diff --git a/contrib/intarray/README.intarray b/contrib/intarray/README.intarray index f0138a148427f46de71d3ece8d63ad0f7086b45e..bac191e65d30b3def9652e6cd36a9573be389044 100644 --- a/contrib/intarray/README.intarray +++ b/contrib/intarray/README.intarray @@ -8,10 +8,42 @@ signature with length of 4096 bits to represent sets). All work was done by Teodor Sigaev (teodor@stack.net) and Oleg Bartunov (oleg@sai.msu.su). See http://www.sai.msu.su/~megera/postgres/gist -for additional information. +for additional information. Andrey Oktyabrski has done a great work on +adding new functions and operations. + + +FUNCTIONS: + + int icount(int[]) - the number of elements in intarray + int[] sort(int[], 'asc' | 'desc') - sort intarray + int[] sort(int[]) - sort in ascending order + int[] sort_asc(int[]),sort_desc(int[]) - shortcuts for sort + int[] uniq(int[]) - returns unique elements + int idx(int[], int item) - returns index of first intarray matching element to item, or + '0' if matching failed. + int[] subarray(int[],int START [, int LEN]) - returns part of intarray starting from + element number START (from 1) and length LEN. + +OPERATIONS: + + int[] && int[] - overlap - returns TRUE if arrays has at least one common elements. + int[] @ int[] - contains - returns TRUE if left array contains right array + int[] ~ int[] - contained - returns TRUE if left array is contained in right array + # int[] - return the number of elements in array + int[] + int - push element to array ( add to end of array) + int[] + int[] - merge of arrays (right array added to the end of left one) + int[] - int - remove entries matched by right argument from array + int[] - int[] - remove left array from right + int[] | int - returns intarray - union of arguments + int[] | int[] - returns intarray as a union of two arrays + int[] & int[] - returns intersection of arrays CHANGES: +August 6, 2002 + 1. Reworked patch from Andrey Oktyabrski (ano@spider.ru) with + functions: icount, sort, sort_asc, uniq, idx, subarray + operations: #, +, -, |, & October 1, 2001 1. Change search method in array to binary September 28, 2001 diff --git a/contrib/intarray/_int.c b/contrib/intarray/_int.c index d956543af5f075eb8fc05e63248a3150e3c9d375..1c0de0bce80a59cc6cc8305b5532deeb0997a0d6 100644 --- a/contrib/intarray/_int.c +++ b/contrib/intarray/_int.c @@ -2217,3 +2217,303 @@ querytree(PG_FUNCTION_ARGS) { PG_RETURN_POINTER( res ); } + +/* +** Additional array functions +*/ +static int32 intarray_match_first(ArrayType *a, int32 elem); +static ArrayType *intarray_add_elem(ArrayType *a, int32 elem); +static ArrayType *intarray_concat_arrays(ArrayType *a, ArrayType *b); +static ArrayType *int_to_intset(int32 elem); + +PG_FUNCTION_INFO_V1( intset ); +PG_FUNCTION_INFO_V1( icount ); +PG_FUNCTION_INFO_V1( sort ); +PG_FUNCTION_INFO_V1( sort_asc ); +PG_FUNCTION_INFO_V1( sort_desc ); +PG_FUNCTION_INFO_V1( uniq ); +PG_FUNCTION_INFO_V1( idx ); +PG_FUNCTION_INFO_V1( subarray ); +PG_FUNCTION_INFO_V1( intarray_push_elem ); +PG_FUNCTION_INFO_V1( intarray_push_array ); +PG_FUNCTION_INFO_V1( intarray_del_elem ); +PG_FUNCTION_INFO_V1( intset_union_elem ); +PG_FUNCTION_INFO_V1( intset_subtract ); +Datum intset(PG_FUNCTION_ARGS); +Datum icount(PG_FUNCTION_ARGS); +Datum sort(PG_FUNCTION_ARGS); +Datum sort_asc(PG_FUNCTION_ARGS); +Datum sort_desc(PG_FUNCTION_ARGS); +Datum uniq(PG_FUNCTION_ARGS); +Datum idx(PG_FUNCTION_ARGS); +Datum subarray(PG_FUNCTION_ARGS); +Datum intarray_push_elem(PG_FUNCTION_ARGS); +Datum intarray_push_array(PG_FUNCTION_ARGS); +Datum intarray_del_elem(PG_FUNCTION_ARGS); +Datum intset_union_elem(PG_FUNCTION_ARGS); +Datum intset_subtract(PG_FUNCTION_ARGS); + +static int32 +intarray_match_first(ArrayType *a, int32 elem) { + int32 *aa, c, i; + c = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + aa = ARRPTR(a); + for (i = 0; i < c; i++) + if (aa[i] == elem) return (i + 1); + return 0; +} + +static ArrayType * +intarray_add_elem(ArrayType *a, int32 elem) { + ArrayType *result; + int32 *r; + int32 c = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + result = new_intArrayType(c + 1); + r = ARRPTR(result); + if (c > 0) memcpy(r, ARRPTR(a), c * sizeof(int32)); + r[c] = elem; + return result; +} + +static ArrayType * +intarray_concat_arrays(ArrayType *a, ArrayType *b) { + ArrayType *result; + int32 ac = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + int32 bc = (ARRISVOID(b)) ? 0 : ARRNELEMS(b); + result = new_intArrayType(ac+bc); + if ( ac ) + memcpy(ARRPTR(result), ARRPTR(a), ac * sizeof(int32)); + if ( bc ) + memcpy(ARRPTR(result) + ac, ARRPTR(b), bc * sizeof(int32)); + return result; +} + +static ArrayType * +int_to_intset(int32 n) { + ArrayType *result; + int32 *aa; + result = new_intArrayType(1); + aa = ARRPTR(result); + aa[0] = n; + return result; +} + +static int +compASC(const void *a, const void *b) { + if ( *(int4*)a == *(int4*)b ) return 0; + return ( *(int4*)a > *(int4*)b ) ? 1 : -1; +} + +static int +compDESC(const void *a, const void *b) { + if ( *(int4*)a == *(int4*)b ) return 0; + return ( *(int4*)a < *(int4*)b ) ? 1 : -1; +} + +#define QSORT(a, direction) \ +if (ARRNELEMS(a) > 1) \ + qsort((void*)ARRPTR(a), ARRNELEMS(a),sizeof(int4), \ + (direction) ? compASC : compDESC ) + +#define UNIX_UNIQ(a) a = resize_intArrayType(a, unix_uniq(ARRPTR(a), ARRNELEMS(a))) + +static int32 +unix_uniq(int32 *array, int32 count) { + register int32 i, k = 0; + for (i = 1; i < count; i++) + if (array[k] != array[i]) { + k++; + if (i > k) array[k] = array[i]; + } + return (k+1); +} + +Datum +intset(PG_FUNCTION_ARGS) { + PG_RETURN_POINTER(int_to_intset(PG_GETARG_INT32(0))); +} + +Datum +icount(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + int32 count = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + PG_FREE_IF_COPY(a, 0); + PG_RETURN_INT32(count); +} + +Datum +sort(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + text *dirstr = ( fcinfo->nargs==2 ) ? PG_GETARG_TEXT_P(1) : NULL; + int32 dc = ( dirstr ) ? VARSIZE(dirstr)-VARHDRSZ : 0; + char *d = ( dirstr ) ? VARDATA(dirstr) : NULL; + int dir = -1; + if (ARRISVOID(a) || ARRNELEMS(a) < 2) PG_RETURN_POINTER(a); + + if (dirstr==NULL || (dc == 3 + && (d[0] == 'A' || d[0] == 'a') + && (d[1] == 'S' || d[1] == 's') + && (d[2] == 'C' || d[2] == 'c'))) + dir = 1; + else if (dc == 4 + && (d[0] == 'D' || d[0] == 'd') + && (d[1] == 'E' || d[1] == 'e') + && (d[2] == 'S' || d[2] == 's') + && (d[3] == 'C' || d[3] == 'c')) + dir = 0; + if (dir == -1) + elog(ERROR, "Invalid second parameter in function sort. It must be 'ASC' or 'DESC'."); + QSORT(a, dir); + PG_RETURN_POINTER(a); +} + +Datum +sort_asc(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + if (ARRISVOID(a)) PG_RETURN_POINTER(a); + QSORT(a, 1); + PG_RETURN_POINTER(a); +} + +Datum +sort_desc(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + if (ARRISVOID(a)) PG_RETURN_POINTER(a); + QSORT(a, 0); + PG_RETURN_POINTER(a); +} + +Datum +uniq(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + if (ARRISVOID(a) || ARRNELEMS(a) < 2) PG_RETURN_POINTER(a); + UNIX_UNIQ(a); + PG_RETURN_POINTER(a); +} + +Datum +idx(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + int32 result = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + if (result) result = intarray_match_first(a, PG_GETARG_INT32(1)); + PG_FREE_IF_COPY(a, 0); + PG_RETURN_INT32(result); +} + +Datum +subarray(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + ArrayType *result; + int32 start = ( PG_GETARG_INT32(1) > 0 ) ? PG_GETARG_INT32(1)-1 : PG_GETARG_INT32(1); + int32 len = ( fcinfo->nargs==3 ) ? PG_GETARG_INT32(2) : 0; + int32 end = 0; + int32 c; + + if ( ARRISVOID(a) ) { + PG_FREE_IF_COPY(a, 0); + PG_RETURN_POINTER( new_intArrayType(0) ); + } + + c = ARRNELEMS(a); + + if ( start < 0 ) + start = c + start; + + if ( len < 0 ) + end = c + len; + else if ( len == 0 ) + end = c; + else + end = start + len; + + if ( end > c ) + end = c; + + if ( start < 0 ) + start = 0; + + if ( start >= end || end <= 0 ) { + PG_FREE_IF_COPY(a, 0); + PG_RETURN_POINTER( new_intArrayType(0) ); + } + + + result = new_intArrayType(end-start); + if (end-start > 0) + memcpy(ARRPTR(result), ARRPTR(a) + start, (end-start) * sizeof(int32)); + PG_FREE_IF_COPY(a, 0); + PG_RETURN_POINTER(result); +} + +Datum +intarray_push_elem(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + ArrayType *result; + result = intarray_add_elem(a, PG_GETARG_INT32(1)); + PG_FREE_IF_COPY(a, 0); + PG_RETURN_POINTER(result); +} + +Datum +intarray_push_array(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + ArrayType *b = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(1))); + ArrayType *result; + result = intarray_concat_arrays(a, b); + PG_FREE_IF_COPY(a, 0); + PG_FREE_IF_COPY(b, 1); + PG_RETURN_POINTER(result); +} + +Datum +intarray_del_elem(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + int32 c = (ARRISVOID(a)) ? 0 : ARRNELEMS(a); + int32 *aa = ARRPTR(a); + int32 n = 0, i; + int32 elem = PG_GETARG_INT32(1); + for (i = 0; i < c; i++) + if (aa[i] != elem) { + if (i > n) aa[n++] = aa[i]; + else n++; + } + if (c > 0) a = resize_intArrayType(a, n); + PG_RETURN_POINTER(a); +} + +Datum +intset_union_elem(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM(PG_GETARG_DATUM(0))); + ArrayType *result; + result = intarray_add_elem(a, PG_GETARG_INT32(1)); + PG_FREE_IF_COPY(a, 0); + QSORT(result, 1); + UNIX_UNIQ(result); + PG_RETURN_POINTER(result); +} + +Datum +intset_subtract(PG_FUNCTION_ARGS) { + ArrayType *a = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(0))); + ArrayType *b = (ArrayType*)DatumGetPointer(PG_DETOAST_DATUM_COPY(PG_GETARG_DATUM(1))); + ArrayType *result; + int32 ca = ARRISVOID(a); + int32 cb = ARRISVOID(b); + int32 *aa, *bb, *r; + int32 n = 0, i = 0, k = 0; + QSORT(a, 1); UNIX_UNIQ(a); ca = ARRNELEMS(a); + QSORT(b, 1); UNIX_UNIQ(b); cb = ARRNELEMS(b); + result = new_intArrayType(ca); + aa = ARRPTR(a); + bb = ARRPTR(b); + r = ARRPTR(result); + while (i < ca) { + if (k == cb || aa[i] < bb[k]) r[n++] = aa[i++]; + else if (aa[i] == bb[k]) { i++; k++; } + else k++; + } + result = resize_intArrayType(result, n); + pfree(a); pfree(b); + PG_RETURN_POINTER(result); +} + diff --git a/contrib/intarray/_int.sql.in b/contrib/intarray/_int.sql.in index 24a87cb0356ec5158bbf94e830143bec0535ff85..fe1a6f9d389b05efc26bdceef635afb9673a03e5 100644 --- a/contrib/intarray/_int.sql.in +++ b/contrib/intarray/_int.sql.in @@ -123,6 +123,81 @@ CREATE OPERATOR ~ ( COMMUTATOR = '@', RESTRICT = contsel, JOIN = contjoinsel ); +-------------- +CREATE FUNCTION intset(int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION icount(_int4) RETURNS int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR # ( + RIGHTARG = _int4, PROCEDURE = icount +); + +CREATE FUNCTION sort(_int4, text) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION sort(_int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION sort_asc(_int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION sort_desc(_int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION uniq(_int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION idx(_int4, int4) RETURNS int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR # ( + LEFTARG = _int4, RIGHTARG = int4, PROCEDURE = idx +); + +CREATE FUNCTION subarray(_int4, int4, int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION subarray(_int4, int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); + +CREATE FUNCTION intarray_push_elem(_int4, int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR + ( + LEFTARG = _int4, RIGHTARG = int4, PROCEDURE = intarray_push_elem +); + +CREATE FUNCTION intarray_push_array(_int4, _int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR + ( + LEFTARG = _int4, RIGHTARG = _int4, COMMUTATOR = +, PROCEDURE = intarray_push_array +); + +CREATE FUNCTION intarray_del_elem(_int4, int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR - ( + LEFTARG = _int4, RIGHTARG = int4, PROCEDURE = intarray_del_elem +); + +CREATE FUNCTION intset_union_elem(_int4, int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR | ( + LEFTARG = _int4, RIGHTARG = int4, PROCEDURE = intset_union_elem +); + +CREATE OPERATOR | ( + LEFTARG = _int4, RIGHTARG = _int4, COMMUTATOR = |, PROCEDURE = _int_union +); + +CREATE FUNCTION intset_subtract(_int4, _int4) RETURNS _int4 + AS 'MODULE_PATHNAME' LANGUAGE 'c' WITH (isStrict, isCachable); +CREATE OPERATOR - ( + LEFTARG = _int4, RIGHTARG = _int4, PROCEDURE = intset_subtract +); + +CREATE OPERATOR & ( + LEFTARG = _int4, RIGHTARG = _int4, COMMUTATOR = &, PROCEDURE = _int_inter +); +-------------- -- define the GiST support methods CREATE FUNCTION g_int_consistent(opaque,_int4,int4) RETURNS bool diff --git a/contrib/intarray/expected/_int.out b/contrib/intarray/expected/_int.out index d1b440f5a9be3f97a42d6456b3f16690fb39a389..87298630be45a2f1a4dd490e2449d9eb6de60ba7 100644 --- a/contrib/intarray/expected/_int.out +++ b/contrib/intarray/expected/_int.out @@ -3,6 +3,144 @@ -- does not depend on contents of seg.sql. -- \set ECHO none +select intset(1234); + intset +-------- + {1234} +(1 row) + +select icount('{1234234,234234}'); + icount +-------- + 2 +(1 row) + +select sort('{1234234,-30,234234}'); + sort +---------------------- + {-30,234234,1234234} +(1 row) + +select sort('{1234234,-30,234234}','asc'); + sort +---------------------- + {-30,234234,1234234} +(1 row) + +select sort('{1234234,-30,234234}','desc'); + sort +---------------------- + {1234234,234234,-30} +(1 row) + +select sort_asc('{1234234,-30,234234}'); + sort_asc +---------------------- + {-30,234234,1234234} +(1 row) + +select sort_desc('{1234234,-30,234234}'); + sort_desc +---------------------- + {1234234,234234,-30} +(1 row) + +select uniq('{1234234,-30,-30,234234,-30}'); + uniq +-------------------------- + {1234234,-30,234234,-30} +(1 row) + +select uniq(sort_asc('{1234234,-30,-30,234234,-30}')); + uniq +---------------------- + {-30,234234,1234234} +(1 row) + +select idx('{1234234,-30,-30,234234,-30}',-30); + idx +----- + 2 +(1 row) + +select subarray('{1234234,-30,-30,234234,-30}',2,3); + subarray +------------------ + {-30,-30,234234} +(1 row) + +select subarray('{1234234,-30,-30,234234,-30}',-1,1); + subarray +---------- + {-30} +(1 row) + +select subarray('{1234234,-30,-30,234234,-30}',0,-1); + subarray +-------------------------- + {1234234,-30,-30,234234} +(1 row) + +select #'{1234234,234234}'::int[]; + ?column? +---------- + 2 +(1 row) + +select '{123,623,445}'::int[] + 1245; + ?column? +-------------------- + {123,623,445,1245} +(1 row) + +select '{123,623,445}'::int[] + 445; + ?column? +------------------- + {123,623,445,445} +(1 row) + +select '{123,623,445}'::int[] + '{1245,87,445}'; + ?column? +--------------------------- + {123,623,445,1245,87,445} +(1 row) + +select '{123,623,445}'::int[] - 623; + ?column? +----------- + {123,445} +(1 row) + +select '{123,623,445}'::int[] - '{1623,623}'; + ?column? +----------- + {123,445} +(1 row) + +select '{123,623,445}'::int[] | 623; + ?column? +--------------- + {123,445,623} +(1 row) + +select '{123,623,445}'::int[] | 1623; + ?column? +-------------------- + {123,445,623,1623} +(1 row) + +select '{123,623,445}'::int[] | '{1623,623}'; + ?column? +-------------------- + {123,445,623,1623} +(1 row) + +select '{123,623,445}'::int[] & '{1623,623}'; + ?column? +---------- + {623} +(1 row) + --test query_int select '1'::query_int; query_int diff --git a/contrib/intarray/sql/_int.sql b/contrib/intarray/sql/_int.sql index 52313ef7a1746ec30eec41642192abd36db24de3..2b0b0d10194ff1d9f8cec9e9fb2eb1c71510b586 100644 --- a/contrib/intarray/sql/_int.sql +++ b/contrib/intarray/sql/_int.sql @@ -6,6 +6,32 @@ \i _int.sql \set ECHO all +select intset(1234); +select icount('{1234234,234234}'); +select sort('{1234234,-30,234234}'); +select sort('{1234234,-30,234234}','asc'); +select sort('{1234234,-30,234234}','desc'); +select sort_asc('{1234234,-30,234234}'); +select sort_desc('{1234234,-30,234234}'); +select uniq('{1234234,-30,-30,234234,-30}'); +select uniq(sort_asc('{1234234,-30,-30,234234,-30}')); +select idx('{1234234,-30,-30,234234,-30}',-30); +select subarray('{1234234,-30,-30,234234,-30}',2,3); +select subarray('{1234234,-30,-30,234234,-30}',-1,1); +select subarray('{1234234,-30,-30,234234,-30}',0,-1); + +select #'{1234234,234234}'::int[]; +select '{123,623,445}'::int[] + 1245; +select '{123,623,445}'::int[] + 445; +select '{123,623,445}'::int[] + '{1245,87,445}'; +select '{123,623,445}'::int[] - 623; +select '{123,623,445}'::int[] - '{1623,623}'; +select '{123,623,445}'::int[] | 623; +select '{123,623,445}'::int[] | 1623; +select '{123,623,445}'::int[] | '{1623,623}'; +select '{123,623,445}'::int[] & '{1623,623}'; + + --test query_int select '1'::query_int; select ' 1'::query_int;