Files
calc/assocfunc.c
Landon Curt Noll a31078bbec Remove all RCS @(#) lines and RCS strings
Some folks might think: “you still use RCS”?!?  And we will say,
hey, at least we switched from SCCS to RCS back in … I think it was
around 1994 ... at least we are keeping up! :-) :-) :-)

Logs say that SCCS version 18 became RCS version 19 on 1994 March 18.

RCS served us well.  But now it is time to move on.   And so we are
switching to git.

Calc releases produce a lot of file changes.  In the 125 releases
of calc since 1996, when I started managing calc releases, there
have been 15473 file mods!
2017-05-23 01:33:23 -07:00

536 lines
12 KiB
C

/*
* assocfunc - association table routines
*
* Copyright (C) 1999-2007 David I. Bell
*
* Calc is open software; you can redistribute it and/or modify it under
* the terms of the version 2.1 of the GNU Lesser General Public License
* as published by the Free Software Foundation.
*
* Calc is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
* or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General
* Public License for more details.
*
* A copy of version 2.1 of the GNU Lesser General Public License is
* distributed with calc under the filename COPYING-LGPL. You should have
* received a copy with calc; if not, write to Free Software Foundation, Inc.
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*
* Under source code control: 1993/07/20 23:04:27
* File existed as early as: 1993
*
* Share and enjoy! :-) http://www.isthe.com/chongo/tech/comp/calc/
*/
/*
* Association table routines.
* An association table is a type of value which can be "indexed" by
* one or more arbitrary values. Each element in the table is thus an
* association between a particular set of index values and a result value.
* The elements in an association table are stored in a hash table for
* quick access.
*/
#include "value.h"
#define MINHASHSIZE 31 /* minimum size of hash tables */
#define GROWHASHSIZE 50 /* approximate growth for hash tables */
#define CHAINLENGTH 10 /* desired number of elements on a hash chain */
#define ELEMSIZE(n) (sizeof(ASSOCELEM) + (sizeof(VALUE) * ((n) - 1)))
S_FUNC ASSOCELEM *elemindex(ASSOC *ap, long index);
S_FUNC BOOL compareindices(VALUE *v1, VALUE *v2, long dim);
S_FUNC void resize(ASSOC *ap, long newsize);
S_FUNC void assoc_elemfree(ASSOCELEM *ep);
/*
* Return the address of the value specified by normal indexing of
* an association. The create flag is TRUE if a value is going to be
* assigned into the specified indexing location. If create is FALSE and
* the index value doesn't exist, a pointer to a NULL value is returned.
*
* given:
* ap association to index into
* create whether to create the index value
* dim dimension of the indexing
* indices table of values being indexed by
*/
VALUE *
associndex(ASSOC *ap, BOOL create, long dim, VALUE *indices)
{
ASSOCELEM **listhead;
ASSOCELEM *ep;
STATIC VALUE val;
QCKHASH hash;
int i;
if (dim < 0) {
math_error("Negative dimension for indexing association");
/*NOTREACHED*/
}
/*
* Calculate the hash value to use for this set of indices
* so that we can first select the correct hash chain, and
* also so we can quickly compare each element for a match.
*/
hash = QUICKHASH_BASIS;
for (i = 0; i < dim; i++)
hash = hashvalue(&indices[i], hash);
/*
* Search the correct hash chain for the specified set of indices.
* If found, return the address of the found element's value.
*/
listhead = &ap->a_table[hash % ap->a_size];
for (ep = *listhead; ep; ep = ep->e_next) {
if ((ep->e_hash != hash) || (ep->e_dim != dim))
continue;
if (compareindices(ep->e_indices, indices, dim))
return &ep->e_value;
}
/*
* The set of indices was not found.
* Either return a pointer to a NULL value for a read reference,
* or allocate a new element in the list for a write reference.
*/
if (!create) {
val.v_type = V_NULL;
val.v_subtype = V_NOSUBTYPE;
return &val;
}
ep = (ASSOCELEM *) malloc(ELEMSIZE(dim));
if (ep == NULL) {
math_error("Cannot allocate association element");
/*NOTREACHED*/
}
ep->e_dim = dim;
ep->e_hash = hash;
ep->e_value.v_type = V_NULL;
ep->e_value.v_subtype = V_NOSUBTYPE;
for (i = 0; i < dim; i++)
copyvalue(&indices[i], &ep->e_indices[i]);
ep->e_next = *listhead;
*listhead = ep;
ap->a_count++;
resize(ap, ap->a_count / CHAINLENGTH);
return &ep->e_value;
}
/*
* Search an association for the specified value starting at the
* specified index. Returns 0 and stores index if value found,
* otherwise returns 1.
*/
int
assocsearch(ASSOC *ap, VALUE *vp, long i, long j, ZVALUE *index)
{
ASSOCELEM *ep;
if (i < 0 || j > ap->a_count) {
math_error("This should not happen in assocsearch");
/*NOTREACHED*/
}
while (i < j) {
ep = elemindex(ap, i);
if (ep == NULL) {
math_error("This should not happen in assocsearch");
/*NOTREACHED*/
}
if (acceptvalue(&ep->e_value, vp)) {
utoz(i, index);
return 0;
}
i++;
}
return 1;
}
/*
* Search an association backwards for the specified value starting at the
* specified index. Returns 0 and stores the index if the value is
* found; otherwise returns 1.
*/
int
assocrsearch(ASSOC *ap, VALUE *vp, long i, long j, ZVALUE *index)
{
ASSOCELEM *ep;
if (i < 0 || j > ap->a_count) {
math_error("This should not happen in assocsearch");
/*NOTREACHED*/
}
j--;
while (j >= i) {
ep = elemindex(ap, j);
if (ep == NULL) {
math_error("This should not happen in assocsearch");
/*NOTREACHED*/
}
if (acceptvalue(&ep->e_value, vp)) {
utoz(j, index);
return 0;
}
j--;
}
return 1;
}
/*
* Return the address of an element of an association indexed by the
* double-bracket operation.
*
* given:
* ap association to index into
* index index of desired element
*/
S_FUNC ASSOCELEM *
elemindex(ASSOC *ap, long index)
{
ASSOCELEM *ep;
int i;
if ((index < 0) || (index > ap->a_count))
return NULL;
/*
* This loop should be made more efficient by remembering
* previously requested locations within the association.
*/
for (i = 0; i < ap->a_size; i++) {
for (ep = ap->a_table[i]; ep; ep = ep->e_next) {
if (index-- == 0)
return ep;
}
}
return NULL;
}
/*
* Return the address of the value specified by double-bracket indexing
* of an association. Returns NULL if there is no such element.
*
* given:
* ap association to index into
* index index of desired element
*/
VALUE *
assocfindex(ASSOC *ap, long index)
{
ASSOCELEM *ep;
ep = elemindex(ap, index);
if (ep == NULL)
return NULL;
return &ep->e_value;
}
/*
* Returns the list of indices for an association element with specified
* double-bracket index.
*/
LIST *
associndices(ASSOC *ap, long index)
{
ASSOCELEM *ep;
LIST *lp;
int i;
ep = elemindex(ap, index);
if (ep == NULL)
return NULL;
lp = listalloc();
for (i = 0; i < ep->e_dim; i++)
insertlistlast(lp, &ep->e_indices[i]);
return lp;
}
/*
* Compare two associations to see if they are identical.
* Returns TRUE if they are different.
*/
BOOL
assoccmp(ASSOC *ap1, ASSOC *ap2)
{
ASSOCELEM **table1;
ASSOCELEM *ep1;
ASSOCELEM *ep2;
long size1;
long size2;
QCKHASH hash;
long dim;
if (ap1 == ap2)
return FALSE;
if (ap1->a_count != ap2->a_count)
return TRUE;
table1 = ap1->a_table;
size1 = ap1->a_size;
size2 = ap2->a_size;
while (size1-- > 0) {
for (ep1 = *table1++; ep1; ep1 = ep1->e_next) {
hash = ep1->e_hash;
dim = ep1->e_dim;
for (ep2 = ap2->a_table[hash % size2]; ;
ep2 = ep2->e_next) {
if (ep2 == NULL)
return TRUE;
if (ep2->e_hash != hash)
continue;
if (ep2->e_dim != dim)
continue;
if (compareindices(ep1->e_indices,
ep2->e_indices, dim))
break;
}
if (comparevalue(&ep1->e_value, &ep2->e_value))
return TRUE;
}
}
return FALSE;
}
/*
* Copy an association value.
*/
ASSOC *
assoccopy(ASSOC *oldap)
{
ASSOC *ap;
ASSOCELEM *oldep;
ASSOCELEM *ep;
ASSOCELEM **listhead;
int oldhi;
int i;
ap = assocalloc(oldap->a_count / CHAINLENGTH);
ap->a_count = oldap->a_count;
for (oldhi = 0; oldhi < oldap->a_size; oldhi++) {
for (oldep = oldap->a_table[oldhi]; oldep;
oldep = oldep->e_next) {
ep = (ASSOCELEM *) malloc(ELEMSIZE(oldep->e_dim));
if (ep == NULL) {
math_error("Cannot allocate "
"association element");
/*NOTREACHED*/
}
ep->e_dim = oldep->e_dim;
ep->e_hash = oldep->e_hash;
ep->e_value.v_type = V_NULL;
ep->e_value.v_subtype = V_NOSUBTYPE;
for (i = 0; i < ep->e_dim; i++)
copyvalue(&oldep->e_indices[i],
&ep->e_indices[i]);
copyvalue(&oldep->e_value, &ep->e_value);
listhead = &ap->a_table[ep->e_hash % ap->a_size];
ep->e_next = *listhead;
*listhead = ep;
}
}
return ap;
}
/*
* Resize the hash table for an association to be the specified size.
* This is only actually done if the growth from the previous size is
* enough to make this worthwhile.
*/
S_FUNC void
resize(ASSOC *ap, long newsize)
{
ASSOCELEM **oldtable;
ASSOCELEM **newtable;
ASSOCELEM **oldlist;
ASSOCELEM **newlist;
ASSOCELEM *ep;
int i;
if (newsize < ap->a_size + GROWHASHSIZE)
return;
newsize = (long) next_prime((FULL)newsize);
newtable = (ASSOCELEM **) malloc(sizeof(ASSOCELEM *) * newsize);
if (newtable == NULL) {
math_error("No memory to grow association");
/*NOTREACHED*/
}
for (i = 0; i < newsize; i++)
newtable[i] = NULL;
oldtable = ap->a_table;
oldlist = oldtable;
for (i = 0; i < ap->a_size; i++) {
while (*oldlist) {
ep = *oldlist;
*oldlist = ep->e_next;
newlist = &newtable[ep->e_hash % newsize];
ep->e_next = *newlist;
*newlist = ep;
}
oldlist++;
}
ap->a_table = newtable;
ap->a_size = newsize;
free((char *) oldtable);
}
/*
* Free an association element, along with any contained values.
*/
S_FUNC void
assoc_elemfree(ASSOCELEM *ep)
{
int i;
for (i = 0; i < ep->e_dim; i++)
freevalue(&ep->e_indices[i]);
freevalue(&ep->e_value);
ep->e_dim = 0;
ep->e_next = NULL;
free((char *) ep);
}
/*
* Allocate a new association value with an initial hash table.
* The hash table size is set at specified (but at least a minimum size).
*/
ASSOC *
assocalloc(long initsize)
{
register ASSOC *ap;
int i;
if (initsize < MINHASHSIZE)
initsize = MINHASHSIZE;
ap = (ASSOC *) malloc(sizeof(ASSOC));
if (ap == NULL) {
math_error("No memory for association");
/*NOTREACHED*/
}
ap->a_count = 0;
ap->a_size = initsize;
ap->a_table = (ASSOCELEM **) malloc(sizeof(ASSOCELEM *) * initsize);
if (ap->a_table == NULL) {
free((char *) ap);
math_error("No memory for association");
/*NOTREACHED*/
}
for (i = 0; i < initsize; i++)
ap->a_table[i] = NULL;
return ap;
}
/*
* Free an association value, along with all of its elements.
*/
void
assocfree(ASSOC *ap)
{
ASSOCELEM **listhead;
ASSOCELEM *ep;
ASSOCELEM *nextep;
int i;
listhead = ap->a_table;
for (i = 0; i < ap->a_size; i++) {
nextep = *listhead;
*listhead = NULL;
while (nextep) {
ep = nextep;
nextep = ep->e_next;
assoc_elemfree(ep);
}
listhead++;
}
free((char *) ap->a_table);
ap->a_table = NULL;
free((char *) ap);
}
/*
* Print out an association along with the specified number of
* its elements. The elements are printed out in shortened form.
*/
void
assocprint(ASSOC *ap, long max_print)
{
ASSOCELEM *ep;
long index;
long i;
int savemode;
if (max_print <= 0) {
math_fmt("assoc (%ld element%s)", ap->a_count,
((ap->a_count == 1) ? "" : "s"));
return;
}
math_fmt("\n assoc (%ld element%s):\n", ap->a_count,
((ap->a_count == 1) ? "" : "s"));
for (index = 0; ((index < max_print) && (index < ap->a_count));
index++) {
ep = elemindex(ap, index);
if (ep == NULL)
continue;
math_str(" [");
for (i = 0; i < ep->e_dim; i++) {
if (i)
math_chr(',');
savemode = math_setmode(MODE_FRAC);
printvalue(&ep->e_indices[i],
(PRINT_SHORT | PRINT_UNAMBIG));
math_setmode(savemode);
}
math_str("] = ");
printvalue(&ep->e_value, PRINT_SHORT | PRINT_UNAMBIG);
math_chr('\n');
}
if (max_print < ap->a_count)
math_str(" ...\n");
}
/*
* Compare two lists of index values to see if they are identical.
* Returns TRUE if they are the same.
*/
S_FUNC BOOL
compareindices(VALUE *v1, VALUE *v2, long dim)
{
int i;
for (i = 0; i < dim; i++)
if (v1[i].v_type != v2[i].v_type)
return FALSE;
while (dim-- > 0)
if (comparevalue(v1++, v2++))
return FALSE;
return TRUE;
}