Package PGSQL extension with Spanish VAT check

2023-01-21 19:31:50 +01:00 · 2023-01-21 19:31:50 +01:00 · a2069e102c
commit a2069e102c
9 changed files with 497 additions and 0 deletions
--- a/.gitignore
+++ b/.gitignore
@ -0,0 +1,4 @@
+*.o
+*.so
+regression.*
+results/
--- a/12
+++ b/12
@ -0,0 +1,12 @@
+MODULE_big = vat
+OBJS = vatin.o es.o
+
+EXTENSION = vat
+DATA = vat--0.0.sql
+PGFILEDESC = "vat - data type for VAT identification numbers"
+
+REGRESS = vat
+
+PG_CONFIG := pg_config
+PGXS := $(shell $(PG_CONFIG) --pgxs)
+include $(PGXS)
--- a/es.c
+++ b/es.c
@ -0,0 +1,148 @@
+/*
+ * SPDX-FileCopyrightText: 2023 jordi fita mas <jfita@peritasoft.com>
+ *
+ * SPDX-License-Identifier: GPL-3.0-only
+ */
+#include "es.h"
+#include <stdbool.h>
+#include <ctype.h>
+#include <postgres.h>
+#include <utils/palloc.h>
+
+static char
+compute_dni_check_digit(const char *start, const char *end)
+{
+	unsigned long int n = 0;
+	const char *p = start;
+	for(; p != end; p++) {
+		n = n * 10 + *p - 48;
+	}
+	return "TRWAGMYFPDXBNJZSQVHLCKE"[n % 23];
+}
+
+static bool
+valid_short_dni(const char *dni) {
+	return dni[8] == compute_dni_check_digit(dni + 1, dni + 8);
+}
+
+static bool
+valid_dni(const char *dni) {
+	return dni[8] == compute_dni_check_digit(dni, dni + 8);
+}
+
+static bool
+valid_nie(char *nie) {
+	bool valid = false;
+	nie[0] -= 'X' - 48;
+	valid = nie[8] == compute_dni_check_digit(nie, nie + 8);
+	nie[0] += 'X' - 48;	
+	return valid;
+}
+
+static bool
+valid_cif(const char *cif) {
+	int odds = 0;
+	int evens = 0;
+	int n = 0;
+	int checksum;
+
+	for (n = 1; n < 8; n++) {
+		if (n % 2 == 0) {
+			evens += cif[n] - 48;
+		} else {
+			int d = cif[n] * 2;
+			int a = d / 10;
+			int b = d % 10;
+			odds += a + b;
+		}
+	}
+	checksum = 10 - (odds + evens) % 10;
+	return cif[8] == checksum + 48 || cif[8] == "JABCDEFGHI"[checksum];
+}
+
+char *
+check_es_vat(const char *str, size_t len) {
+	const char *p = str;
+	const char *end = p + len;
+	char buf[10] = { 0 };
+	int n = 0;
+
+	if (len < 9 ) {
+		return NULL;
+	}
+
+	for (; p != end && n < 10; p++) {
+		char c = *p;
+		if (c == ' ' || c == '-') continue;
+		if (n > 0 && n < 8) {
+			if (!isdigit(c)) {
+				return NULL;
+			}
+			buf[n] = c;
+		} else {
+			buf[n] = toupper(c);
+		}
+		n++;
+	}
+	if (n == 10) {
+		return NULL;
+	}
+	buf[9] = 0;
+
+	switch (buf[0]) {
+		case 'K':
+		case 'L':
+		case 'M':
+			if (!valid_short_dni(buf)) {
+				return NULL;
+			}
+			break;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			if (!valid_dni(buf)) {
+				return NULL;
+			}
+			break;
+		case 'X':
+		case 'Y':
+		case 'Z':
+			if (!valid_nie(buf)) {
+				return NULL;
+			}
+			break;
+		case 'A':
+		case 'B':
+		case 'C':
+		case 'D':
+		case 'E':
+		case 'F':
+		case 'G':
+		case 'H':
+		case 'J':
+		case 'N':
+		case 'P':
+		case 'Q':
+		case 'R':
+		case 'S':
+		case 'U':
+		case 'V':
+		case 'W':
+			if (!valid_cif(buf)) {
+				return NULL;
+			}
+			break;
+		default:
+			return NULL;
+	}
+
+	return pstrdup(buf);
+}
+
--- a/es.h
+++ b/es.h
@ -0,0 +1,21 @@
+/*
+ * SPDX-FileCopyrightText: 2023 jordi fita mas <jfita@peritasoft.com>
+ *
+ * SPDX-License-Identifier: GPL-3.0-only
+ */
+#ifndef VAT_ES_H
+#define VAT_ES_H 1
+
+#include <stddef.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+char * check_es_vat(const char *str, size_t len);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* VAT_ES_H */
--- a/expected/vat.out
+++ b/expected/vat.out
@ -0,0 +1,31 @@
+--
+-- Test VATIN extensions
+--
+CREATE EXTENSION vat;
+-- Test valid conversions
+SELECT 'ES40404040D'::VATIN,
+       'ES40404040-D'::VATIN,
+       ' esx0523821l '::VATIN,
+       'ESM0243487d'::VATIN,
+       ' ESb17616756   '::VATIN
+;
+    vatin    |    vatin    |    vatin    |    vatin    |    vatin    
+-------------+-------------+-------------+-------------+-------------
+ ES40404040D | ES40404040D | ESX0523821L | ESM0243487D | ESB17616756
+(1 row)
+
+-- Test invalid checksums
+SELECT 'ES40414040D'::VATIN;
+ERROR:  invalid input syntax for VAT number: "ES40414040D"
+LINE 1: SELECT 'ES40414040D'::VATIN;
+               ^
+SELECT 'ESM0243487F'::VATIN;
+ERROR:  invalid input syntax for VAT number: "ESM0243487F"
+LINE 1: SELECT 'ESM0243487F'::VATIN;
+               ^
+SELECT 'ESB17616757'::VATIN;
+ERROR:  invalid input syntax for VAT number: "ESB17616757"
+LINE 1: SELECT 'ESB17616757'::VATIN;
+               ^
+-- Cleanup
+DROP EXTENSION vat;
--- a/sql/vat.sql
+++ b/sql/vat.sql
@ -0,0 +1,21 @@
+--
+-- Test VATIN extensions
+--
+
+CREATE EXTENSION vat;
+
+-- Test valid conversions
+SELECT 'ES40404040D'::VATIN,
+       'ES40404040-D'::VATIN,
+       ' esx0523821l '::VATIN,
+       'ESM0243487d'::VATIN,
+       ' ESb17616756   '::VATIN
+;
+
+-- Test invalid checksums
+SELECT 'ES40414040D'::VATIN;
+SELECT 'ESM0243487F'::VATIN;
+SELECT 'ESB17616757'::VATIN;
+
+-- Cleanup
+DROP EXTENSION vat;
--- a/vat--0.0.sql
+++ b/vat--0.0.sql
@ -0,0 +1,187 @@
+/* vatin--0.0.sql */
+
+-- complain if script is sources in psql, rather than via CREATE EXTENSION
+\echo Use "CREATE EXTENSION vat" to load this file. \quit
+
+-- Example:
+--  create table test ( id vatin );
+--  insert into tast values('ES40404040D');
+--
+-- select vatin('ES40404040-D');
+--
+
+--
+-- Input and output functions and data types:
+--
+---------------------------------------------
+CREATE FUNCTION vatin_in(cstring)
+	RETURNS vatin
+	AS 'MODULE_PATHNAME'
+	LANGUAGE C
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE FUNCTION vatin_out(vatin)
+	RETURNS cstring
+	AS 'textout'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE TYPE vatin (
+	INPUT = vatin_in,
+	OUTPUT = vatin_out,
+	LIKE = pg_catalog.text
+);
+COMMENT ON TYPE vatin
+	IS 'Value added tax identification number (VATIN)'
+;
+
+--
+-- Operator functions:
+--
+---------------------------------------------
+CREATE FUNCTION vatinlt(vatin, vatin)
+	RETURNS boolean
+	AS 'text_lt'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE FUNCTION vatinle(vatin, vatin)
+	RETURNS boolean
+	AS 'text_le'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE FUNCTION vatineq(vatin, vatin)
+	RETURNS boolean
+	AS 'texteq'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE;
+CREATE FUNCTION vatinge(vatin, vatin)
+	RETURNS boolean
+	AS 'text_ge'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE FUNCTION vatingt(vatin, vatin)
+	RETURNS boolean
+	AS 'text_gt'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+CREATE FUNCTION vatinne(vatin, vatin)
+	RETURNS boolean
+	AS 'textne'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+
+--
+-- Operators:
+--
+---------------------------------------------
+CREATE OPERATOR < (
+	PROCEDURE = vatinlt,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = >,
+	NEGATOR = >=,
+	RESTRICT = scalarltsel,
+	JOIN = scalarltjoinsel
+);
+CREATE OPERATOR <= (
+	PROCEDURE = vatinle,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = >=,
+	NEGATOR = >,
+	RESTRICT = scalarltsel,
+	JOIN = scalarltjoinsel
+);
+CREATE OPERATOR = (
+	PROCEDURE = vatineq,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = =,
+	NEGATOR = <>,
+	RESTRICT = eqsel,
+	JOIN = eqjoinsel,
+	MERGES,
+	HASHES
+);
+CREATE OPERATOR >= (
+	PROCEDURE = vatinge,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = <=,
+	NEGATOR = <,
+	RESTRICT = scalargtsel,
+	JOIN = scalargtjoinsel
+);
+CREATE OPERATOR > (
+	PROCEDURE = vatingt,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = <,
+	NEGATOR = <=,
+	RESTRICT = scalargtsel,
+	JOIN = scalargtjoinsel
+);
+CREATE OPERATOR <> (
+	PROCEDURE = vatinne,
+	LEFTARG = vatin,
+	RIGHTARG = vatin,
+	COMMUTATOR = <>,
+	NEGATOR = =,
+	RESTRICT = neqsel,
+	JOIN = neqjoinsel
+);
+
+--
+-- Operator families for the various operator classes:
+--
+---------------------------------------------
+CREATE OPERATOR FAMILY vatin_ops USING btree;
+CREATE OPERATOR FAMILY vatin_ops USING hash;
+
+--
+-- Operator classes:
+--
+---------------------------------------------
+CREATE FUNCTION btvatincmp(vatin, vatin)
+	RETURNS int4
+	AS 'bttextcmp'
+	LANGUAGE 'internal'
+	IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+
+CREATE OPERATOR CLASS vatin_ops DEFAULT
+	FOR TYPE vatin USING btree FAMILY vatin_ops AS
+	OPERATOR 1  <,
+	OPERATOR 2  <=,
+	OPERATOR 3  =,
+	OPERATOR 4  >=,
+	OPERATOR 5  >,
+	FUNCTION 1  btvatincmp(vatin, vatin)
+;
+
+CREATE FUNCTION hashvatin(vatin)
+	RETURNS int4
+	AS 'hashtext'
+	LANGUAGE 'internal' IMMUTABLE STRICT
+	PARALLEL SAFE
+;
+
+CREATE OPERATOR CLASS vatin_ops DEFAULT
+	FOR TYPE vatin USING hash FAMILY vatin_ops AS
+	OPERATOR 1  =,
+	FUNCTION 1  hashvatin(vatin)
+;
--- a/vat.control
+++ b/vat.control
@ -0,0 +1,5 @@
+comment = 'data type for VAT registration numbers'
+default_version = '0.0'
+module_pathname = '$libdir/vat'
+relocatable = true
+superuser = false
--- a/vatin.c
+++ b/vatin.c
@ -0,0 +1,68 @@
+/*
+ * SPDX-FileCopyrightText: 2023 jordi fita mas <jfita@peritasoft.com>
+ *
+ * SPDX-License-Identifier: GPL-3.0-only
+ */
+
+#include <postgres.h>
+#include <fmgr.h>
+#include <utils/builtins.h>
+
+#include "es.h"
+
+PG_MODULE_MAGIC;
+
+static bool
+string2vatin(const char *str, text **result) {
+	const char *r = str;
+	char *vatin = NULL;
+	size_t len;
+	char country[2];
+	int i;
+
+	for (; *r == ' '; r++);
+	if (*r == 0) {
+		goto invalidvatin;
+	}
+	len = strlen(r);
+	for (; *(r + len - 1) == ' '; len--);
+	if (len < 3) {
+		goto invalidvatin;
+	}
+
+	for(i = 0; i < 2; i++, r++, len--) {
+		country[i] = toupper(*r);
+	}
+	if (country[0] == 'E' && country[1] == 'S') {
+		vatin = check_es_vat(r, len);
+		if (vatin == NULL) {
+			goto invalidvatin;
+		}
+	} else {
+		// For countries without validation yet assume it is correct
+		vatin = pstrdup(r);
+	}
+	*result = (text *)palloc(2 + len + VARHDRSZ);
+	SET_VARSIZE(*result, 2 + len + VARHDRSZ);
+	memcpy(VARDATA(*result), country, 2);
+	memcpy(VARDATA(*result) + 2, vatin, len);
+	pfree(vatin);
+
+	return true;
+
+invalidvatin:
+	ereport(ERROR, (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION), errmsg("invalid input syntax for VAT number: \"%s\"", str)));
+	return false;
+}
+
+PG_FUNCTION_INFO_V1(vatin_in);
+	Datum
+vatin_in(PG_FUNCTION_ARGS)
+{
+	const char *str = PG_GETARG_CSTRING(0);
+	text *result = NULL;
+	if (!string2vatin(str, &result)) {
+		PG_RETURN_NULL();
+	}
+	PG_RETURN_TEXT_P(result);
+}