xref: /freebsd/bin/dd/conv.c (revision 61afd5bb22d787b0641523e7b9b95c964d669bd5)
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. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  *	This product includes software developed by the University of
20  *	California, Berkeley and its contributors.
21  * 4. Neither the name of the University nor the names of its contributors
22  *    may be used to endorse or promote products derived from this software
23  *    without specific prior written permission.
24  *
25  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35  * SUCH DAMAGE.
36  *
37  *	$Id: conv.c,v 1.5 1996/11/13 19:59:56 phk Exp $
38  */
39 
40 #ifndef lint
41 static char const sccsid[] = "@(#)conv.c	8.3 (Berkeley) 4/2/94";
42 #endif /* not lint */
43 
44 #include <sys/param.h>
45 
46 #include <err.h>
47 #include <string.h>
48 
49 #include "dd.h"
50 #include "extern.h"
51 
52 /*
53  * def --
54  * Copy input to output.  Input is buffered until reaches obs, and then
55  * output until less than obs remains.  Only a single buffer is used.
56  * Worst case buffer calculation is (ibs + obs - 1).
57  */
58 void
59 def()
60 {
61 	int cnt;
62 	u_char *inp, *t;
63 
64 	if ((t = ctab) != NULL)
65 		for (inp = in.dbp - (cnt = in.dbrcnt); cnt--; ++inp)
66 			*inp = t[*inp];
67 
68 	/* Make the output buffer look right. */
69 	out.dbp = in.dbp;
70 	out.dbcnt = in.dbcnt;
71 
72 	if (in.dbcnt >= out.dbsz) {
73 		/* If the output buffer is full, write it. */
74 		dd_out(0);
75 
76 		/*
77 		 * Ddout copies the leftover output to the beginning of
78 		 * the buffer and resets the output buffer.  Reset the
79 		 * input buffer to match it.
80 	 	 */
81 		in.dbp = out.dbp;
82 		in.dbcnt = out.dbcnt;
83 	}
84 }
85 
86 void
87 def_close()
88 {
89 	/* Just update the count, everything is already in the buffer. */
90 	if (in.dbcnt)
91 		out.dbcnt = in.dbcnt;
92 }
93 
94 /*
95  * Copy variable length newline terminated records with a max size cbsz
96  * bytes to output.  Records less than cbs are padded with spaces.
97  *
98  * max in buffer:  MAX(ibs, cbsz)
99  * max out buffer: obs + cbsz
100  */
101 void
102 block()
103 {
104 	static int intrunc;
105 	int ch, cnt, maxlen;
106 	u_char *inp, *outp, *t;
107 
108 	/*
109 	 * Record truncation can cross block boundaries.  If currently in a
110 	 * truncation state, keep tossing characters until reach a newline.
111 	 * Start at the beginning of the buffer, as the input buffer is always
112 	 * left empty.
113 	 */
114 	if (intrunc) {
115 		for (inp = in.db, cnt = in.dbrcnt;
116 		    cnt && *inp++ != '\n'; --cnt);
117 		if (!cnt) {
118 			in.dbcnt = 0;
119 			in.dbp = in.db;
120 			return;
121 		}
122 		intrunc = 0;
123 		/* Adjust the input buffer numbers. */
124 		in.dbcnt = cnt - 1;
125 		in.dbp = inp + cnt - 1;
126 	}
127 
128 	/*
129 	 * Copy records (max cbsz size chunks) into the output buffer.  The
130 	 * translation is done as we copy into the output buffer.
131 	 */
132 	ch = 0;
133 	for (inp = in.dbp - in.dbcnt, outp = out.dbp; in.dbcnt;) {
134 		maxlen = MIN(cbsz, in.dbcnt);
135 		if ((t = ctab) != NULL)
136 			for (cnt = 0;
137 			    cnt < maxlen && (ch = *inp++) != '\n'; ++cnt)
138 				*outp++ = t[ch];
139 		else
140 			for (cnt = 0;
141 			    cnt < maxlen && (ch = *inp++) != '\n'; ++cnt)
142 				*outp++ = ch;
143 		/*
144 		 * Check for short record without a newline.  Reassemble the
145 		 * input block.
146 		 */
147 		if (ch != '\n' && in.dbcnt < cbsz) {
148 			memmove(in.db, in.dbp - in.dbcnt, in.dbcnt);
149 			break;
150 		}
151 
152 		/* Adjust the input buffer numbers. */
153 		in.dbcnt -= cnt;
154 		if (ch == '\n')
155 			--in.dbcnt;
156 
157 		/* Pad short records with spaces. */
158 		if (cnt < cbsz)
159 			(void)memset(outp, ctab ? ctab[' '] : ' ', cbsz - cnt);
160 		else {
161 			/*
162 			 * If the next character wouldn't have ended the
163 			 * block, it's a truncation.
164 			 */
165 			if (!in.dbcnt || *inp != '\n')
166 				++st.trunc;
167 
168 			/* Toss characters to a newline. */
169 			for (; in.dbcnt && *inp++ != '\n'; --in.dbcnt);
170 			if (!in.dbcnt)
171 				intrunc = 1;
172 			else
173 				--in.dbcnt;
174 		}
175 
176 		/* Adjust output buffer numbers. */
177 		out.dbp += cbsz;
178 		if ((out.dbcnt += cbsz) >= out.dbsz)
179 			dd_out(0);
180 		outp = out.dbp;
181 	}
182 	in.dbp = in.db + in.dbcnt;
183 }
184 
185 void
186 block_close()
187 {
188 	/*
189 	 * Copy any remaining data into the output buffer and pad to a record.
190 	 * Don't worry about truncation or translation, the input buffer is
191 	 * always empty when truncating, and no characters have been added for
192 	 * translation.  The bottom line is that anything left in the input
193 	 * buffer is a truncated record.  Anything left in the output buffer
194 	 * just wasn't big enough.
195 	 */
196 	if (in.dbcnt) {
197 		++st.trunc;
198 		memmove(out.dbp, in.dbp - in.dbcnt, in.dbcnt);
199 		(void)memset(out.dbp + in.dbcnt,
200 		    ctab ? ctab[' '] : ' ', cbsz - in.dbcnt);
201 		out.dbcnt += cbsz;
202 	}
203 }
204 
205 /*
206  * Convert fixed length (cbsz) records to variable length.  Deletes any
207  * trailing blanks and appends a newline.
208  *
209  * max in buffer:  MAX(ibs, cbsz) + cbsz
210  * max out buffer: obs + cbsz
211  */
212 void
213 unblock()
214 {
215 	int cnt;
216 	u_char *inp, *t;
217 
218 	/* Translation and case conversion. */
219 	if ((t = ctab) != NULL)
220 		for (cnt = in.dbrcnt, inp = in.dbp; cnt--;)
221 			*--inp = t[*inp];
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 		if (t >= inp) {
230 			cnt = t - inp + 1;
231 			memmove(out.dbp, inp, cnt);
232 			out.dbp += cnt;
233 			out.dbcnt += cnt;
234 		}
235 		++out.dbcnt;
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()
247 {
248 	int cnt;
249 	u_char *t;
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 		if (t >= in.db) {
255 			cnt = t - in.db + 1;
256 			memmove(out.dbp, in.db, cnt);
257 			out.dbp += cnt;
258 			out.dbcnt += cnt;
259 		}
260 		++out.dbcnt;
261 		*out.dbp++ = '\n';
262 	}
263 }
264