| 1 |
/* |
| 2 |
$Id: index.h,v 1.52 2002/08/29 13:45:39 jmruiz Exp $ |
| 3 |
** |
| 4 |
** This program and library is free software; you can redistribute it and/or |
| 5 |
** modify it under the terms of the GNU (Library) General Public License |
| 6 |
** as published by the Free Software Foundation; either version 2 |
| 7 |
** of the License, or any later version. |
| 8 |
** |
| 9 |
** This program is distributed in the hope that it will be useful, |
| 10 |
** but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 |
** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 12 |
** GNU (Library) General Public License for more details. |
| 13 |
** |
| 14 |
** You should have received a copy of the GNU (Library) General Public License |
| 15 |
** along with this program; if not, write to the Free Software |
| 16 |
** Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. |
| 17 |
*/ |
| 18 |
|
| 19 |
#ifndef __HasSeenModule_Index |
| 20 |
#define __HasSeenModule_Index 1 |
| 21 |
|
| 22 |
struct dev_ino |
| 23 |
{ |
| 24 |
dev_t dev; |
| 25 |
ino_t ino; |
| 26 |
struct dev_ino *next; |
| 27 |
}; |
| 28 |
|
| 29 |
struct IgnoreLimitPositions |
| 30 |
{ |
| 31 |
int n; /* Number of entries per file */ |
| 32 |
int *pos; /* Store metaID1,position1, metaID2,position2 ..... */ |
| 33 |
}; |
| 34 |
|
| 35 |
/* This is used to build a list of the metaIDs that are currently in scope when indexing words */ |
| 36 |
|
| 37 |
typedef struct |
| 38 |
{ |
| 39 |
int *array; /* list of metaIDs that need to be indexed */ |
| 40 |
int max; /* max size of table */ |
| 41 |
int num; /* number in list */ |
| 42 |
int defaultID; /* default metaID (should always be one, I suppose) */ |
| 43 |
} |
| 44 |
METAIDTABLE; |
| 45 |
|
| 46 |
|
| 47 |
/* |
| 48 |
-- module data |
| 49 |
*/ |
| 50 |
|
| 51 |
|
| 52 |
struct MOD_Index |
| 53 |
{ |
| 54 |
/* entry vars */ |
| 55 |
METAIDTABLE metaIDtable; |
| 56 |
ENTRYARRAY *entryArray; |
| 57 |
ENTRY *hashentries[VERYBIGHASHSIZE]; |
| 58 |
char hashentriesdirty[VERYBIGHASHSIZE]; /* just a 0/1 flag */ |
| 59 |
|
| 60 |
/* Compression Work buffer while compression locations in index ** proccess */ |
| 61 |
unsigned char *compression_buffer; |
| 62 |
int len_compression_buffer; |
| 63 |
|
| 64 |
unsigned char *worddata_buffer; /* Buffer to store worddata */ |
| 65 |
int len_worddata_buffer; /* Max size of the buffer */ |
| 66 |
int sz_worddata_buffer; /* Space being used in worddata_buffer */ |
| 67 |
|
| 68 |
/* File counter */ |
| 69 |
int filenum; |
| 70 |
|
| 71 |
/* index tmp (both FS and HTTP methods) */ |
| 72 |
char *tmpdir; |
| 73 |
|
| 74 |
/* Filenames of the swap files */ |
| 75 |
char *swap_location_name[MAX_LOC_SWAP_FILES]; /* Location info file */ |
| 76 |
|
| 77 |
/* handlers for both files */ |
| 78 |
FILE *fp_loc_write[MAX_LOC_SWAP_FILES]; /* Location (writing) */ |
| 79 |
FILE *fp_loc_read[MAX_LOC_SWAP_FILES]; /* Location (reading) */ |
| 80 |
|
| 81 |
struct dev_ino *inode_hash[BIGHASHSIZE]; |
| 82 |
|
| 83 |
/* Buffers used by indexstring */ |
| 84 |
int lenswishword; |
| 85 |
char *swishword; |
| 86 |
int lenword; |
| 87 |
char *word; |
| 88 |
|
| 89 |
/* Economic mode (-e) */ |
| 90 |
int swap_locdata; /* swap location data */ |
| 91 |
|
| 92 |
/* Pointer to swap functions */ |
| 93 |
long (*swap_tell) (FILE *); |
| 94 |
size_t(*swap_write) (const void *, size_t, size_t, FILE *); |
| 95 |
int (*swap_seek) (FILE *, long, int); |
| 96 |
size_t(*swap_read) (void *, size_t, size_t, FILE *); |
| 97 |
int (*swap_close) (FILE *); |
| 98 |
int (*swap_putc) (int, FILE *); |
| 99 |
int (*swap_getc) (FILE *); |
| 100 |
|
| 101 |
/* IgnoreLimit option values */ |
| 102 |
int plimit; |
| 103 |
int flimit; |
| 104 |
/* Number of words from IgnoreLimit */ |
| 105 |
int nIgnoreLimitWords; |
| 106 |
/* Positions from stopwords from IgnoreLimit */ |
| 107 |
struct IgnoreLimitPositions **IgnoreLimitPositionsArray; |
| 108 |
|
| 109 |
/* Index in blocks of chunk_size files */ |
| 110 |
int chunk_size; |
| 111 |
|
| 112 |
/* Variable to control the size of the zone used for store locations during chunk proccesing */ |
| 113 |
int optimalChunkLocZoneSize; |
| 114 |
|
| 115 |
/* variable to handle free memory space for locations inside currentChunkLocZone */ |
| 116 |
|
| 117 |
LOCATION *freeLocMemChain; |
| 118 |
|
| 119 |
MEM_ZONE *perDocTmpZone; |
| 120 |
MEM_ZONE *currentChunkLocZone; |
| 121 |
MEM_ZONE *totalLocZone; |
| 122 |
MEM_ZONE *entryZone; |
| 123 |
}; |
| 124 |
|
| 125 |
void initModule_Index(SWISH *); |
| 126 |
void freeModule_Index(SWISH *); |
| 127 |
int configModule_Index(SWISH *, StringList *); |
| 128 |
|
| 129 |
|
| 130 |
void do_index_file(SWISH * sw, FileProp * fprop); |
| 131 |
|
| 132 |
ENTRY *getentry(SWISH * , char *); |
| 133 |
void addentry(SWISH *, ENTRY *, int, int, int, int); |
| 134 |
|
| 135 |
void addCommonProperties(SWISH * sw, FileProp * fprop, FileRec * fi, char *title, char *summary, int start); |
| 136 |
|
| 137 |
|
| 138 |
int getfilecount(IndexFILE *); |
| 139 |
|
| 140 |
int getNumberOfIgnoreLimitWords(SWISH *); |
| 141 |
void getPositionsFromIgnoreLimitWords(SWISH * sw); |
| 142 |
|
| 143 |
char *ruleparse(SWISH *, char *); |
| 144 |
void stripIgnoreFirstChars(INDEXDATAHEADER *, char *); |
| 145 |
void stripIgnoreLastChars(INDEXDATAHEADER *, char *); |
| 146 |
|
| 147 |
#define isIgnoreFirstChar(header,c) (header)->ignorefirstcharlookuptable[(int)((unsigned char)c)] |
| 148 |
#define isIgnoreLastChar(header,c) (header)->ignorelastcharlookuptable[(int)((unsigned char)c)] |
| 149 |
#define isBumpPositionCounterChar(header,c) (header)->bumpposcharslookuptable[(int)((unsigned char)c)] |
| 150 |
|
| 151 |
|
| 152 |
void computehashentry(ENTRY **, ENTRY *); |
| 153 |
|
| 154 |
void sort_words(SWISH *, IndexFILE *); |
| 155 |
void sortChunkLocations(SWISH *, IndexFILE *, ENTRY *); |
| 156 |
|
| 157 |
int indexstring(SWISH * sw, char *s, int filenum, int structure, int numMetaNames, int *metaID, int *position); |
| 158 |
|
| 159 |
void addsummarytofile(IndexFILE *, int, char *); |
| 160 |
|
| 161 |
void BuildSortedArrayOfWords(SWISH *, IndexFILE *); |
| 162 |
|
| 163 |
|
| 164 |
|
| 165 |
void PrintHeaderLookupTable(int ID, int table[], int table_size, FILE * fp); |
| 166 |
void coalesce_all_word_locations(SWISH * sw, IndexFILE * indexf); |
| 167 |
void coalesce_word_locations(SWISH * sw, IndexFILE * indexf, ENTRY * e); |
| 168 |
|
| 169 |
void adjustWordPositions(unsigned char *worddata, int *sz_worddata, int n_files, struct IgnoreLimitPositions **ilp); |
| 170 |
|
| 171 |
#endif |
Parent Directory
| 
Revision Log
| 
Revision Graph