,---------------.
| Contributions |
`---------------´


,------------------------------------------
| 2024-07-29 13:31:50  kumaym  wrote:
| requestClarification - compile semantics or compile action of literal word
| see: https://forth-standard.org/standard/core/LITERAL#contribution-353
`------------------------------------------
it's said compile semantics ( x -- ) of literal is to compile the rutime semantics of place x on stack ( -- x ) ,  in other words the action literal is supposed to do in compilation state is to compile a literal in the definition of currently defining word, and the literal compiled is obtained from the stack at the moment literal is executed (or compiled since it is a immediate word).

so,  
```
: foo [ 4 ] literal ;
```
is supposed to define a word foo as having 4 in its body, and that's true

```
see foo
: foo 4 ; ok
```

following the compilation of foo step by step we first create a word foo then switch to interpretation mode and push 4 to the stack and then go back to compilation mode to compile literal witch being an immediate word gets executed just to compile 4 (what is on the stack) to the definition of foo and then ; ends the definiton, getting a word containing 4 in its body.

But, then shouldn't be equivalent these two definitions?
```
: foo [ 4 ] literal ;
4 : foo literal ;
```

so strange they're not, the first one is ok but the last one gives this error in gforth:
```
*the terminal*:20:17: error: Control structure mismatch
4 : foo literal >>>;<<<
Backtrace:
kernel/cond.fs:114:26                    0 $6FFFFF810200 throw
glocals.fs:635:5                         1 $6FFFFF821AB0 ?struc
kernel/comp.fs:773:5                     2 $6FFFFF806BF0 ;-hook
```

why?


,---------.
| Replies |
`---------´


,------------------------------------------
| 2024-07-29 17:43:04  ruv  replies:
| requestClarification - compile semantics or compile action of literal word
| see: https://forth-standard.org/standard/core/LITERAL#reply-1267
`------------------------------------------
> shouldn't be equivalent these two definitions?
> ```
> : foo [ 4 ] literal ;
> 4 : foo literal ;
> ```

They are not equivalent because the [word `:`](https://forth-standard.org/standard/core/Colon) (colon) leaves _colon-sys_ on the stack, and the [word `;`](https://forth-standard.org/standard/core/Semi) (semicolon) consumes this _colon-sys_ from the stack.

Therefore, in the latter case, before the compilation semantics for `literal` are performed, the top of the stack is _( 4 colon-sys )_.
See: [3.1.5.1 System-compilation types](https://forth-standard.org/standard/usage#subsubsection.3.1.5.1).

In some Forth systems _colon-sys_ occupies zero items in the data stack. In these systems your two definitions for `foo` are equivalent. But the latter case is not standard compliant.