From: Mart Lubbers Date: Wed, 2 Feb 2022 08:24:30 +0000 (+0100) Subject: document (all but ast), test script, test output X-Git-Url: https://git.martlubbers.net/?a=commitdiff_plain;h=589160805c4bce9e5a59dd8fbf23286361bb0f3a;p=ccc.git document (all but ast), test script, test output --- diff --git a/.gitignore b/.gitignore index f9d23a1..827f8bf 100644 --- a/.gitignore +++ b/.gitignore @@ -14,3 +14,5 @@ tests/*.c callgrind.out.* massif.out.* + +doc diff --git a/Doxyfile b/Doxyfile index e5b9be5..2b8c0aa 100644 --- a/Doxyfile +++ b/Doxyfile @@ -836,7 +836,7 @@ WARN_NO_PARAMDOC = NO # Possible values are: NO, YES and FAIL_ON_WARNINGS. # The default value is: NO. -WARN_AS_ERROR = NO +WARN_AS_ERROR = YES # The WARN_FORMAT tag determines the format of the warning messages that doxygen # can produce. The string should contain the $file, $line, and $text tags, which @@ -943,7 +943,7 @@ FILE_PATTERNS = *.c \ # be searched for input files as well. # The default value is: NO. -RECURSIVE = NO +RECURSIVE = YES # The EXCLUDE tag can be used to specify files and/or directories that should be # excluded from the INPUT source files. This way you can easily exclude a @@ -968,7 +968,7 @@ EXCLUDE_SYMLINKS = NO # Note that the wildcards are matched against the file with absolute path, so to # exclude all test directories for example use the pattern */test/* -EXCLUDE_PATTERNS = +EXCLUDE_PATTERNS = */scan.h/* */parse.h/* # The EXCLUDE_SYMBOLS tag can be used to specify one or more symbol names # (namespaces, classes, functions, etc.) that should be excluded from the diff --git a/Makefile b/Makefile index cc4661b..adc8f9f 100644 --- a/Makefile +++ b/Makefile @@ -3,3 +3,4 @@ all: clean: $(MAKE) -C src clean + $(RM) a.c a.ssm a.out diff --git a/compilerunc.bash b/compilerunc.bash deleted file mode 100755 index df4e01e..0000000 --- a/compilerunc.bash +++ /dev/null @@ -1,28 +0,0 @@ -#!/bin/bash -usage() { - echo "Usage: $0 CSOURCE [-o OFILE]" >&2 - exit 1 -} -if [ $# -lt 1 ] -then - usage -fi -if [ $# -eq 3 ] -then - if [ $2 != "-o" ] - then - usage - fi - OFILE=$3 -else - OFILE=a.out -fi -make -./splc < "$1" -CFLAGS=${CFLAGS:-} -LDLIBS=${LDLIBS:-} -LDFLAGS=${LDFLAGS:-} -CC=${CC:-gcc} -set -xe -"$CC" $CFLAGS a.c $LDFLAGS rts.c $LDLIBS -o "$OFILE" -./"$OFILE" diff --git a/rts.c b/rts.c index 0d5e5cb..8a1093a 100644 --- a/rts.c +++ b/rts.c @@ -18,7 +18,7 @@ void splc_free(void *ptr) // if (REFC(ptr) == 0) // free(ptr-1); } -struct splc_tuple *splc_tuple(WORD fst, WORD snd) { +struct splc_tuple *splc_tuple_real(WORD fst, WORD snd) { struct splc_tuple *res = splc_malloc(sizeof(struct splc_tuple)); res->fst = fst; res->snd = snd; @@ -43,3 +43,8 @@ WORD hd(struct splc_list *l) } return l->hd; } +int main() +{ + realmain(); + return 0; +} diff --git a/rts.h b/rts.h index 32ddb99..d1f2d11 100644 --- a/rts.h +++ b/rts.h @@ -11,20 +11,24 @@ struct splc_tuple { WORD fst; WORD snd; }; struct splc_list { WORD hd; struct splc_list *tl; }; -struct splc_tuple *splc_tuple(WORD fst, WORD snd); +#define splc_tuple(l, r) splc_tuple_real((intptr_t)(l), (intptr_t)(r)) +struct splc_tuple *splc_tuple_real(WORD fst, WORD snd); struct splc_list *splc_cons(WORD hd, struct splc_list *tl); WORD hd(struct splc_list *l); struct splc_list *tl(struct splc_list *l); #define die(...) { fprintf(stderr, __VA_ARGS__); exit(1); } #define isEmpty(l) ((l) == NULL) #define tl(l) (l)->tl -#define fst(l) (l)->fst -#define snd(l) (l)->snd -#define print_Int(l) printf("%ld", l); -#define print_Char(l) printf("%c", l); +#define fst(l) ((struct splc_tuple *)(l))->fst +#define snd(l) ((struct splc_tuple *)(l))->snd +#define print_Int(l) printf("%ld", (long)l); +#define print_Char(l) printf("%c", (char)l); #define print_Bool(l) printf("%s", (l) ? "true" : "false"); #define eq_Int(x, y) ((x)==(y)) #define eq_Char(x, y) ((x)==(y)) #define eq_Bool(x, y) ((x)==(y)) +void realmain(void); +int main(void); + #endif diff --git a/src/array.h b/src/array.h index 37a5878..5793645 100644 --- a/src/array.h +++ b/src/array.h @@ -32,9 +32,8 @@ * @param iter Name of the iterator variable * @param array Pointer to the array */ -/** Iterate over the indices and elements of an array */ -#define ARRAY_ITER(type, x, iter, a) ARRAY_ITERI (iter, a) {\ - type (x) = ARRAY_EL(type, a, iter); +#define ARRAY_ITER(type, x, iter, array) ARRAY_ITERI (iter, array) {\ + type (x) = ARRAY_EL(type, array, iter); /** * Macro to end an array iteration * @@ -146,7 +145,7 @@ void array_insert(struct array *array, size_t idx, void *x); * @param array Pointer to the array * @param idx Index of removal */ -void array_remove(struct array *a, size_t idx); +void array_remove(struct array *array, size_t idx); /** * Move an item from a position to another, moving the other items. @@ -155,7 +154,7 @@ void array_remove(struct array *a, size_t idx); * @param from Original index of the item * @param to New index of the item */ -void array_move(struct array *a, size_t from, size_t to); +void array_move(struct array *array, size_t from, size_t to); /** * Insert an item into a sorted array. @@ -164,6 +163,6 @@ void array_move(struct array *a, size_t from, size_t to); * @param array Pointer to the array * @param cmp Comparison function */ -void array_binsert(void *key, struct array *a, int (*cmp)(const void *, const void *)); +void array_binsert(void *key, struct array *array, int (*cmp)(const void *, const void *)); #endif diff --git a/src/ast.c b/src/ast.c index 5ae01f2..dca36d4 100644 --- a/src/ast.c +++ b/src/ast.c @@ -6,7 +6,6 @@ #include "util.h" #include "ast.h" #include "type.h" -#include "list.h" #include "parse.h" const char *binop_str[] = { diff --git a/src/gen.h b/src/gen.h index ef13f19..dc9c723 100644 --- a/src/gen.h +++ b/src/gen.h @@ -7,8 +7,8 @@ /** Datatype that stores the overloading information */ struct overload { - struct array print; /** overloaded print calls, @type struct overload_entry * */ - struct array eq; /** overloaded equality calls, @type struct overload_entry * */ + struct array print; /** overloaded print calls, type: struct overload_entry * */ + struct array eq; /** overloaded equality calls, type: struct overload_entry * */ }; /** Single overloading entry */ struct overload_entry { diff --git a/src/gen/c.c b/src/gen/c.c index a366bd1..0e0340b 100644 --- a/src/gen/c.c +++ b/src/gen/c.c @@ -6,6 +6,13 @@ #include "../sem.h" #include "../gen.h" +static const char *fun_name(const char *name) +{ + if (strcmp(name, "abs") == 0) + return "splc_abs"; + return name; +} + static void expr_genc(struct expr *expr, FILE *cout); static void binop_genc(char *fun, struct expr *l, struct expr *r, FILE *cout) @@ -60,7 +67,7 @@ static void expr_genc(struct expr *expr, FILE *cout) safe_fprintf(cout, "print_"); overloaded_type(expr->loc, expr->data.efuncall.type, cout); } else { - safe_fprintf(cout, "%s", expr->data.efuncall.ident); + safe_fprintf(cout, "%s", fun_name(expr->data.efuncall.ident)); } safe_fprintf(cout, "("); ARRAY_ITER(struct expr *, e, i, &expr->data.efuncall.args) { @@ -142,6 +149,14 @@ static void vardecl_genc(struct vardecl *vardecl, int indent, FILE *cout) safe_fprintf(cout, ";\n"); } +static void fields_genc(struct array *fields, const char *ident, FILE *cout) +{ + safe_fprintf(cout, "%s", ident); + ARRAY_ITER(char *, f, i, fields) + safe_fprintf(cout, "->%s", f); + AIEND +} + static void stmt_genc(struct stmt *stmt, int indent, FILE *cout) { if (stmt == NULL) @@ -149,10 +164,7 @@ static void stmt_genc(struct stmt *stmt, int indent, FILE *cout) switch(stmt->type) { case sassign: pindent(indent, cout); - safe_fprintf(cout, "%s", stmt->data.sassign.ident); - ARRAY_ITER(char *, f, i, &stmt->data.sassign.fields) - safe_fprintf(cout, "->%s", f); - AIEND + fields_genc(&stmt->data.sassign.fields, stmt->data.sassign.ident, cout); safe_fprintf(cout, " = "); expr_genc(stmt->data.sassign.expr, cout); safe_fprintf(cout, ";\n"); @@ -210,7 +222,7 @@ static void stmt_genc(struct stmt *stmt, int indent, FILE *cout) static void fundecl_sig(struct fundecl *decl, FILE *cout) { type_genc(decl->rtype, cout); - safe_fprintf(cout, "%s (", decl->ident); + safe_fprintf(cout, "%s (", fun_name(decl->ident)); ARRAY_ITER(char *, a, i, &decl->args) { if (i >= ARRAY_SIZE(decl->atypes)) die("function with unmatched type\n"); @@ -239,7 +251,7 @@ static void initialise_non_constant_globals(const struct ast *ast, FILE *cout) static void fundecl_genc(const struct ast *ast, struct fundecl *decl, FILE *cout) { if (strcmp(decl->ident, "main") == 0) - safe_fprintf(cout, "int main()"); + safe_fprintf(cout, "void real_main()"); else fundecl_sig(decl, cout); safe_fprintf(cout, "{\n"); @@ -248,12 +260,9 @@ static void fundecl_genc(const struct ast *ast, struct fundecl *decl, FILE *cout ARRAY_ITER(struct stmt *, s, i, &decl->body) stmt_genc(s, 1, cout); AIEND - if (strcmp(decl->ident, "main") == 0) - safe_fprintf(cout, "\treturn 0;\n"); safe_fprintf(cout, "}\n"); } - static void decl_genc(const struct ast *ast, struct decl *decl, FILE *cout) { switch (decl->type) { @@ -295,9 +304,9 @@ static void generate_eq(struct type *type, FILE *cout) safe_fprintf(cout, "\twhile(t != NULL) {\n"); safe_fprintf(cout, "\t\tif (!eq_"); overloaded_type(loc, type->data.tlist, cout); - safe_fprintf(cout, "(x->hd, y->hd));\n"); + safe_fprintf(cout, "(hd(x), hd(y)));\n"); safe_fprintf(cout, "\t\t\treturn false;\n"); - safe_fprintf(cout, "\t\tt = t->tl;\n"); + safe_fprintf(cout, "\t\tt = tl(t);\n"); safe_fprintf(cout, "\t}\n"); safe_fprintf(cout, "\treturn true;\n"); break; @@ -307,7 +316,7 @@ static void generate_eq(struct type *type, FILE *cout) safe_fprintf(cout, "(x->fy->fst)"); safe_fprintf(cout, " && eq_"); overloaded_type(loc, type->data.ttuple.r, cout); - safe_fprintf(cout, "(x->snd, y->snd);"); + safe_fprintf(cout, "(snd(x), snd(y));"); break; default: die("cannot compare anything else than tuples and lists"); @@ -332,10 +341,12 @@ static void generate_print(struct type *type, FILE *cout) safe_fprintf(cout, "\twhile(t != NULL) {\n"); safe_fprintf(cout, "\t\tprint_"); overloaded_type(loc, type->data.tlist, cout); - safe_fprintf(cout, "(t->hd);\n"); - safe_fprintf(cout, "\t\tif (t->tl != NULL)\n"); + safe_fprintf(cout, "(("); + type_genc(type->data.tlist, cout); + safe_fprintf(cout, ")hd(t));\n"); + safe_fprintf(cout, "\t\tif (tl(t) != NULL)\n"); safe_fprintf(cout, "\t\t\tprintf(\", \");\n"); - safe_fprintf(cout, "\t\tt = t->tl;\n"); + safe_fprintf(cout, "\t\tt = tl(t);\n"); safe_fprintf(cout, "\t}\n"); safe_fprintf(cout, "\tprintf(\"]\");\n"); break; @@ -343,11 +354,15 @@ static void generate_print(struct type *type, FILE *cout) safe_fprintf(cout, "\tprintf(\"(\");\n"); safe_fprintf(cout, "\tprint_"); overloaded_type(loc, type->data.ttuple.l, cout); - safe_fprintf(cout, "(t->fst);\n"); + safe_fprintf(cout, "(("); + type_genc(type->data.ttuple.l, cout); + safe_fprintf(cout, ")fst(t));\n"); safe_fprintf(cout, "\tprintf(\",\");\n"); safe_fprintf(cout, "\tprint_"); overloaded_type(loc, type->data.ttuple.r, cout); - safe_fprintf(cout, "(t->snd);\n"); + safe_fprintf(cout, "(("); + type_genc(type->data.ttuple.r, cout); + safe_fprintf(cout, ")snd(t));\n"); safe_fprintf(cout, "\tprintf(\")\");\n"); break; default: diff --git a/src/gen/c.h b/src/gen/c.h index 7b400b2..167c3ae 100644 --- a/src/gen/c.h +++ b/src/gen/c.h @@ -6,6 +6,13 @@ #include "../ast.h" #include "../gen.h" +/** + * Generate C code for the abstract syntax tree + * + * @param res abstract syntax tree + * @param ol dictionary of overloaded function calls + * @param cout stream to output the result to + */ void genc(const struct ast *res, const struct overload ol, FILE *cout); #endif diff --git a/src/gen/ssm.h b/src/gen/ssm.h index 64ae852..70bbd16 100644 --- a/src/gen/ssm.h +++ b/src/gen/ssm.h @@ -6,6 +6,13 @@ #include "../ast.h" #include "../gen.h" +/** + * Generate SSM code for the abstract syntax tree + * + * @param res abstract syntax tree + * @param ol dictionary of overloaded function calls + * @param cout stream to output the result to + */ void genssm(const struct ast *res, const struct overload ol, FILE *cout); #endif diff --git a/src/ident.h b/src/ident.h index d334003..7ecb6ec 100644 --- a/src/ident.h +++ b/src/ident.h @@ -48,7 +48,7 @@ struct ident ident_dup(struct ident i); * Print an identifier * * @param i identifier - * @param stream to print to + * @param cout to print to */ void ident_print(struct ident i, FILE *cout); /** diff --git a/src/list.c b/src/list.c deleted file mode 100644 index d58ad78..0000000 --- a/src/list.c +++ /dev/null @@ -1,51 +0,0 @@ -#include -#include - -#include "list.h" -#include "util.h" - -struct list *list_cons(void *el, struct list *tail) -{ - struct list *res = xalloc(1, struct list); - res->el = el; - res->tail = tail; - return res; -} - -void list_free(struct list *head, void (*freefun)(void *)) -{ - while (head != NULL) { - struct list *t = head; - if (freefun != NULL) - freefun(head->el); - head = head->tail; - free(t); - } -} - -void **list_to_array(struct list *list, int *num, bool reverse, int extra) -{ - int i = list_length(list); - *num = i; - void **ptr = xalloc(i+extra, void *); - - struct list *r = list; - while(i > 0) { - if (reverse) - ptr[--i] = r->el; - else - ptr[*num-(--i)-1] = r->el; - struct list *t = r; - r = r->tail; - free(t); - } - return ptr; -} - -int list_length(struct list *r) -{ - int i = 0; - FOREACH(e, r) - i++; - return i; -} diff --git a/src/list.h b/src/list.h deleted file mode 100644 index 6e87129..0000000 --- a/src/list.h +++ /dev/null @@ -1,18 +0,0 @@ -#ifndef LIST_H -#define LIST_H - -#include - -#define FOREACH(x, l) for(struct list *x = l; x != NULL; x = x->tail) - -struct list { - void *el; - struct list *tail; -}; - -struct list *list_cons(void *el, struct list *tail); -void list_free(struct list *head, void (*freefun)(void *)); -void **list_to_array(struct list *list, int *num, bool reverse, int extra); -int list_length(struct list *head); - -#endif diff --git a/src/parse.y b/src/parse.y index 8008a0e..7e29cde 100644 --- a/src/parse.y +++ b/src/parse.y @@ -3,7 +3,6 @@ #include "array.h" #include "ast.h" -#include "list.h" #include "parse.h" int yylex(void); diff --git a/src/sem.h b/src/sem.h index 3c1a929..9bc7895 100644 --- a/src/sem.h +++ b/src/sem.h @@ -5,7 +5,22 @@ #include "ast.h" +/** + * Emit a type error + * + * @param l location + * @param d also exit with error code 1 + * @param msg format string + * @param ... format arguments + */ void type_error(YYLTYPE l, bool d, const char *msg, ...); + +/** + * Perform all semantical analyses. + * + * @param ast pointer to the unchecked abstract syntax tree + * @result pointer to the checked abstract syntax tree + */ struct ast *sem(struct ast *ast); #endif diff --git a/src/sem/constant.h b/src/sem/constant.h index bdaf5b5..ed11d0e 100644 --- a/src/sem/constant.h +++ b/src/sem/constant.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Check that all global variable initialisers are constant + * + * @param ast abstract syntax tree + */ void sem_check_constant(struct ast *ast); #endif diff --git a/src/sem/hm.h b/src/sem/hm.h index eab8cb9..fe7bedd 100644 --- a/src/sem/hm.h +++ b/src/sem/hm.h @@ -7,11 +7,48 @@ #include "hm/scheme.h" #include "../ident.h" +/** + * Infer all types in the AST + * + * @param ast abstract syntax tree + * @result abstract syntax tree with all types inferred + */ struct ast *infer(struct ast *ast); +/** + * Unify a type + * + * @param loc location of the type to localise possible errors + * @param l left-hand side + * @param r right-hand side + */ struct subst *unify(YYLTYPE loc, struct type *l, struct type *r); +/** + * Infer an expression + * + * @param gamma environment + * @param expr expression to infer + * @param type σ, the type to infer to + * @result substitution + */ struct subst *infer_expr(struct gamma *gamma, struct expr *expr, struct type *type); +/** + * Infer a statement + * + * @param gamma environment + * @param stmt statement to infer + * @param type σ, the type to infer to + * @result substitution + */ struct subst *infer_stmt(struct gamma *gamma, struct stmt *stmt, struct type *type); -struct subst *infer_fundecl(struct gamma *gamma, struct fundecl *fundecl, struct type *ftype); +/** + * Infer a function declaration + * + * @param gamma environment + * @param fundecl function declaration to infer + * @param type σ, the type to infer to + * @result substitution + */ +struct subst *infer_fundecl(struct gamma *gamma, struct fundecl *fundecl, struct type *type); #endif diff --git a/src/sem/hm/gamma.h b/src/sem/hm/gamma.h index c477771..08424a2 100644 --- a/src/sem/hm/gamma.h +++ b/src/sem/hm/gamma.h @@ -7,21 +7,94 @@ #include "../hm.h" #include "../../ident.h" +/** abstract type representing the environment */ struct gamma; +/** abstract type representing an entry in the environment */ struct gamma_entry; +/** + * Initialise an empty environment + * + * @result environment + */ struct gamma *gamma_init(); +/** + * Insert an entry in the environment. + * + * @param gamma environment + * @param ident identifier + * @param scheme type scheme + */ void gamma_insert(struct gamma *gamma, struct ident ident, struct scheme *scheme); +/** + * Increment the internal scope of gamma, used to separate namespaces and + * scopes. + * + * @param gamma environment + */ void gamma_increment_scope(struct gamma *gamma); +/** + * Decrement the internal scope of gamma, used to separate namespaces and + * scopes. + * + * @param gamma environment + */ void gamma_decrement_scope(struct gamma *gamma); -void gamma_iter(struct gamma *gamma, void *st, void (*iter)(struct ident, struct scheme *, void *)); +/** + * Function to call for each entry while iterating over gamma. + * + * @param i type variable + * @param s type scheme + * @param st user state + */ +typedef void (*gamma_iterfun)(struct ident i, struct scheme *s, void *st); +/** + * Iterate over all entries in gamma. + * + * @param gamma environment + * @param st pointer to the initial state + * @param iter pointer to the iteration function + */ +void gamma_iter(struct gamma *gamma, void *st, gamma_iterfun iter); + +/** + * Lookup an identifier in the environment + * + * @param gamma environment + * @param ident identifier + * @result pointer to the type scheme, NULL if not found + */ struct scheme *gamma_lookup(struct gamma *gamma, struct ident ident); + +/** + * Check whether \p ident is free in gamma. + * + * @param gamma environment + * @param ident identifier + * @result result + */ bool gamma_free_in(struct gamma *gamma, struct ident ident); +/** + * Generate a fresh type variable. + * + * @param gamma environment + * @result fresh type variable + */ struct type *gamma_fresh(struct gamma *gamma); - +/** + * Print the entire enviroment + * + * @param gamma environment + * @param out output stream + */ void gamma_print(struct gamma *gamma, FILE *out); +/** + * Recursively free the enviroment + * + * @param gamma environment + */ void gamma_free(struct gamma *gamma); #endif diff --git a/src/sem/hm/scheme.h b/src/sem/hm/scheme.h index b194e00..57dc0d7 100644 --- a/src/sem/hm/scheme.h +++ b/src/sem/hm/scheme.h @@ -7,18 +7,58 @@ struct gamma; #include "../hm.h" #include "../../ident.h" +/** Definition of a type scheme */ struct scheme { - struct type *type; - int nvar; - struct ident *var; + struct type *type; /** Type */ + int nvar; /** Number of quantified type variables */ + struct ident *var; /** array of quantified type variables */ }; +/** + * Instantiate a type scheme, i.e. replace the quantified type variables by + * fresh type variables. + * + * @param gamma environment + * @param s scheme + * @result type + */ struct type *scheme_instantiate(struct gamma *gamma, struct scheme *s); +/** + * Create a scheme from a type and assume no type variable is quantified. + * + * @param t type + * @result scheme + */ struct scheme *scheme_create(struct type *t); +/** + * Create a scheme by generalising a type, i.e. quantify all free type variables + * that are not free in gamma. + * + * @param gamma environment + * @param t type + * @result scheme + */ struct scheme *scheme_generalise(struct gamma *gamma, struct type *t); +/** + * Check whether \p ident is free in the type scheme. + * + * @param scheme type scheme + * @param ident type variable + * @result result + */ bool scheme_free_in(struct scheme *scheme, struct ident ident); - +/** + * Print a type scheme. + * + * @param scheme type scheme + * @param out output stream + */ void scheme_print(struct scheme *scheme, FILE *out); +/** + * Free a type scheme. + * + * @param scheme type scheme + */ void scheme_free(struct scheme *scheme); #endif diff --git a/src/sem/hm/subst.h b/src/sem/hm/subst.h index 11d7c98..101dbff 100644 --- a/src/sem/hm/subst.h +++ b/src/sem/hm/subst.h @@ -5,18 +5,78 @@ #include "../hm.h" #include "../../ident.h" +/** Abstract type representing a substitution, a map from type variable to type */ struct subst; +/** + * The identity substitution + * + * @result the empty substitution + */ struct subst *subst_id(); +/** + * Extend a substitution. + * + * @param s substitution to extend + * @param ident type variable + * @param t substitute type + * @result the extended substitution + */ struct subst *subst_insert(struct subst *s, struct ident ident, struct type *t); +/** + * Create a singleton substitution. + * i.e. substs_singleton(i, t) == subst_insert(subst_id(), i, t) + * + * @param ident type variable + * @param t substitute type + * @result substitution + */ struct subst *subst_singleton(struct ident ident, struct type *t); +/** + * Take the union of two substitutions. This is destructive, \p l is freed, \p r is + * extended with the substitutions from \p l. + * + * @param l left-hand side substitution + * @param r right-hand side substitution + * @result union substitution ( + */ struct subst *subst_union(struct subst *l, struct subst *r); - -struct type *subst_apply_t(struct subst *subst, struct type *l); +/** + * Apply a substitution to a type. + * + * @param subst substitution + * @param t type + * @result type + */ +struct type *subst_apply_t(struct subst *subst, struct type *t); +/** + * Apply a substitution to a type scheme + * + * @param subst substitution + * @param scheme type scheme + * @result type scheme + */ struct scheme *subst_apply_s(struct subst *subst, struct scheme *scheme); +/** + * Apply a substitution to an environment + * + * @param subst substitution + * @param gamma environment + * @result environment + */ struct gamma *subst_apply_g(struct subst *subst, struct gamma *gamma); - +/** + * Print a substitution. + * + * @param s substitution + * @param out output stream + */ void subst_print(struct subst *s, FILE *out); +/** + * Free a substitution. + * + * @param s substitution + */ void subst_free(struct subst *s); #endif diff --git a/src/sem/main.h b/src/sem/main.h index 63a6cc0..6bfa61f 100644 --- a/src/sem/main.h +++ b/src/sem/main.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Check that there is a main function and that it has the correct type. + * + * @param ast abstract syntax tree + */ void sem_check_main(struct ast *ast); #endif diff --git a/src/sem/return.h b/src/sem/return.h index 7ccc0b9..1e8291e 100644 --- a/src/sem/return.h +++ b/src/sem/return.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Check that all functions return that have a non-void return type + * + * @param ast abstract syntax tree + */ void sem_check_return(struct ast *ast); #endif diff --git a/src/sem/scc.c b/src/sem/scc.c index 93327a9..cffc7e0 100644 --- a/src/sem/scc.c +++ b/src/sem/scc.c @@ -107,10 +107,11 @@ static int ptrcmp(const void *l, const void *r) * * Returns NULL when there are invalid edges * - * @param number of nodes - * @param data of the nodes - * @param number of edges - * @param data of edges + * @param nnodes number of nodes + * @param nodedata data of the nodes + * @param nedges number of edges + * @param edgedata data of edges + * @result list of components */ struct components *tarjans( int nnodes, void *nodedata[], diff --git a/src/sem/scc.h b/src/sem/scc.h index e948435..89b21e3 100644 --- a/src/sem/scc.h +++ b/src/sem/scc.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Group mutually recursive functions in components. + * + * @param ast abstract syntax tree + */ void sem_check_scc(struct ast *ast); #endif diff --git a/src/sem/type.c b/src/sem/type.c index a69d3be..ead0b40 100644 --- a/src/sem/type.c +++ b/src/sem/type.c @@ -206,6 +206,5 @@ void sem_check_types(struct ast *ast) break; } } - gamma_print(gamma, stdout); gamma_free(gamma); } diff --git a/src/sem/type.h b/src/sem/type.h index 352cbb5..645a294 100644 --- a/src/sem/type.h +++ b/src/sem/type.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Infer and check the intire AST + * + * @param ast abstract syntax tree + */ void sem_check_types(struct ast *ast); #endif diff --git a/src/sem/vardecl.h b/src/sem/vardecl.h index 8f72e2c..224045a 100644 --- a/src/sem/vardecl.h +++ b/src/sem/vardecl.h @@ -3,6 +3,11 @@ #include "../ast.h" +/** + * Check that vardecls aren't mutually recursive and order them + * + * @param ast abstract syntax tree + */ void sem_check_vardecls(struct ast *ast); #endif diff --git a/src/type.c b/src/type.c index 5a61d4a..4519b80 100644 --- a/src/type.c +++ b/src/type.c @@ -229,7 +229,7 @@ int type_cmp(struct type *l, struct type *r) int res = 0; switch(l->type) { case tarrow: - if ((res = type_cmp(l->data.tarrow.l, r->data.tarrow.l)) != 0) + if ((res = type_cmp(l->data.tarrow.l, r->data.tarrow.l)) == 0) res = type_cmp(l->data.tarrow.r, r->data.tarrow.r); break; case tbasic: @@ -239,7 +239,7 @@ int type_cmp(struct type *l, struct type *r) res = type_cmp(l->data.tlist, r->data.tlist); break; case ttuple: - if ((res = type_cmp(l->data.ttuple.l, r->data.ttuple.l)) != 0) + if ((res = type_cmp(l->data.ttuple.l, r->data.ttuple.l)) == 0) res = type_cmp(l->data.ttuple.r, r->data.ttuple.r); break; case tvar: diff --git a/src/type.h b/src/type.h index 467696b..35a0d12 100644 --- a/src/type.h +++ b/src/type.h @@ -5,40 +5,133 @@ #include "ast.h" #include "ident.h" +/** + * Export the string representations of the basic types enumeration + * + * @see basictype + */ extern const char *basictype_str[]; +/** The basic types */ enum basictype {btbool, btchar, btint, btvoid}; +/** Representation of a type */ struct type { - enum {tarrow,tbasic,tlist,ttuple,tvar} type; + enum {tarrow,tbasic,tlist,ttuple,tvar} type; /** tag of the tagged union */ union { struct { - struct type *l; - struct type *r; - } tarrow; - enum basictype tbasic; - struct type *tlist; + struct type *l; /** rhs of the arrow */ + struct type *r; /** rhs of the arrow */ + } tarrow; /** Arrow type: (l -> r) */ + enum basictype tbasic; /** Basic type */ + struct type *tlist; /** List type: [t] */ struct { - struct type *l; - struct type *r; - } ttuple; - struct ident tvar; + struct type *l; /** rhs of the tuple */ + struct type *r; /** rhs of the tuple */ + } ttuple; /** Tuple type: (l, r) */ + struct ident tvar; /** Type variable */ } data; }; +/** + * Constructor for the arrow type + * + * @param left left-hand side + * @param right right-hand side + * @result arrow type + */ struct type *type_arrow(struct type *left, struct type *right); +/** + * Constructor for a basic type + * + * @param type tag of the basic type + * @result basic type + */ struct type *type_basic(enum basictype type); +/** + * Constructor for a list type + * + * @param type type of the elements + * @result list type + */ struct type *type_list(struct type *type); +/** + * Constructor for the tuple type + * + * @param left left-hand side + * @param right right-hand side + * @result tuple type + */ struct type *type_tuple(struct type *left, struct type *right); +/** + * Constructor for a type variable + * + * @param ident identifier of the variable + * @result type variable + */ struct type *type_var(struct ident ident); +/** + * Constructor for a type variable from a string + * + * @param s string identifier of the variable + * @result type variable + */ struct type *type_var_str(char *s); +/** + * Constructor for a type variable from an integer + * + * @param i integer identifier of the variable + * @result type variable + */ struct type *type_var_int(int i); +/** + * Print a type to the output stream + * + * @param type type + * @param stream stream to print to + */ void type_print(struct type *type, FILE *stream); +/** + * Recursively free a type + * + * @param type type + */ void type_free(struct type *type); +/** + * Compare a type + * + * @param l left-hand side + * @param r right-hand side + * @result comparison + */ int type_cmp(struct type *l, struct type *r); +/** + * Duplicate a type + * + * @param t type + * @result duplicated type + */ struct type *type_dup(struct type *t); + +/** + * Find the free type variables in a type. The provided buffer will be + * reallocated whenever a free type variable is found. + * + * @param r type to search in + * @param nftv pointer to an integer representing the number of found free + * type variables + * @param ftv pointer to a pointer to a buffer for storing the free type + * variables + */ void type_ftv(struct type *r, int *nftv, struct ident **ftv); +/** + * Predicate to test whether an identifier is free in a type. + * + * @param type type to search in + * @param ident identifier + * @result free + */ bool type_free_in(struct type *type, struct ident ident); #endif diff --git a/src/util.h b/src/util.h index bfdd21c..2b9dd6f 100644 --- a/src/util.h +++ b/src/util.h @@ -5,29 +5,81 @@ #include #include -/* exit with an error message */ +/** + * Emit the error message and exit with return code 1 + * + * @param msg format string + * @param ... format arguments + */ void die(const char *msg, ...); -/* exit with the system's error message prefixed by msg */ +/** + * Emit the system's error message prefixed with \p msg and exit with return code + * 1 + * + * @param msg message prefix + */ void pdie(const char *msg); -/* if buf == NULL, a fresh buffer is allocated */ +/** + * Escape a character, if \p buf == NULL, a fresh buffer is allocated + * + * @param c character to escape or the name of the escape (e.g. n for newline) + * @param buf buffer to store the escape sequence in + * @param str string or in a character context (escape " or ') + * @result buffer containing the escape sequence + */ char *escape_char(char c, char *buf, bool str); -/* unescaped character will be in position 0 and the rest from position 1 on */ +/** + * Unescape the given escape sequence + * + * @param c buffer starting with the escape sequence + * @result same buffer but now unescaped + */ char *unescape_char(char *c); -/* Remove the last and first character from the string */ +/** + * Remove the last and first character from a string + * + * @param c string + * @result trimmed string + */ char *trimquotes(char *c); -/* Print indentation */ +/** + * Print \p indent indentation level + * + * @param indent indentation level + * @param out output stream + */ void pindent(int indent, FILE *out); -/* Safe wrappers around syscalls */ +/** Call vfprintf but bail out if it fails */ void safe_vfprintf(FILE *out, const char *msg, va_list ap); +/** Call fprintf but bail out if it fails */ void safe_fprintf(FILE *out, const char *msg, ...); +/** Call malloc but bail out if it fails */ void *safe_malloc(size_t size); +/** + * Typed variant of safe_malloc + * + * @param nmemb number of members + * @param type type of the members + * @result pointer to the allocated buffer + */ #define xalloc(nmemb, type) ((type *)safe_malloc((nmemb)*sizeof(type))) +/** Call remalloc but bail out if it fails */ void *safe_realloc(void *ptr, size_t size); +/** + * Typed variant of safe_remalloc + * + * @param ptr pointer to the buffer that possibly needs reallocation + * @param nmemb number of members + * @param type type of the members + * @result pointer to the reallocated buffer + */ #define xrealloc(ptr, nmemb, type) ((type *)safe_realloc(ptr, (nmemb)*sizeof(type))) void *safe_strdup(const char *c); +/** Call fopen but bail out if it fails */ FILE *safe_fopen(const char *path, const char *mode); +/** Call fclose but bail out if it fails */ void safe_fclose(FILE *file); #endif diff --git a/test.bash b/test.bash index 04d4be6..ba9e576 100644 --- a/test.bash +++ b/test.bash @@ -23,7 +23,7 @@ do fail=$((fail+1)) else "$SPLC" $1 -o "$ccode" - "$CC" -I"$RTSDIR" "$ccode" -o "$base" + "$CC" -I"$RTSDIR" "$RTSDIR"/rts.c "$ccode" -o "$base" ./"$base" > "$out" if ! diff "$out" "$expected" then diff --git a/tests/2D.expected b/tests/2D.expected new file mode 100644 index 0000000..be15b51 --- /dev/null +++ b/tests/2D.expected @@ -0,0 +1 @@ +(2,2)(39,44)(84,126) diff --git a/tests/2D.spl b/tests/2D.spl index 2903ef9..80e25dc 100644 --- a/tests/2D.spl +++ b/tests/2D.spl @@ -16,4 +16,5 @@ main() { print(foo(42)); print(transpose((38, 42), (1, 2))); print(scale((28, 42), 3)); + print('\n'); } diff --git a/tests/3D.expected b/tests/3D.expected new file mode 100644 index 0000000..537cb72 --- /dev/null +++ b/tests/3D.expected @@ -0,0 +1 @@ +(39,(44,19))(114,(126,48)) diff --git a/tests/3D.spl b/tests/3D.spl index e040099..1f2effa 100644 --- a/tests/3D.spl +++ b/tests/3D.spl @@ -5,3 +5,8 @@ transpose (p1,p2) :: (Int, (Int, Int)) (Int, (Int, Int)) -> (Int, (Int, Int)) scale(p, scalar) :: (Int, (Int, Int)) Int -> (Int, (Int, Int)) { return (p.fst * scalar, (p.snd.fst * scalar, p.snd.snd * scalar)); } +main() { + print(transpose((38, (42, 16)), (1, (2, 3)))); + print(scale((38, (42, 16)), 3)); + print('\n'); +} diff --git a/tests/a_bit_of_everything.expected b/tests/a_bit_of_everything.expected new file mode 100644 index 0000000..e69de29 diff --git a/tests/arguments.expected b/tests/arguments.expected new file mode 100644 index 0000000..e69de29 diff --git a/tests/arguments.shouldfail b/tests/arguments.shouldfail new file mode 100644 index 0000000..e69de29 diff --git a/tests/assignment_to_builtin.expected b/tests/assignment_to_builtin.expected new file mode 100644 index 0000000..e69de29 diff --git a/tests/assignment_to_builtin.shouldfail b/tests/assignment_to_builtin.shouldfail new file mode 100644 index 0000000..e69de29 diff --git a/tests/assignment_to_builtin.spl b/tests/assignment_to_builtin.spl index dc53a37..28de5eb 100644 --- a/tests/assignment_to_builtin.spl +++ b/tests/assignment_to_builtin.spl @@ -11,3 +11,9 @@ f() // Adapt this line if your isEmpty is called differently. isEmpty = blaat; } + +main() +{ + f(); + +}