Using Recital SQL
Contents
- 1 Using Recital SQL
- 1.1 Creating a Recital Database
- 1.2 Opening a Recital Database
- 1.3 Creating a Recital Table using Recital SQL
- 1.4 Inserting Records into a Recital Table using Recital SQL
- 1.5 Selecting Records from a Recital Table Using Recital SQL
- 1.6 Updating Records in a Recital Table using Recital SQL
- 1.7 Deleting Records in a Recital Table using Recital SQL
- 1.8 Using a Cursor to Select, Update and Delete Records
- 1.9 Obtaining Information about a Recital Table using SQL
- 1.10 Executing Procedures and Prepared Statements using Recital SQL
Using Recital SQL
Creating a Recital Database
The create database statement is used to create a new database. This will create a sub-directory in the DB_DATADIR directory and a database catalog with the same name as the database specified.
create database [if not exists] <database>
A database, its tables and other associated files can be physically deleted using the drop database statement.
drop database [if exists] <database>
Example
create database test drop database test
Opening a Recital Database
To open a database and set it as the default database for any subsequent SQL statements, use the open database statement.
open database <database> [EXCLUSIVE | SHARED] [NOUPDATE] [VALIDATE]
The open database command triggers the DBC_OPENDATA database event. If a dbc_opendata.prg program file exists in the database's directory, this will be run. If the dbc_opendata.prg program returns .F. (False), the open database operation will be abandoned.
Databases can also have an associated procedure library that is activated automatically when the database is opened. The procedure library must follow the naming convention dbc_<database>_library.prg and must exist in the database's directory. When the database is closed, the procedure library is also closed.
Tables from a database that is not currently open can be accessed by preceding the table name by the database name and a !.
select * from southwind!shippers
Recital also supports the MySQL compatible use statement to open a database.
use <database>
Creating a Recital Table using Recital SQL
The create table statement creates a new table in the current database, or a new free table if no database is open.
Table column names, data types and constraints can be defined. Table constraints can also be specified.
create table [if not exists] <table> (<column> <datatype> [<column constraint> [...]][,...] [<table constraint> [...]])
Data Types
Recital supports the following data types. For full details, please see Appendix A.
- BIGINT
- BIT
- CHAR
- CURRENCY
- DATE
- DATETIME
- DECIMAL
- DOUBLE
- FLOAT
- GENERAL
- INTEGER
- LOGICAL
- LONG VARCHAR
- LONG VARBINARY
- MEDIUMINT
- MEMO
- NUMERIC
- OBJECT
- REAL
- SHORT
- SMALLINT
- TEXT
- TIME
- TIMESTAMP
- TINYINT
- VARCHAR
Table Constraints
CHECK | ERROR | FOREIGN KEY |
INDEX | KEY | ONCLOSE |
ONDELETE | ONINSERT | ONOPEN |
ONROLLBACK | ONUPDATE | PRIMARY KEY |
UNIQUE |
Column Constraints
AUTO_INCREMENT | AUTOINC | CALCULATED |
CHECK | DEFAULT | DESCRIPTION |
ERROR | FOREIGN KEY | NOCPTRANS |
NOT NULL | NULL | PRIMARY KEY |
RANGE | RECALCULATE | REFERENCES |
SET CHECK | UNIQUE |
Example
create table customer; (account_no char(5) description "Account Code"; default strzero(seqno(),5),; title char(3) description "Personal Title", last_name char(16) description "Customer's Last Name",; first_name char(10) description "Customer's Given Name",; initial char(2) description "Customer's Middle Initial",; street char(25) description "Street Number and Name",; city char(12) description "City",; state char(2) description "State Abbreviation"; check rlookup(customer.state,state); error "Invalid State",; zip char(10) description "Zip Code",; limit decimal(11,2) description "Credit Limit"; recalculate,; balance decimal(11,2) description "Credit Balance"; recalculate,; available decimal(11,2) description "Credit Available"; calculated limit-balance,; notes long varchar description "Customer Notes",; start_date date description "Customer Start Date"; default date(),; onopen "customer")
Loading Metadata from an Array or XML File
The create table statement also allows a new table to be created from metadata stored in an array or an XML file. In addition, data can optionally be loaded from the specified XML file.
create table [if not exists] <table> [from] xml <.xml file> [load] | from array <array>
Examples
// Create table from XML file select * from customer; save as xml cust.xml create table customer2; from xml cust.xml // Create table from array create table newtable from array meta1
Modifying a Table
The alter table statement allows columns in an existing table to be added, modified, renamed or dropped. Both table and column constraints may be added or dropped.
Add column(s), column constraint(s) or table constraint
alter table <table> add [column] (<column> <datatype> [<column constraints>] [,...]) | <table constraint>
Alter column(s), column constraint(s) or table constraint
alter table <table> alter [column] <column> [set default <value> | drop default] | (<column> <datatype> [<column constraint>] [,...]) | constraint (<column> set <column constraint> <value> [,...]) | <table constraint>
Drop column(s), column constraint(s) or table constraint
alter table <table> drop [column] <column> | (<column> [,...]) | constraint (<column> <column constraint> [,...]) | <table constraint>
Add check table constraint
alter table <table> set check <condition> [error <message>]
Rename column
alter table <table> rename (<column>,<new column>)
Examples
// Add new column with column constraints alter table customer add column timeref char(8); check validtime(timeref); error "Not a valid time string" // Alter existing columns to add column constraints alter table customer; alter column available calculated limit-balance; alter column limit recalculate; alter column balance recalculate //or alter table customer; alter (available calculated limit-balance,; limit recalculate,; balance recalculate) // Add new column, add column constraint, // modify column datatype and drop constraints then drop column alter table customer add (timeref char(8)) alter table customer; alter constraint ; (timeref set check validime(timeref); error "Not a valid time string") alter table customer; alter (timeref datetime); drop constraint (timeref check, timeref error) alter table customer drop (timeref) // Add an onupdate table constraint alter table customer; add onupdate "do check_update" // Add and then remove check table constraint alter table customer set check checkit() error "Invalid operation" alter table customer drop check // Rename column alter table customer rename(first_name,forename)
Deleting a Table
The drop table statement is used to delete a table.
drop table <table>
Example
drop table example
Creating an Index
Indexes may be created using the column and table constraints (foreign key, primary key, unique) and the create table or alter table statements or with the create index statement.
create [unique] index <index> on <table> (<column> [asc | desc] [,...]) create [unique] index <index> on <table> (<expression> [asc | desc])
Example
create index lname; on example (lower(last_name))
Rebuilding or Removing an Index
Existing indexes may be rebuilt using the alter index statement.
alter index <index> on <table> rebuild
Example
alter index lname; on example; rebuild
The drop index statement is used to remove an index.
drop index <index> on <table>
Example
drop index lname; on example
Inserting Records into a Recital Table using Recital SQL
Records are added into a Recital table using the insert statement. A single row can be inserted by specifying the row contents in the values clause. Specifying a list of column names determines the columns whose values will be set.
insert into <table> [(<column> [,...])] values (<expr> [,...])
Example
insert into example; (account_no, last_name); values("01001","Brown")
One or more rows can be inserted by specifying a sub-query as the data source. Again the columns to be set can be restricted.
insert into <table> [(<column> [,...])] <sub-query>
Example
insert into example; (account_no, last_name); select * from temp
Data can also be loaded from an array, from a collection of memory variables with the same names as the table columns or from an object with property names that match the table columns.
insert into <table> from array <array> | from memvar | from name <object>
Examples
account_no = "01002" title = "Ms" last_name = "Smith" insert into example; from memvar arr1 = {"01003","Mr","Jones"} insert into example; from array arr1 obj1 = object("account_no" => "01004","title" => "Mrs","last_name" => "Green") insert into example; from name obj1
Selecting Records from a Recital Table Using Recital SQL
The SQL select statement is used to query the data in one or more tables.
The columns to be returned can be defined individually, or the * can be used to return all columns. Constants, functions and aggregate functions (avg(), count(), max(), min(), sum()) can also be included and each element to be returned may be given an alias name.
select * | <column>|<"column"> | <constant> | <function> | <aggregate> [[as] <alias>] [,...] [, *] from <table-definition>
Examples
// Return all columns select * from products // Return specified columns select 'Product Details: ', productid, upper(productname) as Name,; unitprice*unitsinstock as stockholding; from products // Return aggregate function results for the whole table select avg(unitprice) Average, min(unitprice) Minimum,; max(unitprice) Maximum, count(unitprice) Count,; sum(unitprice) Sum; from products
Rows can be restricted in number and duplicates excluded:
select [all | distinct | distinctrow | top <expN> [percent]] <column-definition> from <table-definition>
Examples
// Return one row for each group of rows with an identical contactname column select distinct contactname from customers // Return one row for each group of identical rows select distinctrow contactname from customers // Return the first 10 rows select top 10 * from customers // Return the first 50% of rows select top 50 percent * from customers
Rows can be queried from tables and views. These data sources can be related using inner joins, cross joins or outer joins (left, right and full). Each data source may be given an alias name.
select <column-definition> from <table> | <view> [[as] <t_alias>] [, ...]
select <column-definition> from <table|view> inner join | outer join | left [outer] join | right [outer] join <table2|view2> on <table|view>.<column> = <table2|view2>.<column> [, ...]
select <column-definition> from <table> | <view> cross join | full [outer] join <table2|view2>
- order by
- group by
- having
- into...
- save as...
- limit
- nested selects
- aggregates?
- predicates?
Updating Records in a Recital Table using Recital SQL
The update statement is used to update records. The individual columns and their new values are listed and an optional condition can be specified to restrict the records affected.
update <table> set <column> = <expr> [,...] [where <condition>]
Example
update products; set unitsinstock = unitsinstock + unitsonorder, unitsonorder = 0; where supplierid = 1
Columns can also be updated based on values in an XML file:
update <table> from xml <xml filename>
Example
update products from xml prices.xml
Updates, along with insertions and deletions, can be processed as a transaction within a begin...end transaction block. Nested transactions are supported.
begin transaction [<transaction>] <statements> end transaction [<transaction>]
As are save points:
save transaction <savepoint>
Transactions can be manually committed:
commit [transaction <transaction>]
or rolled back:
rollback [transaction <transaction> | <savepoint>] [to savepoint <savepoint>]
Examples
begin transaction trans1 update orders; set requireddate = date()+30; where orderid = 10248 update order_details set discount = 0.05; where orderid = 10248 begin transaction trans2 update order_details set discount = 0.10; where orderid = 10248 rollback transaction trans1 end transaction begin transaction trans3 update orders; set requireddate = date()+30; where orderid = 10248 update order_details set discount = 0.05; where orderid = 10248 commit transaction trans3 end transaction begin transaction trans4 update orders; set requireddate = date()+30; where orderid = 10248 save transaction parent_updated update order_details set discount = 0.10; where orderid = 10248 rollback transaction parent_updated end transaction begin transaction trans5 update orders; set requireddate = date()+30; where orderid = 10248 savepoint parent_updated update order_details set discount = 0.15; where orderid = 10248 rollback transaction parent_updated end transaction
Deleting Records in a Recital Table using Recital SQL
The delete statement is used to delete records. It executes an immediate physical deletion of the specified records: records are permanently removed from the table and cannot be recalled.
delete from <table> [where <condition>]
Example
delete from currorders where shippeddate < date()
Using a Cursor to Select, Update and Delete Records
A cursor is a temporary selection of rows that can be fetched one at a time and processed further. The update and delete statements can both operate on the most recently fetched row from a cursor.
update <table> set <column> = <expr> [,...] where current of <cursor> delete from <table> where current of <cursor>
A cursor must be declared:
declare <cursor> cursor for select <statement>
then opened:
open <cursor>
then fetched into memory variables or an XML file:
fetch [next | previous | first | last | absolute <row> | current | relative <row>] <cursor> into [<variable> [, ...] | xml <xml filename>]
Once processing is complete, the cursor can be closed:
close <cursor>
then released:
drop cursor <cursor>
Example
open database southwind // Declare the cursor to select records from the orders table declare cursor1; cursor for; select orderid, customerid, orderdate; from orders; where requireddate < date() // Open the cursor open cursor1 // Fetch records one at a time from the cursor and update them fetch cursor1; into m_order, m_customer, m_orderdate do while sqlcode = 0 if not empty(m_order) and empty(shippeddate) update orders; set shippeddate = date(); where current of cursor1 endif fetch cursor1; into m_order, m_customer, m_orderdate enddo // Close the cursor and free up any resources used for the cursor close cursor1 drop cursor cursor1 close databases
Obtaining Information about a Recital Table using SQL
System defined read-only system tables can be queried using the select statment to provide a range of information about Recital tables.
System tables include the following:
- sysbestrowidentifier - description of a table’s optimal set of columns that uniquely identifies a row
- syscolumnconstraints - description of the constraints for a table’s columns
- syscolumnprivileges - description of the access rights for a table’s columns
- syscolumns - description of the table columns available in the catalog
- syscrossreference - description of how one table imports the keys of another table
- sysexportedkeys - description of the foreign key columns that reference the primary key columns
- sysimportedkeys - description of the primary key columns that are referenced by the foreign key
- sysindexinfo - description of a table’s indices and statistics
- sysprimarykeys - description of the primary key columns in the table
- systableconstraints - description of the constraints for each table available in the catalog
- systableprivileges - description of the access rights for each table available in the catalog
- systables - description of the tables available in the catalog
- systabletypes - table types available in the database system
- sysversioncolumns - description of the columns in a table that are automatically updated when any row is updated
Example
open database southwind select * from syscolumns; where table_cat = "southwind" and table_name = "orders" select * from sysindexinfo; where table_cat = "southwind" and table_name = "orders"
Executing Procedures and Prepared Statements using Recital SQL
The create procedure statement allows for the creation of a stored procedure in a Recital database.
create procedure <procedure> as <procedure source code> endcreate
To remove a stored procedure from a database, use the drop procedure statement.
drop procedure <procedure>
Example
open database southwind create procedure creaxml as select orders.orderid, orders.customerid, employees.employeeid,; employees.lastname, employees.firstname, orders.orderdate,; orders.freight, orders.requireddate, orders.shippeddate,; orders.shipvia, orders.shipname, orders.shipaddress, orders.shipcity,; orders.shipregion, orders.shippostalcode, orders.shipcountry,; customers.companyname, customers.address, customers.city,; customers.region, customers.postalcode, customers.country; from orders inner join customers; on customers.customerid = orders.customerid,; orders inner join employees; on orders.employeeid = employees.employeeid; save as xml orderinfo endcreate drop procedure creaxml
The prepare statement is used to set up an SQL statement for subsequent execution.
prepare <statement> from :<variable>
Variables used as parameters in the SQL statement can then be set just before executing it using the execute statement.
execute <statement> using :<variable>[,<variable2>[,...]]
Examples
open database southwind stmtbuf = "select * from example where account_no = ?" prepare mystmt from :stmtbuf m_acc = "00002" execute mystmt using :m_acc m_acc = "00003" execute mystmt using :m_acc stmtbuf = "insert into example (account_no, last_name, forename) values (?,?,?)" prepare mystmt from :stmtbuf buf1 = "99999" buf2 = "Smith" buf3 = "John" execute mystmt using :buf1, :buf2, :buf3 stmtbuf = "delete from example where account_no = ?" prepare mystmt from :stmtbuf buf1 = "99999" execute mystmt using :buf1
The execute immediate statement is used to prepare and execute immediately a delete, insert, or update SQL statement.
m_acc = "00003" execute immediate <statement>
Example
execute immediate select * from example where account_no = m_acc
The sysresultset system table returns the singleton result from any Recital expression.
select set("exclusive") as excl, time() as time from sysresultset