1/*	$NetBSD: content-bozo.c,v 1.22 2023/06/07 20:12:31 mrg Exp $	*/
2
3/*	$eterna: content-bozo.c,v 1.17 2011/11/18 09:21:15 mrg Exp $	*/
4
5/*
6 * Copyright (c) 1997-2023 Matthew R. Green
7 * All rights reserved.
8 *
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
11 * are met:
12 * 1. Redistributions of source code must retain the above copyright
13 *    notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 *    notice, this list of conditions and the following disclaimer and
16 *    dedication in the documentation and/or other materials provided
17 *    with the distribution.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
20 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
21 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
22 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
23 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
24 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
25 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
26 * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
27 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 *
31 */
32
33/* this code implements content-type handling for bozohttpd */
34
35#include <sys/param.h>
36
37#include <errno.h>
38#include <string.h>
39
40#include "bozohttpd.h"
41
42/*
43 * this map and the functions below map between filenames and the
44 * content type and content encoding definitions.  this should become
45 * a configuration file, perhaps like apache's mime.types (but that
46 * has less info per-entry).
47 */
48
49static bozo_content_map_t static_content_map[] = {
50	{ ".html",	"text/html",			"",		"", NULL },
51	{ ".htm",	"text/html",			"",		"", NULL },
52	{ ".gif",	"image/gif",			"",		"", NULL },
53	{ ".jpeg",	"image/jpeg",			"",		"", NULL },
54	{ ".jpg",	"image/jpeg",			"",		"", NULL },
55	{ ".jpe",	"image/jpeg",			"",		"", NULL },
56	{ ".png",	"image/png",			"",		"", NULL },
57	{ ".mp3",	"audio/mpeg",			"",		"", NULL },
58	{ ".css",	"text/css",			"",		"", NULL },
59	{ ".txt",	"text/plain",			"",		"", NULL },
60	{ ".swf",	"application/x-shockwave-flash","",		"", NULL },
61	{ ".dcr",	"application/x-director",	"",		"", NULL },
62	{ ".pac",	"application/x-ns-proxy-autoconfig", "",	"", NULL },
63	{ ".pa",	"application/x-ns-proxy-autoconfig", "",	"", NULL },
64	{ ".tar",	"multipart/x-tar",		"",		"", NULL },
65	{ ".gtar",	"application/x-gtar-compressed", "",		"", NULL },
66	{ ".tar.gz",	"application/x-gtar-compressed", "",		"", NULL },
67	{ ".tar.bz2",	"application/x-gtar-compressed", "",		"", NULL },
68	{ ".tar.xz",	"application/x-gtar-compressed", "",		"", NULL },
69	{ ".tar.lz",	"application/x-gtar-compressed", "",		"", NULL },
70	{ ".tar.zst",	"application/x-gtar-compressed", "",		"", NULL },
71	{ ".tgz",	"application/x-gtar-compressed", "",		"", NULL },
72	{ ".tbz2",	"application/x-gtar-compressed", "",		"", NULL },
73	{ ".txz",	"application/x-gtar-compressed", "",		"", NULL },
74	{ ".tlz",	"application/x-gtar-compressed", "",		"", NULL },
75	{ ".tar.z",	"application/x-gtar-compressed", "",		"", NULL },
76	{ ".tar.Z",	"application/x-gtar-compressed", "",		"", NULL },
77	{ ".taz",	"application/x-gtar-compressed", "",		"", NULL },
78	{ ".Z",		"application/x-compress",	"",		"", NULL },
79	{ ".gz",	"application/x-gzip",		"",		"", NULL },
80	{ ".z",		"unknown",			"",		"", NULL },
81	{ ".bz2",	"application/x-bzip2",		"",		"", NULL },
82	{ ".ogg",	"application/x-ogg",		"",		"", NULL },
83	{ ".mkv",	"video/x-matroska",		"",		"", NULL },
84	{ ".xbel",	"text/xml",			"",		"", NULL },
85	{ ".xml",	"text/xml",			"",		"", NULL },
86	{ ".xsl",	"text/xml",			"",		"", NULL },
87	{ ".hqx",	"application/mac-binhex40",	"",		"", NULL },
88	{ ".cpt",	"application/mac-compactpro",	"",		"", NULL },
89	{ ".doc",	"application/msword",		"",		"", NULL },
90	{ ".bin",	"application/octet-stream",	"",		"", NULL },
91	{ ".dms",	"application/octet-stream",	"",		"", NULL },
92	{ ".lha",	"application/octet-stream",	"",		"", NULL },
93	{ ".lzh",	"application/octet-stream",	"",		"", NULL },
94	{ ".exe",	"application/octet-stream",	"",		"", NULL },
95	{ ".class",	"application/octet-stream",	"",		"", NULL },
96	{ ".oda",	"application/oda",		"",		"", NULL },
97	{ ".pdf",	"application/pdf",		"",		"", NULL },
98	{ ".ai",	"application/postscript",	"",		"", NULL },
99	{ ".eps",	"application/postscript",	"",		"", NULL },
100	{ ".ps",	"application/postscript",	"",		"", NULL },
101	{ ".ppt",	"application/powerpoint",	"",		"", NULL },
102	{ ".rtf",	"application/rtf",		"",		"", NULL },
103	{ ".bcpio",	"application/x-bcpio",		"",		"", NULL },
104	{ ".torrent",	"application/x-bittorrent",	"",		"", NULL },
105	{ ".vcd",	"application/x-cdlink",		"",		"", NULL },
106	{ ".cpio",	"application/x-cpio",		"",		"", NULL },
107	{ ".csh",	"application/x-csh",		"",		"", NULL },
108	{ ".dir",	"application/x-director",	"",		"", NULL },
109	{ ".dxr",	"application/x-director",	"",		"", NULL },
110	{ ".dvi",	"application/x-dvi",		"",		"", NULL },
111	{ ".hdf",	"application/x-hdf",		"",		"", NULL },
112	{ ".cgi",	"application/x-httpd-cgi",	"",		"", NULL },
113	{ ".skp",	"application/x-koan",		"",		"", NULL },
114	{ ".skd",	"application/x-koan",		"",		"", NULL },
115	{ ".skt",	"application/x-koan",		"",		"", NULL },
116	{ ".skm",	"application/x-koan",		"",		"", NULL },
117	{ ".latex",	"application/x-latex",		"",		"", NULL },
118	{ ".mif",	"application/x-mif",		"",		"", NULL },
119	{ ".nc",	"application/x-netcdf",		"",		"", NULL },
120	{ ".cdf",	"application/x-netcdf",		"",		"", NULL },
121	{ ".patch",	"application/x-patch",		"",		"", NULL },
122	{ ".sh",	"application/x-sh",		"",		"", NULL },
123	{ ".shar",	"application/x-shar",		"",		"", NULL },
124	{ ".sit",	"application/x-stuffit",	"",		"", NULL },
125	{ ".sv4cpio",	"application/x-sv4cpio",	"",		"", NULL },
126	{ ".sv4crc",	"application/x-sv4crc",		"",		"", NULL },
127	{ ".tar",	"application/x-tar",		"",		"", NULL },
128	{ ".tcl",	"application/x-tcl",		"",		"", NULL },
129	{ ".tex",	"application/x-tex",		"",		"", NULL },
130	{ ".texinfo",	"application/x-texinfo",	"",		"", NULL },
131	{ ".texi",	"application/x-texinfo",	"",		"", NULL },
132	{ ".t",		"application/x-troff",		"",		"", NULL },
133	{ ".tr",	"application/x-troff",		"",		"", NULL },
134	{ ".roff",	"application/x-troff",		"",		"", NULL },
135	{ ".man",	"application/x-troff-man",	"",		"", NULL },
136	{ ".me",	"application/x-troff-me",	"",		"", NULL },
137	{ ".ms",	"application/x-troff-ms",	"",		"", NULL },
138	{ ".ustar",	"application/x-ustar",		"",		"", NULL },
139	{ ".src",	"application/x-wais-source",	"",		"", NULL },
140	{ ".zip",	"application/zip",		"",		"", NULL },
141	{ ".zipx",	"application/zip",		"",		"", NULL },
142	{ ".xz",	"application/x-xz",		"",		"", NULL },
143	{ ".zst",	"application/x-zstd",		"",		"", NULL },
144	{ ".sz",	"application/x-snappy-framed",	"",		"", NULL },
145	{ ".lz",	"application/x-lzip",		"",		"", NULL },
146	{ ".lzma",	"application/x-lzma",		"",		"", NULL },
147	{ ".lzo",	"application/x-lzop",		"",		"", NULL },
148	{ ".7z",	"application/x-7z-compressed",	"",		"", NULL },
149	{ ".lzo",	"application/x-lzop",		"",		"", NULL },
150	{ ".cab",	"application/vnd.ms-cab-compressed", "",	"", NULL },
151	{ ".dmg",	"application/x-apple-diskimage","",		"", NULL },
152	{ ".jar",	"application/java-archive",	"",		"", NULL },
153	{ ".rar",	"application/x-rar-compressed",	"",		"", NULL },
154	{ ".au",	"audio/basic",			"",		"", NULL },
155	{ ".snd",	"audio/basic",			"",		"", NULL },
156	{ ".mpga",	"audio/mpeg",			"",		"", NULL },
157	{ ".mp2",	"audio/mpeg",			"",		"", NULL },
158	{ ".m4a",	"audio/mp4",			"",		"", NULL },
159	{ ".aif",	"audio/x-aiff",			"",		"", NULL },
160	{ ".aiff",	"audio/x-aiff",			"",		"", NULL },
161	{ ".aifc",	"audio/x-aiff",			"",		"", NULL },
162	{ ".ram",	"audio/x-pn-realaudio",		"",		"", NULL },
163	{ ".rpm",	"audio/x-pn-realaudio-plugin",	"",		"", NULL },
164	{ ".ra",	"audio/x-realaudio",		"",		"", NULL },
165	{ ".wav",	"audio/x-wav",			"",		"", NULL },
166	{ ".pdb",	"chemical/x-pdb",		"",		"", NULL },
167	{ ".xyz",	"chemical/x-pdb",		"",		"", NULL },
168	{ ".ief",	"image/ief",			"",		"", NULL },
169	{ ".tiff",	"image/tiff",			"",		"", NULL },
170	{ ".tif",	"image/tiff",			"",		"", NULL },
171	{ ".ras",	"image/x-cmu-raster",		"",		"", NULL },
172	{ ".pnm",	"image/x-portable-anymap",	"",		"", NULL },
173	{ ".pbm",	"image/x-portable-bitmap",	"",		"", NULL },
174	{ ".pgm",	"image/x-portable-graymap",	"",		"", NULL },
175	{ ".ppm",	"image/x-portable-pixmap",	"",		"", NULL },
176	{ ".rgb",	"image/x-rgb",			"",		"", NULL },
177	{ ".xbm",	"image/x-xbitmap",		"",		"", NULL },
178	{ ".xpm",	"image/x-xpixmap",		"",		"", NULL },
179	{ ".xwd",	"image/x-xwindowdump",		"",		"", NULL },
180	{ ".rtx",	"text/richtext",		"",		"", NULL },
181	{ ".tsv",	"text/tab-separated-values",	"",		"", NULL },
182	{ ".etx",	"text/x-setext",		"",		"", NULL },
183	{ ".sgml",	"text/x-sgml",			"",		"", NULL },
184	{ ".sgm",	"text/x-sgml",			"",		"", NULL },
185	{ ".mpeg",	"video/mpeg",			"",		"", NULL },
186	{ ".mpg",	"video/mpeg",			"",		"", NULL },
187	{ ".mpe",	"video/mpeg",			"",		"", NULL },
188	{ ".ts",	"video/mpeg",			"",		"", NULL },
189	{ ".vob",	"video/mpeg",			"",		"", NULL },
190	{ ".mp4",	"video/mp4",			"",		"", NULL },
191	{ ".m4v",	"video/mp4",			"",		"", NULL },
192	{ ".qt",	"video/quicktime",		"",		"", NULL },
193	{ ".mov",	"video/quicktime",		"",		"", NULL },
194	{ ".avi",	"video/x-msvideo",		"",		"", NULL },
195	{ ".movie",	"video/x-sgi-movie",		"",		"", NULL },
196	{ ".ice",	"x-conference/x-cooltalk",	"",		"", NULL },
197	{ ".wrl",	"x-world/x-vrml",		"",		"", NULL },
198	{ ".vrml",	"x-world/x-vrml",		"",		"", NULL },
199	{ ".svg",	"image/svg+xml",		"",		"", NULL },
200	{ ".mobi",	"application/x-mobipocket-ebook", "",		"", NULL },
201	{ ".iso",	"application/octet-stream",	"",		"", NULL },
202	{ NULL,		NULL,		NULL,		NULL, NULL }
203};
204
205static bozo_content_map_t *
206search_map(bozo_content_map_t *map, const char *name, size_t len)
207{
208
209	for ( ; map && map->name; map++) {
210		const size_t namelen = strlen(map->name);
211
212		if (namelen < len &&
213		    strcasecmp(map->name, name + (len - namelen)) == 0)
214			return map;
215	}
216	return NULL;
217}
218
219/* match a suffix on a file - dynamiconly means no static content search */
220bozo_content_map_t *
221bozo_match_content_map(bozohttpd_t *httpd, const char *name,
222		       const int dynamiconly)
223{
224	bozo_content_map_t	*map;
225	size_t			 len;
226
227	len = strlen(name);
228	map = search_map(httpd->dynamic_content_map, name, len);
229	if (map == NULL && !dynamiconly)
230		map = search_map(static_content_map, name, len);
231
232	return map;
233}
234
235/*
236 * given the file name, return a valid Content-Type: value.
237 */
238const char *
239bozo_content_type(bozo_httpreq_t *request, const char *file)
240{
241	bozohttpd_t *httpd = request->hr_httpd;
242	bozo_content_map_t	*map;
243
244	map = bozo_match_content_map(httpd, file, 0);
245	if (map)
246		return map->type;
247	return httpd->consts.text_plain;
248}
249
250/*
251 * given the file name, return a valid Content-Encoding: value.
252 */
253const char *
254bozo_content_encoding(bozo_httpreq_t *request, const char *file)
255{
256	bozohttpd_t *httpd = request->hr_httpd;
257	bozo_content_map_t	*map;
258
259	map = bozo_match_content_map(httpd, file, 0);
260	if (map)
261		return (request->hr_proto == httpd->consts.http_11) ?
262		    map->encoding11 : map->encoding;
263	return NULL;
264}
265
266#ifndef NO_DYNAMIC_CONTENT
267
268bozo_content_map_t *
269bozo_get_content_map(bozohttpd_t *httpd, const char *name)
270{
271	bozo_content_map_t	*map;
272
273	if ((map = bozo_match_content_map(httpd, name, 1)) != NULL)
274		return map;
275
276	httpd->dynamic_content_map_size++;
277	httpd->dynamic_content_map = bozorealloc(httpd,
278		httpd->dynamic_content_map,
279		(httpd->dynamic_content_map_size + 1) * sizeof *map);
280	if (httpd->dynamic_content_map == NULL)
281		bozoerr(httpd, 1, "out of memory allocating content map");
282	map = &httpd->dynamic_content_map[httpd->dynamic_content_map_size];
283	map->name = map->type = map->encoding = map->encoding11 =
284		map->cgihandler = NULL;
285	map--;
286
287	return map;
288}
289
290/*
291 * mime content maps look like:
292 *	".name type encoding encoding11"
293 * where any of type, encoding or encoding11 a dash "-" means "".
294 * eg the .gtar, .tar.Z from above  could be written like:
295 *	".gtar multipart/x-gtar - -"
296 *	".tar.Z multipart/x-tar x-compress compress"
297 * or
298 *	".gtar multipart/x-gtar"
299 *	".tar.Z multipart/x-tar x-compress compress"
300 * NOTE: we destroy 'arg'
301 */
302void
303bozo_add_content_map_mime(bozohttpd_t *httpd, const char *cmap0,
304		const char *cmap1, const char *cmap2, const char *cmap3)
305{
306	bozo_content_map_t *map;
307
308	debug((httpd, DEBUG_FAT,
309		"add_content_map: name %s type %s enc %s enc11 %s ",
310		cmap0, cmap1, cmap2, cmap3));
311
312	map = bozo_get_content_map(httpd, cmap0);
313#define CHECKMAP(s)	(!s || ((s)[0] == '-' && (s)[1] == '\0') ? "" : (s))
314	map->name = CHECKMAP(cmap0);
315	map->type = CHECKMAP(cmap1);
316	map->encoding = CHECKMAP(cmap2);
317	map->encoding11 = CHECKMAP(cmap3);
318#undef CHECKMAP
319	map->cgihandler = NULL;
320}
321#endif /* NO_DYNAMIC_CONTENT */
322