firebird-mirror/src/isql/extract.epp

/*
 *	PROGRAM:	Interactive SQL utility
 *	MODULE:		extract.epp
 *	DESCRIPTION:	Definition extract routines
 *
 * The contents of this file are subject to the Interbase Public
 * License Version 1.0 (the "License"); you may not use this file
 * except in compliance with the License. You may obtain a copy
 * of the License at http://www.Inprise.com/IPL.html
 *
 * Software distributed under the License is distributed on an
 * "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, either express
 * or implied. See the License for the specific language governing
 * rights and limitations under the License.
 *
 * The Original Code was created by Inprise Corporation
 * and its predecessors. Portions created by Inprise Corporation are
 * Copyright (C) Inprise Corporation.
 *
 * All Rights Reserved.
 * Contributor(s): ______________________________________.
 * $Id: extract.epp,v 1.24 2004-01-21 07:16:17 skidder Exp $
 * Revision 1.3  2000/11/22 17:07:25  patrickgriffin
 * In get_procedure_args change comment style from // to c style
 *
 * ...pat
 *
 * 2001.09.09 Claudio Valderrama: procedure's parameter names may need
 *   double quotes if they are in dialect 3 and have special characters.
 * 2001.09.21 Claudio Valderrama: Show correct mechanism for UDF parameters
 *   and support the RETURNS PARAMETER <n> syntax.
 * 2001.10.01 Claudio Valderrama: list_all_grants2() and EXTRACT_list_grants()
 *   to better organize the code that should be called to handle SHOW GRANTS.
 *
 * Revision 1.2  2000/11/18 16:49:24  fsg
 * Increased PRINT_BUFFER_LENGTH to 2048 to show larger plans
 * Fixed Bug #122563 in extract.e get_procedure_args
 * Apparently this has to be done in show.e also,
 * but that is for another day :-)
 *
 * 2003.02.04 Dmitry Yemanov: support for universal triggers
 */

#include "firebird.h"
#include "../jrd/ib_stdio.h"
#include <string.h>
#include <math.h>
#include <ctype.h>				// isdigit 
#include "../jrd/common.h"
#include "../jrd/y_ref.h"
#include "../jrd/ibase.h"
#include "../dsql/dsql.h"
#include "../isql/isql.h"
#include "../jrd/constants.h"
#include "../isql/extra_proto.h"
#include "../isql/isql_proto.h"
#include "../isql/show_proto.h"
#include "../jrd/gds_proto.h"
#include "../jrd/obj.h"
#include "../jrd/ods.h"
#include "../common/utils_proto.h"

DATABASE DB = EXTERN COMPILETIME "yachts.lnk";

extern USHORT major_ods;
extern USHORT minor_ods;
extern bool V33;

static void list_all_grants();
static processing_state list_all_grants2(bool, const SCHAR*);
static void list_all_procs();
static void list_all_tables(SSHORT, SSHORT);
static void list_all_triggers();
static void list_check();
static void list_create_db();
static void list_domain_table(const SCHAR*, SSHORT);
static void list_domains(SSHORT);
static void list_exception();
static void list_filters();
static void list_foreign();
static void list_functions();
static void list_generators();
static void list_index();
static void list_views();

static void get_procedure_args(char*);

#define MAX_INTSUBTYPES	2
#define MAXSUBTYPES	8
#define MAX_UDFPARAM_TYPES 4

extern IB_FILE* Out;

// Command terminator and procedure terminator 

extern const sqltypes Column_types[];
extern const SCHAR* Integral_subtypes[];
extern const SCHAR* Sub_types[];
extern const SCHAR* UDF_param_types[];

extern const char* trigger_action(int);

static SCHAR* Procterm = "^";	// TXNN: script use only 

/* Maybe 512 would be really enough as Print_buffer size, but
   as we have PRINT_BUFFER_LENGTH in isql.h, we should use it
   FSG 17.Nov.2000
*/

static TEXT Print_buffer[PRINT_BUFFER_LENGTH];
static TEXT SQL_identifier[BUFFER_LENGTH128];
static TEXT SQL_identifier2[BUFFER_LENGTH128];



int EXTRACT_ddl(int flag,
				   SCHAR* tabname)
{
/**************************************
 *
 *	E X T R A C T _ d d l
 *
 **************************************
 *
 * Functional description
 *	Extract all sql information
 *	0 flag means SQL only tables. 1 flag means all tables
 *
 **************************************/
	SCHAR errbuf[MSG_LENGTH];
	int ret_code = FINI_OK;
	bool did_attach = false;
	bool did_start = false;

	if (!DB)
	{
		if (isc_attach_database(gds_status, 0, global_Db_name, &DB, 0, NULL))
		{
			ISQL_errmsg(gds_status);
			return FINI_ERROR;
		}
		did_attach = true;
	}

	ISQL_get_version(false);
	if (SQL_dialect != db_SQL_dialect)
	{
		sprintf(Print_buffer,
				"/*=========================================================*/%s",
				NEWLINE);
		ISQL_printf(Out, Print_buffer);

		sprintf(Print_buffer,
				"/*=                                                      ==*/%s",
				NEWLINE);
		ISQL_printf(Out, Print_buffer);

		sprintf(Print_buffer,
				"/*=     Command Line -sqldialect %d is overwritten by    ==*/%s",
				SQL_dialect, NEWLINE);
		ISQL_printf(Out, Print_buffer);

		sprintf(Print_buffer,
				"/*=     Database SQL Dialect %d.                          ==*/%s",
				db_SQL_dialect, NEWLINE);
		ISQL_printf(Out, Print_buffer);

		sprintf(Print_buffer,
				"/*=                                                      ==*/%s",
				NEWLINE);
		ISQL_printf(Out, Print_buffer);

		sprintf(Print_buffer,
				"/*=========================================================*/%s",
				NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
	sprintf(Print_buffer, " %s", NEWLINE);
	ISQL_printf(Out, Print_buffer);

	sprintf(Print_buffer, "SET SQL DIALECT %d; %s", db_SQL_dialect, NEWLINE);
	ISQL_printf(Out, Print_buffer);

	sprintf(Print_buffer, " %s", NEWLINE);
	ISQL_printf(Out, Print_buffer);

	if (!gds_trans)
	{
		if (isc_start_transaction(gds_status, &gds_trans, 1, &DB, 0, NULL))
		{
			ISQL_errmsg(gds_status);
			return FINI_ERROR;
		}
		did_start = true;
	}

	const SSHORT default_char_set_id = ISQL_get_default_char_set_id();

// If a table name was passed, extract only that table and domains 
	if (*tabname)
	{
		if (EXTRACT_list_table(tabname, NULL, true, default_char_set_id))
		{
			gds__msg_format(NULL, ISQL_MSG_FAC, NOT_FOUND, sizeof(errbuf),
							errbuf, tabname, NULL, NULL, NULL, NULL);
			STDERROUT(errbuf, 1);
			ret_code = FINI_ERROR;
		}
	}
	else
	{
		list_create_db();
		list_filters();
		list_functions();
		list_generators();
		list_domains(default_char_set_id);
		list_all_tables(flag, default_char_set_id);
		list_index();
		list_foreign();
		list_views();
		list_check();
		list_exception();
		list_all_procs();
		list_all_triggers();
		list_all_grants();
	}

	if (gds_trans && did_start)
		if (isc_commit_transaction(gds_status, &gds_trans))
		{
			ISQL_errmsg(gds_status);
			return FINI_ERROR;
		}

	if (DB && did_attach)
	{
		if (isc_detach_database(gds_status, &DB))
		{
			ISQL_errmsg(gds_status);
			return FINI_ERROR;
		}
		DB = NULL;
	}

	return ret_code;
}


processing_state EXTRACT_list_grants(const SCHAR* terminator)
{
/**************************************
 *
 *	E X T R A C T _ l i s t _ g r a n t s
 *
 **************************************
 *
 * Functional description
 *	Shows columns, types, info for a given table name
 *	and text of views.
 *	Use a GDML query to get the info and print it.
 *	If a new_name is passed, substitute it for relation_name
 *
 **************************************/
	return list_all_grants2(false, terminator);
}


int EXTRACT_list_table(const SCHAR* relation_name,
						  const SCHAR* new_name,
						  bool domain_flag,
						  SSHORT default_char_set_id)
{
/**************************************
 *
 *	E X T R A C T _ l i s t _ t a b l e
 *
 **************************************
 *
 * Functional description
 *	Shows columns, types, info for a given table name
 *	and text of views.
 *	Use a GDML query to get the info and print it.
 *	If a new_name is passed, substitute it for relation_name
 *
 *	relation_name -- Name of table to investigate
 *	new_name -- Name of a new name for a replacement table
 *	domain_flag -- extract needed domains before the table
 *	default_char_set_id -- character set def to supress
 *
 **************************************/
/**************************************
 *	default_char_set_id warrents special
 *	consideration.  If the metadata for a
 *	table is being extracted when there is
 *	really no need to redundantly and repeatedly
 *	list the databases default character set
 *	for every field.
 *
 *	At the same time there is a need to list
 *	the character set NONE when it is not
 *	the default character set for the database.
 *
 *	EXCEPT!  If the metadata is being extracted
 *	with the intention of coping that tables structure
 *	into another database, and it is not possible
 *	to know the default character set for the
 *	target database, then list every fields
 *	character set.  This includes the character
 *	set NONE.
 *
 *	Fields with no character set definition will
 *	not have any character set listed.
 *
 *	Use -1 as the default_char_set_id
 *	in this case.
 *
 *	POTENTIAL TRAP!  Consider the following:
 *	When copying a table from one database
 *	to another how should fields using the
 *	default character set be handled?
 *
 *	If both databases have the same default
 *	character set, then there is no problem
 *	or confusion.
 *
 *	If the databases have different default
 *	character sets then should fields using
 *	the default is the source database use
 *	the default of the target database?
 *
 **************************************/

	bool first = true;
	SSHORT collation;
	SSHORT char_set_id;
	SSHORT i;
	SCHAR char_sets[86];
	SSHORT subtype;
	bool intchar = false;
	// CVC: shouldn't this var be initialized inside the FOR statement instead?

// Query to obtain relation detail information 

	FOR REL IN RDB$RELATIONS CROSS
		RFR IN RDB$RELATION_FIELDS CROSS
		FLD IN RDB$FIELDS WITH
		RFR.RDB$FIELD_SOURCE EQ FLD.RDB$FIELD_NAME AND
		RFR.RDB$RELATION_NAME EQ REL.RDB$RELATION_NAME AND
		REL.RDB$RELATION_NAME EQ relation_name
		SORTED BY RFR.RDB$FIELD_POSITION, RFR.RDB$FIELD_NAME

		if (first)
		{
			first = false;
			// Do we need to print domains 
			if (domain_flag)
				list_domain_table (relation_name, default_char_set_id);

			fb_utils::fb_exact_name(REL.RDB$OWNER_NAME);
			sprintf (Print_buffer, "%s/* Table: %s, Owner: %s */%s",
					 NEWLINE,
					 relation_name,
					 REL.RDB$OWNER_NAME,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{
				if (new_name)
					ISQL_copy_SQL_id (new_name, SQL_identifier, DBL_QUOTE);
				else
					ISQL_copy_SQL_id (relation_name, SQL_identifier, DBL_QUOTE);
				sprintf (Print_buffer, "CREATE TABLE %s ", SQL_identifier);
			}
			else
				sprintf (Print_buffer, "CREATE TABLE %s ",
						 new_name ? new_name : relation_name);
			ISQL_printf (Out, Print_buffer);
			if (!REL.RDB$EXTERNAL_FILE.NULL)
			{
				ISQL_copy_SQL_id (REL.RDB$EXTERNAL_FILE, SQL_identifier2,
								  SINGLE_QUOTE);
				sprintf (Print_buffer, "EXTERNAL FILE %s ", SQL_identifier2);
				ISQL_printf (Out, Print_buffer);
			}
			ISQL_printf (Out, "(");
		}
		else
		{
			sprintf (Print_buffer, ",%s%s", NEWLINE, TAB_AS_SPACES);
			ISQL_printf (Out, Print_buffer);
		}

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (fb_utils::fb_exact_name(RFR.RDB$FIELD_NAME),
							  SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, "%s ", SQL_identifier);
		}
		else
			sprintf (Print_buffer, "%s ", fb_utils::fb_exact_name(RFR.RDB$FIELD_NAME));
		ISQL_printf (Out, Print_buffer);

	    /*
	    ** Check first for computed fields, then domains.
	    ** If this is a known domain, then just print the domain rather than type
	    ** Domains won't have length, array, or blob definitions, but they
	    ** may have not null, default and check overriding their definitions
	    */

		if (!FLD.RDB$COMPUTED_BLR.NULL)
		{
			ISQL_printf (Out, "COMPUTED BY ");
			if (!FLD.RDB$COMPUTED_SOURCE.NULL)
				ISQL_print_validation (Out, &FLD.RDB$COMPUTED_SOURCE, true, gds_trans);
		}
		else if (!((strncmp(FLD.RDB$FIELD_NAME, "RDB$", 4) == 0) &&
				isdigit (FLD.RDB$FIELD_NAME[4]) &&
				FLD.RDB$SYSTEM_FLAG != 1))
		{
			fb_utils::fb_exact_name(FLD.RDB$FIELD_NAME);
			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{
				ISQL_copy_SQL_id (FLD.RDB$FIELD_NAME, SQL_identifier, DBL_QUOTE);
				sprintf (Print_buffer, "%s", SQL_identifier);
			}
			else
				sprintf (Print_buffer, "%s", FLD.RDB$FIELD_NAME);
			ISQL_printf (Out, Print_buffer);
			// International character sets 
			if ((FLD.RDB$FIELD_TYPE == FCHAR || FLD.RDB$FIELD_TYPE == VARCHAR)
				&& !RFR.RDB$COLLATION_ID.NULL && RFR.RDB$COLLATION_ID)
			{
				char_sets[0] = '\0';
				collation = RFR.RDB$COLLATION_ID;
				ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID, collation,
										 true, char_sets);
				if (char_sets[0])
					ISQL_printf (Out, char_sets);
			}
		}
		else
		{
			// Look through types array 
			for (i = 0; Column_types[i].type; i++)
				if (FLD.RDB$FIELD_TYPE == Column_types[i].type)
				{
					bool precision_known = false;

					if (major_ods >= ODS_VERSION10)
					{
						// Handle Integral subtypes NUMERIC and DECIMAL 
						if ((FLD.RDB$FIELD_TYPE == SMALLINT) ||
							(FLD.RDB$FIELD_TYPE == INTEGER)  ||
							(FLD.RDB$FIELD_TYPE == blr_int64))
						{
							FOR FLD1 IN RDB$FIELDS WITH
								FLD1.RDB$FIELD_NAME EQ FLD.RDB$FIELD_NAME

								/* We are ODS >= 10 and could be any Dialect */
								if (!FLD1.RDB$FIELD_PRECISION.NULL)
								{
									/* We are Dialect >=3 since FIELD_PRECISION is non-NULL */
									if (FLD1.RDB$FIELD_SUB_TYPE > 0 &&
										FLD1.RDB$FIELD_SUB_TYPE <= MAX_INTSUBTYPES)
									{
										sprintf (Print_buffer, "%s(%d, %d)",
												 Integral_subtypes[FLD1.RDB$FIELD_SUB_TYPE],
												 FLD1.RDB$FIELD_PRECISION,
												 -FLD1.RDB$FIELD_SCALE);
										precision_known = true;
									}
								}
							END_FOR
							ON_ERROR
								ISQL_errmsg (isc_status);
								return ERR;
							END_ERROR;
						}
					}

					if (!precision_known)
					{

						// Take a stab at numerics and decimals 
						if ((FLD.RDB$FIELD_TYPE == SMALLINT) &&
							(FLD.RDB$FIELD_SCALE < 0))
						{
							sprintf (Print_buffer, "NUMERIC(4, %d)",
									 -FLD.RDB$FIELD_SCALE);
						}
						else if ((FLD.RDB$FIELD_TYPE == INTEGER) &&
								 (FLD.RDB$FIELD_SCALE < 0))
						{
							sprintf (Print_buffer, "NUMERIC(9, %d)",
									 -FLD.RDB$FIELD_SCALE);
						}
						else if ((FLD.RDB$FIELD_TYPE == DOUBLE_PRECISION) &&
								 (FLD.RDB$FIELD_SCALE < 0))
						{
							sprintf (Print_buffer, "NUMERIC(15, %d)",
									 -FLD.RDB$FIELD_SCALE);
						}
						else
							sprintf (Print_buffer, "%s",
									 Column_types[i].type_name);
					}
					ISQL_printf (Out, Print_buffer);
					break;
				}


			if ((FLD.RDB$FIELD_TYPE == FCHAR) || (FLD.RDB$FIELD_TYPE == VARCHAR))
				if (FLD.RDB$CHARACTER_LENGTH.NULL)
				{

					sprintf (Print_buffer, "(%d)", FLD.RDB$FIELD_LENGTH);
					ISQL_printf (Out, Print_buffer);
				}
				else
				{
					sprintf (Print_buffer, "(%d)", FLD.RDB$CHARACTER_LENGTH);
					ISQL_printf (Out, Print_buffer);
				}

			// Catch arrays after printing the type  

			if (!FLD.RDB$DIMENSIONS.NULL)
				ISQL_array_dimensions (FLD.RDB$FIELD_NAME);

			if (FLD.RDB$FIELD_TYPE == BLOB)
			{
				subtype = FLD.RDB$FIELD_SUB_TYPE;
				ISQL_printf (Out, " SUB_TYPE ");
				if ((subtype > 0) && (subtype <= MAXSUBTYPES))
					ISQL_printf (Out, const_cast<char*>(Sub_types[subtype]));
				else
				{
					sprintf (Print_buffer, "%d", subtype);
					ISQL_printf (Out, Print_buffer);
				}
				sprintf (Print_buffer, " SEGMENT SIZE %u",
						 (USHORT) FLD.RDB$SEGMENT_LENGTH);
				ISQL_printf (Out, Print_buffer);
			}

			// International character sets 
			if ((FLD.RDB$FIELD_TYPE == FCHAR || FLD.RDB$FIELD_TYPE == VARCHAR ||
				 FLD.RDB$FIELD_TYPE == BLOB) &&
				!FLD.RDB$CHARACTER_SET_ID.NULL)
			{

				char_sets[0] = '\0';

				/* Override rdb$fields id with relation_fields if present */

				collation = 0;
				if (!RFR.RDB$COLLATION_ID.NULL)
					collation = RFR.RDB$COLLATION_ID;
				else if (!FLD.RDB$COLLATION_ID.NULL)
					collation = FLD.RDB$COLLATION_ID;

				char_set_id = 0;
				if (!FLD.RDB$CHARACTER_SET_ID.NULL)
					char_set_id = FLD.RDB$CHARACTER_SET_ID;
				if ((char_set_id != default_char_set_id) || collation)
					ISQL_get_character_sets (char_set_id, 0, false, char_sets);
				if (char_sets[0])
					ISQL_printf (Out, char_sets);
				if (!char_set_id)
					intchar = true;
			}
		}


		// Handle defaults for columns 

		if (!RFR.RDB$DEFAULT_SOURCE.NULL)
		{

			ISQL_printf (Out, " ");
			SHOW_print_metadata_text_blob (Out, &RFR.RDB$DEFAULT_SOURCE);
		}


	    /* The null flag is either 1 or null (for nullable) .  if there is
	    **  a constraint name, print that too.  Domains cannot have named
	    **  constraints.  The column name is in rdb$trigger_name in
	    **  rdb$check_constraints.  We hope we get at most one row back.
	    */

	    if (RFR.RDB$NULL_FLAG == 1)
		{
			FOR RCO IN RDB$RELATION_CONSTRAINTS CROSS
				CON IN RDB$CHECK_CONSTRAINTS WITH
				CON.RDB$TRIGGER_NAME = RFR.RDB$FIELD_NAME AND
				CON.RDB$CONSTRAINT_NAME = RCO.RDB$CONSTRAINT_NAME AND
				RCO.RDB$CONSTRAINT_TYPE EQ "NOT NULL" AND
				RCO.RDB$RELATION_NAME = RFR.RDB$RELATION_NAME

				if (strncmp(CON.RDB$CONSTRAINT_NAME, "INTEG", 5))
				{
					fb_utils::fb_exact_name(CON.RDB$CONSTRAINT_NAME);
					if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
					{

						ISQL_copy_SQL_id (CON.RDB$CONSTRAINT_NAME, SQL_identifier,
										  DBL_QUOTE);
						sprintf (Print_buffer, " CONSTRAINT %s", SQL_identifier);
					}
					else
					sprintf (Print_buffer, " CONSTRAINT %s",
							 CON.RDB$CONSTRAINT_NAME);
					ISQL_printf (Out, Print_buffer);
				}
			END_FOR
			ON_ERROR
				ISQL_errmsg (gds_status);
				return (SSHORT)FINI_ERROR;
			END_ERROR;


			ISQL_printf (Out, " NOT NULL");
		}

		if ((FLD.RDB$FIELD_TYPE == FCHAR || FLD.RDB$FIELD_TYPE == VARCHAR ||
			 FLD.RDB$FIELD_TYPE == BLOB) &&
			!FLD.RDB$CHARACTER_SET_ID.NULL && FLD.RDB$CHARACTER_SET_ID && intchar)
		{
			collation = 0;
			if (!RFR.RDB$COLLATION_ID.NULL)
				collation = RFR.RDB$COLLATION_ID;
			else if (!FLD.RDB$COLLATION_ID.NULL)
				collation = FLD.RDB$COLLATION_ID;

			char_set_id = 0;
			if (!FLD.RDB$CHARACTER_SET_ID.NULL)
				char_set_id = FLD.RDB$CHARACTER_SET_ID;

			if (collation)
			{
				char_sets[0] = '\0';
				ISQL_get_character_sets (char_set_id, collation, true, char_sets);
				if (char_sets[0])
					ISQL_printf (Out, char_sets);
			}
		}

	END_FOR
	ON_ERROR
		if (!V33)
		{
			ISQL_errmsg(gds_status);
			return (SSHORT) FINI_ERROR;
		}
	END_ERROR;

// Do primary and unique keys only. references come later 

	SCHAR* collist = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH512 * 2);
	if (!collist)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		return ((SSHORT) FINI_ERROR);
	}

	FOR RELC IN RDB$RELATION_CONSTRAINTS WITH
		(RELC.RDB$CONSTRAINT_TYPE EQ "PRIMARY KEY" OR
		RELC.RDB$CONSTRAINT_TYPE EQ "UNIQUE") AND
		RELC.RDB$RELATION_NAME EQ relation_name
		SORTED BY RELC.RDB$CONSTRAINT_NAME

		sprintf (Print_buffer, ",%s", NEWLINE);
		ISQL_printf (Out, Print_buffer);
		// If the name of the constraint is not INTEG..., print it 
		if (strncmp(RELC.RDB$CONSTRAINT_NAME, "INTEG", 5))
		{
			fb_utils::fb_exact_name(RELC.RDB$CONSTRAINT_NAME);
			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{
				ISQL_copy_SQL_id (RELC.RDB$CONSTRAINT_NAME, SQL_identifier,
								  DBL_QUOTE);
				sprintf (Print_buffer, "CONSTRAINT %s ", SQL_identifier);
			}
			else
				sprintf (Print_buffer, "CONSTRAINT %s ", RELC.RDB$CONSTRAINT_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		if (!strncmp (RELC.RDB$CONSTRAINT_TYPE, "PRIMARY", 7))
		{
			ISQL_get_index_segments (collist, RELC.RDB$INDEX_NAME, true);
			sprintf (Print_buffer, "PRIMARY KEY (%s)", collist);
			ISQL_printf (Out, Print_buffer);
		}

		else if (!strncmp (RELC.RDB$CONSTRAINT_TYPE, "UNIQUE", 6))
		{
			ISQL_get_index_segments (collist, RELC.RDB$INDEX_NAME, true);
			sprintf (Print_buffer, "UNIQUE (%s)", collist);
			ISQL_printf (Out, Print_buffer);
		}

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		if (collist)
			ISQL_FREE(collist);
		return (SSHORT) FINI_ERROR;
	END_ERROR;

// Check constaints are now deferred

	sprintf(Print_buffer, ")%s%s", global_Term, NEWLINE);
	ISQL_printf(Out, Print_buffer);

	if (collist)
		ISQL_FREE(collist);

	return (SSHORT) FINI_OK;
}


static void get_procedure_args(char* proc_name)
{
/**************************************
 *
 *	g e t _ p r o c e d u r e _ a r g s
 *
 **************************************
 *
 * Functional description
 *	This function extract the procedure parameters and adds it to the
 *	extract file
 *
 **************************************/
	SSHORT i;
	SCHAR char_sets[86];

// query to retrieve the parameters. 
	{


/* placed the two identical code blocks into one
   for loop as suggested by Ann H. and Claudio C.
   FSG 18.Nov.2000
*/

		/*  Parameter types 0 = input  */
		/*  Parameter types 1 = return */

		for (SSHORT ptype = 0; ptype < 2; ptype++)
		{
			bool first_time = true;

			FOR PRM IN RDB$PROCEDURE_PARAMETERS CROSS
				FLD IN RDB$FIELDS WITH
				PRM.RDB$PROCEDURE_NAME = proc_name AND
				PRM.RDB$FIELD_SOURCE EQ FLD.RDB$FIELD_NAME AND
				PRM.RDB$PARAMETER_TYPE = ptype
				SORTED BY PRM.RDB$PARAMETER_NUMBER

				if (first_time)
				{
					first_time = false;
					if (ptype == 0)
					{ // this is the input part 
						ISQL_printf (Out, "(");
					}
					else
					{ // we are in the output part 
						ISQL_printf (Out, "RETURNS (");
					}
				}
				else
				{
					sprintf (Print_buffer, ",%s", NEWLINE);
					ISQL_printf (Out, Print_buffer);
				}

				fb_utils::fb_exact_name(PRM.RDB$PARAMETER_NAME);

				// CVC: Parameter names need check for dialect 3, too. 

				if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
					ISQL_copy_SQL_id (PRM.RDB$PARAMETER_NAME, SQL_identifier, DBL_QUOTE);
				else
					strcpy (SQL_identifier, PRM.RDB$PARAMETER_NAME);

				sprintf (Print_buffer, "%s ", SQL_identifier);
				ISQL_printf (Out, Print_buffer);

				// Get column type name to print 
				for (i = 0; Column_types[i].type; i++)
					if (FLD.RDB$FIELD_TYPE == Column_types[i].type)
					{
						bool precision_known = false;

						if (major_ods >= ODS_VERSION10)
						{
							// Handle Integral subtypes NUMERIC and DECIMAL 
							if ((FLD.RDB$FIELD_TYPE == SMALLINT) ||
								(FLD.RDB$FIELD_TYPE == INTEGER) ||
								(FLD.RDB$FIELD_TYPE == blr_int64))
							{
								/* We are ODS >= 10 and could be any Dialect */

								FOR FLD1 IN RDB$FIELDS WITH
									FLD1.RDB$FIELD_NAME EQ FLD.RDB$FIELD_NAME

									if (!FLD1.RDB$FIELD_PRECISION.NULL)
									{
										/* We are Dialect >=3 since FIELD_PRECISION is non-NULL */
										if (FLD1.RDB$FIELD_SUB_TYPE > 0 &&
											FLD1.RDB$FIELD_SUB_TYPE <= MAX_INTSUBTYPES)
										{
											sprintf (Print_buffer, "%s(%d, %d)",
													Integral_subtypes[FLD1.RDB$FIELD_SUB_TYPE],
													FLD1.RDB$FIELD_PRECISION,
													-FLD1.RDB$FIELD_SCALE);
											precision_known = true;
										}
									}
								END_FOR
								ON_ERROR
									ISQL_errmsg (isc_status);
									return;
								END_ERROR;
							}
						}
						if (!precision_known)
						{
							// Take a stab at numerics and decimals 
							if ((FLD.RDB$FIELD_TYPE == SMALLINT) &&
								(FLD.RDB$FIELD_SCALE < 0))
							{
								sprintf (Print_buffer, "NUMERIC(4, %d)",
										-FLD.RDB$FIELD_SCALE);
							}
							else if ((FLD.RDB$FIELD_TYPE == INTEGER) &&
									 (FLD.RDB$FIELD_SCALE < 0))
							{
								sprintf (Print_buffer, "NUMERIC(9, %d)",
										-FLD.RDB$FIELD_SCALE);
							}
							else if ((FLD.RDB$FIELD_TYPE == DOUBLE_PRECISION) &&
									 (FLD.RDB$FIELD_SCALE < 0))
							{
								sprintf (Print_buffer, "NUMERIC(15, %d)",
										-FLD.RDB$FIELD_SCALE);
							}
							else
							{
								sprintf (Print_buffer, "%s",
										 Column_types[i].type_name);
							}
						}
						ISQL_printf (Out, Print_buffer);
						break;
					}

				/* Changed this to return RDB$CHARACTER_LENGTH if available
				   Fix for Bug #122563
				   FSG 18.Nov.2000
				*/
				if ((FLD.RDB$FIELD_TYPE == FCHAR) || (FLD.RDB$FIELD_TYPE == VARCHAR))
					if (FLD.RDB$CHARACTER_LENGTH.NULL)
					{

						sprintf (Print_buffer, "(%d)", FLD.RDB$FIELD_LENGTH);
						ISQL_printf (Out, Print_buffer);
					}
					else
					{
						sprintf (Print_buffer, "(%d)", FLD.RDB$CHARACTER_LENGTH);
						ISQL_printf (Out, Print_buffer);
					}



				// Show international character sets and collations 
				if (!FLD.RDB$COLLATION_ID.NULL || !FLD.RDB$CHARACTER_SET_ID.NULL)
				{
					char_sets[0] = 0;
					if (FLD.RDB$COLLATION_ID.NULL)
						FLD.RDB$COLLATION_ID = 0;

					if (FLD.RDB$CHARACTER_SET_ID.NULL)
						FLD.RDB$CHARACTER_SET_ID = 0;

					ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID,
											 FLD.RDB$COLLATION_ID, false, char_sets);
					if (char_sets[0])
						ISQL_printf (Out, char_sets);
				}
			END_FOR
			ON_ERROR
				if (!V33)
				{
					ISQL_errmsg(gds_status);
					return;
				}
			END_ERROR;

			// If there was at least one param, close parens 

			if (!first_time)
			{
				sprintf(Print_buffer, ")%s", NEWLINE);
				ISQL_printf(Out, Print_buffer);
			}

		}						// end for ptype 
		sprintf(Print_buffer, "AS %s", NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
}


static void list_all_grants()
{
/**************************************
 *
 *	l i s t _ a l l _ g r a n t s
 *
 **************************************
 *
 * Functional description
 *	Print the permissions on all user tables.
 *
 *	Wrapper around list_all_grants2().
 *
 **************************************/
	list_all_grants2(true, global_Term);
}


static processing_state list_all_grants2(bool show_role_list,
							const SCHAR* terminator)
{
/**************************************
 *
 *	l i s t _ a l l _ g r a n t s
 *
 **************************************
 *
 * Functional description
 *	Print the permissions on all user tables.
 *
 *	Get separate permissions on table/views and then procedures
 *
 **************************************/
	bool first = true;
	TEXT prev_owner[44];

/******************************************************************
**
**	process GRANT roles
**
*******************************************************************/

	if (major_ods >= ODS_VERSION9 && show_role_list)
	{
		prev_owner[0] = '\0';

		FOR XX IN RDB$ROLES
			SORTED BY XX.RDB$ROLE_NAME

			if (first)
			{
				sprintf (Print_buffer, "%s/* Grant role for this database */%s",
						 NEWLINE,
						 NEWLINE);
				ISQL_printf (Out, Print_buffer);
			}

			first = false;

			// Null terminate name string 
			fb_utils::fb_exact_name(XX.RDB$ROLE_NAME);
			fb_utils::fb_exact_name(XX.RDB$OWNER_NAME);

			if (strcmp (prev_owner, XX.RDB$OWNER_NAME) != 0)
			{
				sprintf (Print_buffer, "%s/* Role: %s, Owner: %s */%s",
						 NEWLINE,
						 XX.RDB$ROLE_NAME,
						 XX.RDB$OWNER_NAME,
						 NEWLINE);
				ISQL_printf (Out, Print_buffer);
				strcpy (prev_owner, XX.RDB$OWNER_NAME);
			}

			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{

				ISQL_copy_SQL_id (XX.RDB$ROLE_NAME, SQL_identifier, DBL_QUOTE);
				sprintf (Print_buffer, "CREATE ROLE %s;%s",
						 SQL_identifier, NEWLINE);
			}
			else
				sprintf (Print_buffer, "CREATE ROLE %s;%s",
						 XX.RDB$ROLE_NAME, NEWLINE);
			ISQL_printf (Out, Print_buffer);

		END_FOR
		ON_ERROR
			ISQL_errmsg(gds_status);
		    return OBJECT_NOT_FOUND;
		END_ERROR;
	}

/* This version of cursor gets only sql tables identified by security class
   and misses views, getting only null view_source */

	first = true;

	FOR REL IN RDB$RELATIONS WITH
		(REL.RDB$SYSTEM_FLAG NE 1 OR REL.RDB$SYSTEM_FLAG MISSING) AND
		REL.RDB$SECURITY_CLASS STARTING "SQL$"
		SORTED BY REL.RDB$RELATION_NAME

		if (first) {
			sprintf (Print_buffer, "%s/* Grant permissions for this database */%s",
					 NEWLINE,
					 NEWLINE);
			if (show_role_list) {
				ISQL_printf (Out, Print_buffer);
				first = false;
			}
		}

		// Null terminate name string 

		fb_utils::fb_exact_name(REL.RDB$RELATION_NAME);

		const processing_state rc =
			SHOW_grants2 (REL.RDB$RELATION_NAME, terminator, obj_relation,
						   first ? Print_buffer : 0);
		if (rc == SKIP) {
			first = false;
		}

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return OBJECT_NOT_FOUND;
	END_ERROR;

	if (first) {
		sprintf(Print_buffer, "%s/* Grant permissions for this database */%s",
				NEWLINE, NEWLINE);
		if (show_role_list) {
			ISQL_printf (Out, Print_buffer);
			first = false;
		}
	}

	if (first)
		SHOW_grant_roles2 (terminator, &first, Print_buffer);
	else
		SHOW_grant_roles (terminator, 0);


// Again for stored procedures 
	FOR PRC IN RDB$PROCEDURES
		SORTED BY PRC.RDB$PROCEDURE_NAME

		if (first) {
			sprintf (Print_buffer, "%s/* Grant permissions for this database */%s",
					 NEWLINE, NEWLINE);
			if (show_role_list) {
				ISQL_printf (Out, Print_buffer);
				first = false;
			}
		}

		// Null terminate name string 

		fb_utils::fb_exact_name(PRC.RDB$PROCEDURE_NAME);

		const processing_state rc =
			SHOW_grants2 (PRC.RDB$PROCEDURE_NAME, terminator, obj_procedure,
						   first ? Print_buffer: 0);
		if (rc == SKIP)
			first = false;

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return OBJECT_NOT_FOUND;
	END_ERROR;

	return first ? OBJECT_NOT_FOUND : SKIP;
}


static void list_all_procs()
{
/**************************************
 *
 *	l i s t _ a l l _ p r o c s
 *
 **************************************
 *
 * Functional description
 *	Shows text of a stored procedure given a name.
 *	or lists procedures if no argument.
 *	Since procedures may reference each other, we will create all
 *	dummy procedures of the correct name, then alter these to their
 *	correct form.
 *      Add the parameter names when these procedures are created.
 *
 *	procname -- Name of procedure to investigate
 *
 **************************************/
	bool header = true;
	TEXT msg[MSG_LENGTH];
	static const SCHAR* create_procedure_str1 = "CREATE PROCEDURE %s ";
	static const SCHAR* create_procedure_str2 = "BEGIN EXIT; END %s%s";


//  First the dummy procedures 
// create the procedures with their parameters 

	FOR PRC IN RDB$PROCEDURES
		SORTED BY PRC.RDB$PROCEDURE_NAME
		if (header)
		{
			sprintf (Print_buffer, "COMMIT WORK;%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
			sprintf (Print_buffer, "SET AUTODDL OFF;%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
			sprintf (Print_buffer, "SET TERM %s %s%s", Procterm, global_Term, NEWLINE);
			ISQL_printf (Out, Print_buffer);
			sprintf (Print_buffer, "%s/* Stored procedures */%s", NEWLINE, NEWLINE);
			ISQL_printf (Out, Print_buffer);
			header = false;
		}
		fb_utils::fb_exact_name(PRC.RDB$PROCEDURE_NAME);
		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (PRC.RDB$PROCEDURE_NAME, SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, create_procedure_str1,
					 SQL_identifier);
			ISQL_printf (Out, Print_buffer);
			get_procedure_args (SQL_identifier);
			sprintf (Print_buffer, create_procedure_str2, Procterm, NEWLINE);
		}
		else
		{
			sprintf (Print_buffer, create_procedure_str1,
					 PRC.RDB$PROCEDURE_NAME);
			ISQL_printf (Out, Print_buffer);
			get_procedure_args (PRC.RDB$PROCEDURE_NAME);
			sprintf (Print_buffer, create_procedure_str2, Procterm, NEWLINE);
		}

		ISQL_printf (Out, Print_buffer);
	END_FOR
	ON_ERROR
		if (!V33)
		{
			ISQL_errmsg(gds_status);
			return;
		}
	END_ERROR;

/* This query gets the procedure name and the source.  We then nest a query
   to retrieve the parameters. Alter is used, because the procedures are already there*/
	FOR PRC IN RDB$PROCEDURES
		SORTED BY PRC.RDB$PROCEDURE_NAME

		fb_utils::fb_exact_name(PRC.RDB$PROCEDURE_NAME);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (PRC.RDB$PROCEDURE_NAME, SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, "%sALTER PROCEDURE %s ", NEWLINE,
					 SQL_identifier);
		}
		else
			sprintf (Print_buffer, "%sALTER PROCEDURE %s ", NEWLINE,
					 PRC.RDB$PROCEDURE_NAME);
		ISQL_printf (Out, Print_buffer);
		get_procedure_args (PRC.RDB$PROCEDURE_NAME);

		// Print the procedure body 

		if (!PRC.RDB$PROCEDURE_SOURCE.NULL)
			SHOW_print_metadata_text_blob (Out, &PRC.RDB$PROCEDURE_SOURCE);

		sprintf (Print_buffer, " %s%s", Procterm, NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_msg_get(GEN_ERR, msg, (TEXT*) (IPTR) isc_sqlcode(gds_status), NULL,
					 NULL, NULL, NULL);
		STDERROUT(msg, 1);			/* Statement failed, SQLCODE = %d\n\n */
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;

// Only reset the terminators is there were procs to print 
	if (!header)
	{
		sprintf(Print_buffer, "SET TERM %s %s%s", global_Term, Procterm, NEWLINE);
		ISQL_printf(Out, Print_buffer);
		sprintf(Print_buffer, "COMMIT WORK %s%s", global_Term, NEWLINE);
		ISQL_printf(Out, Print_buffer);
		sprintf(Print_buffer, "SET AUTODDL ON;%s", NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
}


static void list_all_tables(SSHORT flag,
							SSHORT default_char_set_id)
{
/**************************************
 *
 *	l i s t _ a l l _ t a b l e s
 *
 **************************************
 *
 * Functional description
 *	Extract the names of all user tables from
 *	rdb$relations.  Filter SQL tables by
 *	security class after we fetch them
 *	Parameters:  flag -- 0, get all tables
 *
 **************************************/

/* This version of cursor gets only sql tables identified by security class
   and misses views, getting only null view_source */

	FOR REL IN RDB$RELATIONS WITH
		(REL.RDB$SYSTEM_FLAG NE 1 OR REL.RDB$SYSTEM_FLAG MISSING) AND
		REL.RDB$VIEW_BLR MISSING
		SORTED BY REL.RDB$RELATION_NAME

		// If this is not an SQL table and we aren't doing ALL objects 
		if ((REL.RDB$FLAGS.NULL || !(REL.RDB$FLAGS & REL_sql)) && (flag != ALL_objects) )
			continue;
		// Null terminate name string 

		fb_utils::fb_exact_name(REL.RDB$RELATION_NAME);

		if (flag || !strncmp (REL.RDB$SECURITY_CLASS, "SQL$", 4))
			EXTRACT_list_table (REL.RDB$RELATION_NAME, NULL, false,default_char_set_id);
	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		ROLLBACK;
		return;
	END_ERROR;

}


static void list_all_triggers()
{
/**************************************
 *
 *	l i s t _ a l l _ t r i g g e r s
 *
 **************************************
 *
 * Functional description
 *	Lists triggers in general on non-system
 *	tables with sql source only.
 *
 **************************************/
	bool header = true;


/* Query gets the trigger info for non-system triggers with
   source that are not part of an SQL constraint */

	FOR TRG IN RDB$TRIGGERS CROSS REL IN RDB$RELATIONS OVER RDB$RELATION_NAME
		WITH (REL.RDB$SYSTEM_FLAG NE 1 OR REL.RDB$SYSTEM_FLAG MISSING) AND
		NOT (ANY CHK IN RDB$CHECK_CONSTRAINTS WITH
		TRG.RDB$TRIGGER_NAME EQ CHK.RDB$TRIGGER_NAME)
		SORTED BY TRG.RDB$RELATION_NAME, TRG.RDB$TRIGGER_TYPE,
			TRG.RDB$TRIGGER_SEQUENCE, TRG.RDB$TRIGGER_NAME

		if (header)
		{
			sprintf (Print_buffer, "SET TERM %s %s%s", Procterm, global_Term, NEWLINE);
			ISQL_printf (Out, Print_buffer);
			sprintf (Print_buffer,
					 "%s/* Triggers only will work for SQL triggers */%s",
					 NEWLINE,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
			header = false;
		}
		fb_utils::fb_exact_name(TRG.RDB$TRIGGER_NAME);
		fb_utils::fb_exact_name(TRG.RDB$RELATION_NAME);

		if (TRG.RDB$TRIGGER_INACTIVE.NULL)
			TRG.RDB$TRIGGER_INACTIVE = 0;

		//  If trigger is not SQL put it in comments 
		if (TRG.RDB$FLAGS != TRG_sql)
			ISQL_printf (Out, "/* ");

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (TRG.RDB$TRIGGER_NAME, SQL_identifier,  DBL_QUOTE);
			ISQL_copy_SQL_id (TRG.RDB$RELATION_NAME, SQL_identifier2, DBL_QUOTE);
		}
		else
		{
			strcpy (SQL_identifier,  TRG.RDB$TRIGGER_NAME);
			strcpy (SQL_identifier2, TRG.RDB$RELATION_NAME);
		}

		sprintf (Print_buffer, "CREATE TRIGGER %s FOR %s %s%s %s POSITION %d %s",
				 SQL_identifier, SQL_identifier2, NEWLINE,
				 (TRG.RDB$TRIGGER_INACTIVE ? "INACTIVE" : "ACTIVE"),
				 trigger_action (TRG.RDB$TRIGGER_TYPE), TRG.RDB$TRIGGER_SEQUENCE,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);

		if (!TRG.RDB$TRIGGER_SOURCE.NULL)
			SHOW_print_metadata_text_blob (Out, &TRG.RDB$TRIGGER_SOURCE);

		sprintf (Print_buffer, " %s%s", Procterm, NEWLINE);
		ISQL_printf (Out, Print_buffer);
		sprintf (Print_buffer, " %s", NEWLINE);
		ISQL_printf (Out, Print_buffer);

		if (TRG.RDB$FLAGS != TRG_sql)
		{
			sprintf (Print_buffer, "*/%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;

	if (!header)
	{
		sprintf(Print_buffer, "COMMIT WORK %s%s", Procterm, NEWLINE);
		ISQL_printf(Out, Print_buffer);
		sprintf(Print_buffer, "SET TERM %s %s%s", global_Term, Procterm, NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
}


static void list_check()
{
/**************************************
 *
 *	l i s t _ c h e c k
 *
 **************************************
 *
 * Functional description
 *	List check constraints for all objects to allow forward references
 *
 **************************************/

// Query gets the check clauses for triggers stored for check constraints 

	FOR TRG IN RDB$TRIGGERS CROSS
		CHK IN RDB$CHECK_CONSTRAINTS WITH
		TRG.RDB$TRIGGER_TYPE EQ 1 AND
		TRG.RDB$TRIGGER_NAME EQ CHK.RDB$TRIGGER_NAME AND
		(ANY RELC IN RDB$RELATION_CONSTRAINTS WITH
		CHK.RDB$CONSTRAINT_NAME EQ RELC.RDB$CONSTRAINT_NAME)
		REDUCED TO CHK.RDB$CONSTRAINT_NAME
		SORTED BY CHK.RDB$CONSTRAINT_NAME


		fb_utils::fb_exact_name(TRG.RDB$RELATION_NAME);

		sprintf (Print_buffer, " %s ", NEWLINE);
		ISQL_printf (Out, Print_buffer);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (TRG.RDB$RELATION_NAME, SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, "ALTER TABLE %s ADD %s%s",
					 SQL_identifier, NEWLINE, TAB_AS_SPACES);
		}
		else
			sprintf (Print_buffer, "ALTER TABLE %s ADD %s%s",
					 TRG.RDB$RELATION_NAME, NEWLINE, TAB_AS_SPACES);
		ISQL_printf (Out, Print_buffer);

		// If the name of the constraint is not INTEG..., print it 
		if (strncmp(CHK.RDB$CONSTRAINT_NAME, "INTEG", 5))
		{
			fb_utils::fb_exact_name(CHK.RDB$CONSTRAINT_NAME);
			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{
				ISQL_copy_SQL_id (CHK.RDB$CONSTRAINT_NAME, SQL_identifier,
								  DBL_QUOTE);
				sprintf (Print_buffer, "CONSTRAINT %s ", SQL_identifier);
			}
			else
				sprintf (Print_buffer, "CONSTRAINT %s ", CHK.RDB$CONSTRAINT_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		if (!TRG.RDB$TRIGGER_SOURCE.NULL)
			SHOW_print_metadata_text_blob (Out, &TRG.RDB$TRIGGER_SOURCE);

		sprintf (Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;
}


static void print_set(bool* set_used)
{
/**************************************
 *
 *	p r i n t _ s e t
 *
 **************************************
 *
 * Functional description
 *    print (using ISQL_printf) the word "SET"
 *    if the first line of the ALTER DATABASE
 *    settings options. Also, add trailing
 *    comma for end of prior line if needed.
 *
 * uses Print_buffer, a global
 *
 **************************************/
	if (!*set_used)
	{
		sprintf(Print_buffer, "  SET ");
		ISQL_printf(Out, Print_buffer);
		*set_used = true;
	}
	else
	{
		sprintf(Print_buffer, ", %s      ", NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
}


static void list_create_db()
{
/**************************************
 *
 *	l i s t _ c r e a t e _ d b
 *
 **************************************
 *
 * Functional description
 *	Print the create database command if requested.  At least put
 *	the page size in a comment with the extracted db name
 *
 **************************************/
	static const SCHAR page_items[] = {
		isc_info_page_size,
		isc_info_end
	};

// Comment out the create database if no db param was specified
	bool nodb = false;
	if (!*global_Target_db)
	{
		ISQL_printf(Out, "/* ");
		strcpy(global_Target_db, global_Db_name);
		nodb = true;
	}
	sprintf(Print_buffer, "CREATE DATABASE '%s'", global_Target_db);
	ISQL_printf(Out, Print_buffer);

	// Get the page size from db_info call 
	SCHAR info_buf[20];
	const bool translate = false; // ???
	if (!SHOW_dbb_parameters(DB, info_buf, page_items, sizeof(page_items),
						translate))
	{
		sprintf(Print_buffer, " %s", info_buf);
		ISQL_printf(Out, Print_buffer);
	}

	FOR DBP IN RDB$DATABASE
		WITH DBP.RDB$CHARACTER_SET_NAME NOT MISSING
		AND DBP.RDB$CHARACTER_SET_NAME != " ";
		sprintf (Print_buffer, " DEFAULT CHARACTER SET %s",
		fb_utils::fb_exact_name(DBP.RDB$CHARACTER_SET_NAME));
		ISQL_printf (Out, Print_buffer);
	END_FOR
	ON_ERROR
		if (!V33)
		{
			ISQL_errmsg(gds_status);
			return;
		}
	END_ERROR;

	if (nodb)
	{
		sprintf(Print_buffer, " */%s", NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}
	else
	{
		sprintf(Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf(Out, Print_buffer);
	}

// List secondary files and shadows as alter db and create shadow in comment 

	bool first = true;
	
	FOR FIL IN RDB$FILES
		SORTED BY FIL.RDB$SHADOW_NUMBER, FIL.RDB$FILE_SEQUENCE

		if (first)
		{
			sprintf (Print_buffer, "%s/* Add secondary files in comments %s",
					 NEWLINE,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}

		first = false;
		// reset nulls to zero 

		if (FIL.RDB$FILE_FLAGS.NULL)
			FIL.RDB$FILE_FLAGS = 0;
		if (FIL.RDB$FILE_LENGTH.NULL)
			FIL.RDB$FILE_LENGTH = 0;
		if (FIL.RDB$FILE_SEQUENCE.NULL)
			FIL.RDB$FILE_SEQUENCE = 0;
		if (FIL.RDB$FILE_START.NULL)
			FIL.RDB$FILE_START = 0;
		fb_utils::fb_exact_name(FIL.RDB$FILE_NAME);

		// Pure secondary files 
		if (FIL.RDB$FILE_FLAGS == 0)
		{
			sprintf (Print_buffer, "%sALTER DATABASE ADD FILE '%s'",
					 NEWLINE,
					 FIL.RDB$FILE_NAME);
			ISQL_printf (Out, Print_buffer);
			if (FIL.RDB$FILE_START)
			{
				sprintf (Print_buffer, " STARTING %ld", FIL.RDB$FILE_START);
				ISQL_printf (Out, Print_buffer);
			}
			if (FIL.RDB$FILE_LENGTH)
			{
				sprintf (Print_buffer, " LENGTH %ld", FIL.RDB$FILE_LENGTH);
				ISQL_printf (Out, Print_buffer);
			}
			sprintf (Print_buffer, "%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}

		if (FIL.RDB$FILE_FLAGS & FILE_cache)
		{
			sprintf (Print_buffer, "%sALTER DATABASE ADD CACHE '%s' LENGTH %ld%s",
					NEWLINE,
					FIL.RDB$FILE_NAME,
					FIL.RDB$FILE_LENGTH,
					NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}

		if (FIL.RDB$FILE_FLAGS & FILE_shadow)
		{
			if (FIL.RDB$FILE_SEQUENCE)
			{
				sprintf (Print_buffer, "%sFILE '%s' ", TAB_AS_SPACES, FIL.RDB$FILE_NAME);
				ISQL_printf (Out, Print_buffer);
			}
			else
			{
				sprintf (Print_buffer, "%sCREATE SHADOW %d '%s' ",
						 NEWLINE,
						 FIL.RDB$SHADOW_NUMBER,
						 FIL.RDB$FILE_NAME);
				ISQL_printf (Out, Print_buffer);
				if (FIL.RDB$FILE_FLAGS & FILE_inactive)
					ISQL_printf (Out, "INACTIVE ");
				if (FIL.RDB$FILE_FLAGS & FILE_manual)
					ISQL_printf (Out, "MANUAL ");
				else
					ISQL_printf (Out, "AUTO ");
				if (FIL.RDB$FILE_FLAGS & FILE_conditional)
					ISQL_printf (Out, "CONDITIONAL ");
			}
			if (FIL.RDB$FILE_LENGTH)
			{
				sprintf (Print_buffer, "LENGTH %ld ", FIL.RDB$FILE_LENGTH);
				ISQL_printf (Out, Print_buffer);
			}
			if (FIL.RDB$FILE_START)
			{
				sprintf (Print_buffer, "STARTING %ld ", FIL.RDB$FILE_START);
				ISQL_printf (Out, Print_buffer);
			}
			ISQL_printf (Out, NEWLINE);
		}

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;

	bool first_file = true;
	first = true; // CVC: I'm not sure if this line should be added here.
	bool has_wal = false;

	FOR WAL IN RDB$LOG_FILES
		SORTED BY WAL.RDB$FILE_FLAGS, WAL.RDB$FILE_SEQUENCE

		has_wal = true;
		if (first)
		{
			if (nodb)
			{
				sprintf (Print_buffer, "/*%s", NEWLINE);
				ISQL_printf (Out, Print_buffer);
			}
			sprintf (Print_buffer, "%sALTER DATABASE ADD ", NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}
		first = false;

		if (first_file)
			ISQL_printf (Out, "LOGFILE ");

		if (WAL.RDB$FILE_FLAGS & LOG_default)
			;

		// Overflow files also have the serial bit set 
		else if (WAL.RDB$FILE_FLAGS & LOG_overflow)
		{
			sprintf (Print_buffer, ")%s   OVERFLOW '%s'",
					 NEWLINE,
					 WAL.RDB$FILE_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		else if (WAL.RDB$FILE_FLAGS & LOG_serial)
		{
			sprintf (Print_buffer, "%s  BASE_NAME '%s'",
					 NEWLINE,
					 WAL.RDB$FILE_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		/* Since we are fetching order by FILE_FLAGS, the LOG_0verflow will
		**  be last.  It will only appear if there were named round robin,
		**  so we must close the parens first
		*/

		// We have round robin and overflow file specifications 
		else
		{
			if (first_file)
				ISQL_printf (Out, "(");
			else
			{
				sprintf (Print_buffer, ",%s  ", NEWLINE);
				ISQL_printf (Out, Print_buffer);
			}
			first_file = false;

			sprintf (Print_buffer, "'%s'", WAL.RDB$FILE_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		// Any file can have a length 
		if (!WAL.RDB$FILE_LENGTH.NULL && WAL.RDB$FILE_LENGTH)
		{
			sprintf (Print_buffer, " SIZE %ld ", WAL.RDB$FILE_LENGTH);
			ISQL_printf (Out, Print_buffer);
		}

	END_FOR
	ON_ERROR
		if (!V33)
		{
			ISQL_errmsg(gds_status);
			return;
		}
	END_ERROR;

	ISQL_printf(Out, NEWLINE);
/*************************************************************
** isc_info_num_wal_buffers         for NUM_LOG_BUFFERS
** isc_info_wal_buffer_size         for LOG_BUFFER_SIZE
** isc_info_wal_grpc_wait_usecs     for GROUP_COMMIT_WAIT_TIME
** isc_info_wal_ckpt_length         for CHECK_POINT_LENGTH
**
**************************************************************/

	bool set_used = false;
	if (has_wal)
	{
		char* buffer = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH128);
		if (!buffer)
		{
			isc_status[0] = isc_arg_gds;
			isc_status[1] = isc_virmemexh;
			isc_status[2] = isc_arg_end;
			ISQL_errmsg(isc_status);
			return;
		}

		if (!isc_database_info
			(gds_status, &DB, sizeof(wal_items), wal_items, BUFFER_LENGTH128,
			 buffer))
		{
			for (const char* d = buffer; *d != isc_info_end;)
			{
				const char item = *d++;
				const int length = isc_vax_integer(d, 2);
				d += 2;
				const SLONG value_out = isc_vax_integer(d, length);
				switch (item)
				{
				case isc_info_end:
					break;

				case isc_info_num_wal_buffers:
					print_set(&set_used);
					sprintf(Print_buffer, "NUM_LOG_BUFFERS = %ld", value_out);
					ISQL_printf(Out, Print_buffer);
					break;

				case isc_info_wal_buffer_size:
					print_set(&set_used);
					sprintf(Print_buffer, "LOG_BUFFER_SIZE = %ld", value_out);
					ISQL_printf(Out, Print_buffer);
					break;

				case isc_info_wal_grpc_wait_usecs:
					print_set(&set_used);
					sprintf(Print_buffer, "GROUP_COMMIT_WAIT_TIME = %ld",
							value_out);
					ISQL_printf(Out, Print_buffer);
					break;

				case isc_info_wal_ckpt_length:
					print_set(&set_used);
					sprintf(Print_buffer, "CHECK_POINT_LENGTH = %ld",
							value_out);
					ISQL_printf(Out, Print_buffer);
					break;

				default:
					break;
				}
				d += length;
			}
		}
		if (buffer)
			ISQL_FREE(buffer);
	}

	if (!first)
	{
		if (nodb)
		{
			sprintf(Print_buffer, "%s */%s", NEWLINE, NEWLINE);
			ISQL_printf(Out, Print_buffer);
		}
		else
		{
			sprintf(Print_buffer, "%s%s%s", global_Term, NEWLINE, NEWLINE);
			ISQL_printf(Out, Print_buffer);
		}
	}
}


static void list_domain_table(const SCHAR* table_name,
							  SSHORT default_char_set_id)
{
/**************************************
 *
 *	l i s t _ d o m a i n _ t a b l e
 *
 **************************************
 *
 * Functional description
 *	List domains as identified by fields with any constraints on them
 *	for the named table
 *
 *	Parameters:  table_name == only extract domains for this table
 *	default_char_set_id -- character set def to supress
 *
 **************************************/
	bool first = true;
	SSHORT i;
	SCHAR char_sets[86];
	SSHORT subtype;

	FOR FLD IN RDB$FIELDS CROSS
		RFR IN RDB$RELATION_FIELDS WITH
		RFR.RDB$FIELD_SOURCE EQ FLD.RDB$FIELD_NAME AND
		RFR.RDB$RELATION_NAME EQ table_name
		SORTED BY FLD.RDB$FIELD_NAME

		// Skip over artifical domains 
		if (strncmp (FLD.RDB$FIELD_NAME, "RDB$", 4) == 0 &&
			isdigit (FLD.RDB$FIELD_NAME[4]) &&
			FLD.RDB$SYSTEM_FLAG != 1)
		{
			continue;
		}

		if (first)
		{
			sprintf (Print_buffer, "/* Domain definitions */%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
			first = false;
		}
		fb_utils::fb_exact_name(FLD.RDB$FIELD_NAME);

		sprintf (Print_buffer, "CREATE DOMAIN %s AS ", FLD.RDB$FIELD_NAME);
		ISQL_printf (Out, Print_buffer);

		for (i = 0; Column_types[i].type; i++)
			if (FLD.RDB$FIELD_TYPE == Column_types[i].type)
			{
				bool precision_known = false;

				if (major_ods >= ODS_VERSION10)
				{
					// Handle Integral subtypes NUMERIC and DECIMAL 
					if ((FLD.RDB$FIELD_TYPE == SMALLINT) ||
						(FLD.RDB$FIELD_TYPE == INTEGER) ||
						(FLD.RDB$FIELD_TYPE == blr_int64))
					{
						/* We are ODS >= 10 and could be any Dialect */

						FOR FLD1 IN RDB$FIELDS WITH
							FLD1.RDB$FIELD_NAME EQ FLD.RDB$FIELD_NAME
							if (!FLD1.RDB$FIELD_PRECISION.NULL)
							{
								/* We are Dialect >=3 since FIELD_PRECISION is non-NULL */
								if (FLD1.RDB$FIELD_SUB_TYPE > 0 &&
									FLD1.RDB$FIELD_SUB_TYPE <= MAX_INTSUBTYPES)
								{
									sprintf (Print_buffer, "%s(%d, %d)",
											 Integral_subtypes[FLD1.RDB$FIELD_SUB_TYPE],
											 FLD1.RDB$FIELD_PRECISION,
												 -FLD1.RDB$FIELD_SCALE);
									precision_known = true;
								}
							}
						END_FOR
						ON_ERROR
							ISQL_errmsg (isc_status);
							return;
						END_ERROR;
					}
				}

				if (!precision_known)
				{
					// Take a stab at numerics and decimals 
					if ((FLD.RDB$FIELD_TYPE == SMALLINT) && (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(4, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else if ((FLD.RDB$FIELD_TYPE == INTEGER) && (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(9, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else if ((FLD.RDB$FIELD_TYPE == DOUBLE_PRECISION) &&
							 (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(15, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else
					{
						sprintf (Print_buffer, "%s", Column_types[i].type_name);
					}
				}
				ISQL_printf (Out, Print_buffer);
				break;
			}


		if (FLD.RDB$FIELD_TYPE == BLOB)
		{
			subtype = FLD.RDB$FIELD_SUB_TYPE;
			ISQL_printf (Out, " SUB_TYPE ");
			if ((subtype > 0) && (subtype <= MAXSUBTYPES))
			{
				sprintf (Print_buffer, "%s", Sub_types[subtype]);
				ISQL_printf (Out, Print_buffer);
			}
			else
			{
				sprintf (Print_buffer, "%d", subtype);
				ISQL_printf (Out, Print_buffer);
			}
			sprintf (Print_buffer, " SEGMENT SIZE %u", (USHORT) FLD.RDB$SEGMENT_LENGTH);
			ISQL_printf (Out, Print_buffer);
		}

		else if ((FLD.RDB$FIELD_TYPE == FCHAR) || (FLD.RDB$FIELD_TYPE == VARCHAR))
		{
			// Length for chars 
			sprintf (Print_buffer, "(%d)", ISQL_get_field_length(FLD.RDB$FIELD_NAME));
			ISQL_printf (Out, Print_buffer);
		};

		/* Bug 8261: do not show the collation information just yet!  If you
		   do, then the domain syntax when printed is not correct */

		/* since the character set is part of the field type, display that
		   information now. */
		if (!FLD.RDB$CHARACTER_SET_ID.NULL)
		{

			char_sets[0] = 0;
			if ((FLD.RDB$CHARACTER_SET_ID != default_char_set_id) ||
				(!FLD.RDB$COLLATION_ID.NULL && FLD.RDB$COLLATION_ID != 0))
				ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID, 0, false, char_sets);
			if (char_sets[0])
				ISQL_printf (Out, char_sets);
		}

		if (!FLD.RDB$DIMENSIONS.NULL)
			ISQL_array_dimensions (FLD.RDB$FIELD_NAME);

		if (!FLD.RDB$DEFAULT_SOURCE.NULL)
		{
			sprintf (Print_buffer, "%s%s ", NEWLINE, TAB_AS_SPACES);
			ISQL_printf (Out, Print_buffer);
			SHOW_print_metadata_text_blob (Out, &FLD.RDB$DEFAULT_SOURCE);
		}
		if (!FLD.RDB$VALIDATION_SOURCE.NULL)
		{
			sprintf (Print_buffer, "%s%s ", NEWLINE, TAB_AS_SPACES);
			ISQL_printf (Out, Print_buffer);
			ISQL_print_validation (Out, &FLD.RDB$VALIDATION_SOURCE, false, gds_trans);
		}
		if (FLD.RDB$NULL_FLAG == 1)
			ISQL_printf (Out, " NOT NULL");

		// Bug 8261:  Now show the collation order information 
		/* Show the collation order if one has been specified.  If the collation
		   order is the default for the character set being used, then no collation
		   order will be shown ( because it isn't needed ).

	       If the collation id is 0, then the default for the character set is
	       being used so there is no need to retrieve the collation information.*/

		if (!FLD.RDB$COLLATION_ID.NULL && FLD.RDB$COLLATION_ID != 0)
		{
			char_sets[0] = 0;
			ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID, FLD.RDB$COLLATION_ID,
									 true, char_sets);
			if (char_sets[0])
				ISQL_printf (Out, char_sets);
		}

		sprintf (Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf (Out, Print_buffer);
	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;
}


static void list_domains(SSHORT default_char_set_id)
{
/**************************************
 *
 *	l i s t _ d o m a i n s
 *
 **************************************
 *
 * Functional description
 *	List domains
 *
 *	Parameters:
 *	default_char_set_id -- character set def to supress
 *
 **************************************/
	bool first = true;
	SSHORT i;
	SCHAR char_sets[86];
	SSHORT subtype;

	FOR FLD IN RDB$FIELDS WITH
		FLD.RDB$FIELD_NAME NOT MATCHING "RDB$+" USING "+=[0-9][0-9]* *"
		AND FLD.RDB$SYSTEM_FLAG NE 1
		SORTED BY FLD.RDB$FIELD_NAME

		if (first)
		{
			sprintf (Print_buffer, "/* Domain definitions */%s", NEWLINE);
			ISQL_printf (Out, Print_buffer);
			first = false;
		}
		fb_utils::fb_exact_name(FLD.RDB$FIELD_NAME);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id(FLD.RDB$FIELD_NAME, SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, "CREATE DOMAIN %s AS ", SQL_identifier);
		}
		else
			sprintf (Print_buffer, "CREATE DOMAIN %s AS ", FLD.RDB$FIELD_NAME);

		ISQL_printf (Out, Print_buffer);

		// Get domain type 
		for (i = 0; Column_types[i].type; i++)
			if (FLD.RDB$FIELD_TYPE == Column_types[i].type)
			{
				bool precision_known = false;

				if (major_ods >= ODS_VERSION10)
				{
					// Handle Integral subtypes NUMERIC and DECIMAL 
					if ((FLD.RDB$FIELD_TYPE == SMALLINT) ||
						(FLD.RDB$FIELD_TYPE == INTEGER) ||
						(FLD.RDB$FIELD_TYPE == blr_int64))
					{
						/* We are ODS >= 10 and could be any Dialect */
						FOR FLD1 IN RDB$FIELDS WITH
							FLD1.RDB$FIELD_NAME EQ FLD.RDB$FIELD_NAME

							if (!FLD1.RDB$FIELD_PRECISION.NULL)
							{
								/* We are Dialect >=3 since FIELD_PRECISION is non-NULL */
								if (FLD1.RDB$FIELD_SUB_TYPE > 0 &&
									FLD1.RDB$FIELD_SUB_TYPE <= MAX_INTSUBTYPES)
								{
									sprintf (Print_buffer, "%s(%d, %d)",
											 Integral_subtypes[FLD1.RDB$FIELD_SUB_TYPE],
											 FLD1.RDB$FIELD_PRECISION,
											 -FLD1.RDB$FIELD_SCALE);
									precision_known = true;
								}
							}
						END_FOR
						ON_ERROR
							ISQL_errmsg (isc_status);
							return;
						END_ERROR;
					}
				}

				if (!precision_known)
				{
					// Take a stab at numerics and decimals 
					if ((FLD.RDB$FIELD_TYPE == SMALLINT) && (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(4, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else if ((FLD.RDB$FIELD_TYPE == INTEGER) && (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(9, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else if ((FLD.RDB$FIELD_TYPE == DOUBLE_PRECISION) &&
							 (FLD.RDB$FIELD_SCALE < 0))
					{
						sprintf (Print_buffer, "NUMERIC(15, %d)", -FLD.RDB$FIELD_SCALE);
					}
					else
					{
						sprintf (Print_buffer, "%s", Column_types[i].type_name);
					}
				}
				ISQL_printf (Out, Print_buffer);
				break;
			}

		if (FLD.RDB$FIELD_TYPE == BLOB)
		{
			subtype = FLD.RDB$FIELD_SUB_TYPE;
			ISQL_printf (Out, " SUB_TYPE ");

			if ((subtype > 0) && (subtype <= MAXSUBTYPES))
			{
				sprintf (Print_buffer, "%s", Sub_types[subtype]);
				ISQL_printf (Out, Print_buffer);
			}
			else
			{
				sprintf (Print_buffer, "%d", subtype);
				ISQL_printf (Out, Print_buffer);
			}
			sprintf (Print_buffer, " SEGMENT SIZE %u", (USHORT) FLD.RDB$SEGMENT_LENGTH);
			ISQL_printf (Out, Print_buffer);
		}

		else if ((FLD.RDB$FIELD_TYPE == FCHAR) || (FLD.RDB$FIELD_TYPE == VARCHAR))
		{
			// Length for chars 
			sprintf (Print_buffer, "(%d)", ISQL_get_field_length(FLD.RDB$FIELD_NAME));
			ISQL_printf (Out, Print_buffer);
		};

		/* Bug 8261: do not show the collation information just yet!  If you
		   do, then the domain syntax when printed is not correct */

		/* since the character set is part of the field type, display that
		   information now. */
		if (!FLD.RDB$CHARACTER_SET_ID.NULL)
		{
			char_sets[0] = 0;
			if ((FLD.RDB$CHARACTER_SET_ID != default_char_set_id) ||
				(!FLD.RDB$COLLATION_ID.NULL && FLD.RDB$COLLATION_ID != 0))
			{
				ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID, 0, false, char_sets);
			}
			if (char_sets[0])
				ISQL_printf (Out, char_sets);
		}

		if (!FLD.RDB$DIMENSIONS.NULL)
			ISQL_array_dimensions (FLD.RDB$FIELD_NAME);

		if (!FLD.RDB$DEFAULT_SOURCE.NULL)
		{
			sprintf (Print_buffer, "%s%s ", NEWLINE, TAB_AS_SPACES);
			ISQL_printf (Out, Print_buffer);
			SHOW_print_metadata_text_blob (Out, &FLD.RDB$DEFAULT_SOURCE);
		}
		if (!FLD.RDB$VALIDATION_SOURCE.NULL)
		{
			sprintf (Print_buffer, "%s%s ", NEWLINE, TAB_AS_SPACES);
			ISQL_printf (Out, Print_buffer);
			ISQL_print_validation (Out, &FLD.RDB$VALIDATION_SOURCE, false, gds_trans);
		}
		if (FLD.RDB$NULL_FLAG == 1)
			ISQL_printf (Out, " NOT NULL");

		// Bug 8261:  Now show the collation order information 
		/* Show the collation order if one has been specified.  If the collation
		   order is the default for the character set being used, then no collation
		   order will be shown ( because it isn't needed

		   If the collation id is 0, then the default for the character set is
		   being used so there is no need to retrieve the collation information.*/

		if (!FLD.RDB$COLLATION_ID.NULL && FLD.RDB$COLLATION_ID != 0)
		{
			char_sets[0] = 0;
			ISQL_get_character_sets (FLD.RDB$CHARACTER_SET_ID, FLD.RDB$COLLATION_ID,
									 true, char_sets);
			if (char_sets[0])
				ISQL_printf (Out, char_sets);
		}

		sprintf (Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf (Out, Print_buffer);
	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;
}


static void list_exception()
{
/**************************************
 *
 *	l i s t _ e x c e p t i o n
 *
 **************************************
 *
 * Functional description
 *	List all exceptions defined in the database
 *
 *	Parameters:  none
 *
 **************************************/
	bool first = true;

	FOR EXC IN RDB$EXCEPTIONS
		SORTED BY EXC.RDB$EXCEPTION_NAME

		if (first)
		{
			sprintf (Print_buffer, "%s/*  Exceptions */%s", NEWLINE, NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}
		first = false;
		fb_utils::fb_exact_name(EXC.RDB$EXCEPTION_NAME);

		ISQL_copy_SQL_id (EXC.RDB$MESSAGE, SQL_identifier2, SINGLE_QUOTE);
		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (EXC.RDB$EXCEPTION_NAME, SQL_identifier, DBL_QUOTE);
			sprintf (Print_buffer, "CREATE EXCEPTION %s %s%s%s",
					 SQL_identifier, SQL_identifier2, global_Term, NEWLINE);
		}
		else
			sprintf (Print_buffer, "CREATE EXCEPTION %s %s%s%s",
					 EXC.RDB$EXCEPTION_NAME, SQL_identifier2, global_Term, NEWLINE);

		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;
}


static void list_filters()
{
/**************************************
 *
 *	l i s t _ f i l t e r s
 *
 **************************************
 *
 * Functional description
 *	List all blob filters
 *
 *	Parameters:  none
 * Results in
 * DECLARE FILTER <fname> INPUT_TYPE <blob_sub_type> OUTPUT_TYPE <blob_subtype>
 *		 ENTRY_POINT <string> MODULE_NAME <string>
 **************************************/
	bool first = true;

	FOR FIL IN RDB$FILTERS
		SORTED BY FIL.RDB$FUNCTION_NAME

		fb_utils::fb_exact_name(FIL.RDB$FUNCTION_NAME);
		fb_utils::fb_exact_name(FIL.RDB$MODULE_NAME);
		fb_utils::fb_exact_name(FIL.RDB$ENTRYPOINT);

		if (first)
		{
			sprintf (Print_buffer, "%s/*  BLOB Filter declarations */%s",
					 NEWLINE,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}
		first = false;

		sprintf (Print_buffer, "DECLARE FILTER %s INPUT_TYPE %d OUTPUT_TYPE %d%s%sENTRY_POINT '%s' MODULE_NAME '%s'%s%s%s",
				 FIL.RDB$FUNCTION_NAME, FIL.RDB$INPUT_SUB_TYPE, FIL.RDB$OUTPUT_SUB_TYPE,
				 NEWLINE, TAB_AS_SPACES, FIL.RDB$ENTRYPOINT, FIL.RDB$MODULE_NAME, global_Term, NEWLINE,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;

}


static void list_foreign()
{
/**************************************
 *
 *	l i s t _ f o r e i g n
 *
 **************************************
 *
 * Functional description
 *	List all foreign key constraints and alter the tables
 *
 **************************************/
	SCHAR* collist = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH512 * 2);
	if (!collist)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		return;
	}

/* Static queries for obtaining foreign constraints, where RELC1 is the
   foreign key constraints, RELC2 is the primary key lookup and REFC
   is the join table */

	FOR RELC1 IN RDB$RELATION_CONSTRAINTS CROSS
		RELC2 IN RDB$RELATION_CONSTRAINTS CROSS
		REFC IN RDB$REF_CONSTRAINTS WITH
		RELC1.RDB$CONSTRAINT_TYPE EQ "FOREIGN KEY" AND
		REFC.RDB$CONST_NAME_UQ EQ RELC2.RDB$CONSTRAINT_NAME AND
		REFC.RDB$CONSTRAINT_NAME EQ RELC1.RDB$CONSTRAINT_NAME AND
		(RELC2.RDB$CONSTRAINT_TYPE EQ "UNIQUE" OR
		RELC2.RDB$CONSTRAINT_TYPE EQ "PRIMARY KEY")
		SORTED BY RELC1.RDB$RELATION_NAME, RELC1.RDB$CONSTRAINT_NAME

		fb_utils::fb_exact_name(RELC1.RDB$RELATION_NAME);
		fb_utils::fb_exact_name(RELC2.RDB$RELATION_NAME);

		ISQL_get_index_segments (collist, RELC1.RDB$INDEX_NAME, true);

		sprintf (Print_buffer, " %s", NEWLINE);
		ISQL_printf (Out, Print_buffer);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (RELC1.RDB$RELATION_NAME, SQL_identifier,
							  DBL_QUOTE);
			sprintf (Print_buffer, "ALTER TABLE %s ADD ", SQL_identifier);
		}
		else
			sprintf (Print_buffer, "ALTER TABLE %s ADD ",
					 RELC1.RDB$RELATION_NAME);
		ISQL_printf (Out, Print_buffer);

		/* If the name of the constraint is not INTEG..., print it.
	      INTEG... are internally generated names. */
		if (!RELC1.RDB$CONSTRAINT_NAME.NULL &&
			strncmp(RELC1.RDB$CONSTRAINT_NAME, "INTEG", 5))
		{
			ISQL_truncate_term (RELC1.RDB$CONSTRAINT_NAME,
								strlen(RELC1.RDB$CONSTRAINT_NAME));
			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			{
				ISQL_copy_SQL_id (RELC1.RDB$CONSTRAINT_NAME, SQL_identifier,
								  DBL_QUOTE);
				sprintf (Print_buffer, "CONSTRAINT %s ", SQL_identifier);
			}
			else
				sprintf (Print_buffer, "CONSTRAINT %s ",
						 RELC1.RDB$CONSTRAINT_NAME);
			ISQL_printf (Out, Print_buffer);
		}

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (RELC2.RDB$RELATION_NAME, SQL_identifier,
							  DBL_QUOTE);
			sprintf (Print_buffer, "FOREIGN KEY (%s) REFERENCES %s ",
					 collist, SQL_identifier);
		}
		else
			sprintf (Print_buffer, "FOREIGN KEY (%s) REFERENCES %s ",
					 collist,
					 RELC2.RDB$RELATION_NAME);
		ISQL_printf (Out, Print_buffer);

		// Get the column list for the primary key 

		ISQL_get_index_segments (collist, RELC2.RDB$INDEX_NAME, true);

		sprintf (Print_buffer, "(%s)", collist);
		ISQL_printf (Out, Print_buffer);

		// Add the referential actions, if any 
		if (!REFC.RDB$UPDATE_RULE.NULL)
		{
			ISQL_truncate_term (REFC.RDB$UPDATE_RULE,
								strlen(REFC.RDB$UPDATE_RULE));
			ISQL_ri_action_print (REFC.RDB$UPDATE_RULE, " ON UPDATE", true);
		}

		if (!REFC.RDB$DELETE_RULE.NULL)
		{
			ISQL_truncate_term (REFC.RDB$DELETE_RULE,
								strlen(REFC.RDB$DELETE_RULE));
			ISQL_ri_action_print (REFC.RDB$DELETE_RULE, " ON DELETE", true);
		}

		sprintf (Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		ISQL_FREE(collist);
		return;
	END_ERROR;

	ISQL_FREE(collist);

}


static void list_functions()
{
/**************************************
 *
 *	l i s t _ f u n c t i o n s
 *
 **************************************
 *
 * Functional description
 *	List all external functions
 *
 *	Parameters:  none
 * Results in
 * DECLARE EXTERNAL FUNCTION function_name
 *               CHAR(256) , INTEGER, ....
 *               RETURNS INTEGER BY VALUE
 *               ENTRY_POINT entrypoint MODULE_NAME module;
 **************************************/
	bool first = true;
	bool firstarg = true;
	bool printarg = true;
	SSHORT i, ptype;
	SCHAR* return_buffer;
	SCHAR* type_buffer;
	SCHAR* buffer;

	buffer = NULL;
	type_buffer = NULL;
	return_buffer = NULL;

	buffer = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH360);
	if (!buffer)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		return;
	}

	type_buffer = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH128);
	if (!type_buffer)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		ISQL_FREE(buffer);
		return;
	}

	return_buffer = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH128);
	if (!return_buffer)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		ISQL_FREE(type_buffer);
		ISQL_FREE(buffer);
		return;
	}

	FOR FUN IN RDB$FUNCTIONS
		SORTED BY FUN.RDB$FUNCTION_NAME

		fb_utils::fb_exact_name(FUN.RDB$FUNCTION_NAME);
		fb_utils::fb_exact_name(FUN.RDB$MODULE_NAME);
		fb_utils::fb_exact_name(FUN.RDB$ENTRYPOINT);
		if (first)
		{
			sprintf (Print_buffer, "%s/*  External Function declarations */%s",
					 NEWLINE,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
			first = false;
		}

		// Start new function declaration 
		sprintf (Print_buffer, "DECLARE EXTERNAL FUNCTION %s %s",
				 FUN.RDB$FUNCTION_NAME,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);

		firstarg = true;

		FOR FNA IN RDB$FUNCTION_ARGUMENTS WITH
			FUN.RDB$FUNCTION_NAME EQ FNA.RDB$FUNCTION_NAME
			SORTED BY FNA.RDB$ARGUMENT_POSITION

			// Find parameter type 
			i = 0;
			while (FNA.RDB$FIELD_TYPE != Column_types[i].type)
				i++;

			// Print length where appropriate 
			if ((FNA.RDB$FIELD_TYPE == FCHAR)   ||
				(FNA.RDB$FIELD_TYPE == VARCHAR) ||
				(FNA.RDB$FIELD_TYPE == CSTRING))
			{
				bool did_charset = false;
				FOR CHARSET IN RDB$CHARACTER_SETS
					WITH CHARSET.RDB$CHARACTER_SET_ID = FNA.RDB$CHARACTER_SET_ID

					did_charset = true;
					fb_utils::fb_exact_name(CHARSET.RDB$CHARACTER_SET_NAME);
					sprintf (type_buffer, "%s(%d) CHARACTER SET %s",
							 Column_types[i].type_name,
							 (FNA.RDB$FIELD_LENGTH / MAX (1, CHARSET.RDB$BYTES_PER_CHARACTER)),
							 CHARSET.RDB$CHARACTER_SET_NAME);

				END_FOR
				ON_ERROR
					ISQL_errmsg (gds_status);
					return;
				END_ERROR;

				if (!did_charset)
					sprintf (type_buffer, "%s(%d)",  Column_types[i].type_name,
							 FNA.RDB$FIELD_LENGTH);
			}
			else
			{
				bool precision_known = false;

				if ( (major_ods >= ODS_VERSION10) &&
					((FNA.RDB$FIELD_TYPE == SMALLINT) ||
					(FNA.RDB$FIELD_TYPE == INTEGER) ||
					(FNA.RDB$FIELD_TYPE == blr_int64)))
				{
					FOR FNA1 IN RDB$FUNCTION_ARGUMENTS WITH
						FNA1.RDB$FUNCTION_NAME EQ FNA.RDB$FUNCTION_NAME AND
						FNA1.RDB$ARGUMENT_POSITION EQ FNA.RDB$ARGUMENT_POSITION

						/* We are ODS >= 10 and could be any Dialect */
						if (!FNA1.RDB$FIELD_PRECISION.NULL)
						{
							/* We are Dialect >=3 since FIELD_PRECISION is non-NULL */
							if (FNA1.RDB$FIELD_SUB_TYPE > 0 &&
								FNA1.RDB$FIELD_SUB_TYPE <= MAX_INTSUBTYPES)
							{
								sprintf (type_buffer, "%s(%d, %d)",
										Integral_subtypes[FNA1.RDB$FIELD_SUB_TYPE],
										FNA1.RDB$FIELD_PRECISION,
										-FNA1.RDB$FIELD_SCALE);
								precision_known = true;
							}
						} // if field_precision is not null 
					END_FOR
					ON_ERROR
						ISQL_errmsg (isc_status);
						return;
					END_ERROR;
				} /* if major_ods >= ods_version10 && */

				if (!precision_known)
				{
					// Take a stab at numerics and decimals 
					if ((FNA.RDB$FIELD_TYPE == SMALLINT) &&
						(FNA.RDB$FIELD_SCALE < 0))
					{
						sprintf (type_buffer, "NUMERIC(4, %d)",
								 -FNA.RDB$FIELD_SCALE);
					}
					else if ((FNA.RDB$FIELD_TYPE == INTEGER) &&
							 (FNA.RDB$FIELD_SCALE < 0))
					{
						sprintf (type_buffer, "NUMERIC(9, %d)",
								 -FNA.RDB$FIELD_SCALE);
					}
					else if ((FNA.RDB$FIELD_TYPE == DOUBLE_PRECISION) &&
							 (FNA.RDB$FIELD_SCALE < 0))
					{
						sprintf (type_buffer, "NUMERIC(15, %d)",
								 -FNA.RDB$FIELD_SCALE);
					}
					else
						sprintf (type_buffer, "%s",
								 Column_types[i].type_name);
				} /* if !precision_known */
			} // if FCHAR or VARCHAR or CSTRING ... else 

			// If a return argument, save it for the end, otherwise print 

			/*
			 * Changed the following to not return a BLOB by value.
			 * To be sincere, this code doesn't cater for the RETURNS PARAMETER syntax but
			 * it would require more surgery than I'm willing to do, since I'm sick of isql
			 * so I started my own metadata extraction utility based on IBO that does this
			 * trick and others.
			 * Claudio Valderrama (by way of) MOD 23-Apr-2001

			 CVC: Finally enhanced the UDF metadata extraction.
			*/

			ptype = (SSHORT) abs (FNA.RDB$MECHANISM);
			if (ptype > MAX_UDFPARAM_TYPES + 1 || ptype < 0)
				ptype = MAX_UDFPARAM_TYPES + 1;

			printarg = true;


			if (FUN.RDB$RETURN_ARGUMENT == FNA.RDB$ARGUMENT_POSITION) {
				if (FUN.RDB$RETURN_ARGUMENT) {
					sprintf (return_buffer, "RETURNS PARAMETER %d",
							 FUN.RDB$RETURN_ARGUMENT);
				}
				else {
					sprintf (return_buffer, "RETURNS %s%s %s", type_buffer,
							 UDF_param_types[ptype],
							 (FNA.RDB$MECHANISM < 0 ? "FREE_IT" : ""));
					printarg = false;
				}

			}


			if (printarg) {
				// First arg needs no comma 
				sprintf (Print_buffer, "%s%s%s", (firstarg ? "" : ", "), type_buffer,
						 UDF_param_types[ptype]);
				ISQL_printf (Out, Print_buffer);
				firstarg = false;
			}

		END_FOR
		ON_ERROR
			ISQL_errmsg (gds_status);
			ISQL_FREE (type_buffer);
			ISQL_FREE (return_buffer);
			ISQL_FREE (buffer);
			return;
		END_ERROR;

		// Print the return type -- All functions return a type 
		sprintf (Print_buffer, "%s%s%s",
				 NEWLINE,
				 return_buffer,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);

		// Print out entrypoint information 
		sprintf (Print_buffer, "ENTRY_POINT '%s' MODULE_NAME '%s'%s%s%s",
				 FUN.RDB$ENTRYPOINT,
				 FUN.RDB$MODULE_NAME,
				 global_Term,
				 NEWLINE,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		ISQL_FREE(type_buffer);
		ISQL_FREE(return_buffer);
		ISQL_FREE(buffer);
		return;
	END_ERROR;

	ISQL_FREE(type_buffer);
	ISQL_FREE(return_buffer);
	ISQL_FREE(buffer);
}




static void list_generators()
{
/**************************************
 *
 *	l i s t _ g e n e r a t o r s
 *
 **************************************
 *
 * Functional description
 *	Re create all non-system generators
 *
 **************************************/

	ISQL_printf(Out, NEWLINE);

	FOR GEN IN RDB$GENERATORS WITH
		GEN.RDB$GENERATOR_NAME NOT MATCHING "RDB$+" USING "+=[0-9][0-9]* *" AND
		GEN.RDB$GENERATOR_NAME NOT MATCHING "SQL$+" USING "+=[0-9][0-9]* *" AND
		(GEN.RDB$SYSTEM_FLAG MISSING OR GEN.RDB$SYSTEM_FLAG NE 1)
		SORTED BY GEN.RDB$GENERATOR_NAME

		fb_utils::fb_exact_name(GEN.RDB$GENERATOR_NAME);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			ISQL_copy_SQL_id (GEN.RDB$GENERATOR_NAME, SQL_identifier, DBL_QUOTE);
		else
			strcpy (SQL_identifier, GEN.RDB$GENERATOR_NAME);

		sprintf (Print_buffer, "CREATE GENERATOR %s%s%s",
				 SQL_identifier,
				 global_Term,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);
	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;

	ISQL_printf(Out, NEWLINE);
}


static void list_index()
{
/**************************************
 *
 *	l i s t _ i n d e x
 *
 **************************************
 *
 * Functional description
 *	Define all non-constraint indices
 *	Use a static SQL query to get the info and print it.
 *
 *	Uses get_index_segment to provide a key list for each index
 *
 **************************************/
	bool first = true;

	char* collist = (SCHAR*) ISQL_ALLOC(BUFFER_LENGTH512 * 2);
	if (!collist)
	{
		isc_status[0] = isc_arg_gds;
		isc_status[1] = isc_virmemexh;
		isc_status[2] = isc_arg_end;
		ISQL_errmsg(isc_status);
		return;
	}

	FOR IDX IN RDB$INDICES CROSS RELC IN RDB$RELATIONS
		OVER RDB$RELATION_NAME
		WITH (RELC.RDB$SYSTEM_FLAG NE 1 OR RELC.RDB$SYSTEM_FLAG MISSING)
		AND NOT (ANY RC IN RDB$RELATION_CONSTRAINTS
		   WITH RC.RDB$INDEX_NAME EQ IDX.RDB$INDEX_NAME)
		SORTED BY IDX.RDB$RELATION_NAME, IDX.RDB$INDEX_NAME


		if (first)
		{
			sprintf (Print_buffer,
					 "%s/*  Index definitions for all user tables */%s",
					 NEWLINE,
					 NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}
		first = false;
	// Strip trailing blanks 

		fb_utils::fb_exact_name(IDX.RDB$RELATION_NAME);
		fb_utils::fb_exact_name(IDX.RDB$INDEX_NAME);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
		{
			ISQL_copy_SQL_id (IDX.RDB$INDEX_NAME,    SQL_identifier,  DBL_QUOTE);
			ISQL_copy_SQL_id (IDX.RDB$RELATION_NAME, SQL_identifier2, DBL_QUOTE);
			sprintf (Print_buffer, "CREATE%s%s INDEX %s ON %s(",
					 (IDX.RDB$UNIQUE_FLAG ? " UNIQUE" : ""),
					 (IDX.RDB$INDEX_TYPE ? " DESCENDING" : ""),
					 SQL_identifier,
					 SQL_identifier2);
		}
		else
			sprintf (Print_buffer, "CREATE%s%s INDEX %s ON %s(",
					 (IDX.RDB$UNIQUE_FLAG ? " UNIQUE" : ""),
					 (IDX.RDB$INDEX_TYPE ? " DESCENDING" : ""),
					 IDX.RDB$INDEX_NAME,
					 IDX.RDB$RELATION_NAME);
		ISQL_printf (Out, Print_buffer);

		// Get column names 

		if (ISQL_get_index_segments (collist, IDX.RDB$INDEX_NAME, true))
		{
			sprintf (Print_buffer, "%s)%s%s", collist, global_Term, NEWLINE);
			ISQL_printf (Out, Print_buffer);
		}

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		ISQL_FREE(collist);
		return;
	END_ERROR;

	ISQL_FREE(collist);
}


static void list_views()
{
/**************************************
 *
 *	l i s t _ v i e w s
 *
 **************************************
 *
 * Functional description
 *	Show text of views.
 *	Use a SQL query to get the info and print it.
 *	Note: This should also contain check option
 *
 **************************************/
	bool first;

// If this is a view, use print_blob to print the view text 

	FOR REL IN RDB$RELATIONS WITH
		(REL.RDB$SYSTEM_FLAG NE 1 OR REL.RDB$SYSTEM_FLAG MISSING) AND
		REL.RDB$VIEW_BLR NOT MISSING AND
		REL.RDB$FLAGS = REL_sql
		SORTED BY REL.RDB$RELATION_ID

		first = true;
		fb_utils::fb_exact_name(REL.RDB$RELATION_NAME);

		if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
			ISQL_copy_SQL_id (REL.RDB$RELATION_NAME, SQL_identifier, DBL_QUOTE);
		else
			strcpy (SQL_identifier, REL.RDB$RELATION_NAME);

		fb_utils::fb_exact_name(REL.RDB$OWNER_NAME);

		sprintf (Print_buffer, "%s/* View: %s, Owner: %s */%s",
				 NEWLINE,
				 REL.RDB$RELATION_NAME,
				 REL.RDB$OWNER_NAME,
				 NEWLINE);
		ISQL_printf (Out, Print_buffer);
		sprintf (Print_buffer, "CREATE VIEW %s (", SQL_identifier);
		ISQL_printf (Out, Print_buffer);

		// Get column list 
		FOR RFR IN RDB$RELATION_FIELDS WITH
			RFR.RDB$RELATION_NAME = REL.RDB$RELATION_NAME
			SORTED BY RFR.RDB$FIELD_POSITION

			fb_utils::fb_exact_name(RFR.RDB$FIELD_NAME);

			if (db_SQL_dialect > SQL_DIALECT_V6_TRANSITION)
				ISQL_copy_SQL_id (RFR.RDB$FIELD_NAME, SQL_identifier, DBL_QUOTE);
			else
				strcpy (SQL_identifier, RFR.RDB$FIELD_NAME);

			sprintf (Print_buffer, "%s%s", (first ? "" : ", "), SQL_identifier);
			ISQL_printf (Out, Print_buffer);
			first = false;

		END_FOR
		ON_ERROR
			ISQL_errmsg (gds_status);
			return;
		END_ERROR;
		sprintf (Print_buffer, ") AS%s", NEWLINE);
		ISQL_printf (Out, Print_buffer);

		if (!REL.RDB$VIEW_SOURCE.NULL)
			SHOW_print_metadata_text_blob (Out, &REL.RDB$VIEW_SOURCE);

		sprintf (Print_buffer, "%s%s", global_Term, NEWLINE);
		ISQL_printf (Out, Print_buffer);

	END_FOR
	ON_ERROR
		ISQL_errmsg(gds_status);
		return;
	END_ERROR;
}