xref: /freebsd/bin/dd/conv.c (revision 4f52dfbb8d6c4d446500c5b097e3806ec219fbd4)
1 /*-
2  * SPDX-License-Identifier: BSD-3-Clause
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 #ifndef lint
37 #if 0
38 static char sccsid[] = "@(#)conv.c	8.3 (Berkeley) 4/2/94";
39 #endif
40 #endif /* not lint */
41 #include <sys/cdefs.h>
42 __FBSDID("$FreeBSD$");
43 
44 #include <sys/param.h>
45 
46 #include <err.h>
47 #include <inttypes.h>
48 #include <string.h>
49 
50 #include "dd.h"
51 #include "extern.h"
52 
53 /*
54  * def --
55  * Copy input to output.  Input is buffered until reaches obs, and then
56  * output until less than obs remains.  Only a single buffer is used.
57  * Worst case buffer calculation is (ibs + obs - 1).
58  */
59 void
60 def(void)
61 {
62 	u_char *inp;
63 	const u_char *t;
64 	size_t cnt;
65 
66 	if ((t = ctab) != NULL)
67 		for (inp = in.dbp - (cnt = in.dbrcnt); cnt--; ++inp)
68 			*inp = t[*inp];
69 
70 	/* Make the output buffer look right. */
71 	out.dbp = in.dbp;
72 	out.dbcnt = in.dbcnt;
73 
74 	if (in.dbcnt >= out.dbsz) {
75 		/* If the output buffer is full, write it. */
76 		dd_out(0);
77 
78 		/*
79 		 * dd_out copies the leftover output to the beginning of
80 		 * the buffer and resets the output buffer.  Reset the
81 		 * input buffer to match it.
82 	 	 */
83 		in.dbp = out.dbp;
84 		in.dbcnt = out.dbcnt;
85 	}
86 }
87 
88 void
89 def_close(void)
90 {
91 	/* Just update the count, everything is already in the buffer. */
92 	if (in.dbcnt)
93 		out.dbcnt = in.dbcnt;
94 }
95 
96 /*
97  * Copy variable length newline terminated records with a max size cbsz
98  * bytes to output.  Records less than cbs are padded with spaces.
99  *
100  * max in buffer:  MAX(ibs, cbsz)
101  * max out buffer: obs + cbsz
102  */
103 void
104 block(void)
105 {
106 	u_char *inp, *outp;
107 	const u_char *t;
108 	size_t cnt, maxlen;
109 	static int intrunc;
110 	int ch;
111 
112 	/*
113 	 * Record truncation can cross block boundaries.  If currently in a
114 	 * truncation state, keep tossing characters until reach a newline.
115 	 * Start at the beginning of the buffer, as the input buffer is always
116 	 * left empty.
117 	 */
118 	if (intrunc) {
119 		for (inp = in.db, cnt = in.dbrcnt; cnt && *inp++ != '\n'; --cnt)
120 			;
121 		if (!cnt) {
122 			in.dbcnt = 0;
123 			in.dbp = in.db;
124 			return;
125 		}
126 		intrunc = 0;
127 		/* Adjust the input buffer numbers. */
128 		in.dbcnt = cnt - 1;
129 		in.dbp = inp + cnt - 1;
130 	}
131 
132 	/*
133 	 * Copy records (max cbsz size chunks) into the output buffer.  The
134 	 * translation is done as we copy into the output buffer.
135 	 */
136 	ch = 0;
137 	for (inp = in.dbp - in.dbcnt, outp = out.dbp; in.dbcnt;) {
138 		maxlen = MIN(cbsz, (size_t)in.dbcnt);
139 		if ((t = ctab) != NULL)
140 			for (cnt = 0; cnt < maxlen && (ch = *inp++) != '\n';
141 			    ++cnt)
142 				*outp++ = t[ch];
143 		else
144 			for (cnt = 0; cnt < maxlen && (ch = *inp++) != '\n';
145 			    ++cnt)
146 				*outp++ = ch;
147 		/*
148 		 * Check for short record without a newline.  Reassemble the
149 		 * input block.
150 		 */
151 		if (ch != '\n' && (size_t)in.dbcnt < cbsz) {
152 			(void)memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
153 			break;
154 		}
155 
156 		/* Adjust the input buffer numbers. */
157 		in.dbcnt -= cnt;
158 		if (ch == '\n')
159 			--in.dbcnt;
160 
161 		/* Pad short records with spaces. */
162 		if (cnt < cbsz)
163 			(void)memset(outp, ctab ? ctab[' '] : ' ', cbsz - cnt);
164 		else {
165 			/*
166 			 * If the next character wouldn't have ended the
167 			 * block, it's a truncation.
168 			 */
169 			if (!in.dbcnt || *inp != '\n')
170 				++st.trunc;
171 
172 			/* Toss characters to a newline. */
173 			for (; in.dbcnt && *inp++ != '\n'; --in.dbcnt)
174 				;
175 			if (!in.dbcnt)
176 				intrunc = 1;
177 			else
178 				--in.dbcnt;
179 		}
180 
181 		/* Adjust output buffer numbers. */
182 		out.dbp += cbsz;
183 		if ((out.dbcnt += cbsz) >= out.dbsz)
184 			dd_out(0);
185 		outp = out.dbp;
186 	}
187 	in.dbp = in.db + in.dbcnt;
188 }
189 
190 void
191 block_close(void)
192 {
193 	/*
194 	 * Copy any remaining data into the output buffer and pad to a record.
195 	 * Don't worry about truncation or translation, the input buffer is
196 	 * always empty when truncating, and no characters have been added for
197 	 * translation.  The bottom line is that anything left in the input
198 	 * buffer is a truncated record.  Anything left in the output buffer
199 	 * just wasn't big enough.
200 	 */
201 	if (in.dbcnt) {
202 		++st.trunc;
203 		(void)memmove(out.dbp, in.dbp - in.dbcnt, in.dbcnt);
204 		(void)memset(out.dbp + in.dbcnt, ctab ? ctab[' '] : ' ',
205 		    cbsz - in.dbcnt);
206 		out.dbcnt += cbsz;
207 	}
208 }
209 
210 /*
211  * Convert fixed length (cbsz) records to variable length.  Deletes any
212  * trailing blanks and appends a newline.
213  *
214  * max in buffer:  MAX(ibs, cbsz) + cbsz
215  * max out buffer: obs + cbsz
216  */
217 void
218 unblock(void)
219 {
220 	u_char *inp;
221 	const u_char *t;
222 	size_t cnt;
223 
224 	/* Translation and case conversion. */
225 	if ((t = ctab) != NULL)
226 		for (inp = in.dbp - (cnt = in.dbrcnt); cnt--; ++inp)
227 			*inp = t[*inp];
228 	/*
229 	 * Copy records (max cbsz size chunks) into the output buffer.  The
230 	 * translation has to already be done or we might not recognize the
231 	 * spaces.
232 	 */
233 	for (inp = in.db; (size_t)in.dbcnt >= cbsz; inp += cbsz, in.dbcnt -= cbsz) {
234 		for (t = inp + cbsz - 1; t >= inp && *t == ' '; --t)
235 			;
236 		if (t >= inp) {
237 			cnt = t - inp + 1;
238 			(void)memmove(out.dbp, inp, cnt);
239 			out.dbp += cnt;
240 			out.dbcnt += cnt;
241 		}
242 		*out.dbp++ = '\n';
243 		if (++out.dbcnt >= out.dbsz)
244 			dd_out(0);
245 	}
246 	if (in.dbcnt)
247 		(void)memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
248 	in.dbp = in.db + in.dbcnt;
249 }
250 
251 void
252 unblock_close(void)
253 {
254 	u_char *t;
255 	size_t cnt;
256 
257 	if (in.dbcnt) {
258 		warnx("%s: short input record", in.name);
259 		for (t = in.db + in.dbcnt - 1; t >= in.db && *t == ' '; --t)
260 			;
261 		if (t >= in.db) {
262 			cnt = t - in.db + 1;
263 			(void)memmove(out.dbp, in.db, cnt);
264 			out.dbp += cnt;
265 			out.dbcnt += cnt;
266 		}
267 		++out.dbcnt;
268 		*out.dbp++ = '\n';
269 	}
270 }
271