xref: /dragonfly/bin/dd/conv.c (revision 8a7bdfea)
1 /*-
2  * Copyright (c) 1991, 1993, 1994
3  *	The Regents of the University of California.  All rights reserved.
4  *
5  * This code is derived from software contributed to Berkeley by
6  * Keith Muller of the University of California, San Diego and Lance
7  * Visser of Convex Computer Corporation.
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 in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. Neither the name of the University nor the names of its contributors
18  *    may be used to endorse or promote products derived from this software
19  *    without specific prior written permission.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31  * SUCH DAMAGE.
32  *
33  * @(#)conv.c	8.3 (Berkeley) 4/2/94
34  * $FreeBSD: src/bin/dd/conv.c,v 1.15 1999/09/13 21:47:10 green Exp $
35  * $DragonFly: src/bin/dd/conv.c,v 1.6 2008/01/28 16:08:02 matthias Exp $
36  */
37 
38 #include <sys/param.h>
39 
40 #include <err.h>
41 #include <string.h>
42 
43 #include "dd.h"
44 #include "extern.h"
45 
46 /*
47  * def --
48  * Copy input to output.  Input is buffered until reaches obs, and then
49  * output until less than obs remains.  Only a single buffer is used.
50  * Worst case buffer calculation is (ibs + obs - 1).
51  */
52 void
53 def(void)
54 {
55 	u_char *inp;
56 	const u_char *t;
57 	size_t cnt;
58 
59 	if ((t = ctab) != NULL)
60 		for (inp = in.dbp - (cnt = in.dbrcnt); cnt--; ++inp)
61 			*inp = t[*inp];
62 
63 	/* Make the output buffer look right. */
64 	out.dbp = in.dbp;
65 	out.dbcnt = in.dbcnt;
66 
67 	if (in.dbcnt >= out.dbsz) {
68 		/* If the output buffer is full, write it. */
69 		dd_out(0);
70 
71 		/*
72 		 * Ddout copies the leftover output to the beginning of
73 		 * the buffer and resets the output buffer.  Reset the
74 		 * input buffer to match it.
75 	 	 */
76 		in.dbp = out.dbp;
77 		in.dbcnt = out.dbcnt;
78 	}
79 }
80 
81 void
82 def_close(void)
83 {
84 	/* Just update the count, everything is already in the buffer. */
85 	if (in.dbcnt)
86 		out.dbcnt = in.dbcnt;
87 }
88 
89 /*
90  * Copy variable length newline terminated records with a max size cbsz
91  * bytes to output.  Records less than cbs are padded with spaces.
92  *
93  * max in buffer:  MAX(ibs, cbsz)
94  * max out buffer: obs + cbsz
95  */
96 void
97 block(void)
98 {
99 	u_char *inp, *outp;
100 	const u_char *t;
101 	size_t cnt, maxlen;
102 	static int intrunc;
103 	int ch;
104 
105 	/*
106 	 * Record truncation can cross block boundaries.  If currently in a
107 	 * truncation state, keep tossing characters until reach a newline.
108 	 * Start at the beginning of the buffer, as the input buffer is always
109 	 * left empty.
110 	 */
111 	if (intrunc) {
112 		for (inp = in.db, cnt = in.dbrcnt; cnt && *inp++ != '\n'; --cnt)
113 			;
114 		if (!cnt) {
115 			in.dbcnt = 0;
116 			in.dbp = in.db;
117 			return;
118 		}
119 		intrunc = 0;
120 		/* Adjust the input buffer numbers. */
121 		in.dbcnt = cnt - 1;
122 		in.dbp = inp + cnt - 1;
123 	}
124 
125 	/*
126 	 * Copy records (max cbsz size chunks) into the output buffer.  The
127 	 * translation is done as we copy into the output buffer.
128 	 */
129 	ch = 0;
130 	for (inp = in.dbp - in.dbcnt, outp = out.dbp; in.dbcnt;) {
131 		maxlen = MIN(cbsz, in.dbcnt);
132 		if ((t = ctab) != NULL)
133 			for (cnt = 0; cnt < maxlen && (ch = *inp++) != '\n';
134 			    ++cnt)
135 				*outp++ = t[ch];
136 		else
137 			for (cnt = 0; cnt < maxlen && (ch = *inp++) != '\n';
138 			    ++cnt)
139 				*outp++ = ch;
140 		/*
141 		 * Check for short record without a newline.  Reassemble the
142 		 * input block.
143 		 */
144 		if (ch != '\n' && in.dbcnt < cbsz) {
145 			memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
146 			break;
147 		}
148 
149 		/* Adjust the input buffer numbers. */
150 		in.dbcnt -= cnt;
151 		if (ch == '\n')
152 			--in.dbcnt;
153 
154 		/* Pad short records with spaces. */
155 		if (cnt < cbsz)
156 			memset(outp, ctab ? ctab[' '] : ' ', cbsz - cnt);
157 		else {
158 			/*
159 			 * If the next character wouldn't have ended the
160 			 * block, it's a truncation.
161 			 */
162 			if (!in.dbcnt || *inp != '\n')
163 				++st.trunc;
164 
165 			/* Toss characters to a newline. */
166 			for (; in.dbcnt && *inp++ != '\n'; --in.dbcnt);
167 			if (!in.dbcnt)
168 				intrunc = 1;
169 			else
170 				--in.dbcnt;
171 		}
172 
173 		/* Adjust output buffer numbers. */
174 		out.dbp += cbsz;
175 		if ((out.dbcnt += cbsz) >= out.dbsz)
176 			dd_out(0);
177 		outp = out.dbp;
178 	}
179 	in.dbp = in.db + in.dbcnt;
180 }
181 
182 void
183 block_close(void)
184 {
185 	/*
186 	 * Copy any remaining data into the output buffer and pad to a record.
187 	 * Don't worry about truncation or translation, the input buffer is
188 	 * always empty when truncating, and no characters have been added for
189 	 * translation.  The bottom line is that anything left in the input
190 	 * buffer is a truncated record.  Anything left in the output buffer
191 	 * just wasn't big enough.
192 	 */
193 	if (in.dbcnt) {
194 		++st.trunc;
195 		memmove(out.dbp, in.dbp - in.dbcnt, in.dbcnt);
196 		memset(out.dbp + in.dbcnt, ctab ? ctab[' '] : ' ',
197 		    cbsz - in.dbcnt);
198 		out.dbcnt += cbsz;
199 	}
200 }
201 
202 /*
203  * Convert fixed length (cbsz) records to variable length.  Deletes any
204  * trailing blanks and appends a newline.
205  *
206  * max in buffer:  MAX(ibs, cbsz) + cbsz
207  * max out buffer: obs + cbsz
208  */
209 void
210 unblock(void)
211 {
212 	u_char *inp;
213 	const u_char *t;
214 	size_t cnt;
215 
216 	/* Translation and case conversion. */
217 	if ((t = ctab) != NULL)
218 		for (cnt = in.dbrcnt, inp = in.dbp; cnt--;) {
219 			inp--;
220 			*inp = t[*inp];
221 		}
222 	/*
223 	 * Copy records (max cbsz size chunks) into the output buffer.  The
224 	 * translation has to already be done or we might not recognize the
225 	 * spaces.
226 	 */
227 	for (inp = in.db; in.dbcnt >= cbsz; inp += cbsz, in.dbcnt -= cbsz) {
228 		for (t = inp + cbsz - 1; t >= inp && *t == ' '; --t)
229 			;
230 		if (t >= inp) {
231 			cnt = t - inp + 1;
232 			memmove(out.dbp, inp, cnt);
233 			out.dbp += cnt;
234 			out.dbcnt += cnt;
235 		}
236 		*out.dbp++ = '\n';
237 		if (++out.dbcnt >= out.dbsz)
238 			dd_out(0);
239 	}
240 	if (in.dbcnt)
241 		memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
242 	in.dbp = in.db + in.dbcnt;
243 }
244 
245 void
246 unblock_close(void)
247 {
248 	u_char *t;
249 	size_t cnt;
250 
251 	if (in.dbcnt) {
252 		warnx("%s: short input record", in.name);
253 		for (t = in.db + in.dbcnt - 1; t >= in.db && *t == ' '; --t)
254 			;
255 		if (t >= in.db) {
256 			cnt = t - in.db + 1;
257 			memmove(out.dbp, in.db, cnt);
258 			out.dbp += cnt;
259 			out.dbcnt += cnt;
260 		}
261 		++out.dbcnt;
262 		*out.dbp++ = '\n';
263 	}
264 }
265