In Common Lisp I have defind this macro:
(defmacro gen-update-func (func-name table id keyparam)
`(defun ,func-name (db ,id)
(let ((data))
(keyparams ,keyparam)
(clsql:update-records ',table
:database db
:where [= ',id id]
:av-pairs data))))
I call it with:
(gen-update-func add-sp spender spenderid (vorname name))
macroexpand-1 returns this:
(defun add-sp (db spenderid)
(let ((data))
;(keyparams (vorname name))
(clsql-sys:update-records 'spender :database db :where [= 'spenderid id]
:av-pairs data)))
This is expected.
When I call it, it resolves the line ":where [= ',id id]" to:
:WHERE SPENDENDB::[= 'SPENDENDB::SPENDERID SPENDENDB::ID]
This is wrong. I want to insert [= literally without leading SPENDENDB::. This is intended:
:WHERE [= 'SPENDENDB::SPENDERID SPENDENDB::ID]
How can I achieve that?
The syntax [= ... ] is a special SQL syntax from clsql.
In the file where you define your macro, you are using the [= ...] syntax but you are probably not in the appropriate readtable to do so. This is why [= is parsed as a symbol, which is why it is printed with the package prefix later.
According to the CLSQL documentation - locally enable sql reader syntax, you should be able to enable the syntax when defining the macro:
#.(locally-enable-sql-reader-syntax)
(defmacro ...)
#.(restore-sql-reader-syntax-state)
Alternatively, you enable it inside a file:
(in-package :my-package)
(clsql:file-enable-sql-reader-syntax)
When trying to use this syntax, I wrote:
* '[= 'a b]
(CLSQL-SYS:SQL-= 'A B)
As a consequence, you could directly use the equality operator (clsql-sys:sql-= ...) without relying on the helper syntax. Looking at the documentation it is a bit more portable apparently to do it as follows:
`(clsql-sys:sql-operation '= ,a ,b)