xref: /minix/bin/dd/conv.c (revision 9f988b79)
1 /*	$NetBSD: conv.c,v 1.17 2003/08/07 09:05:10 agc Exp $	*/
2 
3 /*-
4  * Copyright (c) 1991, 1993, 1994
5  *	The Regents of the University of California.  All rights reserved.
6  *
7  * This code is derived from software contributed to Berkeley by
8  * Keith Muller of the University of California, San Diego and Lance
9  * Visser of Convex Computer Corporation.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  * 3. Neither the name of the University nor the names of its contributors
20  *    may be used to endorse or promote products derived from this software
21  *    without specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33  * SUCH DAMAGE.
34  */
35 
36 #include <sys/cdefs.h>
37 #ifndef lint
38 #if 0
39 static char sccsid[] = "@(#)conv.c	8.3 (Berkeley) 4/2/94";
40 #else
41 __RCSID("$NetBSD: conv.c,v 1.17 2003/08/07 09:05:10 agc Exp $");
42 #endif
43 #endif /* not lint */
44 
45 #include <sys/param.h>
46 #include <sys/time.h>
47 
48 #include <err.h>
49 #include <string.h>
50 #include <stdlib.h>
51 
52 #include "dd.h"
53 #include "extern.h"
54 
55 /*
56  * def --
57  * Copy input to output.  Input is buffered until reaches obs, and then
58  * output until less than obs remains.  Only a single buffer is used.
59  * Worst case buffer calculation is (ibs + obs - 1).
60  */
61 void
62 def(void)
63 {
64 	uint64_t cnt;
65 	u_char *inp;
66 	const u_char *t;
67 
68 	if ((t = ctab) != NULL)
69 		for (inp = in.dbp - (cnt = in.dbrcnt); cnt--; ++inp)
70 			*inp = t[*inp];
71 
72 	/* Make the output buffer look right. */
73 	out.dbp = in.dbp;
74 	out.dbcnt = in.dbcnt;
75 
76 	if (in.dbcnt >= out.dbsz) {
77 		/* If the output buffer is full, write it. */
78 		dd_out(0);
79 
80 		/*
81 		 * Ddout copies the leftover output to the beginning of
82 		 * the buffer and resets the output buffer.  Reset the
83 		 * input buffer to match it.
84 	 	 */
85 		in.dbp = out.dbp;
86 		in.dbcnt = out.dbcnt;
87 	}
88 }
89 
90 void
91 def_close(void)
92 {
93 
94 	/* Just update the count, everything is already in the buffer. */
95 	if (in.dbcnt)
96 		out.dbcnt = in.dbcnt;
97 }
98 
99 #ifdef	NO_CONV
100 /* Build a smaller version (i.e. for a miniroot) */
101 /* These can not be called, but just in case...  */
102 static const char no_block[] = "unblock and -DNO_CONV?";
103 void block(void)		{ errx(EXIT_FAILURE, "%s", no_block + 2); }
104 void block_close(void)		{ errx(EXIT_FAILURE, "%s", no_block + 2); }
105 void unblock(void)		{ errx(EXIT_FAILURE, "%s", no_block); }
106 void unblock_close(void)	{ errx(EXIT_FAILURE, "%s", no_block); }
107 #else	/* NO_CONV */
108 
109 /*
110  * Copy variable length newline terminated records with a max size cbsz
111  * bytes to output.  Records less than cbs are padded with spaces.
112  *
113  * max in buffer:  MAX(ibs, cbsz)
114  * max out buffer: obs + cbsz
115  */
116 void
117 block(void)
118 {
119 	static int intrunc;
120 	int ch = 0;	/* pacify gcc */
121 	uint64_t cnt, maxlen;
122 	u_char *inp, *outp;
123 	const u_char *t;
124 
125 	/*
126 	 * Record truncation can cross block boundaries.  If currently in a
127 	 * truncation state, keep tossing characters until reach a newline.
128 	 * Start at the beginning of the buffer, as the input buffer is always
129 	 * left empty.
130 	 */
131 	if (intrunc) {
132 		for (inp = in.db, cnt = in.dbrcnt;
133 		    cnt && *inp++ != '\n'; --cnt);
134 		if (!cnt) {
135 			in.dbcnt = 0;
136 			in.dbp = in.db;
137 			return;
138 		}
139 		intrunc = 0;
140 		/* Adjust the input buffer numbers. */
141 		in.dbcnt = cnt - 1;
142 		in.dbp = inp + cnt - 1;
143 	}
144 
145 	/*
146 	 * Copy records (max cbsz size chunks) into the output buffer.  The
147 	 * translation is done as we copy into the output buffer.
148 	 */
149 	for (inp = in.dbp - in.dbcnt, outp = out.dbp; in.dbcnt;) {
150 		maxlen = MIN(cbsz, in.dbcnt);
151 		if ((t = ctab) != NULL)
152 			for (cnt = 0;
153 			    cnt < maxlen && (ch = *inp++) != '\n'; ++cnt)
154 				*outp++ = t[ch];
155 		else
156 			for (cnt = 0;
157 			    cnt < maxlen && (ch = *inp++) != '\n'; ++cnt)
158 				*outp++ = ch;
159 		/*
160 		 * Check for short record without a newline.  Reassemble the
161 		 * input block.
162 		 */
163 		if (ch != '\n' && in.dbcnt < cbsz) {
164 			(void)memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
165 			break;
166 		}
167 
168 		/* Adjust the input buffer numbers. */
169 		in.dbcnt -= cnt;
170 		if (ch == '\n')
171 			--in.dbcnt;
172 
173 		/* Pad short records with spaces. */
174 		if (cnt < cbsz)
175 			(void)memset(outp, ctab ? ctab[' '] : ' ', cbsz - cnt);
176 		else {
177 			/*
178 			 * If the next character wouldn't have ended the
179 			 * block, it's a truncation.
180 			 */
181 			if (!in.dbcnt || *inp != '\n')
182 				++st.trunc;
183 
184 			/* Toss characters to a newline. */
185 			for (; in.dbcnt && *inp++ != '\n'; --in.dbcnt);
186 			if (!in.dbcnt)
187 				intrunc = 1;
188 			else
189 				--in.dbcnt;
190 		}
191 
192 		/* Adjust output buffer numbers. */
193 		out.dbp += cbsz;
194 		if ((out.dbcnt += cbsz) >= out.dbsz)
195 			dd_out(0);
196 		outp = out.dbp;
197 	}
198 	in.dbp = in.db + in.dbcnt;
199 }
200 
201 void
202 block_close(void)
203 {
204 
205 	/*
206 	 * Copy any remaining data into the output buffer and pad to a record.
207 	 * Don't worry about truncation or translation, the input buffer is
208 	 * always empty when truncating, and no characters have been added for
209 	 * translation.  The bottom line is that anything left in the input
210 	 * buffer is a truncated record.  Anything left in the output buffer
211 	 * just wasn't big enough.
212 	 */
213 	if (in.dbcnt) {
214 		++st.trunc;
215 		(void)memmove(out.dbp, in.dbp - in.dbcnt, in.dbcnt);
216 		(void)memset(out.dbp + in.dbcnt,
217 		    ctab ? ctab[' '] : ' ', cbsz - in.dbcnt);
218 		out.dbcnt += cbsz;
219 	}
220 }
221 
222 /*
223  * Convert fixed length (cbsz) records to variable length.  Deletes any
224  * trailing blanks and appends a newline.
225  *
226  * max in buffer:  MAX(ibs, cbsz) + cbsz
227  * max out buffer: obs + cbsz
228  */
229 void
230 unblock(void)
231 {
232 	uint64_t cnt;
233 	u_char *inp;
234 	const u_char *t;
235 
236 	/* Translation and case conversion. */
237 	if ((t = ctab) != NULL)
238 		for (cnt = in.dbrcnt, inp = in.dbp - 1; cnt--; inp--)
239 			*inp = t[*inp];
240 	/*
241 	 * Copy records (max cbsz size chunks) into the output buffer.  The
242 	 * translation has to already be done or we might not recognize the
243 	 * spaces.
244 	 */
245 	for (inp = in.db; in.dbcnt >= cbsz; inp += cbsz, in.dbcnt -= cbsz) {
246 		for (t = inp + cbsz - 1; t >= inp && *t == ' '; --t);
247 		if (t >= inp) {
248 			cnt = t - inp + 1;
249 			(void)memmove(out.dbp, inp, cnt);
250 			out.dbp += cnt;
251 			out.dbcnt += cnt;
252 		}
253 		++out.dbcnt;
254 		*out.dbp++ = '\n';
255 		if (out.dbcnt >= out.dbsz)
256 			dd_out(0);
257 	}
258 	if (in.dbcnt)
259 		(void)memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
260 	in.dbp = in.db + in.dbcnt;
261 }
262 
263 void
264 unblock_close(void)
265 {
266 	uint64_t cnt;
267 	u_char *t;
268 
269 	if (in.dbcnt) {
270 		warnx("%s: short input record", in.name);
271 		for (t = in.db + in.dbcnt - 1; t >= in.db && *t == ' '; --t);
272 		if (t >= in.db) {
273 			cnt = t - in.db + 1;
274 			(void)memmove(out.dbp, in.db, cnt);
275 			out.dbp += cnt;
276 			out.dbcnt += cnt;
277 		}
278 		++out.dbcnt;
279 		*out.dbp++ = '\n';
280 	}
281 }
282 
283 #endif	/* NO_CONV */
284