1 # Ocean Interpreter - Jamison Creek version
3 Ocean is intended to be a compiled language, so this interpreter is
4 not targeted at being the final product. It is, rather, an intermediate
5 stage and fills that role in two distinct ways.
7 Firstly, it exists as a platform to experiment with the early language
8 design. An interpreter is easy to write and easy to get working, so
9 the barrier for entry is lower if I aim to start with an interpreter.
11 Secondly, the plan for the Ocean compiler is to write it in the
12 [Ocean language](http://ocean-lang.org). To achieve this we naturally
13 need some sort of boot-strap process and this interpreter - written in
14 portable C - will fill that role. It will be used to bootstrap the
17 Two features that are not needed to fill either of these roles are
18 performance and completeness. The interpreter only needs to be fast
19 enough to run small test programs and occasionally to run the compiler
20 on itself. It only needs to be complete enough to test aspects of the
21 design which are developed before the compiler is working, and to run
22 the compiler on itself. Any features not used by the compiler when
23 compiling itself are superfluous. They may be included anyway, but
26 Nonetheless, the interpreter should end up being reasonably complete,
27 and any performance bottlenecks which appear and are easily fixed, will
32 This third version of the interpreter exists to test out some initial
33 ideas relating to types. Particularly it adds arrays (indexed from
34 zero) and simple structures. Basic control flow and variable scoping
35 are already fairly well established, as are basic numerical and
38 Some operators that have only recently been added, and so have not
39 generated all that much experience yet are "and then" and "or else" as
40 short-circuit Boolean operators, and the "if ... else" trinary
41 operator which can select between two expressions based on a third
42 (which appears syntactically in the middle).
44 The "func" clause currently only allows a "main" function to be
45 declared. That will be extended when proper function support is added.
47 An element that is present purely to make a usable language, and
48 without any expectation that they will remain, is the "print" statement
49 which performs simple output.
51 The current scalar types are "number", "Boolean", and "string".
52 Boolean will likely stay in its current form, the other two might, but
53 could just as easily be changed.
57 Versions of the interpreter which obviously do not support a complete
58 language will be named after creeks and streams. This one is Jamison
61 Once we have something reasonably resembling a complete language, the
62 names of rivers will be used.
63 Early versions of the compiler will be named after seas. Major
64 releases of the compiler will be named after oceans. Hopefully I will
65 be finished once I get to the Pacific Ocean release.
69 As well as parsing and executing a program, the interpreter can print
70 out the program from the parsed internal structure. This is useful
71 for validating the parsing.
72 So the main requirements of the interpreter are:
74 - Parse the program, possibly with tracing,
75 - Analyse the parsed program to ensure consistency,
77 - Execute the "main" function in the program, if no parsing or
78 consistency errors were found.
80 This is all performed by a single C program extracted with
83 There will be two formats for printing the program: a default and one
84 that uses bracketing. So a `--bracket` command line option is needed
85 for that. Normally the first code section found is used, however an
86 alternate section can be requested so that a file (such as this one)
87 can contain multiple programs. This is effected with the `--section`
90 This code must be compiled with `-fplan9-extensions` so that anonymous
91 structures can be used.
93 ###### File: oceani.mk
95 myCFLAGS := -Wall -g -fplan9-extensions
96 CFLAGS := $(filter-out $(myCFLAGS),$(CFLAGS)) $(myCFLAGS)
97 myLDLIBS:= libparser.o libscanner.o libmdcode.o -licuuc
98 LDLIBS := $(filter-out $(myLDLIBS),$(LDLIBS)) $(myLDLIBS)
100 all :: $(LDLIBS) oceani
101 oceani.c oceani.h : oceani.mdc parsergen
102 ./parsergen -o oceani --LALR --tag Parser oceani.mdc
103 oceani.mk: oceani.mdc md2c
106 oceani: oceani.o $(LDLIBS)
107 $(CC) $(CFLAGS) -o oceani oceani.o $(LDLIBS)
109 ###### Parser: header
111 struct parse_context;
113 struct parse_context {
114 struct token_config config;
122 #define container_of(ptr, type, member) ({ \
123 const typeof( ((type *)0)->member ) *__mptr = (ptr); \
124 (type *)( (char *)__mptr - offsetof(type,member) );})
126 #define config2context(_conf) container_of(_conf, struct parse_context, \
129 ###### Parser: reduce
130 struct parse_context *c = config2context(config);
138 #include <sys/mman.h>
157 static char Usage[] =
158 "Usage: oceani --trace --print --noexec --brackets --section=SectionName prog.ocn\n";
159 static const struct option long_options[] = {
160 {"trace", 0, NULL, 't'},
161 {"print", 0, NULL, 'p'},
162 {"noexec", 0, NULL, 'n'},
163 {"brackets", 0, NULL, 'b'},
164 {"section", 1, NULL, 's'},
167 const char *options = "tpnbs";
169 static void pr_err(char *msg) // NOTEST
171 fprintf(stderr, "%s\n", msg); // NOTEST
174 int main(int argc, char *argv[])
179 struct section *s = NULL, *ss;
180 char *section = NULL;
181 struct parse_context context = {
183 .ignored = (1 << TK_mark),
184 .number_chars = ".,_+- ",
189 int doprint=0, dotrace=0, doexec=1, brackets=0;
191 while ((opt = getopt_long(argc, argv, options, long_options, NULL))
194 case 't': dotrace=1; break;
195 case 'p': doprint=1; break;
196 case 'n': doexec=0; break;
197 case 'b': brackets=1; break;
198 case 's': section = optarg; break;
199 default: fprintf(stderr, Usage);
203 if (optind >= argc) {
204 fprintf(stderr, "oceani: no input file given\n");
207 fd = open(argv[optind], O_RDONLY);
209 fprintf(stderr, "oceani: cannot open %s\n", argv[optind]);
212 context.file_name = argv[optind];
213 len = lseek(fd, 0, 2);
214 file = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, 0);
215 s = code_extract(file, file+len, pr_err);
217 fprintf(stderr, "oceani: could not find any code in %s\n",
222 ## context initialization
225 for (ss = s; ss; ss = ss->next) {
226 struct text sec = ss->section;
227 if (sec.len == strlen(section) &&
228 strncmp(sec.txt, section, sec.len) == 0)
232 fprintf(stderr, "oceani: cannot find section %s\n",
239 fprintf(stderr, "oceani: no code found in requested section\n"); // NOTEST
240 goto cleanup; // NOTEST
243 parse_oceani(ss->code, &context.config, dotrace ? stderr : NULL);
245 resolve_consts(&context);
246 prepare_types(&context);
247 if (!context.parse_error && !analyse_funcs(&context)) {
248 fprintf(stderr, "oceani: type error in program - not running.\n");
249 context.parse_error += 1;
257 if (doexec && !context.parse_error)
258 interp_main(&context, argc - optind, argv + optind);
261 struct section *t = s->next;
266 // FIXME parser should pop scope even on error
267 while (context.scope_depth > 0)
271 ## free context types
272 ## free context storage
273 exit(context.parse_error ? 1 : 0);
278 The four requirements of parse, analyse, print, interpret apply to
279 each language element individually so that is how most of the code
282 Three of the four are fairly self explanatory. The one that requires
283 a little explanation is the analysis step.
285 The current language design does not require the types of variables to
286 be declared, but they must still have a single type. Different
287 operations impose different requirements on the variables, for example
288 addition requires both arguments to be numeric, and assignment
289 requires the variable on the left to have the same type as the
290 expression on the right.
292 Analysis involves propagating these type requirements around and
293 consequently setting the type of each variable. If any requirements
294 are violated (e.g. a string is compared with a number) or if a
295 variable needs to have two different types, then an error is raised
296 and the program will not run.
298 If the same variable is declared in both branchs of an 'if/else', or
299 in all cases of a 'switch' then the multiple instances may be merged
300 into just one variable if the variable is referenced after the
301 conditional statement. When this happens, the types must naturally be
302 consistent across all the branches. When the variable is not used
303 outside the if, the variables in the different branches are distinct
304 and can be of different types.
306 Undeclared names may only appear in "use" statements and "case" expressions.
307 These names are given a type of "label" and a unique value.
308 This allows them to fill the role of a name in an enumerated type, which
309 is useful for testing the `switch` statement.
311 As we will see, the condition part of a `while` statement can return
312 either a Boolean or some other type. This requires that the expected
313 type that gets passed around comprises a type and a flag to indicate
314 that `Tbool` is also permitted.
316 As there are, as yet, no distinct types that are compatible, there
317 isn't much subtlety in the analysis. When we have distinct number
318 types, this will become more interesting.
322 When analysis discovers an inconsistency it needs to report an error;
323 just refusing to run the code ensures that the error doesn't cascade,
324 but by itself it isn't very useful. A clear understanding of the sort
325 of error message that are useful will help guide the process of
328 At a simplistic level, the only sort of error that type analysis can
329 report is that the type of some construct doesn't match a contextual
330 requirement. For example, in `4 + "hello"` the addition provides a
331 contextual requirement for numbers, but `"hello"` is not a number. In
332 this particular example no further information is needed as the types
333 are obvious from local information. When a variable is involved that
334 isn't the case. It may be helpful to explain why the variable has a
335 particular type, by indicating the location where the type was set,
336 whether by declaration or usage.
338 Using a recursive-descent analysis we can easily detect a problem at
339 multiple locations. In "`hello:= "there"; 4 + hello`" the addition
340 will detect that one argument is not a number and the usage of `hello`
341 will detect that a number was wanted, but not provided. In this
342 (early) version of the language, we will generate error reports at
343 multiple locations, so the use of `hello` will report an error and
344 explain were the value was set, and the addition will report an error
345 and say why numbers are needed. To be able to report locations for
346 errors, each language element will need to record a file location
347 (line and column) and each variable will need to record the language
348 element where its type was set. For now we will assume that each line
349 of an error message indicates one location in the file, and up to 2
350 types. So we provide a `printf`-like function which takes a format, a
351 location (a `struct exec` which has not yet been introduced), and 2
352 types. "`%1`" reports the first type, "`%2`" reports the second. We
353 will need a function to print the location, once we know how that is
354 stored. e As will be explained later, there are sometimes extra rules for
355 type matching and they might affect error messages, we need to pass those
358 As well as type errors, we sometimes need to report problems with
359 tokens, which might be unexpected or might name a type that has not
360 been defined. For these we have `tok_err()` which reports an error
361 with a given token. Each of the error functions sets the flag in the
362 context so indicate that parsing failed.
366 static void fput_loc(struct exec *loc, FILE *f);
367 static void type_err(struct parse_context *c,
368 char *fmt, struct exec *loc,
369 struct type *t1, int rules, struct type *t2);
370 static void tok_err(struct parse_context *c, char *fmt, struct token *t);
372 ###### core functions
374 static void type_err(struct parse_context *c,
375 char *fmt, struct exec *loc,
376 struct type *t1, int rules, struct type *t2)
378 fprintf(stderr, "%s:", c->file_name);
379 fput_loc(loc, stderr);
380 for (; *fmt ; fmt++) {
387 case '%': fputc(*fmt, stderr); break; // NOTEST
388 default: fputc('?', stderr); break; // NOTEST
390 type_print(t1, stderr);
393 type_print(t2, stderr);
402 static void tok_err(struct parse_context *c, char *fmt, struct token *t)
404 fprintf(stderr, "%s:%d:%d: %s: %.*s\n", c->file_name, t->line, t->col, fmt,
405 t->txt.len, t->txt.txt);
409 ## Entities: declared and predeclared.
411 There are various "things" that the language and/or the interpreter
412 needs to know about to parse and execute a program. These include
413 types, variables, values, and executable code. These are all lumped
414 together under the term "entities" (calling them "objects" would be
415 confusing) and introduced here. The following section will present the
416 different specific code elements which comprise or manipulate these
421 Executables can be lots of different things. In many cases an
422 executable is just an operation combined with one or two other
423 executables. This allows for expressions and lists etc. Other times an
424 executable is something quite specific like a constant or variable name.
425 So we define a `struct exec` to be a general executable with a type, and
426 a `struct binode` which is a subclass of `exec`, forms a node in a
427 binary tree, and holds an operation. There will be other subclasses,
428 and to access these we need to be able to `cast` the `exec` into the
429 various other types. The first field in any `struct exec` is the type
430 from the `exec_types` enum.
433 #define cast(structname, pointer) ({ \
434 const typeof( ((struct structname *)0)->type) *__mptr = &(pointer)->type; \
435 if (__mptr && *__mptr != X##structname) abort(); \
436 (struct structname *)( (char *)__mptr);})
438 #define new(structname) ({ \
439 struct structname *__ptr = ((struct structname *)calloc(1,sizeof(struct structname))); \
440 __ptr->type = X##structname; \
441 __ptr->line = -1; __ptr->column = -1; \
444 #define new_pos(structname, token) ({ \
445 struct structname *__ptr = ((struct structname *)calloc(1,sizeof(struct structname))); \
446 __ptr->type = X##structname; \
447 __ptr->line = token.line; __ptr->column = token.col; \
456 enum exec_types type;
465 struct exec *left, *right;
470 static int __fput_loc(struct exec *loc, FILE *f)
474 if (loc->line >= 0) {
475 fprintf(f, "%d:%d: ", loc->line, loc->column);
478 if (loc->type == Xbinode)
479 return __fput_loc(cast(binode,loc)->left, f) ||
480 __fput_loc(cast(binode,loc)->right, f); // NOTEST
483 static void fput_loc(struct exec *loc, FILE *f)
485 if (!__fput_loc(loc, f))
486 fprintf(f, "??:??: "); // NOTEST
489 Each different type of `exec` node needs a number of functions defined,
490 a bit like methods. We must be able to free it, print it, analyse it
491 and execute it. Once we have specific `exec` types we will need to
492 parse them too. Let's take this a bit more slowly.
496 The parser generator requires a `free_foo` function for each struct
497 that stores attributes and they will often be `exec`s and subtypes
498 there-of. So we need `free_exec` which can handle all the subtypes,
499 and we need `free_binode`.
503 static void free_binode(struct binode *b)
512 ###### core functions
513 static void free_exec(struct exec *e)
524 static void free_exec(struct exec *e);
526 ###### free exec cases
527 case Xbinode: free_binode(cast(binode, e)); break;
531 Printing an `exec` requires that we know the current indent level for
532 printing line-oriented components. As will become clear later, we
533 also want to know what sort of bracketing to use.
537 static void do_indent(int i, char *str)
544 ###### core functions
545 static void print_binode(struct binode *b, int indent, int bracket)
549 ## print binode cases
553 static void print_exec(struct exec *e, int indent, int bracket)
559 print_binode(cast(binode, e), indent, bracket); break;
564 do_indent(indent, "/* FREE");
565 for (v = e->to_free; v; v = v->next_free) {
566 printf(" %.*s", v->name->name.len, v->name->name.txt);
567 printf("[%d,%d]", v->scope_start, v->scope_end);
568 if (v->frame_pos >= 0)
569 printf("(%d+%d)", v->frame_pos,
570 v->type ? v->type->size:0);
578 static void print_exec(struct exec *e, int indent, int bracket);
582 As discussed, analysis involves propagating type requirements around the
583 program and looking for errors.
585 So `propagate_types` is passed an expected type (being a `struct type`
586 pointer together with some `val_rules` flags) that the `exec` is
587 expected to return, and returns the type that it does return, either of
588 which can be `NULL` signifying "unknown". A `prop_err` flag set is
589 passed by reference. It has `Efail` set when an error is found, and
590 `Eretry` when the type for some element is set via propagation. If
591 any expression cannot be evaluated immediately, `Enoconst` is set.
592 If the expression can be copied, `Emaycopy` is set.
594 If it remains unchanged at `0`, then no more propagation is needed.
598 enum val_rules {Rnolabel = 1<<0, Rboolok = 1<<1, Rnoconstant = 1<<2};
599 enum prop_err {Efail = 1<<0, Eretry = 1<<1, Enoconst = 1<<2,
604 if (rules & Rnolabel)
605 fputs(" (labels not permitted)", stderr);
609 static struct type *propagate_types(struct exec *prog, struct parse_context *c, enum prop_err *perr,
610 struct type *type, int rules);
611 ###### core functions
613 static struct type *__propagate_types(struct exec *prog, struct parse_context *c, enum prop_err *perr,
614 struct type *type, int rules)
621 switch (prog->type) {
624 struct binode *b = cast(binode, prog);
626 ## propagate binode cases
630 ## propagate exec cases
635 static struct type *propagate_types(struct exec *prog, struct parse_context *c, enum prop_err *perr,
636 struct type *type, int rules)
638 int pre_err = c->parse_error;
639 struct type *ret = __propagate_types(prog, c, perr, type, rules);
641 if (c->parse_error > pre_err)
648 Interpreting an `exec` doesn't require anything but the `exec`. State
649 is stored in variables and each variable will be directly linked from
650 within the `exec` tree. The exception to this is the `main` function
651 which needs to look at command line arguments. This function will be
652 interpreted separately.
654 Each `exec` can return a value combined with a type in `struct lrval`.
655 The type may be `Tnone` but must be non-NULL. Some `exec`s will return
656 the location of a value, which can be updated, in `lval`. Others will
657 set `lval` to NULL indicating that there is a value of appropriate type
661 static struct value interp_exec(struct parse_context *c, struct exec *e,
662 struct type **typeret);
663 ###### core functions
667 struct value rval, *lval;
670 /* If dest is passed, dtype must give the expected type, and
671 * result can go there, in which case type is returned as NULL.
673 static struct lrval _interp_exec(struct parse_context *c, struct exec *e,
674 struct value *dest, struct type *dtype);
676 static struct value interp_exec(struct parse_context *c, struct exec *e,
677 struct type **typeret)
679 struct lrval ret = _interp_exec(c, e, NULL, NULL);
681 if (!ret.type) abort();
685 dup_value(ret.type, ret.lval, &ret.rval);
689 static struct value *linterp_exec(struct parse_context *c, struct exec *e,
690 struct type **typeret)
692 struct lrval ret = _interp_exec(c, e, NULL, NULL);
694 if (!ret.type) abort();
698 free_value(ret.type, &ret.rval);
702 /* dinterp_exec is used when the destination type is certain and
703 * the value has a place to go.
705 static void dinterp_exec(struct parse_context *c, struct exec *e,
706 struct value *dest, struct type *dtype,
709 struct lrval ret = _interp_exec(c, e, dest, dtype);
713 free_value(dtype, dest);
715 dup_value(dtype, ret.lval, dest);
717 memcpy(dest, &ret.rval, dtype->size);
720 static struct lrval _interp_exec(struct parse_context *c, struct exec *e,
721 struct value *dest, struct type *dtype)
723 /* If the result is copied to dest, ret.type is set to NULL */
725 struct value rv = {}, *lrv = NULL;
728 rvtype = ret.type = Tnone;
738 struct binode *b = cast(binode, e);
739 struct value left, right, *lleft;
740 struct type *ltype, *rtype;
741 ltype = rtype = Tnone;
743 ## interp binode cases
745 free_value(ltype, &left);
746 free_value(rtype, &right);
756 ## interp exec cleanup
762 Values come in a wide range of types, with more likely to be added.
763 Each type needs to be able to print its own values (for convenience at
764 least) as well as to compare two values, at least for equality and
765 possibly for order. For now, values might need to be duplicated and
766 freed, though eventually such manipulations will be better integrated
769 Rather than requiring every numeric type to support all numeric
770 operations (add, multiply, etc), we allow types to be able to present
771 as one of a few standard types: integer, float, and fraction. The
772 existence of these conversion functions eventually enable types to
773 determine if they are compatible with other types, though such types
774 have not yet been implemented.
776 Named type are stored in a simple linked list. Objects of each type are
777 "values" which are often passed around by value.
779 There are both explicitly named types, and anonymous types. Anonymous
780 cannot be accessed by name, but are used internally and have a name
781 which might be reported in error messages.
788 ## value union fields
795 struct token first_use;
798 void (*init)(struct type *type, struct value *val);
799 int (*prepare_type)(struct parse_context *c, struct type *type, int parse_time);
800 void (*print)(struct type *type, struct value *val, FILE *f);
801 void (*print_type)(struct type *type, FILE *f);
802 int (*cmp_order)(struct type *t1, struct type *t2,
803 struct value *v1, struct value *v2);
804 int (*cmp_eq)(struct type *t1, struct type *t2,
805 struct value *v1, struct value *v2);
806 void (*dup)(struct type *type, struct value *vold, struct value *vnew);
807 int (*test)(struct type *type, struct value *val);
808 void (*free)(struct type *type, struct value *val);
809 void (*free_type)(struct type *t);
810 long long (*to_int)(struct value *v);
811 double (*to_float)(struct value *v);
812 int (*to_mpq)(mpq_t *q, struct value *v);
821 struct type *typelist;
828 static struct type *find_type(struct parse_context *c, struct text s)
830 struct type *t = c->typelist;
832 while (t && (t->anon ||
833 text_cmp(t->name, s) != 0))
838 static struct type *_add_type(struct parse_context *c, struct text s,
839 struct type *proto, int anon)
843 n = calloc(1, sizeof(*n));
850 n->next = c->typelist;
855 static struct type *add_type(struct parse_context *c, struct text s,
858 return _add_type(c, s, proto, 0);
861 static struct type *add_anon_type(struct parse_context *c,
862 struct type *proto, char *name, ...)
868 vasprintf(&t.txt, name, ap);
870 t.len = strlen(t.txt);
871 return _add_type(c, t, proto, 1);
874 static void free_type(struct type *t)
876 /* The type is always a reference to something in the
877 * context, so we don't need to free anything.
881 static void free_value(struct type *type, struct value *v)
885 memset(v, 0x5a, type->size);
889 static void type_print(struct type *type, FILE *f)
892 fputs("*unknown*type*", f); // NOTEST
893 else if (type->name.len && !type->anon)
894 fprintf(f, "%.*s", type->name.len, type->name.txt);
895 else if (type->print_type)
896 type->print_type(type, f);
897 else if (type->name.len && type->anon)
898 fprintf(f, "\"%.*s\"", type->name.len, type->name.txt);
900 fputs("*invalid*type*", f); // NOTEST
903 static void val_init(struct type *type, struct value *val)
905 if (type && type->init)
906 type->init(type, val);
909 static void dup_value(struct type *type,
910 struct value *vold, struct value *vnew)
912 if (type && type->dup)
913 type->dup(type, vold, vnew);
916 static int value_cmp(struct type *tl, struct type *tr,
917 struct value *left, struct value *right)
919 if (tl && tl->cmp_order)
920 return tl->cmp_order(tl, tr, left, right);
921 if (tl && tl->cmp_eq) // NOTEST
922 return tl->cmp_eq(tl, tr, left, right); // NOTEST
926 static void print_value(struct type *type, struct value *v, FILE *f)
928 if (type && type->print)
929 type->print(type, v, f);
931 fprintf(f, "*Unknown*"); // NOTEST
934 static void prepare_types(struct parse_context *c)
938 enum { none, some, cannot } progress = none;
943 for (t = c->typelist; t; t = t->next) {
945 tok_err(c, "error: type used but not declared",
947 if (t->size == 0 && t->prepare_type) {
948 if (t->prepare_type(c, t, 1))
950 else if (progress == cannot)
951 tok_err(c, "error: type has recursive definition",
961 progress = cannot; break;
963 progress = none; break;
970 static void free_value(struct type *type, struct value *v);
971 static int type_compat(struct type *require, struct type *have, int rules);
972 static void type_print(struct type *type, FILE *f);
973 static void val_init(struct type *type, struct value *v);
974 static void dup_value(struct type *type,
975 struct value *vold, struct value *vnew);
976 static int value_cmp(struct type *tl, struct type *tr,
977 struct value *left, struct value *right);
978 static void print_value(struct type *type, struct value *v, FILE *f);
980 ###### free context types
982 while (context.typelist) {
983 struct type *t = context.typelist;
985 context.typelist = t->next;
993 Type can be specified for local variables, for fields in a structure,
994 for formal parameters to functions, and possibly elsewhere. Different
995 rules may apply in different contexts. As a minimum, a named type may
996 always be used. Currently the type of a formal parameter can be
997 different from types in other contexts, so we have a separate grammar
1003 Type -> IDENTIFIER ${
1004 $0 = find_type(c, $ID.txt);
1006 $0 = add_type(c, $ID.txt, NULL);
1007 $0->first_use = $ID;
1012 FormalType -> Type ${ $0 = $<1; }$
1013 ## formal type grammar
1017 Values of the base types can be numbers, which we represent as
1018 multi-precision fractions, strings, Booleans and labels. When
1019 analysing the program we also need to allow for places where no value
1020 is meaningful (type `Tnone`) and where we don't know what type to
1021 expect yet (type is `NULL`).
1023 Values are never shared, they are always copied when used, and freed
1024 when no longer needed.
1026 When propagating type information around the program, we need to
1027 determine if two types are compatible, where type `NULL` is compatible
1028 with anything. There are two special cases with type compatibility,
1029 both related to the Conditional Statement which will be described
1030 later. In some cases a Boolean can be accepted as well as some other
1031 primary type, and in others any type is acceptable except a label (`Vlabel`).
1032 A separate function encoding these cases will simplify some code later.
1034 ###### type functions
1036 int (*compat)(struct type *this, struct type *other);
1038 ###### ast functions
1040 static int type_compat(struct type *require, struct type *have, int rules)
1042 if ((rules & Rboolok) && have == Tbool)
1044 if ((rules & Rnolabel) && have == Tlabel)
1046 if (!require || !have)
1049 if (require->compat)
1050 return require->compat(require, have);
1052 return require == have;
1057 #include "parse_string.h"
1058 #include "parse_number.h"
1061 myLDLIBS := libnumber.o libstring.o -lgmp
1062 LDLIBS := $(filter-out $(myLDLIBS),$(LDLIBS)) $(myLDLIBS)
1064 ###### type union fields
1065 enum vtype {Vnone, Vstr, Vnum, Vbool, Vlabel} vtype;
1067 ###### value union fields
1073 ###### ast functions
1074 static void _free_value(struct type *type, struct value *v)
1078 switch (type->vtype) {
1080 case Vstr: free(v->str.txt); break;
1081 case Vnum: mpq_clear(v->num); break;
1087 ###### value functions
1089 static void _val_init(struct type *type, struct value *val)
1091 switch(type->vtype) {
1092 case Vnone: // NOTEST
1095 mpq_init(val->num); break;
1097 val->str.txt = malloc(1);
1109 static void _dup_value(struct type *type,
1110 struct value *vold, struct value *vnew)
1112 switch (type->vtype) {
1113 case Vnone: // NOTEST
1116 vnew->label = vold->label;
1119 vnew->bool = vold->bool;
1122 mpq_init(vnew->num);
1123 mpq_set(vnew->num, vold->num);
1126 vnew->str.len = vold->str.len;
1127 vnew->str.txt = malloc(vnew->str.len);
1128 memcpy(vnew->str.txt, vold->str.txt, vnew->str.len);
1133 static int _value_cmp(struct type *tl, struct type *tr,
1134 struct value *left, struct value *right)
1138 return tl - tr; // NOTEST
1139 switch (tl->vtype) {
1140 case Vlabel: cmp = left->label == right->label ? 0 : 1; break;
1141 case Vnum: cmp = mpq_cmp(left->num, right->num); break;
1142 case Vstr: cmp = text_cmp(left->str, right->str); break;
1143 case Vbool: cmp = left->bool - right->bool; break;
1144 case Vnone: cmp = 0; // NOTEST
1149 static void _print_value(struct type *type, struct value *v, FILE *f)
1151 switch (type->vtype) {
1152 case Vnone: // NOTEST
1153 fprintf(f, "*no-value*"); break; // NOTEST
1154 case Vlabel: // NOTEST
1155 fprintf(f, "*label-%p*", v->label); break; // NOTEST
1157 fprintf(f, "%.*s", v->str.len, v->str.txt); break;
1159 fprintf(f, "%s", v->bool ? "True":"False"); break;
1164 mpf_set_q(fl, v->num);
1165 gmp_fprintf(f, "%.10Fg", fl);
1172 static void _free_value(struct type *type, struct value *v);
1174 static int bool_test(struct type *type, struct value *v)
1179 static struct type base_prototype = {
1181 .print = _print_value,
1182 .cmp_order = _value_cmp,
1183 .cmp_eq = _value_cmp,
1185 .free = _free_value,
1188 static struct type *Tbool, *Tstr, *Tnum, *Tnone, *Tlabel;
1190 ###### ast functions
1191 static struct type *add_base_type(struct parse_context *c, char *n,
1192 enum vtype vt, int size)
1194 struct text txt = { n, strlen(n) };
1197 t = add_type(c, txt, &base_prototype);
1200 t->align = size > sizeof(void*) ? sizeof(void*) : size;
1201 if (t->size & (t->align - 1))
1202 t->size = (t->size | (t->align - 1)) + 1; // NOTEST
1206 ###### context initialization
1208 Tbool = add_base_type(&context, "Boolean", Vbool, sizeof(char));
1209 Tbool->test = bool_test;
1210 Tstr = add_base_type(&context, "string", Vstr, sizeof(struct text));
1211 Tnum = add_base_type(&context, "number", Vnum, sizeof(mpq_t));
1212 Tnone = add_base_type(&context, "none", Vnone, 0);
1213 Tlabel = add_base_type(&context, "label", Vlabel, sizeof(void*));
1217 We have already met values as separate objects. When manifest constants
1218 appear in the program text, that must result in an executable which has
1219 a constant value. So the `val` structure embeds a value in an
1232 ###### ast functions
1233 struct val *new_val(struct type *T, struct token tk)
1235 struct val *v = new_pos(val, tk);
1246 $0 = new_val(Tbool, $1);
1250 $0 = new_val(Tbool, $1);
1255 $0 = new_val(Tnum, $1);
1256 if (number_parse($0->val.num, tail, $1.txt) == 0)
1257 mpq_init($0->val.num); // UNTESTED
1259 tok_err(c, "error: unsupported number suffix",
1264 $0 = new_val(Tstr, $1);
1265 string_parse(&$1, '\\', &$0->val.str, tail);
1267 tok_err(c, "error: unsupported string suffix",
1272 $0 = new_val(Tstr, $1);
1273 string_parse(&$1, '\\', &$0->val.str, tail);
1275 tok_err(c, "error: unsupported string suffix",
1279 ###### print exec cases
1282 struct val *v = cast(val, e);
1283 if (v->vtype == Tstr)
1285 // FIXME how to ensure numbers have same precision.
1286 print_value(v->vtype, &v->val, stdout);
1287 if (v->vtype == Tstr)
1292 ###### propagate exec cases
1295 struct val *val = cast(val, prog);
1296 if (!type_compat(type, val->vtype, rules))
1297 type_err(c, "error: expected %1%r found %2",
1298 prog, type, rules, val->vtype);
1302 ###### interp exec cases
1304 rvtype = cast(val, e)->vtype;
1305 dup_value(rvtype, &cast(val, e)->val, &rv);
1308 ###### ast functions
1309 static void free_val(struct val *v)
1312 free_value(v->vtype, &v->val);
1316 ###### free exec cases
1317 case Xval: free_val(cast(val, e)); break;
1319 ###### ast functions
1320 // Move all nodes from 'b' to 'rv', reversing their order.
1321 // In 'b' 'left' is a list, and 'right' is the last node.
1322 // In 'rv', left' is the first node and 'right' is a list.
1323 static struct binode *reorder_bilist(struct binode *b)
1325 struct binode *rv = NULL;
1328 struct exec *t = b->right;
1332 b = cast(binode, b->left);
1342 Variables are scoped named values. We store the names in a linked list
1343 of "bindings" sorted in lexical order, and use sequential search and
1350 struct binding *next; // in lexical order
1354 This linked list is stored in the parse context so that "reduce"
1355 functions can find or add variables, and so the analysis phase can
1356 ensure that every variable gets a type.
1358 ###### parse context
1360 struct binding *varlist; // In lexical order
1362 ###### ast functions
1364 static struct binding *find_binding(struct parse_context *c, struct text s)
1366 struct binding **l = &c->varlist;
1371 (cmp = text_cmp((*l)->name, s)) < 0)
1375 n = calloc(1, sizeof(*n));
1382 Each name can be linked to multiple variables defined in different
1383 scopes. Each scope starts where the name is declared and continues
1384 until the end of the containing code block. Scopes of a given name
1385 cannot nest, so a declaration while a name is in-scope is an error.
1387 ###### binding fields
1388 struct variable *var;
1392 struct variable *previous;
1394 struct binding *name;
1395 struct exec *where_decl;// where name was declared
1396 struct exec *where_set; // where type was set
1400 When a scope closes, the values of the variables might need to be freed.
1401 This happens in the context of some `struct exec` and each `exec` will
1402 need to know which variables need to be freed when it completes.
1405 struct variable *to_free;
1407 ####### variable fields
1408 struct exec *cleanup_exec;
1409 struct variable *next_free;
1411 ####### interp exec cleanup
1414 for (v = e->to_free; v; v = v->next_free) {
1415 struct value *val = var_value(c, v);
1416 free_value(v->type, val);
1420 ###### ast functions
1421 static void variable_unlink_exec(struct variable *v)
1423 struct variable **vp;
1424 if (!v->cleanup_exec)
1426 for (vp = &v->cleanup_exec->to_free;
1427 *vp; vp = &(*vp)->next_free) {
1431 v->cleanup_exec = NULL;
1436 While the naming seems strange, we include local constants in the
1437 definition of variables. A name declared `var := value` can
1438 subsequently be changed, but a name declared `var ::= value` cannot -
1441 ###### variable fields
1444 Scopes in parallel branches can be partially merged. More
1445 specifically, if a given name is declared in both branches of an
1446 if/else then its scope is a candidate for merging. Similarly if
1447 every branch of an exhaustive switch (e.g. has an "else" clause)
1448 declares a given name, then the scopes from the branches are
1449 candidates for merging.
1451 Note that names declared inside a loop (which is only parallel to
1452 itself) are never visible after the loop. Similarly names defined in
1453 scopes which are not parallel, such as those started by `for` and
1454 `switch`, are never visible after the scope. Only variables defined in
1455 both `then` and `else` (including the implicit then after an `if`, and
1456 excluding `then` used with `for`) and in all `case`s and `else` of a
1457 `switch` or `while` can be visible beyond the `if`/`switch`/`while`.
1459 Labels, which are a bit like variables, follow different rules.
1460 Labels are not explicitly declared, but if an undeclared name appears
1461 in a context where a label is legal, that effectively declares the
1462 name as a label. The declaration remains in force (or in scope) at
1463 least to the end of the immediately containing block and conditionally
1464 in any larger containing block which does not declare the name in some
1465 other way. Importantly, the conditional scope extension happens even
1466 if the label is only used in one parallel branch of a conditional --
1467 when used in one branch it is treated as having been declared in all
1470 Merge candidates are tentatively visible beyond the end of the
1471 branching statement which creates them. If the name is used, the
1472 merge is affirmed and they become a single variable visible at the
1473 outer layer. If not - if it is redeclared first - the merge lapses.
1475 To track scopes we have an extra stack, implemented as a linked list,
1476 which roughly parallels the parse stack and which is used exclusively
1477 for scoping. When a new scope is opened, a new frame is pushed and
1478 the child-count of the parent frame is incremented. This child-count
1479 is used to distinguish between the first of a set of parallel scopes,
1480 in which declared variables must not be in scope, and subsequent
1481 branches, whether they may already be conditionally scoped.
1483 We need a total ordering of scopes so we can easily compare to variables
1484 to see if they are concurrently in scope. To achieve this we record a
1485 `scope_count` which is actually a count of both beginnings and endings
1486 of scopes. Then each variable has a record of the scope count where it
1487 enters scope, and where it leaves.
1489 To push a new frame *before* any code in the frame is parsed, we need a
1490 grammar reduction. This is most easily achieved with a grammar
1491 element which derives the empty string, and creates the new scope when
1492 it is recognised. This can be placed, for example, between a keyword
1493 like "if" and the code following it.
1497 struct scope *parent;
1501 ###### parse context
1504 struct scope *scope_stack;
1506 ###### variable fields
1507 int scope_start, scope_end;
1509 ###### ast functions
1510 static void scope_pop(struct parse_context *c)
1512 struct scope *s = c->scope_stack;
1514 c->scope_stack = s->parent;
1516 c->scope_depth -= 1;
1517 c->scope_count += 1;
1520 static void scope_push(struct parse_context *c)
1522 struct scope *s = calloc(1, sizeof(*s));
1524 c->scope_stack->child_count += 1;
1525 s->parent = c->scope_stack;
1527 c->scope_depth += 1;
1528 c->scope_count += 1;
1534 OpenScope -> ${ scope_push(c); }$
1536 Each variable records a scope depth and is in one of four states:
1538 - "in scope". This is the case between the declaration of the
1539 variable and the end of the containing block, and also between
1540 the usage with affirms a merge and the end of that block.
1542 The scope depth is not greater than the current parse context scope
1543 nest depth. When the block of that depth closes, the state will
1544 change. To achieve this, all "in scope" variables are linked
1545 together as a stack in nesting order.
1547 - "pending". The "in scope" block has closed, but other parallel
1548 scopes are still being processed. So far, every parallel block at
1549 the same level that has closed has declared the name.
1551 The scope depth is the depth of the last parallel block that
1552 enclosed the declaration, and that has closed.
1554 - "conditionally in scope". The "in scope" block and all parallel
1555 scopes have closed, and no further mention of the name has been seen.
1556 This state includes a secondary nest depth (`min_depth`) which records
1557 the outermost scope seen since the variable became conditionally in
1558 scope. If a use of the name is found, the variable becomes "in scope"
1559 and that secondary depth becomes the recorded scope depth. If the
1560 name is declared as a new variable, the old variable becomes "out of
1561 scope" and the recorded scope depth stays unchanged.
1563 - "out of scope". The variable is neither in scope nor conditionally
1564 in scope. It is permanently out of scope now and can be removed from
1565 the "in scope" stack. When a variable becomes out-of-scope it is
1566 moved to a separate list (`out_scope`) of variables which have fully
1567 known scope. This will be used at the end of each function to assign
1568 each variable a place in the stack frame.
1570 ###### variable fields
1571 int depth, min_depth;
1572 enum { OutScope, PendingScope, CondScope, InScope } scope;
1573 struct variable *in_scope;
1575 ###### parse context
1577 struct variable *in_scope;
1578 struct variable *out_scope;
1580 All variables with the same name are linked together using the
1581 'previous' link. Those variable that have been affirmatively merged all
1582 have a 'merged' pointer that points to one primary variable - the most
1583 recently declared instance. When merging variables, we need to also
1584 adjust the 'merged' pointer on any other variables that had previously
1585 been merged with the one that will no longer be primary.
1587 A variable that is no longer the most recent instance of a name may
1588 still have "pending" scope, if it might still be merged with most
1589 recent instance. These variables don't really belong in the
1590 "in_scope" list, but are not immediately removed when a new instance
1591 is found. Instead, they are detected and ignored when considering the
1592 list of in_scope names.
1594 The storage of the value of a variable will be described later. For now
1595 we just need to know that when a variable goes out of scope, it might
1596 need to be freed. For this we need to be able to find it, so assume that
1597 `var_value()` will provide that.
1599 ###### variable fields
1600 struct variable *merged;
1602 ###### ast functions
1604 static void variable_merge(struct variable *primary, struct variable *secondary)
1608 primary = primary->merged;
1610 for (v = primary->previous; v; v=v->previous)
1611 if (v == secondary || v == secondary->merged ||
1612 v->merged == secondary ||
1613 v->merged == secondary->merged) {
1614 v->scope = OutScope;
1615 v->merged = primary;
1616 if (v->scope_start < primary->scope_start)
1617 primary->scope_start = v->scope_start;
1618 if (v->scope_end > primary->scope_end)
1619 primary->scope_end = v->scope_end; // NOTEST
1620 variable_unlink_exec(v);
1624 ###### forward decls
1625 static struct value *var_value(struct parse_context *c, struct variable *v);
1627 ###### free global vars
1629 while (context.varlist) {
1630 struct binding *b = context.varlist;
1631 struct variable *v = b->var;
1632 context.varlist = b->next;
1635 struct variable *next = v->previous;
1637 if (v->global && v->frame_pos >= 0) {
1638 free_value(v->type, var_value(&context, v));
1639 if (v->depth == 0 && v->type->free == function_free)
1640 // This is a function constant
1641 free_exec(v->where_decl);
1648 #### Manipulating Bindings
1650 When a name is conditionally visible, a new declaration discards the old
1651 binding - the condition lapses. Similarly when we reach the end of a
1652 function (outermost non-global scope) any conditional scope must lapse.
1653 Conversely a usage of the name affirms the visibility and extends it to
1654 the end of the containing block - i.e. the block that contains both the
1655 original declaration and the latest usage. This is determined from
1656 `min_depth`. When a conditionally visible variable gets affirmed like
1657 this, it is also merged with other conditionally visible variables with
1660 When we parse a variable declaration we either report an error if the
1661 name is currently bound, or create a new variable at the current nest
1662 depth if the name is unbound or bound to a conditionally scoped or
1663 pending-scope variable. If the previous variable was conditionally
1664 scoped, it and its homonyms becomes out-of-scope.
1666 When we parse a variable reference (including non-declarative assignment
1667 "foo = bar") we report an error if the name is not bound or is bound to
1668 a pending-scope variable; update the scope if the name is bound to a
1669 conditionally scoped variable; or just proceed normally if the named
1670 variable is in scope.
1672 When we exit a scope, any variables bound at this level are either
1673 marked out of scope or pending-scoped, depending on whether the scope
1674 was sequential or parallel. Here a "parallel" scope means the "then"
1675 or "else" part of a conditional, or any "case" or "else" branch of a
1676 switch. Other scopes are "sequential".
1678 When exiting a parallel scope we check if there are any variables that
1679 were previously pending and are still visible. If there are, then
1680 they weren't redeclared in the most recent scope, so they cannot be
1681 merged and must become out-of-scope. If it is not the first of
1682 parallel scopes (based on `child_count`), we check that there was a
1683 previous binding that is still pending-scope. If there isn't, the new
1684 variable must now be out-of-scope.
1686 When exiting a sequential scope that immediately enclosed parallel
1687 scopes, we need to resolve any pending-scope variables. If there was
1688 no `else` clause, and we cannot determine that the `switch` was exhaustive,
1689 we need to mark all pending-scope variable as out-of-scope. Otherwise
1690 all pending-scope variables become conditionally scoped.
1693 enum closetype { CloseSequential, CloseFunction, CloseParallel, CloseElse };
1695 ###### ast functions
1697 static struct variable *var_decl(struct parse_context *c, struct text s)
1699 struct binding *b = find_binding(c, s);
1700 struct variable *v = b->var;
1702 switch (v ? v->scope : OutScope) {
1704 /* Caller will report the error */
1708 v && v->scope == CondScope;
1710 v->scope = OutScope;
1714 v = calloc(1, sizeof(*v));
1715 v->previous = b->var;
1719 v->min_depth = v->depth = c->scope_depth;
1721 v->in_scope = c->in_scope;
1722 v->scope_start = c->scope_count;
1728 static struct variable *var_ref(struct parse_context *c, struct text s)
1730 struct binding *b = find_binding(c, s);
1731 struct variable *v = b->var;
1732 struct variable *v2;
1734 switch (v ? v->scope : OutScope) {
1737 /* Caller will report the error */
1740 /* All CondScope variables of this name need to be merged
1741 * and become InScope
1743 v->depth = v->min_depth;
1745 for (v2 = v->previous;
1746 v2 && v2->scope == CondScope;
1748 variable_merge(v, v2);
1756 static int var_refile(struct parse_context *c, struct variable *v)
1758 /* Variable just went out of scope. Add it to the out_scope
1759 * list, sorted by ->scope_start
1761 struct variable **vp = &c->out_scope;
1762 while ((*vp) && (*vp)->scope_start < v->scope_start)
1763 vp = &(*vp)->in_scope;
1769 static void var_block_close(struct parse_context *c, enum closetype ct,
1772 /* Close off all variables that are in_scope.
1773 * Some variables in c->scope may already be not-in-scope,
1774 * such as when a PendingScope variable is hidden by a new
1775 * variable with the same name.
1776 * So we check for v->name->var != v and drop them.
1777 * If we choose to make a variable OutScope, we drop it
1780 struct variable *v, **vp, *v2;
1783 for (vp = &c->in_scope;
1784 (v = *vp) && v->min_depth > c->scope_depth;
1785 (v->scope == OutScope || v->name->var != v)
1786 ? (*vp = v->in_scope, var_refile(c, v))
1787 : ( vp = &v->in_scope, 0)) {
1788 v->min_depth = c->scope_depth;
1789 if (v->name->var != v)
1790 /* This is still in scope, but we haven't just
1794 v->min_depth = c->scope_depth;
1795 if (v->scope == InScope)
1796 v->scope_end = c->scope_count;
1797 if (v->scope == InScope && e && !v->global) {
1798 /* This variable gets cleaned up when 'e' finishes */
1799 variable_unlink_exec(v);
1800 v->cleanup_exec = e;
1801 v->next_free = e->to_free;
1806 case CloseParallel: /* handle PendingScope */
1810 if (c->scope_stack->child_count == 1)
1811 /* first among parallel branches */
1812 v->scope = PendingScope;
1813 else if (v->previous &&
1814 v->previous->scope == PendingScope)
1815 /* all previous branches used name */
1816 v->scope = PendingScope;
1817 else if (v->type == Tlabel)
1818 /* Labels remain pending even when not used */
1819 v->scope = PendingScope; // UNTESTED
1821 v->scope = OutScope;
1822 if (ct == CloseElse) {
1823 /* All Pending variables with this name
1824 * are now Conditional */
1826 v2 && v2->scope == PendingScope;
1828 v2->scope = CondScope;
1832 /* Not possible as it would require
1833 * parallel scope to be nested immediately
1834 * in a parallel scope, and that never
1838 /* Not possible as we already tested for
1845 if (v->scope == CondScope)
1846 /* Condition cannot continue past end of function */
1849 case CloseSequential:
1850 if (v->type == Tlabel)
1851 v->scope = PendingScope;
1854 v->scope = OutScope;
1857 /* There was no 'else', so we can only become
1858 * conditional if we know the cases were exhaustive,
1859 * and that doesn't mean anything yet.
1860 * So only labels become conditional..
1863 v2 && v2->scope == PendingScope;
1865 if (v2->type == Tlabel)
1866 v2->scope = CondScope;
1868 v2->scope = OutScope;
1871 case OutScope: break;
1880 The value of a variable is store separately from the variable, on an
1881 analogue of a stack frame. There are (currently) two frames that can be
1882 active. A global frame which currently only stores constants, and a
1883 stacked frame which stores local variables. Each variable knows if it
1884 is global or not, and what its index into the frame is.
1886 Values in the global frame are known immediately they are relevant, so
1887 the frame needs to be reallocated as it grows so it can store those
1888 values. The local frame doesn't get values until the interpreted phase
1889 is started, so there is no need to allocate until the size is known.
1891 We initialize the `frame_pos` to an impossible value, so that we can
1892 tell if it was set or not later.
1894 ###### variable fields
1898 ###### variable init
1901 ###### parse context
1903 short global_size, global_alloc;
1905 void *global, *local;
1907 ###### forward decls
1908 static struct value *global_alloc(struct parse_context *c, struct type *t,
1909 struct variable *v, struct value *init);
1911 ###### ast functions
1913 static struct value *var_value(struct parse_context *c, struct variable *v)
1916 if (!c->local || !v->type)
1917 return NULL; // UNTESTED
1918 if (v->frame_pos + v->type->size > c->local_size) {
1919 printf("INVALID frame_pos\n"); // NOTEST
1922 return c->local + v->frame_pos;
1924 if (c->global_size > c->global_alloc) {
1925 int old = c->global_alloc;
1926 c->global_alloc = (c->global_size | 1023) + 1024;
1927 c->global = realloc(c->global, c->global_alloc);
1928 memset(c->global + old, 0, c->global_alloc - old);
1930 return c->global + v->frame_pos;
1933 static struct value *global_alloc(struct parse_context *c, struct type *t,
1934 struct variable *v, struct value *init)
1937 struct variable scratch;
1939 if (t->prepare_type)
1940 t->prepare_type(c, t, 1); // NOTEST
1942 if (c->global_size & (t->align - 1))
1943 c->global_size = (c->global_size + t->align) & ~(t->align-1); // NOTEST
1948 v->frame_pos = c->global_size;
1950 c->global_size += v->type->size;
1951 ret = var_value(c, v);
1953 memcpy(ret, init, t->size);
1959 As global values are found -- struct field initializers, labels etc --
1960 `global_alloc()` is called to record the value in the global frame.
1962 When the program is fully parsed, each function is analysed, we need to
1963 walk the list of variables local to that function and assign them an
1964 offset in the stack frame. For this we have `scope_finalize()`.
1966 We keep the stack from dense by re-using space for between variables
1967 that are not in scope at the same time. The `out_scope` list is sorted
1968 by `scope_start` and as we process a varible, we move it to an FIFO
1969 stack. For each variable we consider, we first discard any from the
1970 stack anything that went out of scope before the new variable came in.
1971 Then we place the new variable just after the one at the top of the
1974 ###### ast functions
1976 static void scope_finalize(struct parse_context *c, struct type *ft)
1978 int size = ft->function.local_size;
1979 struct variable *next = ft->function.scope;
1980 struct variable *done = NULL;
1983 struct variable *v = next;
1984 struct type *t = v->type;
1991 if (v->frame_pos >= 0)
1993 while (done && done->scope_end < v->scope_start)
1994 done = done->in_scope;
1996 pos = done->frame_pos + done->type->size;
1998 pos = ft->function.local_size;
1999 if (pos & (t->align - 1))
2000 pos = (pos + t->align) & ~(t->align-1);
2002 if (size < pos + v->type->size)
2003 size = pos + v->type->size;
2007 c->out_scope = NULL;
2008 ft->function.local_size = size;
2011 ###### free context storage
2012 free(context.global);
2014 #### Variables as executables
2016 Just as we used a `val` to wrap a value into an `exec`, we similarly
2017 need a `var` to wrap a `variable` into an exec. While each `val`
2018 contained a copy of the value, each `var` holds a link to the variable
2019 because it really is the same variable no matter where it appears.
2020 When a variable is used, we need to remember to follow the `->merged`
2021 link to find the primary instance.
2023 When a variable is declared, it may or may not be given an explicit
2024 type. We need to record which so that we can report the parsed code
2033 struct variable *var;
2036 ###### variable fields
2044 VariableDecl -> IDENTIFIER : ${ {
2045 struct variable *v = var_decl(c, $1.txt);
2046 $0 = new_pos(var, $1);
2051 v = var_ref(c, $1.txt);
2053 type_err(c, "error: variable '%v' redeclared",
2055 type_err(c, "info: this is where '%v' was first declared",
2056 v->where_decl, NULL, 0, NULL);
2059 | IDENTIFIER :: ${ {
2060 struct variable *v = var_decl(c, $1.txt);
2061 $0 = new_pos(var, $1);
2067 v = var_ref(c, $1.txt);
2069 type_err(c, "error: variable '%v' redeclared",
2071 type_err(c, "info: this is where '%v' was first declared",
2072 v->where_decl, NULL, 0, NULL);
2075 | IDENTIFIER : Type ${ {
2076 struct variable *v = var_decl(c, $1.txt);
2077 $0 = new_pos(var, $1);
2083 v->explicit_type = 1;
2085 v = var_ref(c, $1.txt);
2087 type_err(c, "error: variable '%v' redeclared",
2089 type_err(c, "info: this is where '%v' was first declared",
2090 v->where_decl, NULL, 0, NULL);
2093 | IDENTIFIER :: Type ${ {
2094 struct variable *v = var_decl(c, $1.txt);
2095 $0 = new_pos(var, $1);
2102 v->explicit_type = 1;
2104 v = var_ref(c, $1.txt);
2106 type_err(c, "error: variable '%v' redeclared",
2108 type_err(c, "info: this is where '%v' was first declared",
2109 v->where_decl, NULL, 0, NULL);
2114 Variable -> IDENTIFIER ${ {
2115 struct variable *v = var_ref(c, $1.txt);
2116 $0 = new_pos(var, $1);
2118 /* This might be a global const or a label
2119 * Allocate a var with impossible type Tnone,
2120 * which will be adjusted when we find out what it is,
2121 * or will trigger an error.
2123 v = var_decl(c, $1.txt);
2130 cast(var, $0)->var = v;
2133 ###### print exec cases
2136 struct var *v = cast(var, e);
2138 struct binding *b = v->var->name;
2139 printf("%.*s", b->name.len, b->name.txt);
2146 if (loc && loc->type == Xvar) {
2147 struct var *v = cast(var, loc);
2149 struct binding *b = v->var->name;
2150 fprintf(stderr, "%.*s", b->name.len, b->name.txt);
2152 fputs("???", stderr); // NOTEST
2154 fputs("NOTVAR", stderr); // NOTEST
2157 ###### propagate exec cases
2161 struct var *var = cast(var, prog);
2162 struct variable *v = var->var;
2164 type_err(c, "%d:BUG: no variable!!", prog, NULL, 0, NULL); // NOTEST
2165 return Tnone; // NOTEST
2168 if (v->constant && (rules & Rnoconstant)) {
2169 type_err(c, "error: Cannot assign to a constant: %v",
2170 prog, NULL, 0, NULL);
2171 type_err(c, "info: name was defined as a constant here",
2172 v->where_decl, NULL, 0, NULL);
2175 if (v->type == Tnone && v->where_decl == prog)
2176 type_err(c, "error: variable used but not declared: %v",
2177 prog, NULL, 0, NULL);
2178 if (v->type == NULL) {
2179 if (type && !(*perr & Efail)) {
2181 v->where_set = prog;
2184 } else if (!type_compat(type, v->type, rules)) {
2185 type_err(c, "error: expected %1%r but variable '%v' is %2", prog,
2186 type, rules, v->type);
2187 type_err(c, "info: this is where '%v' was set to %1", v->where_set,
2188 v->type, rules, NULL);
2190 if (!v->global || v->frame_pos < 0)
2197 ###### interp exec cases
2200 struct var *var = cast(var, e);
2201 struct variable *v = var->var;
2204 lrv = var_value(c, v);
2209 ###### ast functions
2211 static void free_var(struct var *v)
2216 ###### free exec cases
2217 case Xvar: free_var(cast(var, e)); break;
2222 Now that we have the shape of the interpreter in place we can add some
2223 complex types and connected them in to the data structures and the
2224 different phases of parse, analyse, print, interpret.
2226 Being "complex" the language will naturally have syntax to access
2227 specifics of objects of these types. These will fit into the grammar as
2228 "Terms" which are the things that are combined with various operators to
2229 form "Expression". Where a Term is formed by some operation on another
2230 Term, the subordinate Term will always come first, so for example a
2231 member of an array will be expressed as the Term for the array followed
2232 by an index in square brackets. The strict rule of using postfix
2233 operations makes precedence irrelevant within terms. To provide a place
2234 to put the grammar for each terms of each type, we will start out by
2235 introducing the "Term" grammar production, with contains at least a
2236 simple "Value" (to be explained later).
2240 Term -> Value ${ $0 = $<1; }$
2241 | Variable ${ $0 = $<1; }$
2244 Thus far the complex types we have are arrays and structs.
2248 Arrays can be declared by giving a size and a type, as `[size]type' so
2249 `freq:[26]number` declares `freq` to be an array of 26 numbers. The
2250 size can be either a literal number, or a named constant. Some day an
2251 arbitrary expression will be supported.
2253 As a formal parameter to a function, the array can be declared with a
2254 new variable as the size: `name:[size::number]string`. The `size`
2255 variable is set to the size of the array and must be a constant. As
2256 `number` is the only supported type, it can be left out:
2257 `name:[size::]string`.
2259 Arrays cannot be assigned. When pointers are introduced we will also
2260 introduce array slices which can refer to part or all of an array -
2261 the assignment syntax will create a slice. For now, an array can only
2262 ever be referenced by the name it is declared with. It is likely that
2263 a "`copy`" primitive will eventually be define which can be used to
2264 make a copy of an array with controllable recursive depth.
2266 For now we have two sorts of array, those with fixed size either because
2267 it is given as a literal number or because it is a struct member (which
2268 cannot have a runtime-changing size), and those with a size that is
2269 determined at runtime - local variables with a const size. The former
2270 have their size calculated at parse time, the latter at run time.
2272 For the latter type, the `size` field of the type is the size of a
2273 pointer, and the array is reallocated every time it comes into scope.
2275 We differentiate struct fields with a const size from local variables
2276 with a const size by whether they are prepared at parse time or not.
2278 ###### type union fields
2281 int unspec; // size is unspecified - vsize must be set.
2284 struct variable *vsize;
2285 struct type *member;
2288 ###### value union fields
2289 void *array; // used if not static_size
2291 ###### value functions
2293 static int array_prepare_type(struct parse_context *c, struct type *type,
2296 struct value *vsize;
2298 if (type->array.static_size)
2299 return 1; // UNTESTED
2300 if (type->array.unspec && parse_time)
2301 return 1; // UNTESTED
2302 if (parse_time && type->array.vsize && !type->array.vsize->global)
2303 return 1; // UNTESTED
2305 if (type->array.vsize) {
2306 vsize = var_value(c, type->array.vsize);
2308 return 1; // UNTESTED
2310 mpz_tdiv_q(q, mpq_numref(vsize->num), mpq_denref(vsize->num));
2311 type->array.size = mpz_get_si(q);
2316 if (type->array.member->size <= 0)
2317 return 0; // UNTESTED
2319 type->array.static_size = 1;
2320 type->size = type->array.size * type->array.member->size;
2321 type->align = type->array.member->align;
2326 static void array_init(struct type *type, struct value *val)
2329 void *ptr = val->ptr;
2333 if (!type->array.static_size) {
2334 val->array = calloc(type->array.size,
2335 type->array.member->size);
2338 for (i = 0; i < type->array.size; i++) {
2340 v = (void*)ptr + i * type->array.member->size;
2341 val_init(type->array.member, v);
2345 static void array_free(struct type *type, struct value *val)
2348 void *ptr = val->ptr;
2350 if (!type->array.static_size)
2352 for (i = 0; i < type->array.size; i++) {
2354 v = (void*)ptr + i * type->array.member->size;
2355 free_value(type->array.member, v);
2357 if (!type->array.static_size)
2361 static int array_compat(struct type *require, struct type *have)
2363 if (have->compat != require->compat)
2365 /* Both are arrays, so we can look at details */
2366 if (!type_compat(require->array.member, have->array.member, 0))
2368 if (have->array.unspec && require->array.unspec) {
2369 if (have->array.vsize && require->array.vsize &&
2370 have->array.vsize != require->array.vsize) // UNTESTED
2371 /* sizes might not be the same */
2372 return 0; // UNTESTED
2375 if (have->array.unspec || require->array.unspec)
2376 return 1; // UNTESTED
2377 if (require->array.vsize == NULL && have->array.vsize == NULL)
2378 return require->array.size == have->array.size;
2380 return require->array.vsize == have->array.vsize; // UNTESTED
2383 static void array_print_type(struct type *type, FILE *f)
2386 if (type->array.vsize) {
2387 struct binding *b = type->array.vsize->name;
2388 fprintf(f, "%.*s%s]", b->name.len, b->name.txt,
2389 type->array.unspec ? "::" : "");
2390 } else if (type->array.size)
2391 fprintf(f, "%d]", type->array.size);
2394 type_print(type->array.member, f);
2397 static struct type array_prototype = {
2399 .prepare_type = array_prepare_type,
2400 .print_type = array_print_type,
2401 .compat = array_compat,
2403 .size = sizeof(void*),
2404 .align = sizeof(void*),
2407 ###### declare terminals
2412 | [ NUMBER ] Type ${ {
2418 if (number_parse(num, tail, $2.txt) == 0)
2419 tok_err(c, "error: unrecognised number", &$2);
2421 tok_err(c, "error: unsupported number suffix", &$2);
2424 elements = mpz_get_ui(mpq_numref(num));
2425 if (mpz_cmp_ui(mpq_denref(num), 1) != 0) {
2426 tok_err(c, "error: array size must be an integer",
2428 } else if (mpz_cmp_ui(mpq_numref(num), 1UL << 30) >= 0)
2429 tok_err(c, "error: array size is too large",
2434 $0 = t = add_anon_type(c, &array_prototype, "array[%d]", elements );
2435 t->array.size = elements;
2436 t->array.member = $<4;
2437 t->array.vsize = NULL;
2440 | [ IDENTIFIER ] Type ${ {
2441 struct variable *v = var_ref(c, $2.txt);
2444 tok_err(c, "error: name undeclared", &$2);
2445 else if (!v->constant)
2446 tok_err(c, "error: array size must be a constant", &$2);
2448 $0 = add_anon_type(c, &array_prototype, "array[%.*s]", $2.txt.len, $2.txt.txt);
2449 $0->array.member = $<4;
2451 $0->array.vsize = v;
2456 OptType -> Type ${ $0 = $<1; }$
2459 ###### formal type grammar
2461 | [ IDENTIFIER :: OptType ] Type ${ {
2462 struct variable *v = var_decl(c, $ID.txt);
2468 $0 = add_anon_type(c, &array_prototype, "array[var]");
2469 $0->array.member = $<6;
2471 $0->array.unspec = 1;
2472 $0->array.vsize = v;
2480 | Term [ Expression ] ${ {
2481 struct binode *b = new(binode);
2488 ###### print binode cases
2490 print_exec(b->left, -1, bracket);
2492 print_exec(b->right, -1, bracket);
2496 ###### propagate binode cases
2498 /* left must be an array, right must be a number,
2499 * result is the member type of the array
2501 propagate_types(b->right, c, perr, Tnum, 0);
2502 t = propagate_types(b->left, c, perr, NULL, rules & Rnoconstant);
2503 if (!t || t->compat != array_compat) {
2504 type_err(c, "error: %1 cannot be indexed", prog, t, 0, NULL);
2507 if (!type_compat(type, t->array.member, rules)) {
2508 type_err(c, "error: have %1 but need %2", prog,
2509 t->array.member, rules, type);
2511 return t->array.member;
2515 ###### interp binode cases
2521 lleft = linterp_exec(c, b->left, <ype);
2522 right = interp_exec(c, b->right, &rtype);
2524 mpz_tdiv_q(q, mpq_numref(right.num), mpq_denref(right.num));
2528 if (ltype->array.static_size)
2531 ptr = *(void**)lleft;
2532 rvtype = ltype->array.member;
2533 if (i >= 0 && i < ltype->array.size)
2534 lrv = ptr + i * rvtype->size;
2536 val_init(ltype->array.member, &rv); // UNSAFE
2543 A `struct` is a data-type that contains one or more other data-types.
2544 It differs from an array in that each member can be of a different
2545 type, and they are accessed by name rather than by number. Thus you
2546 cannot choose an element by calculation, you need to know what you
2549 The language makes no promises about how a given structure will be
2550 stored in memory - it is free to rearrange fields to suit whatever
2551 criteria seems important.
2553 Structs are declared separately from program code - they cannot be
2554 declared in-line in a variable declaration like arrays can. A struct
2555 is given a name and this name is used to identify the type - the name
2556 is not prefixed by the word `struct` as it would be in C.
2558 Structs are only treated as the same if they have the same name.
2559 Simply having the same fields in the same order is not enough. This
2560 might change once we can create structure initializers from a list of
2563 Each component datum is identified much like a variable is declared,
2564 with a name, one or two colons, and a type. The type cannot be omitted
2565 as there is no opportunity to deduce the type from usage. An initial
2566 value can be given following an equals sign, so
2568 ##### Example: a struct type
2574 would declare a type called "complex" which has two number fields,
2575 each initialised to zero.
2577 Struct will need to be declared separately from the code that uses
2578 them, so we will need to be able to print out the declaration of a
2579 struct when reprinting the whole program. So a `print_type_decl` type
2580 function will be needed.
2582 ###### type union fields
2591 } *fields; // This is created when field_list is analysed.
2593 struct fieldlist *prev;
2596 } *field_list; // This is created during parsing
2599 ###### type functions
2600 void (*print_type_decl)(struct type *type, FILE *f);
2602 ###### value functions
2604 static void structure_init(struct type *type, struct value *val)
2608 for (i = 0; i < type->structure.nfields; i++) {
2610 v = (void*) val->ptr + type->structure.fields[i].offset;
2611 if (type->structure.fields[i].init)
2612 dup_value(type->structure.fields[i].type,
2613 type->structure.fields[i].init,
2616 val_init(type->structure.fields[i].type, v);
2620 static void structure_free(struct type *type, struct value *val)
2624 for (i = 0; i < type->structure.nfields; i++) {
2626 v = (void*)val->ptr + type->structure.fields[i].offset;
2627 free_value(type->structure.fields[i].type, v);
2631 static void free_fieldlist(struct fieldlist *f)
2635 free_fieldlist(f->prev);
2640 static void structure_free_type(struct type *t)
2643 for (i = 0; i < t->structure.nfields; i++)
2644 if (t->structure.fields[i].init) {
2645 free_value(t->structure.fields[i].type,
2646 t->structure.fields[i].init);
2648 free(t->structure.fields);
2649 free_fieldlist(t->structure.field_list);
2652 static int structure_prepare_type(struct parse_context *c,
2653 struct type *t, int parse_time)
2656 struct fieldlist *f;
2658 if (!parse_time || t->structure.fields)
2661 for (f = t->structure.field_list; f; f=f->prev) {
2665 if (f->f.type->size <= 0)
2667 if (f->f.type->prepare_type)
2668 f->f.type->prepare_type(c, f->f.type, parse_time);
2670 if (f->init == NULL)
2674 propagate_types(f->init, c, &perr, f->f.type, 0);
2675 } while (perr & Eretry);
2677 c->parse_error += 1; // NOTEST
2680 t->structure.nfields = cnt;
2681 t->structure.fields = calloc(cnt, sizeof(struct field));
2682 f = t->structure.field_list;
2684 int a = f->f.type->align;
2686 t->structure.fields[cnt] = f->f;
2687 if (t->size & (a-1))
2688 t->size = (t->size | (a-1)) + 1;
2689 t->structure.fields[cnt].offset = t->size;
2690 t->size += ((f->f.type->size - 1) | (a-1)) + 1;
2694 if (f->init && !c->parse_error) {
2695 struct value vl = interp_exec(c, f->init, NULL);
2696 t->structure.fields[cnt].init =
2697 global_alloc(c, f->f.type, NULL, &vl);
2705 static struct type structure_prototype = {
2706 .init = structure_init,
2707 .free = structure_free,
2708 .free_type = structure_free_type,
2709 .print_type_decl = structure_print_type,
2710 .prepare_type = structure_prepare_type,
2724 ###### free exec cases
2726 free_exec(cast(fieldref, e)->left);
2730 ###### declare terminals
2735 | Term . IDENTIFIER ${ {
2736 struct fieldref *fr = new_pos(fieldref, $2);
2743 ###### print exec cases
2747 struct fieldref *f = cast(fieldref, e);
2748 print_exec(f->left, -1, bracket);
2749 printf(".%.*s", f->name.len, f->name.txt);
2753 ###### ast functions
2754 static int find_struct_index(struct type *type, struct text field)
2757 for (i = 0; i < type->structure.nfields; i++)
2758 if (text_cmp(type->structure.fields[i].name, field) == 0)
2763 ###### propagate exec cases
2767 struct fieldref *f = cast(fieldref, prog);
2768 struct type *st = propagate_types(f->left, c, perr, NULL, 0);
2771 type_err(c, "error: unknown type for field access", f->left, // UNTESTED
2773 else if (st->init != structure_init)
2774 type_err(c, "error: field reference attempted on %1, not a struct",
2775 f->left, st, 0, NULL);
2776 else if (f->index == -2) {
2777 f->index = find_struct_index(st, f->name);
2779 type_err(c, "error: cannot find requested field in %1",
2780 f->left, st, 0, NULL);
2782 if (f->index >= 0) {
2783 struct type *ft = st->structure.fields[f->index].type;
2784 if (!type_compat(type, ft, rules))
2785 type_err(c, "error: have %1 but need %2", prog,
2792 ###### interp exec cases
2795 struct fieldref *f = cast(fieldref, e);
2797 struct value *lleft = linterp_exec(c, f->left, <ype);
2798 lrv = (void*)lleft->ptr + ltype->structure.fields[f->index].offset;
2799 rvtype = ltype->structure.fields[f->index].type;
2803 ###### top level grammar
2804 DeclareStruct -> struct IDENTIFIER FieldBlock Newlines ${ {
2806 t = find_type(c, $ID.txt);
2808 t = add_type(c, $ID.txt, &structure_prototype);
2809 else if (t->size >= 0) {
2810 tok_err(c, "error: type already declared", &$ID);
2811 tok_err(c, "info: this is location of declartion", &t->first_use);
2812 /* Create a new one - duplicate */
2813 t = add_type(c, $ID.txt, &structure_prototype);
2815 struct type tmp = *t;
2816 *t = structure_prototype;
2820 t->structure.field_list = $<FB;
2825 FieldBlock -> { IN OptNL FieldLines OUT OptNL } ${ $0 = $<FL; }$
2826 | { SimpleFieldList } ${ $0 = $<SFL; }$
2827 | IN OptNL FieldLines OUT ${ $0 = $<FL; }$
2828 | SimpleFieldList EOL ${ $0 = $<SFL; }$
2830 FieldLines -> SimpleFieldList Newlines ${ $0 = $<SFL; }$
2831 | FieldLines SimpleFieldList Newlines ${
2836 SimpleFieldList -> Field ${ $0 = $<F; }$
2837 | SimpleFieldList ; Field ${
2841 | SimpleFieldList ; ${
2844 | ERROR ${ tok_err(c, "Syntax error in struct field", &$1); }$
2846 Field -> IDENTIFIER : Type = Expression ${ {
2847 $0 = calloc(1, sizeof(struct fieldlist));
2848 $0->f.name = $ID.txt;
2849 $0->f.type = $<Type;
2853 | IDENTIFIER : Type ${
2854 $0 = calloc(1, sizeof(struct fieldlist));
2855 $0->f.name = $ID.txt;
2856 $0->f.type = $<Type;
2859 ###### forward decls
2860 static void structure_print_type(struct type *t, FILE *f);
2862 ###### value functions
2863 static void structure_print_type(struct type *t, FILE *f)
2867 fprintf(f, "struct %.*s\n", t->name.len, t->name.txt);
2869 for (i = 0; i < t->structure.nfields; i++) {
2870 struct field *fl = t->structure.fields + i;
2871 fprintf(f, " %.*s : ", fl->name.len, fl->name.txt);
2872 type_print(fl->type, f);
2873 if (fl->type->print && fl->init) {
2875 if (fl->type == Tstr)
2876 fprintf(f, "\""); // UNTESTED
2877 print_value(fl->type, fl->init, f);
2878 if (fl->type == Tstr)
2879 fprintf(f, "\""); // UNTESTED
2885 ###### print type decls
2890 while (target != 0) {
2892 for (t = context.typelist; t ; t=t->next)
2893 if (!t->anon && t->print_type_decl &&
2903 t->print_type_decl(t, stdout);
2911 A function is a chunk of code which can be passed parameters and can
2912 return results. Each function has a type which includes the set of
2913 parameters and the return value. As yet these types cannot be declared
2914 separately from the function itself.
2916 The parameters can be specified either in parentheses as a ';' separated
2919 ##### Example: function 1
2921 func main(av:[ac::number]string; env:[envc::number]string)
2924 or as an indented list of one parameter per line (though each line can
2925 be a ';' separated list)
2927 ##### Example: function 2
2930 argv:[argc::number]string
2931 env:[envc::number]string
2935 In the first case a return type can follow the parentheses after a colon,
2936 in the second it is given on a line starting with the word `return`.
2938 ##### Example: functions that return
2940 func add(a:number; b:number): number
2950 Rather than returning a type, the function can specify a set of local
2951 variables to return as a struct. The values of these variables when the
2952 function exits will be provided to the caller. For this the return type
2953 is replaced with a block of result declarations, either in parentheses
2954 or bracketed by `return` and `do`.
2956 ##### Example: functions returning multiple variables
2958 func to_cartesian(rho:number; theta:number):(x:number; y:number)
2971 For constructing the lists we use a `List` binode, which will be
2972 further detailed when Expression Lists are introduced.
2974 ###### type union fields
2977 struct binode *params;
2978 struct type *return_type;
2979 struct variable *scope;
2980 int inline_result; // return value is at start of 'local'
2984 ###### value union fields
2985 struct exec *function;
2987 ###### type functions
2988 void (*check_args)(struct parse_context *c, enum prop_err *perr,
2989 struct type *require, struct exec *args);
2991 ###### value functions
2993 static void function_free(struct type *type, struct value *val)
2995 free_exec(val->function);
2996 val->function = NULL;
2999 static int function_compat(struct type *require, struct type *have)
3001 // FIXME can I do anything here yet?
3005 static void function_check_args(struct parse_context *c, enum prop_err *perr,
3006 struct type *require, struct exec *args)
3008 /* This should be 'compat', but we don't have a 'tuple' type to
3009 * hold the type of 'args'
3011 struct binode *arg = cast(binode, args);
3012 struct binode *param = require->function.params;
3015 struct var *pv = cast(var, param->left);
3017 type_err(c, "error: insufficient arguments to function.",
3018 args, NULL, 0, NULL);
3022 propagate_types(arg->left, c, perr, pv->var->type, 0);
3023 param = cast(binode, param->right);
3024 arg = cast(binode, arg->right);
3027 type_err(c, "error: too many arguments to function.",
3028 args, NULL, 0, NULL);
3031 static void function_print(struct type *type, struct value *val, FILE *f)
3033 print_exec(val->function, 1, 0);
3036 static void function_print_type_decl(struct type *type, FILE *f)
3040 for (b = type->function.params; b; b = cast(binode, b->right)) {
3041 struct variable *v = cast(var, b->left)->var;
3042 fprintf(f, "%.*s%s", v->name->name.len, v->name->name.txt,
3043 v->constant ? "::" : ":");
3044 type_print(v->type, f);
3049 if (type->function.return_type != Tnone) {
3051 if (type->function.inline_result) {
3053 struct type *t = type->function.return_type;
3055 for (i = 0; i < t->structure.nfields; i++) {
3056 struct field *fl = t->structure.fields + i;
3059 fprintf(f, "%.*s:", fl->name.len, fl->name.txt);
3060 type_print(fl->type, f);
3064 type_print(type->function.return_type, f);
3069 static void function_free_type(struct type *t)
3071 free_exec(t->function.params);
3074 static struct type function_prototype = {
3075 .size = sizeof(void*),
3076 .align = sizeof(void*),
3077 .free = function_free,
3078 .compat = function_compat,
3079 .check_args = function_check_args,
3080 .print = function_print,
3081 .print_type_decl = function_print_type_decl,
3082 .free_type = function_free_type,
3085 ###### declare terminals
3095 FuncName -> IDENTIFIER ${ {
3096 struct variable *v = var_decl(c, $1.txt);
3097 struct var *e = new_pos(var, $1);
3104 v = var_ref(c, $1.txt);
3106 type_err(c, "error: function '%v' redeclared",
3108 type_err(c, "info: this is where '%v' was first declared",
3109 v->where_decl, NULL, 0, NULL);
3115 Args -> ArgsLine NEWLINE ${ $0 = $<AL; }$
3116 | Args ArgsLine NEWLINE ${ {
3117 struct binode *b = $<AL;
3118 struct binode **bp = &b;
3120 bp = (struct binode **)&(*bp)->left;
3125 ArgsLine -> ${ $0 = NULL; }$
3126 | Varlist ${ $0 = $<1; }$
3127 | Varlist ; ${ $0 = $<1; }$
3129 Varlist -> Varlist ; ArgDecl ${
3130 $0 = new_pos(binode, $2);
3143 ArgDecl -> IDENTIFIER : FormalType ${ {
3144 struct variable *v = var_decl(c, $ID.txt);
3145 $0 = new_pos(var, $ID);
3152 ##### Function calls
3154 A function call can appear either as an expression or as a statement.
3155 We use a new 'Funcall' binode type to link the function with a list of
3156 arguments, form with the 'List' nodes.
3158 We have already seen the "Term" which is how a function call can appear
3159 in an expression. To parse a function call into a statement we include
3160 it in the "SimpleStatement Grammar" which will be described later.
3166 | Term ( ExpressionList ) ${ {
3167 struct binode *b = new(binode);
3170 b->right = reorder_bilist($<EL);
3174 struct binode *b = new(binode);
3181 ###### SimpleStatement Grammar
3183 | Term ( ExpressionList ) ${ {
3184 struct binode *b = new(binode);
3187 b->right = reorder_bilist($<EL);
3191 ###### print binode cases
3194 do_indent(indent, "");
3195 print_exec(b->left, -1, bracket);
3197 for (b = cast(binode, b->right); b; b = cast(binode, b->right)) {
3200 print_exec(b->left, -1, bracket);
3210 ###### propagate binode cases
3213 /* Every arg must match formal parameter, and result
3214 * is return type of function
3216 struct binode *args = cast(binode, b->right);
3217 struct var *v = cast(var, b->left);
3219 if (!v->var->type || v->var->type->check_args == NULL) {
3220 type_err(c, "error: attempt to call a non-function.",
3221 prog, NULL, 0, NULL);
3225 v->var->type->check_args(c, perr, v->var->type, args);
3226 if (v->var->type->function.inline_result)
3228 return v->var->type->function.return_type;
3231 ###### interp binode cases
3234 struct var *v = cast(var, b->left);
3235 struct type *t = v->var->type;
3236 void *oldlocal = c->local;
3237 int old_size = c->local_size;
3238 void *local = calloc(1, t->function.local_size);
3239 struct value *fbody = var_value(c, v->var);
3240 struct binode *arg = cast(binode, b->right);
3241 struct binode *param = t->function.params;
3244 struct var *pv = cast(var, param->left);
3245 struct type *vtype = NULL;
3246 struct value val = interp_exec(c, arg->left, &vtype);
3248 c->local = local; c->local_size = t->function.local_size;
3249 lval = var_value(c, pv->var);
3250 c->local = oldlocal; c->local_size = old_size;
3251 memcpy(lval, &val, vtype->size);
3252 param = cast(binode, param->right);
3253 arg = cast(binode, arg->right);
3255 c->local = local; c->local_size = t->function.local_size;
3256 if (t->function.inline_result && dtype) {
3257 _interp_exec(c, fbody->function, NULL, NULL);
3258 memcpy(dest, local, dtype->size);
3259 rvtype = ret.type = NULL;
3261 rv = interp_exec(c, fbody->function, &rvtype);
3262 c->local = oldlocal; c->local_size = old_size;
3267 ## Complex executables: statements and expressions
3269 Now that we have types and values and variables and most of the basic
3270 Terms which provide access to these, we can explore the more complex
3271 code that combine all of these to get useful work done. Specifically
3272 statements and expressions.
3274 Expressions are various combinations of Terms. We will use operator
3275 precedence to ensure correct parsing. The simplest Expression is just a
3276 Term - others will follow.
3281 Expression -> Term ${ $0 = $<Term; }$
3282 ## expression grammar
3284 ### Expressions: Conditional
3286 Our first user of the `binode` will be conditional expressions, which
3287 is a bit odd as they actually have three components. That will be
3288 handled by having 2 binodes for each expression. The conditional
3289 expression is the lowest precedence operator which is why we define it
3290 first - to start the precedence list.
3292 Conditional expressions are of the form "value `if` condition `else`
3293 other_value". They associate to the right, so everything to the right
3294 of `else` is part of an else value, while only a higher-precedence to
3295 the left of `if` is the if values. Between `if` and `else` there is no
3296 room for ambiguity, so a full conditional expression is allowed in
3302 ###### declare terminals
3306 ###### expression grammar
3308 | Expression if Expression else Expression $$ifelse ${ {
3309 struct binode *b1 = new(binode);
3310 struct binode *b2 = new(binode);
3320 ###### print binode cases
3323 b2 = cast(binode, b->right);
3324 if (bracket) printf("(");
3325 print_exec(b2->left, -1, bracket);
3327 print_exec(b->left, -1, bracket);
3329 print_exec(b2->right, -1, bracket);
3330 if (bracket) printf(")");
3333 ###### propagate binode cases
3336 /* cond must be Tbool, others must match */
3337 struct binode *b2 = cast(binode, b->right);
3340 propagate_types(b->left, c, perr, Tbool, 0);
3341 t = propagate_types(b2->left, c, perr, type, Rnolabel);
3342 t2 = propagate_types(b2->right, c, perr, type ?: t, Rnolabel);
3346 ###### interp binode cases
3349 struct binode *b2 = cast(binode, b->right);
3350 left = interp_exec(c, b->left, <ype);
3352 rv = interp_exec(c, b2->left, &rvtype); // UNTESTED
3354 rv = interp_exec(c, b2->right, &rvtype);
3360 We take a brief detour, now that we have expressions, to describe lists
3361 of expressions. These will be needed for function parameters and
3362 possibly other situations. They seem generic enough to introduce here
3363 to be used elsewhere.
3365 And ExpressionList will use the `List` type of `binode`, building up at
3366 the end. And place where they are used will probably call
3367 `reorder_bilist()` to get a more normal first/next arrangement.
3369 ###### declare terminals
3372 `List` execs have no implicit semantics, so they are never propagated or
3373 interpreted. The can be printed as a comma separate list, which is how
3374 they are parsed. Note they are also used for function formal parameter
3375 lists. In that case a separate function is used to print them.
3377 ###### print binode cases
3381 print_exec(b->left, -1, bracket);
3384 b = cast(binode, b->right);
3388 ###### propagate binode cases
3389 case List: abort(); // NOTEST
3390 ###### interp binode cases
3391 case List: abort(); // NOTEST
3396 ExpressionList -> ExpressionList , Expression ${
3409 ### Expressions: Boolean
3411 The next class of expressions to use the `binode` will be Boolean
3412 expressions. "`and then`" and "`or else`" are similar to `and` and `or`
3413 have same corresponding precendence. The difference is that they don't
3414 evaluate the second expression if not necessary.
3423 ###### declare terminals
3428 ###### expression grammar
3429 | Expression or Expression ${ {
3430 struct binode *b = new(binode);
3436 | Expression or else Expression ${ {
3437 struct binode *b = new(binode);
3444 | Expression and Expression ${ {
3445 struct binode *b = new(binode);
3451 | Expression and then Expression ${ {
3452 struct binode *b = new(binode);
3459 | not Expression ${ {
3460 struct binode *b = new(binode);
3466 ###### print binode cases
3468 if (bracket) printf("(");
3469 print_exec(b->left, -1, bracket);
3471 print_exec(b->right, -1, bracket);
3472 if (bracket) printf(")");
3475 if (bracket) printf("(");
3476 print_exec(b->left, -1, bracket);
3477 printf(" and then ");
3478 print_exec(b->right, -1, bracket);
3479 if (bracket) printf(")");
3482 if (bracket) printf("(");
3483 print_exec(b->left, -1, bracket);
3485 print_exec(b->right, -1, bracket);
3486 if (bracket) printf(")");
3489 if (bracket) printf("(");
3490 print_exec(b->left, -1, bracket);
3491 printf(" or else ");
3492 print_exec(b->right, -1, bracket);
3493 if (bracket) printf(")");
3496 if (bracket) printf("(");
3498 print_exec(b->right, -1, bracket);
3499 if (bracket) printf(")");
3502 ###### propagate binode cases
3508 /* both must be Tbool, result is Tbool */
3509 propagate_types(b->left, c, perr, Tbool, 0);
3510 propagate_types(b->right, c, perr, Tbool, 0);
3511 if (type && type != Tbool)
3512 type_err(c, "error: %1 operation found where %2 expected", prog,
3516 ###### interp binode cases
3518 rv = interp_exec(c, b->left, &rvtype);
3519 right = interp_exec(c, b->right, &rtype);
3520 rv.bool = rv.bool && right.bool;
3523 rv = interp_exec(c, b->left, &rvtype);
3525 rv = interp_exec(c, b->right, NULL);
3528 rv = interp_exec(c, b->left, &rvtype);
3529 right = interp_exec(c, b->right, &rtype);
3530 rv.bool = rv.bool || right.bool;
3533 rv = interp_exec(c, b->left, &rvtype);
3535 rv = interp_exec(c, b->right, NULL);
3538 rv = interp_exec(c, b->right, &rvtype);
3542 ### Expressions: Comparison
3544 Of slightly higher precedence that Boolean expressions are Comparisons.
3545 A comparison takes arguments of any comparable type, but the two types
3548 To simplify the parsing we introduce an `eop` which can record an
3549 expression operator, and the `CMPop` non-terminal will match one of them.
3556 ###### ast functions
3557 static void free_eop(struct eop *e)
3571 ###### declare terminals
3572 $LEFT < > <= >= == != CMPop
3574 ###### expression grammar
3575 | Expression CMPop Expression ${ {
3576 struct binode *b = new(binode);
3586 CMPop -> < ${ $0.op = Less; }$
3587 | > ${ $0.op = Gtr; }$
3588 | <= ${ $0.op = LessEq; }$
3589 | >= ${ $0.op = GtrEq; }$
3590 | == ${ $0.op = Eql; }$
3591 | != ${ $0.op = NEql; }$
3593 ###### print binode cases
3601 if (bracket) printf("(");
3602 print_exec(b->left, -1, bracket);
3604 case Less: printf(" < "); break;
3605 case LessEq: printf(" <= "); break;
3606 case Gtr: printf(" > "); break;
3607 case GtrEq: printf(" >= "); break;
3608 case Eql: printf(" == "); break;
3609 case NEql: printf(" != "); break;
3610 default: abort(); // NOTEST
3612 print_exec(b->right, -1, bracket);
3613 if (bracket) printf(")");
3616 ###### propagate binode cases
3623 /* Both must match but not be labels, result is Tbool */
3624 t = propagate_types(b->left, c, perr, NULL, Rnolabel);
3626 propagate_types(b->right, c, perr, t, 0);
3628 t = propagate_types(b->right, c, perr, NULL, Rnolabel); // UNTESTED
3630 t = propagate_types(b->left, c, perr, t, 0); // UNTESTED
3632 if (!type_compat(type, Tbool, 0))
3633 type_err(c, "error: Comparison returns %1 but %2 expected", prog,
3634 Tbool, rules, type);
3637 ###### interp binode cases
3646 left = interp_exec(c, b->left, <ype);
3647 right = interp_exec(c, b->right, &rtype);
3648 cmp = value_cmp(ltype, rtype, &left, &right);
3651 case Less: rv.bool = cmp < 0; break;
3652 case LessEq: rv.bool = cmp <= 0; break;
3653 case Gtr: rv.bool = cmp > 0; break;
3654 case GtrEq: rv.bool = cmp >= 0; break;
3655 case Eql: rv.bool = cmp == 0; break;
3656 case NEql: rv.bool = cmp != 0; break;
3657 default: rv.bool = 0; break; // NOTEST
3662 ### Expressions: Arithmetic etc.
3664 The remaining expressions with the highest precedence are arithmetic,
3665 string concatenation, string conversion, and testing. String concatenation
3666 (`++`) has the same precedence as multiplication and division, but lower
3669 Testing comes in two forms. A single question mark (`?`) is a uniary
3670 operator which converts come types into Boolean. The general meaning is
3671 "is this a value value" and there will be more uses as the language
3672 develops. A double questionmark (`??`) is a binary operator (Choose),
3673 with same precedence as multiplication, which returns the LHS if it
3674 tests successfully, else returns the RHS.
3676 String conversion is a temporary feature until I get a better type
3677 system. `$` is a prefix operator which expects a string and returns
3680 `+` and `-` are both infix and prefix operations (where they are
3681 absolute value and negation). These have different operator names.
3683 We also have a 'Bracket' operator which records where parentheses were
3684 found. This makes it easy to reproduce these when printing. Possibly I
3685 should only insert brackets were needed for precedence. Putting
3686 parentheses around an expression converts it into a Term,
3692 Absolute, Negate, Test,
3696 ###### declare terminals
3698 $LEFT * / % ++ ?? Top
3702 ###### expression grammar
3703 | Expression Eop Expression ${ {
3704 struct binode *b = new(binode);
3711 | Expression Top Expression ${ {
3712 struct binode *b = new(binode);
3719 | Uop Expression ${ {
3720 struct binode *b = new(binode);
3728 | ( Expression ) ${ {
3729 struct binode *b = new_pos(binode, $1);
3738 Eop -> + ${ $0.op = Plus; }$
3739 | - ${ $0.op = Minus; }$
3741 Uop -> + ${ $0.op = Absolute; }$
3742 | - ${ $0.op = Negate; }$
3743 | $ ${ $0.op = StringConv; }$
3744 | ? ${ $0.op = Test; }$
3746 Top -> * ${ $0.op = Times; }$
3747 | / ${ $0.op = Divide; }$
3748 | % ${ $0.op = Rem; }$
3749 | ++ ${ $0.op = Concat; }$
3750 | ?? ${ $0.op = Choose; }$
3752 ###### print binode cases
3760 if (bracket) printf("(");
3761 print_exec(b->left, indent, bracket);
3763 case Plus: fputs(" + ", stdout); break;
3764 case Minus: fputs(" - ", stdout); break;
3765 case Times: fputs(" * ", stdout); break;
3766 case Divide: fputs(" / ", stdout); break;
3767 case Rem: fputs(" % ", stdout); break;
3768 case Concat: fputs(" ++ ", stdout); break;
3769 case Choose: fputs(" ?? ", stdout); break;
3770 default: abort(); // NOTEST
3772 print_exec(b->right, indent, bracket);
3773 if (bracket) printf(")");
3779 if (bracket) printf("(");
3781 case Absolute: fputs("+", stdout); break;
3782 case Negate: fputs("-", stdout); break;
3783 case StringConv: fputs("$", stdout); break;
3784 case Test: fputs("?", stdout); break;
3785 default: abort(); // NOTEST
3787 print_exec(b->right, indent, bracket);
3788 if (bracket) printf(")");
3792 print_exec(b->right, indent, bracket);
3796 ###### propagate binode cases
3802 /* both must be numbers, result is Tnum */
3805 /* as propagate_types ignores a NULL,
3806 * unary ops fit here too */
3807 propagate_types(b->left, c, perr, Tnum, 0);
3808 propagate_types(b->right, c, perr, Tnum, 0);
3809 if (!type_compat(type, Tnum, 0))
3810 type_err(c, "error: Arithmetic returns %1 but %2 expected", prog,
3815 /* both must be Tstr, result is Tstr */
3816 propagate_types(b->left, c, perr, Tstr, 0);
3817 propagate_types(b->right, c, perr, Tstr, 0);
3818 if (!type_compat(type, Tstr, 0))
3819 type_err(c, "error: Concat returns %1 but %2 expected", prog,
3824 /* op must be string, result is number */
3825 propagate_types(b->left, c, perr, Tstr, 0);
3826 if (!type_compat(type, Tnum, 0))
3827 type_err(c, // UNTESTED
3828 "error: Can only convert string to number, not %1",
3829 prog, type, 0, NULL);
3833 /* LHS must support ->test, result is Tbool */
3834 t = propagate_types(b->right, c, perr, NULL, 0);
3836 type_err(c, "error: '?' requires a testable value, not %1",
3841 /* LHS and RHS must match and are returned. Must support
3844 t = propagate_types(b->left, c, perr, type, rules);
3845 t = propagate_types(b->right, c, perr, t, rules);
3846 if (t && t->test == NULL)
3847 type_err(c, "error: \"??\" requires a testable value, not %1",
3852 return propagate_types(b->right, c, perr, type, 0);
3854 ###### interp binode cases
3857 rv = interp_exec(c, b->left, &rvtype);
3858 right = interp_exec(c, b->right, &rtype);
3859 mpq_add(rv.num, rv.num, right.num);
3862 rv = interp_exec(c, b->left, &rvtype);
3863 right = interp_exec(c, b->right, &rtype);
3864 mpq_sub(rv.num, rv.num, right.num);
3867 rv = interp_exec(c, b->left, &rvtype);
3868 right = interp_exec(c, b->right, &rtype);
3869 mpq_mul(rv.num, rv.num, right.num);
3872 rv = interp_exec(c, b->left, &rvtype);
3873 right = interp_exec(c, b->right, &rtype);
3874 mpq_div(rv.num, rv.num, right.num);
3879 left = interp_exec(c, b->left, <ype);
3880 right = interp_exec(c, b->right, &rtype);
3881 mpz_init(l); mpz_init(r); mpz_init(rem);
3882 mpz_tdiv_q(l, mpq_numref(left.num), mpq_denref(left.num));
3883 mpz_tdiv_q(r, mpq_numref(right.num), mpq_denref(right.num));
3884 mpz_tdiv_r(rem, l, r);
3885 val_init(Tnum, &rv);
3886 mpq_set_z(rv.num, rem);
3887 mpz_clear(r); mpz_clear(l); mpz_clear(rem);
3892 rv = interp_exec(c, b->right, &rvtype);
3893 mpq_neg(rv.num, rv.num);
3896 rv = interp_exec(c, b->right, &rvtype);
3897 mpq_abs(rv.num, rv.num);
3900 rv = interp_exec(c, b->right, &rvtype);
3903 left = interp_exec(c, b->left, <ype);
3904 right = interp_exec(c, b->right, &rtype);
3906 rv.str = text_join(left.str, right.str);
3909 right = interp_exec(c, b->right, &rvtype);
3913 struct text tx = right.str;
3916 if (tx.txt[0] == '-') {
3917 neg = 1; // UNTESTED
3918 tx.txt++; // UNTESTED
3919 tx.len--; // UNTESTED
3921 if (number_parse(rv.num, tail, tx) == 0)
3922 mpq_init(rv.num); // UNTESTED
3924 mpq_neg(rv.num, rv.num); // UNTESTED
3926 printf("Unsupported suffix: %.*s\n", tx.len, tx.txt); // UNTESTED
3930 right = interp_exec(c, b->right, &rtype);
3932 rv.bool = !!rtype->test(rtype, &right);
3935 left = interp_exec(c, b->left, <ype);
3936 if (ltype->test(ltype, &left)) {
3941 rv = interp_exec(c, b->right, &rvtype);
3944 ###### value functions
3946 static struct text text_join(struct text a, struct text b)
3949 rv.len = a.len + b.len;
3950 rv.txt = malloc(rv.len);
3951 memcpy(rv.txt, a.txt, a.len);
3952 memcpy(rv.txt+a.len, b.txt, b.len);
3956 ### Blocks, Statements, and Statement lists.
3958 Now that we have expressions out of the way we need to turn to
3959 statements. There are simple statements and more complex statements.
3960 Simple statements do not contain (syntactic) newlines, complex statements do.
3962 Statements often come in sequences and we have corresponding simple
3963 statement lists and complex statement lists.
3964 The former comprise only simple statements separated by semicolons.
3965 The later comprise complex statements and simple statement lists. They are
3966 separated by newlines. Thus the semicolon is only used to separate
3967 simple statements on the one line. This may be overly restrictive,
3968 but I'm not sure I ever want a complex statement to share a line with
3971 Note that a simple statement list can still use multiple lines if
3972 subsequent lines are indented, so
3974 ###### Example: wrapped simple statement list
3979 is a single simple statement list. This might allow room for
3980 confusion, so I'm not set on it yet.
3982 A simple statement list needs no extra syntax. A complex statement
3983 list has two syntactic forms. It can be enclosed in braces (much like
3984 C blocks), or it can be introduced by an indent and continue until an
3985 unindented newline (much like Python blocks). With this extra syntax
3986 it is referred to as a block.
3988 Note that a block does not have to include any newlines if it only
3989 contains simple statements. So both of:
3991 if condition: a=b; d=f
3993 if condition { a=b; print f }
3997 In either case the list is constructed from a `binode` list with
3998 `Block` as the operator. When parsing the list it is most convenient
3999 to append to the end, so a list is a list and a statement. When using
4000 the list it is more convenient to consider a list to be a statement
4001 and a list. So we need a function to re-order a list.
4002 `reorder_bilist` serves this purpose.
4004 The only stand-alone statement we introduce at this stage is `pass`
4005 which does nothing and is represented as a `NULL` pointer in a `Block`
4006 list. Other stand-alone statements will follow once the infrastructure
4009 As many statements will use binodes, we declare a binode pointer 'b' in
4010 the common header for all reductions to use.
4012 ###### Parser: reduce
4023 Block -> { IN OptNL Statementlist OUT OptNL } ${ $0 = $<Sl; }$
4024 | { SimpleStatements } ${ $0 = reorder_bilist($<SS); }$
4025 | SimpleStatements ; ${ $0 = reorder_bilist($<SS); }$
4026 | SimpleStatements EOL ${ $0 = reorder_bilist($<SS); }$
4027 | IN OptNL Statementlist OUT ${ $0 = $<Sl; }$
4029 OpenBlock -> OpenScope { IN OptNL Statementlist OUT OptNL } ${ $0 = $<Sl; }$
4030 | OpenScope { SimpleStatements } ${ $0 = reorder_bilist($<SS); }$
4031 | OpenScope SimpleStatements ; ${ $0 = reorder_bilist($<SS); }$
4032 | OpenScope SimpleStatements EOL ${ $0 = reorder_bilist($<SS); }$
4033 | IN OpenScope OptNL Statementlist OUT ${ $0 = $<Sl; }$
4035 UseBlock -> { OpenScope IN OptNL Statementlist OUT OptNL } ${ $0 = $<Sl; }$
4036 | { OpenScope SimpleStatements } ${ $0 = reorder_bilist($<SS); }$
4037 | IN OpenScope OptNL Statementlist OUT ${ $0 = $<Sl; }$
4039 ColonBlock -> { IN OptNL Statementlist OUT OptNL } ${ $0 = $<Sl; }$
4040 | { SimpleStatements } ${ $0 = reorder_bilist($<SS); }$
4041 | : SimpleStatements ; ${ $0 = reorder_bilist($<SS); }$
4042 | : SimpleStatements EOL ${ $0 = reorder_bilist($<SS); }$
4043 | : IN OptNL Statementlist OUT ${ $0 = $<Sl; }$
4045 Statementlist -> ComplexStatements ${ $0 = reorder_bilist($<CS); }$
4047 ComplexStatements -> ComplexStatements ComplexStatement ${
4057 | ComplexStatement ${
4069 ComplexStatement -> SimpleStatements Newlines ${
4070 $0 = reorder_bilist($<SS);
4072 | SimpleStatements ; Newlines ${
4073 $0 = reorder_bilist($<SS);
4075 ## ComplexStatement Grammar
4078 SimpleStatements -> SimpleStatements ; SimpleStatement ${
4084 | SimpleStatement ${
4093 SimpleStatement -> pass ${ $0 = NULL; }$
4094 | ERROR ${ tok_err(c, "Syntax error in statement", &$1); }$
4095 ## SimpleStatement Grammar
4097 ###### print binode cases
4101 if (b->left == NULL) // UNTESTED
4102 printf("pass"); // UNTESTED
4104 print_exec(b->left, indent, bracket); // UNTESTED
4105 if (b->right) { // UNTESTED
4106 printf("; "); // UNTESTED
4107 print_exec(b->right, indent, bracket); // UNTESTED
4110 // block, one per line
4111 if (b->left == NULL)
4112 do_indent(indent, "pass\n");
4114 print_exec(b->left, indent, bracket);
4116 print_exec(b->right, indent, bracket);
4120 ###### propagate binode cases
4123 /* If any statement returns something other than Tnone
4124 * or Tbool then all such must return same type.
4125 * As each statement may be Tnone or something else,
4126 * we must always pass NULL (unknown) down, otherwise an incorrect
4127 * error might occur. We never return Tnone unless it is
4132 for (e = b; e; e = cast(binode, e->right)) {
4133 t = propagate_types(e->left, c, perr, NULL, rules);
4134 if ((rules & Rboolok) && (t == Tbool || t == Tnone))
4136 if (t == Tnone && e->right)
4137 /* Only the final statement *must* return a value
4145 type_err(c, "error: expected %1%r, found %2",
4146 e->left, type, rules, t);
4152 ###### interp binode cases
4154 while (rvtype == Tnone &&
4157 rv = interp_exec(c, b->left, &rvtype);
4158 b = cast(binode, b->right);
4162 ### The Print statement
4164 `print` is a simple statement that takes a comma-separated list of
4165 expressions and prints the values separated by spaces and terminated
4166 by a newline. No control of formatting is possible.
4168 `print` uses `ExpressionList` to collect the expressions and stores them
4169 on the left side of a `Print` binode unlessthere is a trailing comma
4170 when the list is stored on the `right` side and no trailing newline is
4176 ##### declare terminals
4179 ###### SimpleStatement Grammar
4181 | print ExpressionList ${
4182 $0 = b = new_pos(binode, $1);
4185 b->left = reorder_bilist($<EL);
4187 | print ExpressionList , ${ {
4188 $0 = b = new_pos(binode, $1);
4190 b->right = reorder_bilist($<EL);
4194 $0 = b = new_pos(binode, $1);
4200 ###### print binode cases
4203 do_indent(indent, "print");
4205 print_exec(b->right, -1, bracket);
4208 print_exec(b->left, -1, bracket);
4213 ###### propagate binode cases
4216 /* don't care but all must be consistent */
4218 b = cast(binode, b->left);
4220 b = cast(binode, b->right);
4222 propagate_types(b->left, c, perr, NULL, Rnolabel);
4223 b = cast(binode, b->right);
4227 ###### interp binode cases
4231 struct binode *b2 = cast(binode, b->left);
4233 b2 = cast(binode, b->right);
4234 for (; b2; b2 = cast(binode, b2->right)) {
4235 left = interp_exec(c, b2->left, <ype);
4236 print_value(ltype, &left, stdout);
4237 free_value(ltype, &left);
4241 if (b->right == NULL)
4247 ###### Assignment statement
4249 An assignment will assign a value to a variable, providing it hasn't
4250 been declared as a constant. The analysis phase ensures that the type
4251 will be correct so the interpreter just needs to perform the
4252 calculation. There is a form of assignment which declares a new
4253 variable as well as assigning a value. If a name is assigned before
4254 it is declared, and error will be raised as the name is created as
4255 `Tlabel` and it is illegal to assign to such names.
4261 ###### declare terminals
4264 ###### SimpleStatement Grammar
4265 | Term = Expression ${
4266 $0 = b= new(binode);
4271 | VariableDecl = Expression ${
4272 $0 = b= new(binode);
4279 if ($1->var->where_set == NULL) {
4281 "Variable declared with no type or value: %v",
4285 $0 = b = new(binode);
4292 ###### print binode cases
4295 do_indent(indent, "");
4296 print_exec(b->left, indent, bracket);
4298 print_exec(b->right, indent, bracket);
4305 struct variable *v = cast(var, b->left)->var;
4306 do_indent(indent, "");
4307 print_exec(b->left, indent, bracket);
4308 if (cast(var, b->left)->var->constant) {
4310 if (v->explicit_type) {
4311 type_print(v->type, stdout);
4316 if (v->explicit_type) {
4317 type_print(v->type, stdout);
4323 print_exec(b->right, indent, bracket);
4330 ###### propagate binode cases
4334 /* Both must match and not be labels,
4335 * Type must support 'dup',
4336 * For Assign, left must not be constant.
4339 t = propagate_types(b->left, c, perr, NULL,
4340 Rnolabel | (b->op == Assign ? Rnoconstant : 0));
4345 if (propagate_types(b->right, c, perr, t, 0) != t)
4346 if (b->left->type == Xvar)
4347 type_err(c, "info: variable '%v' was set as %1 here.",
4348 cast(var, b->left)->var->where_set, t, rules, NULL);
4350 t = propagate_types(b->right, c, perr, NULL, Rnolabel);
4352 propagate_types(b->left, c, perr, t,
4353 (b->op == Assign ? Rnoconstant : 0));
4355 if (t && t->dup == NULL && !(*perr & Emaycopy))
4356 type_err(c, "error: cannot assign value of type %1", b, t, 0, NULL);
4361 ###### interp binode cases
4364 lleft = linterp_exec(c, b->left, <ype);
4366 dinterp_exec(c, b->right, lleft, ltype, 1);
4372 struct variable *v = cast(var, b->left)->var;
4375 val = var_value(c, v);
4376 if (v->type->prepare_type)
4377 v->type->prepare_type(c, v->type, 0);
4379 dinterp_exec(c, b->right, val, v->type, 0);
4381 val_init(v->type, val);
4385 ### The `use` statement
4387 The `use` statement is the last "simple" statement. It is needed when a
4388 statement block can return a value. This includes the body of a
4389 function which has a return type, and the "condition" code blocks in
4390 `if`, `while`, and `switch` statements.
4395 ###### declare terminals
4398 ###### SimpleStatement Grammar
4400 $0 = b = new_pos(binode, $1);
4403 if (b->right->type == Xvar) {
4404 struct var *v = cast(var, b->right);
4405 if (v->var->type == Tnone) {
4406 /* Convert this to a label */
4409 v->var->type = Tlabel;
4410 val = global_alloc(c, Tlabel, v->var, NULL);
4416 ###### print binode cases
4419 do_indent(indent, "use ");
4420 print_exec(b->right, -1, bracket);
4425 ###### propagate binode cases
4428 /* result matches value */
4429 return propagate_types(b->right, c, perr, type, 0);
4431 ###### interp binode cases
4434 rv = interp_exec(c, b->right, &rvtype);
4437 ### The Conditional Statement
4439 This is the biggy and currently the only complex statement. This
4440 subsumes `if`, `while`, `do/while`, `switch`, and some parts of `for`.
4441 It is comprised of a number of parts, all of which are optional though
4442 set combinations apply. Each part is (usually) a key word (`then` is
4443 sometimes optional) followed by either an expression or a code block,
4444 except the `casepart` which is a "key word and an expression" followed
4445 by a code block. The code-block option is valid for all parts and,
4446 where an expression is also allowed, the code block can use the `use`
4447 statement to report a value. If the code block does not report a value
4448 the effect is similar to reporting `True`.
4450 The `else` and `case` parts, as well as `then` when combined with
4451 `if`, can contain a `use` statement which will apply to some
4452 containing conditional statement. `for` parts, `do` parts and `then`
4453 parts used with `for` can never contain a `use`, except in some
4454 subordinate conditional statement.
4456 If there is a `forpart`, it is executed first, only once.
4457 If there is a `dopart`, then it is executed repeatedly providing
4458 always that the `condpart` or `cond`, if present, does not return a non-True
4459 value. `condpart` can fail to return any value if it simply executes
4460 to completion. This is treated the same as returning `True`.
4462 If there is a `thenpart` it will be executed whenever the `condpart`
4463 or `cond` returns True (or does not return any value), but this will happen
4464 *after* `dopart` (when present).
4466 If `elsepart` is present it will be executed at most once when the
4467 condition returns `False` or some value that isn't `True` and isn't
4468 matched by any `casepart`. If there are any `casepart`s, they will be
4469 executed when the condition returns a matching value.
4471 The particular sorts of values allowed in case parts has not yet been
4472 determined in the language design, so nothing is prohibited.
4474 The various blocks in this complex statement potentially provide scope
4475 for variables as described earlier. Each such block must include the
4476 "OpenScope" nonterminal before parsing the block, and must call
4477 `var_block_close()` when closing the block.
4479 The code following "`if`", "`switch`" and "`for`" does not get its own
4480 scope, but is in a scope covering the whole statement, so names
4481 declared there cannot be redeclared elsewhere. Similarly the
4482 condition following "`while`" is in a scope the covers the body
4483 ("`do`" part) of the loop, and which does not allow conditional scope
4484 extension. Code following "`then`" (both looping and non-looping),
4485 "`else`" and "`case`" each get their own local scope.
4487 The type requirements on the code block in a `whilepart` are quite
4488 unusal. It is allowed to return a value of some identifiable type, in
4489 which case the loop aborts and an appropriate `casepart` is run, or it
4490 can return a Boolean, in which case the loop either continues to the
4491 `dopart` (on `True`) or aborts and runs the `elsepart` (on `False`).
4492 This is different both from the `ifpart` code block which is expected to
4493 return a Boolean, or the `switchpart` code block which is expected to
4494 return the same type as the casepart values. The correct analysis of
4495 the type of the `whilepart` code block is the reason for the
4496 `Rboolok` flag which is passed to `propagate_types()`.
4498 The `cond_statement` cannot fit into a `binode` so a new `exec` is
4499 defined. As there are two scopes which cover multiple parts - one for
4500 the whole statement and one for "while" and "do" - and as we will use
4501 the 'struct exec' to track scopes, we actually need two new types of
4502 exec. One is a `binode` for the looping part, the rest is the
4503 `cond_statement`. The `cond_statement` will use an auxilliary `struct
4504 casepart` to track a list of case parts.
4515 struct exec *action;
4516 struct casepart *next;
4518 struct cond_statement {
4520 struct exec *forpart, *condpart, *thenpart, *elsepart;
4521 struct binode *looppart;
4522 struct casepart *casepart;
4525 ###### ast functions
4527 static void free_casepart(struct casepart *cp)
4531 free_exec(cp->value);
4532 free_exec(cp->action);
4539 static void free_cond_statement(struct cond_statement *s)
4543 free_exec(s->forpart);
4544 free_exec(s->condpart);
4545 free_exec(s->looppart);
4546 free_exec(s->thenpart);
4547 free_exec(s->elsepart);
4548 free_casepart(s->casepart);
4552 ###### free exec cases
4553 case Xcond_statement: free_cond_statement(cast(cond_statement, e)); break;
4555 ###### ComplexStatement Grammar
4556 | CondStatement ${ $0 = $<1; }$
4558 ###### declare terminals
4559 $TERM for then while do
4566 // A CondStatement must end with EOL, as does CondSuffix and
4568 // ForPart, ThenPart, SwitchPart, CasePart are non-empty and
4569 // may or may not end with EOL
4570 // WhilePart and IfPart include an appropriate Suffix
4572 // ForPart, SwitchPart, and IfPart open scopes, o we have to close
4573 // them. WhilePart opens and closes its own scope.
4574 CondStatement -> ForPart OptNL ThenPart OptNL WhilePart CondSuffix ${
4577 $0->thenpart = $<TP;
4578 $0->looppart = $<WP;
4579 var_block_close(c, CloseSequential, $0);
4581 | ForPart OptNL WhilePart CondSuffix ${
4584 $0->looppart = $<WP;
4585 var_block_close(c, CloseSequential, $0);
4587 | WhilePart CondSuffix ${
4589 $0->looppart = $<WP;
4591 | SwitchPart OptNL CasePart CondSuffix ${
4593 $0->condpart = $<SP;
4594 $CP->next = $0->casepart;
4595 $0->casepart = $<CP;
4596 var_block_close(c, CloseSequential, $0);
4598 | SwitchPart : IN OptNL CasePart CondSuffix OUT Newlines ${
4600 $0->condpart = $<SP;
4601 $CP->next = $0->casepart;
4602 $0->casepart = $<CP;
4603 var_block_close(c, CloseSequential, $0);
4605 | IfPart IfSuffix ${
4607 $0->condpart = $IP.condpart; $IP.condpart = NULL;
4608 $0->thenpart = $IP.thenpart; $IP.thenpart = NULL;
4609 // This is where we close an "if" statement
4610 var_block_close(c, CloseSequential, $0);
4613 CondSuffix -> IfSuffix ${
4616 | Newlines CasePart CondSuffix ${
4618 $CP->next = $0->casepart;
4619 $0->casepart = $<CP;
4621 | CasePart CondSuffix ${
4623 $CP->next = $0->casepart;
4624 $0->casepart = $<CP;
4627 IfSuffix -> Newlines ${ $0 = new(cond_statement); }$
4628 | Newlines ElsePart ${ $0 = $<EP; }$
4629 | ElsePart ${$0 = $<EP; }$
4631 ElsePart -> else OpenBlock Newlines ${
4632 $0 = new(cond_statement);
4633 $0->elsepart = $<OB;
4634 var_block_close(c, CloseElse, $0->elsepart);
4636 | else OpenScope CondStatement ${
4637 $0 = new(cond_statement);
4638 $0->elsepart = $<CS;
4639 var_block_close(c, CloseElse, $0->elsepart);
4643 CasePart -> case Expression OpenScope ColonBlock ${
4644 $0 = calloc(1,sizeof(struct casepart));
4647 var_block_close(c, CloseParallel, $0->action);
4651 // These scopes are closed in CondStatement
4652 ForPart -> for OpenBlock ${
4656 ThenPart -> then OpenBlock ${
4658 var_block_close(c, CloseSequential, $0);
4662 // This scope is closed in CondStatement
4663 WhilePart -> while UseBlock OptNL do OpenBlock ${
4668 var_block_close(c, CloseSequential, $0->right);
4669 var_block_close(c, CloseSequential, $0);
4671 | while OpenScope Expression OpenScope ColonBlock ${
4676 var_block_close(c, CloseSequential, $0->right);
4677 var_block_close(c, CloseSequential, $0);
4681 IfPart -> if UseBlock OptNL then OpenBlock ${
4684 var_block_close(c, CloseParallel, $0.thenpart);
4686 | if OpenScope Expression OpenScope ColonBlock ${
4689 var_block_close(c, CloseParallel, $0.thenpart);
4691 | if OpenScope Expression OpenScope OptNL then Block ${
4694 var_block_close(c, CloseParallel, $0.thenpart);
4698 // This scope is closed in CondStatement
4699 SwitchPart -> switch OpenScope Expression ${
4702 | switch UseBlock ${
4706 ###### print binode cases
4708 if (b->left && b->left->type == Xbinode &&
4709 cast(binode, b->left)->op == Block) {
4711 do_indent(indent, "while {\n");
4713 do_indent(indent, "while\n");
4714 print_exec(b->left, indent+1, bracket);
4716 do_indent(indent, "} do {\n");
4718 do_indent(indent, "do\n");
4719 print_exec(b->right, indent+1, bracket);
4721 do_indent(indent, "}\n");
4723 do_indent(indent, "while ");
4724 print_exec(b->left, 0, bracket);
4729 print_exec(b->right, indent+1, bracket);
4731 do_indent(indent, "}\n");
4735 ###### print exec cases
4737 case Xcond_statement:
4739 struct cond_statement *cs = cast(cond_statement, e);
4740 struct casepart *cp;
4742 do_indent(indent, "for");
4743 if (bracket) printf(" {\n"); else printf("\n");
4744 print_exec(cs->forpart, indent+1, bracket);
4747 do_indent(indent, "} then {\n");
4749 do_indent(indent, "then\n");
4750 print_exec(cs->thenpart, indent+1, bracket);
4752 if (bracket) do_indent(indent, "}\n");
4755 print_exec(cs->looppart, indent, bracket);
4759 do_indent(indent, "switch");
4761 do_indent(indent, "if");
4762 if (cs->condpart && cs->condpart->type == Xbinode &&
4763 cast(binode, cs->condpart)->op == Block) {
4768 print_exec(cs->condpart, indent+1, bracket);
4770 do_indent(indent, "}\n");
4772 do_indent(indent, "then\n");
4773 print_exec(cs->thenpart, indent+1, bracket);
4777 print_exec(cs->condpart, 0, bracket);
4783 print_exec(cs->thenpart, indent+1, bracket);
4785 do_indent(indent, "}\n");
4790 for (cp = cs->casepart; cp; cp = cp->next) {
4791 do_indent(indent, "case ");
4792 print_exec(cp->value, -1, 0);
4797 print_exec(cp->action, indent+1, bracket);
4799 do_indent(indent, "}\n");
4802 do_indent(indent, "else");
4807 print_exec(cs->elsepart, indent+1, bracket);
4809 do_indent(indent, "}\n");
4814 ###### propagate binode cases
4816 t = propagate_types(b->right, c, perr, Tnone, 0);
4817 if (!type_compat(Tnone, t, 0))
4818 *perr |= Efail; // UNTESTED
4819 return propagate_types(b->left, c, perr, type, rules);
4821 ###### propagate exec cases
4822 case Xcond_statement:
4824 // forpart and looppart->right must return Tnone
4825 // thenpart must return Tnone if there is a loopart,
4826 // otherwise it is like elsepart.
4828 // be bool if there is no casepart
4829 // match casepart->values if there is a switchpart
4830 // either be bool or match casepart->value if there
4832 // elsepart and casepart->action must match the return type
4833 // expected of this statement.
4834 struct cond_statement *cs = cast(cond_statement, prog);
4835 struct casepart *cp;
4837 t = propagate_types(cs->forpart, c, perr, Tnone, 0);
4838 if (!type_compat(Tnone, t, 0))
4839 *perr |= Efail; // UNTESTED
4842 t = propagate_types(cs->thenpart, c, perr, Tnone, 0);
4843 if (!type_compat(Tnone, t, 0))
4844 *perr |= Efail; // UNTESTED
4846 if (cs->casepart == NULL) {
4847 propagate_types(cs->condpart, c, perr, Tbool, 0);
4848 propagate_types(cs->looppart, c, perr, Tbool, 0);
4850 /* Condpart must match case values, with bool permitted */
4852 for (cp = cs->casepart;
4853 cp && !t; cp = cp->next)
4854 t = propagate_types(cp->value, c, perr, NULL, 0);
4855 if (!t && cs->condpart)
4856 t = propagate_types(cs->condpart, c, perr, NULL, Rboolok); // UNTESTED
4857 if (!t && cs->looppart)
4858 t = propagate_types(cs->looppart, c, perr, NULL, Rboolok); // UNTESTED
4859 // Now we have a type (I hope) push it down
4861 for (cp = cs->casepart; cp; cp = cp->next)
4862 propagate_types(cp->value, c, perr, t, 0);
4863 propagate_types(cs->condpart, c, perr, t, Rboolok);
4864 propagate_types(cs->looppart, c, perr, t, Rboolok);
4867 // (if)then, else, and case parts must return expected type.
4868 if (!cs->looppart && !type)
4869 type = propagate_types(cs->thenpart, c, perr, NULL, rules);
4871 type = propagate_types(cs->elsepart, c, perr, NULL, rules);
4872 for (cp = cs->casepart;
4874 cp = cp->next) // UNTESTED
4875 type = propagate_types(cp->action, c, perr, NULL, rules); // UNTESTED
4878 propagate_types(cs->thenpart, c, perr, type, rules);
4879 propagate_types(cs->elsepart, c, perr, type, rules);
4880 for (cp = cs->casepart; cp ; cp = cp->next)
4881 propagate_types(cp->action, c, perr, type, rules);
4887 ###### interp binode cases
4889 // This just performs one iterration of the loop
4890 rv = interp_exec(c, b->left, &rvtype);
4891 if (rvtype == Tnone ||
4892 (rvtype == Tbool && rv.bool != 0))
4893 // rvtype is Tnone or Tbool, doesn't need to be freed
4894 interp_exec(c, b->right, NULL);
4897 ###### interp exec cases
4898 case Xcond_statement:
4900 struct value v, cnd;
4901 struct type *vtype, *cndtype;
4902 struct casepart *cp;
4903 struct cond_statement *cs = cast(cond_statement, e);
4906 interp_exec(c, cs->forpart, NULL);
4908 while ((cnd = interp_exec(c, cs->looppart, &cndtype)),
4909 cndtype == Tnone || (cndtype == Tbool && cnd.bool != 0))
4910 interp_exec(c, cs->thenpart, NULL);
4912 cnd = interp_exec(c, cs->condpart, &cndtype);
4913 if ((cndtype == Tnone ||
4914 (cndtype == Tbool && cnd.bool != 0))) {
4915 // cnd is Tnone or Tbool, doesn't need to be freed
4916 rv = interp_exec(c, cs->thenpart, &rvtype);
4917 // skip else (and cases)
4921 for (cp = cs->casepart; cp; cp = cp->next) {
4922 v = interp_exec(c, cp->value, &vtype);
4923 if (value_cmp(cndtype, vtype, &v, &cnd) == 0) {
4924 free_value(vtype, &v);
4925 free_value(cndtype, &cnd);
4926 rv = interp_exec(c, cp->action, &rvtype);
4929 free_value(vtype, &v);
4931 free_value(cndtype, &cnd);
4933 rv = interp_exec(c, cs->elsepart, &rvtype);
4940 ### Top level structure
4942 All the language elements so far can be used in various places. Now
4943 it is time to clarify what those places are.
4945 At the top level of a file there will be a number of declarations.
4946 Many of the things that can be declared haven't been described yet,
4947 such as functions, procedures, imports, and probably more.
4948 For now there are two sorts of things that can appear at the top
4949 level. They are predefined constants, `struct` types, and the `main`
4950 function. While the syntax will allow the `main` function to appear
4951 multiple times, that will trigger an error if it is actually attempted.
4953 The various declarations do not return anything. They store the
4954 various declarations in the parse context.
4956 ###### Parser: grammar
4959 Ocean -> OptNL DeclarationList
4961 ## declare terminals
4969 DeclarationList -> Declaration
4970 | DeclarationList Declaration
4972 Declaration -> ERROR Newlines ${
4973 tok_err(c, // UNTESTED
4974 "error: unhandled parse error", &$1);
4980 ## top level grammar
4984 ### The `const` section
4986 As well as being defined in with the code that uses them, constants can
4987 be declared at the top level. These have full-file scope, so they are
4988 always `InScope`, even before(!) they have been declared. The value of
4989 a top level constant can be given as an expression, and this is
4990 evaluated after parsing and before execution.
4992 A function call can be used to evaluate a constant, but it will not have
4993 access to any program state, once such statement becomes meaningful.
4994 e.g. arguments and filesystem will not be visible.
4996 Constants are defined in a section that starts with the reserved word
4997 `const` and then has a block with a list of assignment statements.
4998 For syntactic consistency, these must use the double-colon syntax to
4999 make it clear that they are constants. Type can also be given: if
5000 not, the type will be determined during analysis, as with other
5003 ###### parse context
5004 struct binode *constlist;
5006 ###### top level grammar
5010 DeclareConstant -> const { IN OptNL ConstList OUT OptNL } Newlines
5011 | const { SimpleConstList } Newlines
5012 | const IN OptNL ConstList OUT Newlines
5013 | const SimpleConstList Newlines
5015 ConstList -> ConstList SimpleConstLine
5018 SimpleConstList -> SimpleConstList ; Const
5022 SimpleConstLine -> SimpleConstList Newlines
5023 | ERROR Newlines ${ tok_err(c, "Syntax error in constant", &$1); }$
5026 CType -> Type ${ $0 = $<1; }$
5030 Const -> IDENTIFIER :: CType = Expression ${ {
5032 struct binode *bl, *bv;
5033 struct var *var = new_pos(var, $ID);
5035 v = var_decl(c, $ID.txt);
5037 v->where_decl = var;
5043 v = var_ref(c, $1.txt);
5044 if (v->type == Tnone) {
5045 v->where_decl = var;
5051 tok_err(c, "error: name already declared", &$1);
5052 type_err(c, "info: this is where '%v' was first declared",
5053 v->where_decl, NULL, 0, NULL);
5065 bl->left = c->constlist;
5070 ###### core functions
5071 static void resolve_consts(struct parse_context *c)
5075 enum { none, some, cannot } progress = none;
5077 c->constlist = reorder_bilist(c->constlist);
5080 for (b = cast(binode, c->constlist); b;
5081 b = cast(binode, b->right)) {
5083 struct binode *vb = cast(binode, b->left);
5084 struct var *v = cast(var, vb->left);
5085 if (v->var->frame_pos >= 0)
5089 propagate_types(vb->right, c, &perr,
5091 } while (perr & Eretry);
5093 c->parse_error += 1;
5094 else if (!(perr & Enoconst)) {
5096 struct value res = interp_exec(
5097 c, vb->right, &v->var->type);
5098 global_alloc(c, v->var->type, v->var, &res);
5100 if (progress == cannot)
5101 type_err(c, "error: const %v cannot be resolved.",
5111 progress = cannot; break;
5113 progress = none; break;
5118 ###### print const decls
5123 for (b = cast(binode, context.constlist); b;
5124 b = cast(binode, b->right)) {
5125 struct binode *vb = cast(binode, b->left);
5126 struct var *vr = cast(var, vb->left);
5127 struct variable *v = vr->var;
5133 printf(" %.*s :: ", v->name->name.len, v->name->name.txt);
5134 type_print(v->type, stdout);
5136 print_exec(vb->right, -1, 0);
5141 ###### free const decls
5142 free_binode(context.constlist);
5144 ### Function declarations
5146 The code in an Ocean program is all stored in function declarations.
5147 One of the functions must be named `main` and it must accept an array of
5148 strings as a parameter - the command line arguments.
5150 As this is the top level, several things are handled a bit differently.
5151 The function is not interpreted by `interp_exec` as that isn't passed
5152 the argument list which the program requires. Similarly type analysis
5153 is a bit more interesting at this level.
5155 ###### ast functions
5157 static struct type *handle_results(struct parse_context *c,
5158 struct binode *results)
5160 /* Create a 'struct' type from the results list, which
5161 * is a list for 'struct var'
5163 struct type *t = add_anon_type(c, &structure_prototype,
5168 for (b = results; b; b = cast(binode, b->right))
5170 t->structure.nfields = cnt;
5171 t->structure.fields = calloc(cnt, sizeof(struct field));
5173 for (b = results; b; b = cast(binode, b->right)) {
5174 struct var *v = cast(var, b->left);
5175 struct field *f = &t->structure.fields[cnt++];
5176 int a = v->var->type->align;
5177 f->name = v->var->name->name;
5178 f->type = v->var->type;
5180 f->offset = t->size;
5181 v->var->frame_pos = f->offset;
5182 t->size += ((f->type->size - 1) | (a-1)) + 1;
5185 variable_unlink_exec(v->var);
5187 free_binode(results);
5191 static struct variable *declare_function(struct parse_context *c,
5192 struct variable *name,
5193 struct binode *args,
5195 struct binode *results,
5199 struct value fn = {.function = code};
5201 var_block_close(c, CloseFunction, code);
5202 t = add_anon_type(c, &function_prototype,
5203 "func %.*s", name->name->name.len,
5204 name->name->name.txt);
5206 t->function.params = reorder_bilist(args);
5208 ret = handle_results(c, reorder_bilist(results));
5209 t->function.inline_result = 1;
5210 t->function.local_size = ret->size;
5212 t->function.return_type = ret;
5213 global_alloc(c, t, name, &fn);
5214 name->type->function.scope = c->out_scope;
5219 var_block_close(c, CloseFunction, NULL);
5221 c->out_scope = NULL;
5225 ###### declare terminals
5228 ###### top level grammar
5231 DeclareFunction -> func FuncName ( OpenScope ArgsLine ) Block Newlines ${
5232 $0 = declare_function(c, $<FN, $<Ar, Tnone, NULL, $<Bl);
5234 | func FuncName IN OpenScope Args OUT OptNL do Block Newlines ${
5235 $0 = declare_function(c, $<FN, $<Ar, Tnone, NULL, $<Bl);
5237 | func FuncName NEWLINE OpenScope OptNL do Block Newlines ${
5238 $0 = declare_function(c, $<FN, NULL, Tnone, NULL, $<Bl);
5240 | func FuncName ( OpenScope ArgsLine ) : Type Block Newlines ${
5241 $0 = declare_function(c, $<FN, $<Ar, $<Ty, NULL, $<Bl);
5243 | func FuncName ( OpenScope ArgsLine ) : ( ArgsLine ) Block Newlines ${
5244 $0 = declare_function(c, $<FN, $<AL, NULL, $<AL2, $<Bl);
5246 | func FuncName IN OpenScope Args OUT OptNL return Type Newlines do Block Newlines ${
5247 $0 = declare_function(c, $<FN, $<Ar, $<Ty, NULL, $<Bl);
5249 | func FuncName NEWLINE OpenScope return Type Newlines do Block Newlines ${
5250 $0 = declare_function(c, $<FN, NULL, $<Ty, NULL, $<Bl);
5252 | func FuncName IN OpenScope Args OUT OptNL return IN Args OUT OptNL do Block Newlines ${
5253 $0 = declare_function(c, $<FN, $<Ar, NULL, $<Ar2, $<Bl);
5255 | func FuncName NEWLINE OpenScope return IN Args OUT OptNL do Block Newlines ${
5256 $0 = declare_function(c, $<FN, NULL, NULL, $<Ar, $<Bl);
5259 ###### print func decls
5264 while (target != 0) {
5266 for (v = context.in_scope; v; v=v->in_scope)
5267 if (v->depth == 0 && v->type && v->type->check_args) {
5276 struct value *val = var_value(&context, v);
5277 printf("func %.*s", v->name->name.len, v->name->name.txt);
5278 v->type->print_type_decl(v->type, stdout);
5280 print_exec(val->function, 0, brackets);
5282 print_value(v->type, val, stdout);
5283 printf("/* frame size %d */\n", v->type->function.local_size);
5289 ###### core functions
5291 static int analyse_funcs(struct parse_context *c)
5295 for (v = c->in_scope; v; v = v->in_scope) {
5299 if (v->depth != 0 || !v->type || !v->type->check_args)
5301 ret = v->type->function.inline_result ?
5302 Tnone : v->type->function.return_type;
5303 val = var_value(c, v);
5306 propagate_types(val->function, c, &perr, ret, 0);
5307 } while (!(perr & Efail) && (perr & Eretry));
5308 if (!(perr & Efail))
5309 /* Make sure everything is still consistent */
5310 propagate_types(val->function, c, &perr, ret, 0);
5313 if (!v->type->function.inline_result &&
5314 !v->type->function.return_type->dup) {
5315 type_err(c, "error: function cannot return value of type %1",
5316 v->where_decl, v->type->function.return_type, 0, NULL);
5319 scope_finalize(c, v->type);
5324 static int analyse_main(struct type *type, struct parse_context *c)
5326 struct binode *bp = type->function.params;
5330 struct type *argv_type;
5332 argv_type = add_anon_type(c, &array_prototype, "argv");
5333 argv_type->array.member = Tstr;
5334 argv_type->array.unspec = 1;
5336 for (b = bp; b; b = cast(binode, b->right)) {
5340 propagate_types(b->left, c, &perr, argv_type, 0);
5342 default: /* invalid */ // NOTEST
5343 propagate_types(b->left, c, &perr, Tnone, 0); // NOTEST
5346 c->parse_error += 1;
5349 return !c->parse_error;
5352 static void interp_main(struct parse_context *c, int argc, char **argv)
5354 struct value *progp = NULL;
5355 struct text main_name = { "main", 4 };
5356 struct variable *mainv;
5362 mainv = var_ref(c, main_name);
5364 progp = var_value(c, mainv);
5365 if (!progp || !progp->function) {
5366 fprintf(stderr, "oceani: no main function found.\n");
5367 c->parse_error += 1;
5370 if (!analyse_main(mainv->type, c)) {
5371 fprintf(stderr, "oceani: main has wrong type.\n");
5372 c->parse_error += 1;
5375 al = mainv->type->function.params;
5377 c->local_size = mainv->type->function.local_size;
5378 c->local = calloc(1, c->local_size);
5380 struct var *v = cast(var, al->left);
5381 struct value *vl = var_value(c, v->var);
5391 mpq_set_ui(argcq, argc, 1);
5392 memcpy(var_value(c, t->array.vsize), &argcq, sizeof(argcq));
5393 t->prepare_type(c, t, 0);
5394 array_init(v->var->type, vl);
5395 for (i = 0; i < argc; i++) {
5396 struct value *vl2 = vl->array + i * v->var->type->array.member->size;
5398 arg.str.txt = argv[i];
5399 arg.str.len = strlen(argv[i]);
5400 free_value(Tstr, vl2);
5401 dup_value(Tstr, &arg, vl2);
5405 al = cast(binode, al->right);
5407 v = interp_exec(c, progp->function, &vtype);
5408 free_value(vtype, &v);
5413 ###### ast functions
5414 void free_variable(struct variable *v)
5418 ## And now to test it out.
5420 Having a language requires having a "hello world" program. I'll
5421 provide a little more than that: a program that prints "Hello world"
5422 finds the GCD of two numbers, prints the first few elements of
5423 Fibonacci, performs a binary search for a number, and a few other
5424 things which will likely grow as the languages grows.
5426 ###### File: oceani.mk
5429 @echo "===== DEMO ====="
5430 ./oceani --section "demo: hello" oceani.mdc 55 33
5436 four ::= 2 + 2 ; five ::= 10/2
5437 const pie ::= "I like Pie";
5438 cake ::= "The cake is"
5446 func main(argv:[argc::]string)
5447 print "Hello World, what lovely oceans you have!"
5448 print "Are there", five, "?"
5449 print pi, pie, "but", cake
5451 A := $argv[1]; B := $argv[2]
5453 /* When a variable is defined in both branches of an 'if',
5454 * and used afterwards, the variables are merged.
5460 print "Is", A, "bigger than", B,"? ", bigger
5461 /* If a variable is not used after the 'if', no
5462 * merge happens, so types can be different
5465 double:string = "yes"
5466 print A, "is more than twice", B, "?", double
5469 print "double", B, "is", double
5474 if a > 0 and then b > 0:
5480 print "GCD of", A, "and", B,"is", a
5482 print a, "is not positive, cannot calculate GCD"
5484 print b, "is not positive, cannot calculate GCD"
5489 print "Fibonacci:", f1,f2,
5490 then togo = togo - 1
5498 /* Binary search... */
5503 mid := (lo + hi) / 2
5516 print "Yay, I found", target
5518 print "Closest I found was", lo
5523 // "middle square" PRNG. Not particularly good, but one my
5524 // Dad taught me - the first one I ever heard of.
5525 for i:=1; then i = i + 1; while i < size:
5526 n := list[i-1] * list[i-1]
5527 list[i] = (n / 100) % 10 000
5529 print "Before sort:",
5530 for i:=0; then i = i + 1; while i < size:
5534 for i := 1; then i=i+1; while i < size:
5535 for j:=i-1; then j=j-1; while j >= 0:
5536 if list[j] > list[j+1]:
5540 print " After sort:",
5541 for i:=0; then i = i + 1; while i < size:
5545 if 1 == 2 then print "yes"; else print "no"
5549 bob.alive = (bob.name == "Hello")
5550 print "bob", "is" if bob.alive else "isn't", "alive"