jbe@0: /*
jbe@7:  *  Copyright (c) 2006-2007 Jan Behrens, FlexiGuided GmbH, Berlin
jbe@0:  *
jbe@7:  *  Permission is hereby granted, free of charge, to any person obtaining a
jbe@7:  *  copy of this software and associated documentation files (the "Software"),
jbe@7:  *  to deal in the Software without restriction, including without limitation
jbe@7:  *  the rights to use, copy, modify, merge, publish, distribute, sublicense,
jbe@7:  *  and/or sell copies of the Software, and to permit persons to whom the
jbe@7:  *  Software is furnished to do so, subject to the following conditions:
jbe@0:  *
jbe@7:  *  The above copyright notice and this permission notice shall be included in
jbe@7:  *  all copies or substantial portions of the Software.
jbe@0:  *
jbe@7:  *  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
jbe@7:  *  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
jbe@7:  *  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
jbe@7:  *  AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
jbe@7:  *  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
jbe@7:  *  FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
jbe@7:  *  DEALINGS IN THE SOFTWARE.
jbe@0:  */
jbe@7: 
jbe@0: 
jbe@0: /*
jbe@0:  *  File name:    pgsql/utf8proc_pgsql.c
jbe@7:  *  Version:      1.1.1
jbe@7:  *  Last changed: 2007-07-22
jbe@0:  *
jbe@0:  *  Description:
jbe@7:  *  PostgreSQL extension to provide two functions 'unifold' and 'unistrip',
jbe@7:  *  which can be used to case-fold and normalize index fields and
jbe@7:  *  optionally strip marks (e.g. accents) from strings.
jbe@0:  */
jbe@0: 
jbe@0: 
jbe@0: #include "../utf8proc.c"
jbe@0: 
jbe@0: #include <postgres.h>
jbe@0: #include <utils/elog.h>
jbe@0: #include <fmgr.h>
jbe@1: #include <string.h>
jbe@0: #include <unistd.h>
jbe@0: #include <utils/builtins.h>
jbe@0: 
jbe@5: #ifdef PG_MODULE_MAGIC
jbe@5: PG_MODULE_MAGIC;
jbe@5: #endif
jbe@5: 
jbe@7: #define UTF8PROC_PGSQL_FOLD_OPTS ( UTF8PROC_REJECTNA | UTF8PROC_COMPAT | \
jbe@2:   UTF8PROC_COMPOSE | UTF8PROC_STABLE | UTF8PROC_IGNORE | UTF8PROC_STRIPCC | \
jbe@3:   UTF8PROC_NLF2LF | UTF8PROC_CASEFOLD | UTF8PROC_LUMP )
jbe@7: #define UTF8PROC_PGSQL_STRIP_OPTS ( UTF8PROC_REJECTNA | UTF8PROC_COMPAT | \
jbe@7:   UTF8PROC_COMPOSE | UTF8PROC_STABLE | UTF8PROC_IGNORE | UTF8PROC_STRIPCC | \
jbe@7:   UTF8PROC_NLF2LF | UTF8PROC_CASEFOLD | UTF8PROC_LUMP | UTF8PROC_STRIPMARK )
jbe@1: 
jbe@7: ssize_t utf8proc_pgsql_utf8map(
jbe@7:   text *input_string, text **output_string_ptr, int options
jbe@7: ) {
jbe@0:   ssize_t result;
jbe@7:   text *output_string;
jbe@7:   result = utf8proc_decompose(
jbe@7:     VARDATA(input_string), VARSIZE(input_string) - VARHDRSZ,
jbe@7:     NULL, 0, options
jbe@7:   );
jbe@7:   if (result < 0) return result;
jbe@7:   if (result > (SIZE_MAX-1-VARHDRSZ)/sizeof(int32_t))
jbe@7:     return UTF8PROC_ERROR_OVERFLOW;
jbe@7:   // reserve one extra byte for termination
jbe@7:   *output_string_ptr = palloc(result * sizeof(int32_t) + 1 + VARHDRSZ);
jbe@7:   output_string = *output_string_ptr;
jbe@7:   if (!output_string) return UTF8PROC_ERROR_NOMEM;
jbe@7:   result = utf8proc_decompose(
jbe@7:     VARDATA(input_string), VARSIZE(input_string) - VARHDRSZ,
jbe@7:     (int32_t *)VARDATA(output_string), result, options
jbe@7:   );
jbe@7:   if (result < 0) return result;
jbe@7:   result = utf8proc_reencode(
jbe@7:     (int32_t *)VARDATA(output_string), result, options
jbe@7:   );
jbe@7:   if (result >= 0) VARATT_SIZEP(output_string) = result + VARHDRSZ;
jbe@7:   return result;
jbe@7: }
jbe@7: 
jbe@7: void utf8proc_pgsql_utf8map_errchk(ssize_t result, text *output_string) {
jbe@0:   if (result < 0) {
jbe@0:     int sqlerrcode;
jbe@1:     if (output_string) pfree(output_string);
jbe@0:     switch(result) {
jbe@0:       case UTF8PROC_ERROR_NOMEM:
jbe@0:       sqlerrcode = ERRCODE_OUT_OF_MEMORY; break;
jbe@0:       case UTF8PROC_ERROR_OVERFLOW:
jbe@0:       sqlerrcode = ERRCODE_PROGRAM_LIMIT_EXCEEDED; break;
jbe@0:       case UTF8PROC_ERROR_INVALIDUTF8:
jbe@0:       case UTF8PROC_ERROR_NOTASSIGNED:
jbe@7:       return;
jbe@0:       default:
jbe@0:       sqlerrcode = ERRCODE_INTERNAL_ERROR;
jbe@0:     }
jbe@0:     ereport(ERROR, (
jbe@0:       errcode(sqlerrcode),
jbe@0:       errmsg("%s", utf8proc_errmsg(result))
jbe@0:     ));
jbe@0:   }
jbe@0: }
jbe@0: 
jbe@7: PG_FUNCTION_INFO_V1(utf8proc_pgsql_unifold);
jbe@7: Datum utf8proc_pgsql_unifold(PG_FUNCTION_ARGS) {
jbe@7:   text *input_string;
jbe@7:   text *output_string = NULL;
jbe@7:   ssize_t result;
jbe@7:   input_string = PG_GETARG_TEXT_P(0);
jbe@7:   result = utf8proc_pgsql_utf8map(
jbe@7:     input_string, &output_string, UTF8PROC_PGSQL_FOLD_OPTS
jbe@7:   );
jbe@7:   PG_FREE_IF_COPY(input_string, 0);
jbe@7:   utf8proc_pgsql_utf8map_errchk(result, output_string);
jbe@7:   if (result >= 0) {
jbe@7:     PG_RETURN_TEXT_P(output_string);
jbe@7:   } else {
jbe@7:     PG_RETURN_NULL();
jbe@7:   }
jbe@7: }
jbe@0: 
jbe@7: PG_FUNCTION_INFO_V1(utf8proc_pgsql_unistrip);
jbe@7: Datum utf8proc_pgsql_unistrip(PG_FUNCTION_ARGS) {
jbe@7:   text *input_string;
jbe@7:   text *output_string = NULL;
jbe@7:   ssize_t result;
jbe@7:   input_string = PG_GETARG_TEXT_P(0);
jbe@7:   result = utf8proc_pgsql_utf8map(
jbe@7:     input_string, &output_string, UTF8PROC_PGSQL_STRIP_OPTS
jbe@7:   );
jbe@7:   PG_FREE_IF_COPY(input_string, 0);
jbe@7:   utf8proc_pgsql_utf8map_errchk(result, output_string);
jbe@7:   if (result >= 0) {
jbe@7:     PG_RETURN_TEXT_P(output_string);
jbe@7:   } else {
jbe@7:     PG_RETURN_NULL();
jbe@7:   }
jbe@7: }
jbe@7: