[gimp] Applied changes from SVN version 97 of official version of TinyScheme Updated "hack.txt" with modifi



commit c828ac3de2a5f74c0d9c459a47d4b4d62dab3286
Author: Kevin Cozens <kevin ve3syb ca>
Date:   Sun Apr 14 15:29:06 2013 -0400

    Applied changes from SVN version 97 of official version of TinyScheme
    Updated "hack.txt" with modified version of information from CMarinier.

 plug-ins/script-fu/tinyscheme/hack.txt | 113 ++++++++++++++++-----------------
 1 file changed, 55 insertions(+), 58 deletions(-)
---
diff --git a/plug-ins/script-fu/tinyscheme/hack.txt b/plug-ins/script-fu/tinyscheme/hack.txt
index 5bb8d08..72a3de6 100644
--- a/plug-ins/script-fu/tinyscheme/hack.txt
+++ b/plug-ins/script-fu/tinyscheme/hack.txt
@@ -20,29 +20,29 @@
      
      In the sequel, lines that begin with '>' denote lines to add to the
      code. Lines that begin with '|' are just citations of existing code.
+     Lines that begin with X are deleted.
 
      First of all, we need to assign a typeid to our new type. Typeids
      in TinyScheme are small integers declared in an enum, very close to
-     the top; it begins with T_STRING. Add a new one at the end, say
-     T_MEMBLOCK. There can be at most 31 types, but you don't have to
-     worry about that limit yet.
+     the top of scheme.c; it begins with T_STRING. Add a new one before the
+     end, call it T_MEMBLOCK. Adjust T_LAST_SYSTEM_TYPE.
 
-|    ...
-|      T_PORT,
-|      T_VECTOR,          /* remember to add a comma to the preceding item! */
-|      T_MEMBLOCK
-}     };
+|      T_ENVIRONMENT=14,
+X      T_LAST_SYSTEM_TYPE=14
+>      T_MEMBLOCK=15,
+>      T_LAST_SYSTEM_TYPE=15
+|    };
 
-     Then, some helper macros would be useful. Go to where isstring() and
+     Then, some helper macros would be useful. Go to where is_string() and
      the rest are defined and define:
 
->    int ismemblock(pointer p)      { return (type(p)==T_MEMBLOCK); }
+>    int is_memblock(pointer p)      { return (type(p)==T_MEMBLOCK); }
 
      This actually is a function, because it is meant to be exported by
      scheme.h. If no foreign function will ever manipulate a memory block,
      you can instead define it as a macro
 
->     #define ismemblock(p) (type(p)==T_MEMBLOCK)
+>    #define is_memblock(p) (type(p)==T_MEMBLOCK)
 
      Then we make space for the new type in the main data structure:
      struct cell. As it happens, the _string part of the union _object
@@ -57,7 +57,7 @@
      length. If we couln't reuse existing fields, we could always add other
      alternatives in union _object.
 
-     We then proceed to write the function that actually makes a new block.
+     We then procede to write the function that actually makes a new block.
      For conformance reasons, we name it mk_memblock
 
    static pointer mk_memblock(scheme *sc, int len, char fill) {
@@ -81,62 +81,59 @@
      that staff, function finalize_cell(), currently handling strings only.
 
 |     static void finalize_cell(scheme *sc, pointer a) {
-|          if(isstring(a)) {
+|          if(is_string(a)) {
 |               sc->free(strvalue(a));
-|          }
->          else if(ismemblock(a)) {
->               sc->free(strvalue(x));
->          }
-|     }
+>          else if(is_memblock(a)) {
+>               sc->free(strvalue(a));
+|          } else if(is_port(a)) {
 
-     There are no MEMBLOCK literals, so we don't concern ourselfs with
+     There are no MEMBLOCK literals, so we don't concern ourselves with
      the READER part (yet!). We must cater to the PRINTER, though. We
-     add one case more in printatom().
+     add one case more in atom2str().
 
-|     } else if (iscontinuation(l)) {
-|          p = "#<CONTINUATION>";
->     } else if (ismemblock(l)) {
->          p = "#<MEMORY BLOCK>";
-|     }
+|    } else if (is_foreign(l)) {
+|         p = sc->strbuff;
+|         snprintf(p,STRBUFFSIZE,"#<FOREIGN PROCEDURE %ld>", procnum(l));
+>    } else if (ismemblock(l)) {
+>         p = "#<MEMBLOCK>";
+|    } else if (is_continuation(l)) {
+|         p = "#<CONTINUATION>";
+|    } else {
 
      Whenever a MEMBLOCK is displayed, it will look like that.
+
      Now, we must add the interface functions: constructor, predicate,
-     accessor, modifier. We must in fact create new op-codes for the virtual
-     machine underlying TinyScheme. There is a huge enum with OP_XXX values.
-     That's where the op-codes are declared. For reasons of cohesion, we add
-     the new op-codes right after those for vectors:
-
-|   OP_VECSET,
->   OP_MKBLOCK,
->   OP_MEMBLOCKP,
->   OP_BLOCKLEN,
->   OP_BLOCKREF,
->   OP_BLOCKSET,
-|   OP_NOT,
+     accessor, modifier. We must in fact create new op-codes for the
+     virtual machine underlying TinyScheme. Since version 1.30, TinyScheme
+     uses macros and a single source text to keep the enums and the
+     dispatch table in sync. That's where the op-codes are declared. Note
+     that the opdefines.h file uses unusually long lines to accomodate
+     all the information; adjust your editor to handle this. The file has
+     six columns: A to Z. they contain:
+       - Column A is the name of a routine to handle the scheme function.
+       - Column B is the name the scheme function.
+       - Columns C and D are the minimum and maximum number of arguments
+         that are accepted by the scheme function.
+       - Column E is a set of flags that are used when the interpreter
+         verifies that the passed parameters are of the correct type.
+       - Column F is used to create a set of enums. The enum is used in a
+         switch in the routine listed in column A to get to the code that
+        does the work needed for the scheme function.
+     For reasons of cohesion, we add the new op-codes right after those
+     for vectors:
+
+|   _OP_DEF(opexe_2, "vector-set!",                    3,  3,       TST_VECTOR TST_NATURAL TST_ANY,  
OP_VECSET           )
+>   _OP_DEF(opexe_2, "make-block",                     1,  2,       TST_NATURAL TST_CHAR,            
OP_MKBLOCK          )
+>   _OP_DEF(opexe_2, "block-length",                   1,  1,       T_MEMBLOCK,                      
OP_BLOCKLEN         )
+>   _OP_DEF(opexe_2, "block-ref",                      2,  2,       T_MEMBLOCK TST_NATURAL,          
OP_BLOCKREF         )
+>   _OP_DEF(opexe_2, "block-set!",                     1,  1,       T_MEMBLOCK TST_NATURAL TST_CHAR, 
OP_BLOCKSET         )
+|   _OP_DEF(opexe_3, "not",                            1,  1,       TST_NONE,                        OP_NOT  
            )
 
      We add the predicate along the other predicates:
 
-|   OP_VECTORP,
->   OP_BLOCKP,
-|   OP_EQ,
-
-     Op-codes are really just tags for a huge C switch, only this switch
-     is broke up in a number of different opexe_X functions. The
-     correspondence is made in table "dispatch_table". There, we assign
-     the new op-codes to opexe_2, where the equivalent ones for vectors
-     are situated. We also assign a name for them, and specify the minimum
-     and maximum arity. INF_ARG as a maximum arity means "unlimited".
-
-|     {opexe_2, "vector-set!", 3, 3},    /* OP_VECSET */
->     {opexe_2, "make-block", 1, 2},     /* OP_MKBLOCK */
->     {opexe_2, "block-length", 1, 1},   /* OP_BLOCKLEN */
->     {opexe_2, "block-ref", 2, 2},      /* OP_BLOCKREF */
->     {opexe_2, "block-set!",3 ,3},      /* OP_BLOCKSET */
-
-     The predicate goes with the other predicates, in opexe_3.
-
-|     {opexe_3, "vector?", 1, 1},  /* OP_VECTORP, */
->     {opexe_3, "block?", 1, 1},   /* OP_BLOCKP, */
+|   _OP_DEF(opexe_3, "vector?",                        1,  1,       TST_ANY,                         
OP_VECTORP          )
+>   _OP_DEF(opexe_3, "block?",                         1,  1,       TST_ANY,                         
OP_BLOCKP           )
+|   _OP_DEF(opexe_3, "eq?",                            2,  2,       TST_ANY,                         OP_EQ   
            )
 
      All that remains is to write the actual processing in opexe_2, right
      after OP_VECSET.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]