locate.bigram.c (17592) | locate.bigram.c (17776) |
---|---|
1/* 2 * Copyright (c) 1989, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * James A. Woods. 7 * 8 * Redistribution and use in source and binary forms, with or without --- 18 unchanged lines hidden (view full) --- 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. | 1/* 2 * Copyright (c) 1989, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * James A. Woods. 7 * 8 * Redistribution and use in source and binary forms, with or without --- 18 unchanged lines hidden (view full) --- 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 * $Id$ |
|
35 */ 36 37#ifndef lint 38static char copyright[] = 39"@(#) Copyright (c) 1989, 1993\n\ 40 The Regents of the University of California. All rights reserved.\n"; 41#endif /* not lint */ 42 43#ifndef lint 44static char sccsid[] = "@(#)locate.bigram.c 8.1 (Berkeley) 6/6/93"; 45#endif /* not lint */ 46 47/* | 37 */ 38 39#ifndef lint 40static char copyright[] = 41"@(#) Copyright (c) 1989, 1993\n\ 42 The Regents of the University of California. All rights reserved.\n"; 43#endif /* not lint */ 44 45#ifndef lint 46static char sccsid[] = "@(#)locate.bigram.c 8.1 (Berkeley) 6/6/93"; 47#endif /* not lint */ 48 49/* |
48 * bigram < text > bigrams | 50 * bigram < sorted_file_names | sort -nr | 51 * awk 'NR <= 128 { printf $2 }' > bigrams |
49 * 50 * List bigrams for 'updatedb' script. 51 * Use 'code' to encode a file using this output. 52 */ 53 54#include <stdio.h> 55#include <sys/param.h> /* for MAXPATHLEN */ 56#include <string.h> /* memchr */ 57#include "locate.h" 58 59u_char buf1[MAXPATHLEN] = " "; 60u_char buf2[MAXPATHLEN]; | 52 * 53 * List bigrams for 'updatedb' script. 54 * Use 'code' to encode a file using this output. 55 */ 56 57#include <stdio.h> 58#include <sys/param.h> /* for MAXPATHLEN */ 59#include <string.h> /* memchr */ 60#include "locate.h" 61 62u_char buf1[MAXPATHLEN] = " "; 63u_char buf2[MAXPATHLEN]; |
61unsigned int bigram[UCHAR_MAX][UCHAR_MAX]; | 64u_int bigram[UCHAR_MAX][UCHAR_MAX]; |
62 | 65 |
63 64void main ( ) | 66int 67main(void) |
65{ 66 register u_char *cp; 67 register u_char *oldpath = buf1, *path = buf2; | 68{ 69 register u_char *cp; 70 register u_char *oldpath = buf1, *path = buf2; |
68 register int i, j; | 71 register u_int i, j; |
69 | 72 |
70 /* init bigram buffer */ 71 for (i = 0; i < UCHAR_MAX; i++) 72 for (j = 0; j < UCHAR_MAX; j++) 73 bigram[i][j] = 0; | 73 while (fgets(path, sizeof(buf2), stdin) != NULL) { |
74 | 74 |
75 while ( fgets ( path, sizeof(buf2), stdin ) != NULL ) { 76 | |
77 /* skip empty lines */ 78 if (*path == '\n') 79 continue; 80 81 /* Squelch characters that would botch the decoding. */ 82 for (cp = path; *cp != NULL; cp++) { 83 /* chop newline */ 84 if (*cp == '\n') 85 *cp = NULL; 86 /* range */ 87 else if (*cp < ASCII_MIN || *cp > ASCII_MAX) 88 *cp = '?'; 89 } 90 | 75 /* skip empty lines */ 76 if (*path == '\n') 77 continue; 78 79 /* Squelch characters that would botch the decoding. */ 80 for (cp = path; *cp != NULL; cp++) { 81 /* chop newline */ 82 if (*cp == '\n') 83 *cp = NULL; 84 /* range */ 85 else if (*cp < ASCII_MIN || *cp > ASCII_MAX) 86 *cp = '?'; 87 } 88 |
91 | |
92 /* skip longest common prefix */ | 89 /* skip longest common prefix */ |
93 for (cp = path; *cp == *oldpath && *cp; cp++, oldpath++); | 90 for (cp = path; *cp == *oldpath && *cp != NULL; cp++, oldpath++); |
94 | 91 |
95 /* 96 * output post-residue bigrams only 97 */ 98 99 /* check later for boundary */ 100 while ( *cp != NULL && *(cp + 1) != NULL ) { | 92 while (*cp != NULL && *(cp+1) != NULL) { |
101 bigram[*cp][*(cp+1)]++; 102 cp += 2; 103 } 104 | 93 bigram[*cp][*(cp+1)]++; 94 cp += 2; 95 } 96 |
105 if ( path == buf1 ) /* swap pointers */ 106 path = buf2, oldpath = buf1; 107 else 108 path = buf1, oldpath = buf2; | 97 /* swap pointers */ 98 if (path == buf1) { 99 path = buf2; 100 oldpath = buf1; 101 } else { 102 path = buf1; 103 oldpath = buf2; 104 } |
109 } 110 | 105 } 106 |
111 /* output, boundary check */ | 107 /* output, (paranoid) boundary check */ |
112 for (i = ASCII_MIN; i <= ASCII_MAX; i++) 113 for (j = ASCII_MIN; j <= ASCII_MAX; j++) 114 if (bigram[i][j] != 0) | 108 for (i = ASCII_MIN; i <= ASCII_MAX; i++) 109 for (j = ASCII_MIN; j <= ASCII_MAX; j++) 110 if (bigram[i][j] != 0) |
115 fprintf(stdout, "%4d %c%c\n", 116 bigram[i][j], i, j); | 111 printf("%4u %c%c\n", bigram[i][j], i, j); 112 113 exit(0); |
117} | 114} |