FILE *
85.Fn fetchGet "struct url *URL" "char *flags"
86.Ft FILE *
87.Fn fetchPut "struct url *URL" "char *flags"
88.Ft int
89.Fn fetchStat "struct url *URL" "struct url_stat *us" "char *flags"
90.Ft struct url_ent *
91.Fn fetchList "struct url *" "char *flags"
92.Ft FILE *
93.Fn fetchXGetFile "struct url *u" "struct url_stat *us" "char *flags"
94.Ft FILE *
95.Fn fetchGetFile "struct url *u" "char *flags"
96.Ft FILE *
97.Fn fetchPutFile "struct url *u" "char *flags"
98.Ft int
99.Fn fetchStatFile "struct url *URL" "struct url_stat *us" "char *flags"
100.Ft struct url_ent *
101.Fn fetchListFile "struct url *" "char *flags"
102.Ft FILE *
103.Fn fetchXGetHTTP "struct url *u" "struct url_stat *us" "char *flags"
104.Ft FILE *
105.Fn fetchGetHTTP "struct url *u" "char *flags"
106.Ft FILE *
107.Fn fetchPutHTTP "struct url *u" "char *flags"
108.Ft int
109.Fn fetchStatHTTP "struct url *URL" "struct url_stat *us" "char *flags"
110.Ft struct url_ent *
111.Fn fetchListHTTP "struct url *" "char *flags"
112.Ft FILE *
113.Fn fetchXGetFTP "struct url *u" "struct url_stat *us" "char *flags"
114.Ft FILE *
115.Fn fetchGetFTP "struct url *u" "char *flags"
116.Ft FILE *
117.Fn fetchPutFTP "struct url *u" "char *flags"
118.Ft int
119.Fn fetchStatFTP "struct url *URL" "struct url_stat *us" "char *flags"
120.Ft struct url_ent *
121.Fn fetchListFTP "struct url *" "char *flags"
122.Sh DESCRIPTION
123.Pp
124These functions implement a high-level library for retrieving and
125uploading files using Uniform Resource Locators (URLs).
126.Pp
127.Fn fetchParseURL
128takes a URL in the form of a null-terminated string and splits it into
129its components function according to the Common Internet Scheme Syntax
130detailed in RFC1738.
131A regular expression which produces this syntax is:
132.Bd -literal
133    <scheme>:(//(<user>(:<pwd>)?@)?<host>(:<port>)?)?/(<document>)?
134.Ed
135.Pp
136If the URL does not seem to begin with a scheme name, the following
137syntax is assumed:
138.Bd -literal
139    ((<user>(:<pwd>)?@)?<host>(:<port>)?)?/(<document>)?
140.Ed
141.Pp
142Note that some components of the URL are not necessarily relevant to
143all URL schemes.
144For instance, the file scheme only needs the <scheme> and <document>
145components.
146.Pp
147.Fn fetchMakeURL
148and
149.Fn fetchParseURL
150return a pointer to a
151.Fa url
152structure, which is defined as follows in
153.Aq Pa fetch.h :
154.Bd -literal
155#define URL_SCHEMELEN 16
156#define URL_USERLEN 256
157#define URL_PWDLEN 256
158
159struct url {
160    char	 scheme[URL_SCHEMELEN+1];
161    char	 user[URL_USERLEN+1];
162    char	 pwd[URL_PWDLEN+1];
163    char	 host[MAXHOSTNAMELEN+1];
164    int		 port;
165    char	*doc;
166    off_t	 offset;
167    size_t	 length;
168};
169.Ed
170.Pp
171The pointer returned by
172.Fn fetchMakeURL
173or
174.Fn fetchParseURL
175should be freed using
176.Fn fetchFreeURL .
177.Pp
178.Fn fetchXGetURL ,
179.Fn fetchGetURL ,
180and
181.Fn fetchPutURL
182constitute the recommended interface to the
183.Nm fetch
184library.
185They examine the URL passed to them to determine the transfer
186method, and call the appropriate lower-level functions to perform the
187actual transfer.
188.Fn fetchXGetURL
189also returns the remote document's metadata in the
190.Fa url_stat
191structure pointed to by the
192.Fa us
193argument.
194.Pp
195The
196.Fa flags
197argument is a string of characters which specify transfer options.
198The
199meaning of the individual flags is scheme-dependent, and is detailed
200in the appropriate section below.
201.Pp
202.Fn fetchStatURL
203attempts to obtain the requested document's metadata and fill in the
204structure pointed to by it's second argument.
205The
206.Fa url_stat
207structure is defined as follows in
208.Aq Pa fetch.h :
209.Bd -literal
210struct url_stat {
211    off_t	 size;
212    time_t	 atime;
213    time_t	 mtime;
214};
215.Ed
216.Pp
217If the size could not be obtained from the server, the
218.Fa size
219field is set to -1.
220If the modification time could not be obtained from the server, the
221.Fa mtime
222field is set to the epoch.
223If the access time could not be obtained from the server, the
224.Fa atime
225field is set to the modification time.
226.Pp
227.Fn fetchListURL
228attempts to list the contents of the directory pointed to by the URL
229provided.
230If successful, it returns a malloced array of
231.Fa url_ent
232structures.
233The
234.Fa url_ent
235structure is defined as follows in
236.Aq Pa fetch.h :
237.Bd -literal
238struct url_ent {
239    char         name[MAXPATHLEN];
240    struct url_stat stat;
241};
242.Ed
243.Pp
244The list is terminated by an entry with an empty name.
245.Pp
246The pointer returned by
247.Fn fetchListURL
248should be freed using
249.Fn free .
250.Pp
251.Fn fetchXGet ,
252.Fn fetchGet ,
253.Fn fetchPut
254and
255.Fn fetchStat
256are similar to
257.Fn fetchXGetURL ,
258.Fn fetchGetURL ,
259.Fn fetchPutURL
260and
261.Fn fetchStatURL ,
262except that they expect a pre-parsed URL in the form of a pointer to
263a
264.Fa struct url
265rather than a string.
266.Pp
267All of the
268.Fn fetchXGetXXX ,
269.Fn fetchGetXXX
270and
271.Fn fetchPutXXX
272functions return a pointer to a stream which can be used to read or
273write data from or to the requested document, respectively.
274Note that
275although the implementation details of the individual access methods
276vary, it can generally be assumed that a stream returned by one of the
277.Fn fetchXGetXXX
278or
279.Fn fetchGetXXX
280functions is read-only, and that a stream returned by one of the
281.Fn fetchPutXXX
282functions is write-only.
283.Sh FILE SCHEME
284.Fn fetchXGetFile ,
285.Fn fetchGetFile
286and
287.Fn fetchPutFile
288provide access to documents which are files in a locally mounted file
289system.
290Only the <document> component of the URL is used.
291.Pp
292.Fn fetchXGetFile
293and
294.Fn fetchGetFile
295do not accept any flags.
296.Pp
297.Fn fetchPutFile
298accepts the
299.Fa a
300(append to file) flag.
301If that flag is specified, the data written to
302the stream returned by
303.Fn fetchPutFile
304will be appended to the previous contents of the file, instead of
305replacing them.
306.Sh FTP SCHEME
307.Fn fetchXGetFTP ,
308.Fn fetchGetFTP
309and
310.Fn fetchPutFTP
311implement the FTP protocol as described in RFC959.
312.Pp
313If the
314.Fa p
315(passive) flag is specified, a passive (rather than active) connection
316will be attempted.
317.Pp
318If the
319.Fa h
320(high) flag is specified, data sockets will be allocated in the high
321port range (see
322.Xr ip 4 ) .
323.Pp
324If the
325.Fa d
326(direct) flag is specified,
327.Fn fetchXGetFTP ,
328.Fn fetchGetFTP
329and
330.Fn fetchPutFTP
331will use a direct connection even if a proxy server is defined.
332.Pp
333If no user name or password is given, the
334.Nm fetch
335library will attempt an anonymous login, with user name "anonymous"
336and password "anonymous@<hostname>".
337.Sh HTTP SCHEME
338The
339.Fn fetchXGetHTTP ,
340.Fn fetchGetHTTP
341and
342.Fn fetchPutHTTP
343functions implement the HTTP/1.1 protocol.
344With a little luck, there's
345even a chance that they comply with RFC2068.
346.Pp
347If the
348.Fa d
349(direct) flag is specified,
350.Fn fetchXGetHTTP ,
351.Fn fetchGetHTTP
352and
353.Fn fetchPutHTTP
354will use a direct connection even if a proxy server is defined.
355.Pp
356Since there seems to be no good way of implementing the HTTP PUT
357method in a manner consistent with the rest of the
358.Nm fetch
359library,
360.Fn fetchPutHTTP
361is currently unimplemented.
362.Sh RETURN VALUES
363.Fn fetchParseURL
364returns a pointer to a
365.Fa struct url
366containing the individual components of the URL.
367If it is
368unable to allocate memory, or the URL is syntactically incorrect,
369.Fn fetchParseURL
370returns a NULL pointer.
371.Pp
372The
373.Fn fetchStat
374functions return 0 on success and -1 on failure.
375.Pp
376All other functions return a stream pointer which may be used to
377access the requested document, or NULL if an error occurred.
378.Pp
379The following error codes are defined in
380.Aq Pa fetch.h :
381.Bl -tag -width 18n
382.It Bq Er FETCH_ABORT
383Operation aborted
384.It Bq Er FETCH_AUTH
385Authentication failed
386.It Bq Er FETCH_DOWN
387Service unavailable
388.It Bq Er FETCH_EXISTS
389File exists
390.It Bq Er FETCH_FULL
391File system full
392.It Bq Er FETCH_INFO
393Informational response
394.It Bq Er FETCH_MEMORY
395Insufficient memory
396.It Bq Er FETCH_MOVED
397File has moved
398.It Bq Er FETCH_NETWORK
399Network error
400.It Bq Er FETCH_OK
401No error
402.It Bq Er FETCH_PROTO
403Protocol error
404.It Bq Er FETCH_RESOLV
405Resolver error
406.It Bq Er FETCH_SERVER
407Server error
408.It Bq Er FETCH_TEMP
409Temporary error
410.It Bq Er FETCH_TIMEOUT
411Operation timed out
412.It Bq Er FETCH_UNAVAIL
413File is not available
414.It Bq Er FETCH_UNKNOWN
415Unknown error
416.It Bq Er FETCH_URL
417Invalid URL
418.El
419.Pp
420The accompanying error message includes a protocol-specific error code
421and message, e.g. "File is not available (404 Not Found)"
422.Sh ENVIRONMENT
423.Bl -tag -width FTP_PASSIVE_MODE
424.It Ev FTP_LOGIN
425Default FTP login if none was provided in the URL.
426.It Ev FTP_PASSIVE_MODE
427If set to anything but
428.Ql no ,
429forces the FTP code to use passive mode.
430.It Ev FTP_PASSWORD
431Default FTP password if the remote server requests one and none was
432provided in the URL.
433.It Ev FTP_PROXY
434URL of the proxy to use for FTP requests.
435The document part is ignored.
436FTP and HTTP proxies are supported; if no scheme is specified, FTP is
437assumed.
438If the proxy is an FTP proxy,
439.Nm libfetch
440will send
441.Ql user@host
442as user name to the proxy, where
443.Ql user
444is the real user name, and
445.Ql host
446is the name of the FTP server.
447.Pp
448If this variable is set to an empty string, no proxy will be used for
449FTP requests, even if the
450.Ev HTTP_PROXY
451variable is set.
452.It Ev HTTP_AUTH
453Specifies HTTP authorization parameters as a colon-separated list of
454items.
455The first and second item are the authorization scheme and realm
456respectively; further items are scheme-dependent.
457Currently, only basic authorization is supported.
458.Pp
459Basic authorization requires two parameters: the user name and
460password, in that order.
461.Pp
462This variable is only used if the server requires authorization and
463no user name or password was specified in the URL.
464.It Ev HTTP_PROXY
465URL of the proxy to use for HTTP requests.
466The document part is ignored.
467Only HTTP proxies are supported for HTTP requests.
468If no port number is specified, the default is 3128.
469.Pp
470Note that this proxy will also be used for FTP documents, unless the
471.Ev FTP_PROXY
472variable is set.
473.It Ev HTTP_PROXY_AUTH
474Specifies authorization parameters for the HTTP proxy in the same
475format as the
476.Ev HTTP_AUTH
477variable.
478.Pp
479This variable is used if and only if connected to an HTTP proxy, and
480is ignored if a user and/or a password were specified in the proxy
481URL.
482.El
483.Sh SEE ALSO
484.Xr fetch 1 ,
485.Xr ftpio 3 ,
486.Xr ip 4
487.Rs
488.%A J. Postel
489.%A J. K. Reynolds
490.%D October 1985
491.%B File Transfer Protocol
492.%O RFC959
493.Re
494.Rs
495.%A P. Deutsch
496.%A A. Emtage
497.%A A. Marine.
498.%D May 1994
499.%T How to Use Anonymous FTP
500.%O RFC1635
501.Re
502.Rs
503.%A T. Berners-Lee
504.%A L. Masinter
505.%A M. McCahill
506.%D December 1994
507.%T Uniform Resource Locators (URL)
508.%O RFC1738
509.Re
510.Rs
511.%A R. Fielding
512.%A J. Gettys
513.%A J. Mogul
514.%A H. Frystyk
515.%A T. Berners-Lee
516.%D Januray 1997
517.%B Hypertext Transfer Protocol -- HTTP/1.1
518.%O RFC2068
519.Re
520.Sh HISTORY
521The
522.Nm fetch
523library first appeared in
524.Fx 3.0 .
525.Sh AUTHORS
526.An -nosplit
527The
528.Nm fetch
529library was mostly written by
530.An Dag-Erling Co\(:idan Sm\(/orgrav Aq des@FreeBSD.org
531with numerous suggestions from
532.An Jordan K. Hubbard Aq jkh@FreeBSD.org ,
533.An Eugene Skepner Aq eu@qub.com
534and other
535.Fx
536developers.
537It replaces the older
538.Nm ftpio
539library written by
540.An Poul-Henning Kamp Aq pkh@FreeBSD.org
541and
542.An Jordan K. Hubbard Aq jkh@FreeBSD.org .
543.Pp
544This manual page was written by
545.An Dag-Erling Co\(:idan Sm\(/orgrav Aq des@FreeBSD.org .
546.Sh BUGS
547Some parts of the library are not yet implemented.
548The most notable
549examples of this are
550.Fn fetchPutHTTP ,
551.Fn fetchListHTTP ,
552.Fn fetchListFTP
553and FTP proxy support.
554.Pp
555There's no way to select a proxy at run-time other than setting the
556.Ev HTTP_PROXY
557or
558.Ev FTP_PROXY
559environment variables as appropriate.
560.Pp
561.Nm libfetch
562does not understand or obey 305 (Use Proxy) replies.
563.Pp
564Error numbers are unique only within a certain context; the error
565codes used for FTP and HTTP overlap, as do those used for resolver and
566system errors.
567For instance, error code 202 means "Command not
568implemented, superfluous at this site" in an FTP context and
569"Accepted" in an HTTP context.
570.Pp
571.Fn fetchStatFTP
572does not check that the result of an MDTM command is a valid date.
573.Pp
574The man page is incomplete, poorly written and produces badly
575formatted text.
576.Pp
577The error reporting mechanism is unsatisfactory.
578.Pp
579Some parts of the code are not fully reentrant.
580