lsm/mdbtools
1
0
Fork 0
mirror of https://github.com/mdbtools/mdbtools.git synced 2025-04-05 20:31:00 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix index counting, from Nirgal Vourgère

Browse Source
This commit is contained in:
Brian Bruns 2011-02-07 21:57:13 -05:00
parent 82b28d6845
commit be2a287f5c
2 changed files with 67 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
HACKING
View File

@ -316,7 +316,7 @@ next_pg field.
| ???? | 1 byte | col_name_len| len of the name of the column | | ???? | 1 byte | col_name_len| len of the name of the column |
| ???? | n bytes | col_name | Name of the column | | ???? | n bytes | col_name | Name of the column |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx (30+9 = 39 bytes) | | Iterate for indexes with index_type 0 or 1 (30+9 = 39 bytes) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate 10 times for 10 possible columns (10*3 = 30 bytes) | | Iterate 10 times for 10 possible columns (10*3 = 30 bytes) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
@ -327,7 +327,7 @@ next_pg field.
| ???? | 4 bytes | first_dp | Data pointer of the index page | | ???? | 4 bytes | first_dp | Data pointer of the index page |
| ???? | 1 byte | flags | See flags table for indexes | | ???? | 1 byte | flags | See flags table for indexes |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx (20 bytes) | | Iterate for the number of num_idx (20 bytes) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| ???? | 4 bytes | index_num | Number of the index | | ???? | 4 bytes | index_num | Number of the index |
| | | |(warn: not always in the sequential order)| | | | |(warn: not always in the sequential order)|
@ -338,7 +338,7 @@ next_pg field.
| 0x04 | 2 bytes | ??? | | | 0x04 | 2 bytes | ??? | |
| ???? | 1 byte | primary_key | 0x01 if this index is primary | | ???? | 1 byte | primary_key | 0x01 if this index is primary |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx | | Iterate for the number of num_idx |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| ???? | 1 byte | idx_name_len| len of the name of the index | | ???? | 1 byte | idx_name_len| len of the name of the index |
| ???? | n bytes | idx_name | Name of the index | | ???? | n bytes | idx_name | Name of the index |
@ -394,7 +394,7 @@ next_pg field.
| ???? | 2 bytes | col_name_len| len of the name of the column | | ???? | 2 bytes | col_name_len| len of the name of the column |
| ???? | n bytes | col_name | Name of the column (UCS-2 format) | | ???? | n bytes | col_name | Name of the column (UCS-2 format) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx (30+22 = 52 bytes) | | Iterate for indexes with index_type 0 or 1 (30+22 = 52 bytes) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| ???? | 4 bytes | ??? | | | ???? | 4 bytes | ??? | |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
@ -408,7 +408,7 @@ next_pg field.
| ???? | 1 byte | flags | See flags table for indexes | | ???? | 1 byte | flags | See flags table for indexes |
| ???? | 9 bytes | unknown | | | ???? | 9 bytes | unknown | |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx (28 bytes) | | Iterate for the number of num_idx (28 bytes) |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| ???? | 4 bytes | unknown | matches first unknown definition block | | ???? | 4 bytes | unknown | matches first unknown definition block |
| ???? | 4 bytes | index_num | Number of the index | | ???? | 4 bytes | index_num | Number of the index |
@ -421,7 +421,7 @@ next_pg field.
| ???? | 1 byte | primary_key | 0x01 if this index is primary | | ???? | 1 byte | primary_key | 0x01 if this index is primary |
| ???? | 4 bytes | unknown | | | ???? | 4 bytes | unknown | |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| Iterate for the number of num_real_idx | | Iterate for the number of num_idx |
+-------------------------------------------------------------------------+ +-------------------------------------------------------------------------+
| ???? | 2 bytes | idx_name_len| len of the name of the index | | ???? | 2 bytes | idx_name_len| len of the name of the index |
| ???? | n bytes | idx_name | Name of the index (UCS-2) | | ???? | n bytes | idx_name | Name of the index (UCS-2) |

81
src/libmdb/index.c
View File

@ -69,15 +69,15 @@ mdb_read_indices(MdbTableDef *table)
MdbHandle *mdb = entry->mdb; MdbHandle *mdb = entry->mdb;
MdbFormatConstants *fmt = mdb->fmt; MdbFormatConstants *fmt = mdb->fmt;
MdbIndex *pidx; MdbIndex *pidx;
unsigned int i, j; unsigned int i, j, k;
int idx_num, key_num, col_num; int key_num, col_num, cleaned_col_num;
int cur_pos, name_sz, idx2_sz, type_offset; int cur_pos, name_sz, idx2_sz, type_offset;
int index_start_pg = mdb->cur_pg; int index_start_pg = mdb->cur_pg;
gchar *tmpbuf; gchar *tmpbuf;
table->indices = g_ptr_array_new(); table->indices = g_ptr_array_new();
if (IS_JET4(mdb)) { if (IS_JET4(mdb)) {
cur_pos = table->index_start + 52 * table->num_real_idxs; cur_pos = table->index_start + 52 * table->num_real_idxs;
idx2_sz = 28; idx2_sz = 28;
type_offset = 23; type_offset = 23;
@ -87,15 +87,33 @@ mdb_read_indices(MdbTableDef *table)
type_offset = 19; type_offset = 19;
} }
//fprintf(stderr, "num_idxs:%d num_real_idxs:%d\n", table->num_idxs, table->num_real_idxs);
/* num_real_idxs should be the number of indexes of type 2.
* It's not always the case. Happens on Northwind Orders table.
*/
table->num_real_idxs = 0;
tmpbuf = (gchar *) g_malloc(idx2_sz); tmpbuf = (gchar *) g_malloc(idx2_sz);
for (i=0;i<table->num_idxs;i++) { for (i=0;i<table->num_idxs;i++) {
read_pg_if_n(mdb, tmpbuf, &cur_pos, idx2_sz); read_pg_if_n(mdb, tmpbuf, &cur_pos, idx2_sz);
pidx = (MdbIndex *) g_malloc0(sizeof(MdbIndex)); pidx = (MdbIndex *) g_malloc0(sizeof(MdbIndex));
pidx->table = table; pidx->table = table;
pidx->index_num = mdb_get_int16(tmpbuf, 4); pidx->index_num = mdb_get_int16(tmpbuf, 4);
pidx->index_type = tmpbuf[type_offset]; pidx->index_type = tmpbuf[type_offset];
g_ptr_array_add(table->indices, pidx); g_ptr_array_add(table->indices, pidx);
/*
{
gint32 dumy0 = mdb_get_int32(tmpbuf, 0);
gint8 dumy1 = tmpbuf[8];
gint32 dumy2 = mdb_get_int32(tmpbuf, 9);
gint32 dumy3 = mdb_get_int32(tmpbuf, 13);
gint16 dumy4 = mdb_get_int16(tmpbuf, 17);
fprintf(stderr, "idx #%d: num2:%d type:%d\n", i, pidx->index_num, pidx->index_type);
fprintf(stderr, "idx #%d: %d %d %d %d %d\n", i, dumy0, dumy1, dumy2, dumy3, dumy4);
}*/
if (pidx->index_type!=2)
table->num_real_idxs++;
} }
//fprintf(stderr, "num_idxs:%d num_real_idxs:%d\n", table->num_idxs, table->num_real_idxs);
g_free(tmpbuf); g_free(tmpbuf);
for (i=0;i<table->num_idxs;i++) { for (i=0;i<table->num_idxs;i++) {
@ -109,31 +127,35 @@ mdb_read_indices(MdbTableDef *table)
read_pg_if_n(mdb, tmpbuf, &cur_pos, name_sz); read_pg_if_n(mdb, tmpbuf, &cur_pos, name_sz);
mdb_unicode2ascii(mdb, tmpbuf, name_sz, pidx->name, MDB_MAX_OBJ_NAME); mdb_unicode2ascii(mdb, tmpbuf, name_sz, pidx->name, MDB_MAX_OBJ_NAME);
g_free(tmpbuf); g_free(tmpbuf);
//fprintf(stderr, "index name %s\n", pidx->name); //fprintf(stderr, "index %d type %d name %s\n", pidx->index_num, pidx->index_type, pidx->name);
} }
mdb_read_alt_pg(mdb, entry->table_pg); mdb_read_alt_pg(mdb, entry->table_pg);
mdb_read_pg(mdb, index_start_pg); mdb_read_pg(mdb, index_start_pg);
cur_pos = table->index_start; cur_pos = table->index_start;
idx_num=0;
for (i=0;i<table->num_real_idxs;i++) { for (i=0;i<table->num_real_idxs;i++) {
if (IS_JET4(mdb)) cur_pos += 4; if (IS_JET4(mdb)) cur_pos += 4;
do { /* look for index number i */
pidx = g_ptr_array_index (table->indices, idx_num++); for (j=0; j<table->num_idxs; ++j) {
} while (pidx && pidx->index_type==2); pidx = g_ptr_array_index (table->indices, j);
if (pidx->index_type!=2 && pidx->index_num==i)
/* if there are more real indexes than index entries left after break;
removing type 2's decrement real indexes and continue. Happens
on Northwind Orders table.
*/
if (!pidx) {
table->num_real_idxs--;
continue;
} }
if (j==table->num_idxs)
fprintf(stderr, "ERROR: can't find index #%d.\n", i);
//fprintf(stderr, "index %d #%d (%s) index_type:%d\n", i, pidx->index_num, pidx->name, pidx->index_type);
pidx->num_rows = mdb_get_int32(mdb->alt_pg_buf, pidx->num_rows = mdb_get_int32(mdb->alt_pg_buf,
fmt->tab_cols_start_offset + fmt->tab_cols_start_offset +
(i*fmt->tab_ridx_entry_size)); (pidx->index_num*fmt->tab_ridx_entry_size));
/*
fprintf(stderr, "ridx block1 i:%d data1:0x%08x data2:0x%08x\n",
i,
mdb_get_int32(mdb->pg_buf,
fmt->tab_cols_start_offset + pidx->index_num * fmt->tab_ridx_entry_size),
mdb_get_int32(mdb->pg_buf,
fmt->tab_cols_start_offset + pidx->index_num * fmt->tab_ridx_entry_size +4));
fprintf(stderr, "pidx->num_rows:%d\n", pidx->num_rows);*/
key_num=0; key_num=0;
for (j=0;j<MDB_MAX_IDX_COLS;j++) { for (j=0;j<MDB_MAX_IDX_COLS;j++) {
@ -142,17 +164,36 @@ mdb_read_indices(MdbTableDef *table)
cur_pos++; cur_pos++;
continue; continue;
} }
/* here we have the internal column number that does not
* always match the table columns because of deletions */
cleaned_col_num = -1;
for (k=0; k<=table->num_cols; k++) {
MdbColumn *col = g_ptr_array_index(table->columns,k);
if (col->col_num == col_num) {
cleaned_col_num = k;
break;
}
}
if (cleaned_col_num==-1) {
fprintf(stderr, "CRITICAL: can't find column with internal id %d in index %s\n",
col_num, pidx->name);
cur_pos++;
continue;
}
/* set column number to a 1 based column number and store */ /* set column number to a 1 based column number and store */
pidx->key_col_num[key_num] = col_num + 1; pidx->key_col_num[key_num] = cleaned_col_num + 1;
pidx->key_col_order[key_num] = pidx->key_col_order[key_num] =
(read_pg_if_8(mdb, &cur_pos)) ? MDB_ASC : MDB_DESC; (read_pg_if_8(mdb, &cur_pos)) ? MDB_ASC : MDB_DESC;
//fprintf(stderr, "component %d using column #%d (internally %d)\n", j, cleaned_col_num, col_num);
key_num++; key_num++;
} }
pidx->num_keys = key_num; pidx->num_keys = key_num;
cur_pos += 4; cur_pos += 4;
//fprintf(stderr, "pidx->unknown_pre_first_pg:0x%08x\n", read_pg_if_32(mdb, &cur_pos));
pidx->first_pg = read_pg_if_32(mdb, &cur_pos); pidx->first_pg = read_pg_if_32(mdb, &cur_pos);
pidx->flags = read_pg_if_8(mdb, &cur_pos); pidx->flags = read_pg_if_8(mdb, &cur_pos);
//fprintf(stderr, "pidx->first_pg:%d pidx->flags:0x%02x\n", pidx->first_pg, pidx->flags);
if (IS_JET4(mdb)) cur_pos += 9; if (IS_JET4(mdb)) cur_pos += 9;
} }
return NULL; return NULL;