Home | History | Annotate | Line # | Download | only in more
ch.c revision 1.3
      1 /*	$NetBSD: ch.c,v 1.3 1998/02/04 11:08:41 christos Exp $	*/
      2 
      3 /*
      4  * Copyright (c) 1988 Mark Nudleman
      5  * Copyright (c) 1988, 1993
      6  *	The Regents of the University of California.  All rights reserved.
      7  *
      8  * Redistribution and use in source and binary forms, with or without
      9  * modification, are permitted provided that the following conditions
     10  * are met:
     11  * 1. Redistributions of source code must retain the above copyright
     12  *    notice, this list of conditions and the following disclaimer.
     13  * 2. Redistributions in binary form must reproduce the above copyright
     14  *    notice, this list of conditions and the following disclaimer in the
     15  *    documentation and/or other materials provided with the distribution.
     16  * 3. All advertising materials mentioning features or use of this software
     17  *    must display the following acknowledgement:
     18  *	This product includes software developed by the University of
     19  *	California, Berkeley and its contributors.
     20  * 4. Neither the name of the University nor the names of its contributors
     21  *    may be used to endorse or promote products derived from this software
     22  *    without specific prior written permission.
     23  *
     24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     34  * SUCH DAMAGE.
     35  */
     36 
     37 #include <sys/cdefs.h>
     38 #ifndef lint
     39 #if 0
     40 static char sccsid[] = "@(#)ch.c	8.1 (Berkeley) 6/6/93";
     41 #else
     42 __RCSID("$NetBSD: ch.c,v 1.3 1998/02/04 11:08:41 christos Exp $");
     43 #endif
     44 #endif /* not lint */
     45 
     46 /*
     47  * Low level character input from the input file.
     48  * We use these special purpose routines which optimize moving
     49  * both forward and backward from the current read pointer.
     50  */
     51 
     52 #include <sys/types.h>
     53 #include <sys/file.h>
     54 #include <unistd.h>
     55 #include <stdlib.h>
     56 #include <stdio.h>
     57 #include <err.h>
     58 
     59 #include "less.h"
     60 #include "extern.h"
     61 
     62 int file = -1;		/* File descriptor of the input file */
     63 
     64 /*
     65  * Pool of buffers holding the most recently used blocks of the input file.
     66  */
     67 struct buf {
     68 	struct buf *next, *prev;
     69 	long block;
     70 	int datasize;
     71 	char data[BUFSIZ];
     72 };
     73 int nbufs;
     74 
     75 /*
     76  * The buffer pool is kept as a doubly-linked circular list, in order from
     77  * most- to least-recently used.  The circular list is anchored by buf_anchor.
     78  */
     79 #define	END_OF_CHAIN	((struct buf *)&buf_anchor)
     80 #define	buf_head	buf_anchor.next
     81 #define	buf_tail	buf_anchor.prev
     82 
     83 static struct {
     84 	struct buf *next, *prev;
     85 } buf_anchor = { END_OF_CHAIN, END_OF_CHAIN };
     86 
     87 /*
     88  * Current position in file.
     89  * Stored as a block number and an offset into the block.
     90  */
     91 static long ch_block;
     92 static int ch_offset;
     93 
     94 /* Length of file, needed if input is a pipe. */
     95 static off_t ch_fsize;
     96 
     97 /* Number of bytes read, if input is standard input (a pipe). */
     98 static off_t last_piped_pos;
     99 
    100 /*
    101  * Get the character pointed to by the read pointer.  ch_get() is a macro
    102  * which is more efficient to call than fch_get (the function), in the usual
    103  * case that the block desired is at the head of the chain.
    104  */
    105 #define	ch_get() \
    106 	((buf_head->block == ch_block && \
    107 	    ch_offset < buf_head->datasize) ? \
    108 	    buf_head->data[ch_offset] : fch_get())
    109 
    110 static int fch_get __P((void));
    111 static int buffered __P((long));
    112 
    113 static int
    114 fch_get()
    115 {
    116 	struct buf *bp;
    117 	int n, ch;
    118 	char *p, *t;
    119 	off_t pos;
    120 
    121 	/* look for a buffer holding the desired block. */
    122 	for (bp = buf_head;  bp != END_OF_CHAIN;  bp = bp->next)
    123 		if (bp->block == ch_block) {
    124 			if (ch_offset >= bp->datasize)
    125 				/*
    126 				 * Need more data in this buffer.
    127 				 */
    128 				goto read_more;
    129 			/*
    130 			 * On a pipe, we don't sort the buffers LRU
    131 			 * because this can cause gaps in the buffers.
    132 			 * For example, suppose we've got 12 1K buffers,
    133 			 * and a 15K input stream.  If we read the first 12K
    134 			 * sequentially, then jump to line 1, then jump to
    135 			 * the end, the buffers have blocks 0,4,5,6,..,14.
    136 			 * If we then jump to line 1 again and try to
    137 			 * read sequentially, we're out of luck when we
    138 			 * get to block 1 (we'd get the "pipe error" below).
    139 			 * To avoid this, we only sort buffers on a pipe
    140 			 * when we actually READ the data, not when we
    141 			 * find it already buffered.
    142 			 */
    143 			if (ispipe)
    144 				return(bp->data[ch_offset]);
    145 			goto found;
    146 		}
    147 	/*
    148 	 * Block is not in a buffer.  Take the least recently used buffer
    149 	 * and read the desired block into it.  If the LRU buffer has data
    150 	 * in it, and input is a pipe, then try to allocate a new buffer first.
    151 	 */
    152 	if (ispipe && buf_tail->block != (long)(-1))
    153 		(void)ch_addbuf(1);
    154 	bp = buf_tail;
    155 	bp->block = ch_block;
    156 	bp->datasize = 0;
    157 
    158 read_more:
    159 	pos = (ch_block * BUFSIZ) + bp->datasize;
    160 	if (ispipe) {
    161 		/*
    162 		 * The data requested should be immediately after
    163 		 * the last data read from the pipe.
    164 		 */
    165 		if (pos != last_piped_pos) {
    166 			error("pipe error");
    167 			quit();
    168 		}
    169 	} else
    170 		(void)lseek(file, pos, L_SET);
    171 
    172 	/*
    173 	 * Read the block.
    174 	 * If we read less than a full block, we just return the
    175 	 * partial block and pick up the rest next time.
    176 	 */
    177 	n = iread(file, &bp->data[bp->datasize], BUFSIZ - bp->datasize);
    178 	if (n == READ_INTR)
    179 		return (EOI);
    180 	if (n < 0) {
    181 		error("read error");
    182 		quit();
    183 	}
    184 	if (ispipe)
    185 		last_piped_pos += n;
    186 
    187 	p = &bp->data[bp->datasize];
    188 	bp->datasize += n;
    189 
    190 	/*
    191 	 * Set an EOI marker in the buffered data itself.  Then ensure the
    192 	 * data is "clean": there are no extra EOI chars in the data and
    193 	 * that the "meta" bit (the 0200 bit) is reset in each char;
    194 	 * also translate \r\n sequences to \n if -u flag not set.
    195 	 */
    196 	if (n == 0) {
    197 		ch_fsize = pos;
    198 		bp->data[bp->datasize++] = EOI;
    199 	}
    200 
    201 	if (bs_mode) {
    202 		for (p = &bp->data[bp->datasize]; --n >= 0;) {
    203 			*--p &= 0177;
    204 			if (*p == EOI)
    205 				*p = 0200;
    206 		}
    207 	}
    208 	else {
    209 		for (t = p; --n >= 0; ++p) {
    210 			ch = *p & 0177;
    211 			if (ch == '\r' && n && (p[1] & 0177) == '\n') {
    212 				++p;
    213 				*t++ = '\n';
    214 			}
    215 			else
    216 				*t++ = (ch == EOI) ? 0200 : ch;
    217 		}
    218 		if (p != t) {
    219 			bp->datasize -= p - t;
    220 			if (ispipe)
    221 				last_piped_pos -= p - t;
    222 		}
    223 	}
    224 
    225 found:
    226 	if (buf_head != bp) {
    227 		/*
    228 		 * Move the buffer to the head of the buffer chain.
    229 		 * This orders the buffer chain, most- to least-recently used.
    230 		 */
    231 		bp->next->prev = bp->prev;
    232 		bp->prev->next = bp->next;
    233 
    234 		bp->next = buf_head;
    235 		bp->prev = END_OF_CHAIN;
    236 		buf_head->prev = bp;
    237 		buf_head = bp;
    238 	}
    239 
    240 	if (ch_offset >= bp->datasize)
    241 		/*
    242 		 * After all that, we still don't have enough data.
    243 		 * Go back and try again.
    244 		 */
    245 		goto read_more;
    246 
    247 	return(bp->data[ch_offset]);
    248 }
    249 
    250 /*
    251  * Determine if a specific block is currently in one of the buffers.
    252  */
    253 static int
    254 buffered(block)
    255 	long block;
    256 {
    257 	struct buf *bp;
    258 
    259 	for (bp = buf_head; bp != END_OF_CHAIN; bp = bp->next)
    260 		if (bp->block == block)
    261 			return(1);
    262 	return(0);
    263 }
    264 
    265 /*
    266  * Seek to a specified position in the file.
    267  * Return 0 if successful, non-zero if can't seek there.
    268  */
    269 int
    270 ch_seek(pos)
    271 	off_t pos;
    272 {
    273 	long new_block;
    274 
    275 	new_block = pos / BUFSIZ;
    276 	if (!ispipe || pos == last_piped_pos || buffered(new_block)) {
    277 		/*
    278 		 * Set read pointer.
    279 		 */
    280 		ch_block = new_block;
    281 		ch_offset = pos % BUFSIZ;
    282 		return(0);
    283 	}
    284 	return(1);
    285 }
    286 
    287 /*
    288  * Seek to the end of the file.
    289  */
    290 int
    291 ch_end_seek()
    292 {
    293 	if (!ispipe)
    294 		return(ch_seek(ch_length()));
    295 
    296 	/*
    297 	 * Do it the slow way: read till end of data.
    298 	 */
    299 	while (ch_forw_get() != EOI)
    300 		if (sigs)
    301 			return(1);
    302 	return(0);
    303 }
    304 
    305 /*
    306  * Seek to the beginning of the file, or as close to it as we can get.
    307  * We may not be able to seek there if input is a pipe and the
    308  * beginning of the pipe is no longer buffered.
    309  */
    310 int
    311 ch_beg_seek()
    312 {
    313 	struct buf *bp, *firstbp;
    314 
    315 	/*
    316 	 * Try a plain ch_seek first.
    317 	 */
    318 	if (ch_seek((off_t)0) == 0)
    319 		return(0);
    320 
    321 	/*
    322 	 * Can't get to position 0.
    323 	 * Look thru the buffers for the one closest to position 0.
    324 	 */
    325 	firstbp = bp = buf_head;
    326 	if (bp == END_OF_CHAIN)
    327 		return(1);
    328 	while ((bp = bp->next) != END_OF_CHAIN)
    329 		if (bp->block < firstbp->block)
    330 			firstbp = bp;
    331 	ch_block = firstbp->block;
    332 	ch_offset = 0;
    333 	return(0);
    334 }
    335 
    336 /*
    337  * Return the length of the file, if known.
    338  */
    339 off_t
    340 ch_length()
    341 {
    342 	if (ispipe)
    343 		return(ch_fsize);
    344 	return((off_t)(lseek(file, (off_t)0, L_XTND)));
    345 }
    346 
    347 /*
    348  * Return the current position in the file.
    349  */
    350 off_t
    351 ch_tell()
    352 {
    353 	return(ch_block * BUFSIZ + ch_offset);
    354 }
    355 
    356 /*
    357  * Get the current char and post-increment the read pointer.
    358  */
    359 int
    360 ch_forw_get()
    361 {
    362 	int c;
    363 
    364 	c = ch_get();
    365 	if (c != EOI && ++ch_offset >= BUFSIZ) {
    366 		ch_offset = 0;
    367 		++ch_block;
    368 	}
    369 	return(c);
    370 }
    371 
    372 /*
    373  * Pre-decrement the read pointer and get the new current char.
    374  */
    375 int
    376 ch_back_get()
    377 {
    378 	if (--ch_offset < 0) {
    379 		if (ch_block <= 0 || (ispipe && !buffered(ch_block-1))) {
    380 			ch_offset = 0;
    381 			return(EOI);
    382 		}
    383 		ch_offset = BUFSIZ - 1;
    384 		ch_block--;
    385 	}
    386 	return(ch_get());
    387 }
    388 
    389 /*
    390  * Allocate buffers.
    391  * Caller wants us to have a total of at least want_nbufs buffers.
    392  * keep==1 means keep the data in the current buffers;
    393  * otherwise discard the old data.
    394  */
    395 void
    396 ch_init(want_nbufs, keep)
    397 	int want_nbufs;
    398 	int keep;
    399 {
    400 	struct buf *bp;
    401 	char message[80];
    402 
    403 	cbufs = nbufs;
    404 	if (nbufs < want_nbufs && ch_addbuf(want_nbufs - nbufs)) {
    405 		/*
    406 		 * Cannot allocate enough buffers.
    407 		 * If we don't have ANY, then quit.
    408 		 * Otherwise, just report the error and return.
    409 		 */
    410 		(void)sprintf(message, "cannot allocate %d buffers",
    411 		    want_nbufs - nbufs);
    412 		error(message);
    413 		if (nbufs == 0)
    414 			quit();
    415 		return;
    416 	}
    417 
    418 	if (keep)
    419 		return;
    420 
    421 	/*
    422 	 * We don't want to keep the old data,
    423 	 * so initialize all the buffers now.
    424 	 */
    425 	for (bp = buf_head;  bp != END_OF_CHAIN;  bp = bp->next)
    426 		bp->block = (long)(-1);
    427 	last_piped_pos = (off_t)0;
    428 	ch_fsize = NULL_POSITION;
    429 	(void)ch_seek((off_t)0);
    430 }
    431 
    432 /*
    433  * Allocate some new buffers.
    434  * The buffers are added to the tail of the buffer chain.
    435  */
    436 int
    437 ch_addbuf(nnew)
    438 	int nnew;
    439 {
    440 	struct buf *bp;
    441 	struct buf *newbufs;
    442 
    443 	/*
    444 	 * We don't have enough buffers.
    445 	 * Allocate some new ones.
    446 	 */
    447 	newbufs = (struct buf *)calloc((u_int)nnew, sizeof(struct buf));
    448 	if (newbufs == NULL)
    449 		return(1);
    450 
    451 	/*
    452 	 * Initialize the new buffers and link them together.
    453 	 * Link them all onto the tail of the buffer list.
    454 	 */
    455 	nbufs += nnew;
    456 	cbufs = nbufs;
    457 	for (bp = &newbufs[0];  bp < &newbufs[nnew];  bp++) {
    458 		bp->next = bp + 1;
    459 		bp->prev = bp - 1;
    460 		bp->block = (long)(-1);
    461 	}
    462 	newbufs[nnew-1].next = END_OF_CHAIN;
    463 	newbufs[0].prev = buf_tail;
    464 	buf_tail->next = &newbufs[0];
    465 	buf_tail = &newbufs[nnew-1];
    466 	return(0);
    467 }
    468