[Ecls-list] LOAD-TIME-VALUE and literal object externalisation.

[Samium Gromoff]

At Tue, 22 Apr 2008 21:26:54 +0200,
Juan Jose Garcia-Ripoll wrote:
> 
> On Tue, Apr 22, 2008 at 7:58 PM, Samium Gromoff
> <_deepfire at feelingofgreen.ru> wrote:
> >  It seems I've discovered yet another angle which makes it fail:
> 
> You are trying to create a binary with a hash table. I am not sure
> this can be done or is expected to work by the ANSI standard. Does it
> work in CMUCL/SBCL? It would not be difficult to implement: you just
> have to write a method for MAKE-LOAD-FORM

I just checked, and yes, it works in SBCL -- in fact these test cases
come from minimisation of my spec-driven parser generator I'm trying
to port from SBCL..

Regarding the expectability of that wrt. ANSI, the relevant passages
seem to be 3.2.4.1 Externalizable Objects:

> An object that can be used as a literal object in code to be processed
> by the file compiler is called an externalizable object.
>
> We define that two objects are similar if they satisfy a two-place
> conceptual equivalence predicate (defined below), which is independent
> of the Lisp image so that the two objects in different Lisp images can be
> understood to be equivalent under this predicate. Further, by inspecting
> the definition of this conceptual predicate, the programmer can anticipate
> what aspects of an object are reliably preserved by file compilation.

<snip>

> The set of objects that are externalizable objects are those for which
> the new conceptual term ``similar'' is defined, such that when a compiled
> file is loaded, an object can be constructed which can be shown to be
> similar to the original object which existed at the time the file compiler
> was operating.

Under this definition, it /seems/ that hash-tables fall under externalizable
objects, as 3.2.4.2.2 defines ``similarity'' for them:

> hash-table
>
>  Two hash tables S and C are similar if they meet the following three 
>  requirements:
>
>   1. They both have the same test (e.g., they are both eql hash tables).
>
>   2. There is a unique one-to-one correspondence between the keys of the
>      two hash tables, such that the corresponding keys are similar.
>
>   3. For all keys, the values associated with two corresponding keys are
>      similar.
>
>      If there is more than one possible one-to-one correspondence between
>      the keys of S and C, the consequences are unspecified. A conforming
>      program cannot use a table such as S as an externalizable constant.

But, also, this raises another question -- functions are not defined
to be externalizable, yet we routinely store them in fasls?


Honestly, the line between literal and non-literal objects becomes
somewhat blurry in my head..


regards, Samium Gromoff