add tiny expression library and make it a dice expression parser

Use tiny expression library (from Github) and expand it to provide
dice expression parsing. This needs more testing obviously.

Makefile.am

@@ -5,7 +5,8 @@ lib_LTLIBRARIES =	libdice.la
 libdice_la_SOURCES =	lib/dice_lexer.l \
 			lib/dice_parse.y \
 			lib/dice.h \
-			lib/dice.c
+			lib/dice.c \
+			lib/diceexpr.c
 
 AM_CFLAGS =		-Ilib
 

configure.ac

@@ -18,4 +18,8 @@ AC_CHECK_FUNCS([arc4random arc4random_uniform strlcat strlcpy],,
 
 AC_CONFIG_FILES([Makefile tests/Makefile])
 
+AC_SEARCH_LIBS([cos], [m], [], [
+  AC_MSG_ERROR([unable to find the cos() function])
+])
+
 AC_OUTPUT

lib/dice.c

@@ -8,7 +8,7 @@
 #include <bsd/stdlib.h>
 #endif
 
-extern int yylex_init(void **state);
+extern int yylex_init_extra(void *extra, void **state);
 extern int yylex_destroy(void *state);
 extern void yylex(void *state);
 extern void yy_switch_to_buffer(void *buffer, void *scanner);
@@ -22,7 +22,7 @@ struct dice_
     uint32_t amount;
     uint32_t sides;
 
-    char *errmsg;
+    char *error;
 };
 
 static dice_t dice_new(void)
@@ -42,6 +42,7 @@ void dice_free(dice_t t)
         return;
     }
 
+    free(t->error);
     free(t);
 }
 
@@ -93,7 +94,7 @@ dice_t dice_parse(char const *s)
         return NULL;
     }
 
-    yylex_init(&scanner);
+    yylex_init_extra(d, &scanner);
     buffer = yy_scan_string(s, scanner);
     yy_switch_to_buffer(buffer, scanner);
 
@@ -118,6 +119,12 @@ bool dice_set(dice_t d, dice_option_t opt, ...)
     switch (opt) {
     case DICEOPTION_AMOUNT: d->amount = va_arg(lst, uint32_t); break;
     case DICEOPTION_SIDES: d->sides = va_arg(lst, uint32_t); break;
+    case DICEOPTION_ERROR:
+    {
+        char const *err = va_arg(lst, char const *);
+        free(d->error);
+        d->error = strdup(err);
+    } break;
 
     default: return false;
     }
@@ -144,6 +151,12 @@ bool dice_get(dice_t d, dice_option_t opt, ...)
         *ptr = d->sides;
     } break;
 
+    case DICEOPTION_ERROR:
+    {
+        char **ptr = va_arg(lst, char **);
+        *ptr = d->error;
+    } break;
+
     default: return false;
     }
     va_end(lst);
@@ -189,7 +202,16 @@ bool dice_evaluate(dice_t d, dice_result_t **res, size_t *reslen)
     return true;
 }
 
-void dice_update_consumned(dice_t d, int offset)
+void dice_update_consumed(dice_t d, int offset)
 {
     d->consumed += offset;
 }
+
+int dice_consumed(dice_t d)
+{
+    if (d == NULL) {
+        return 0;
+    }
+
+    return d->consumed;
+}

lib/dice.h

@@ -9,6 +9,9 @@
 struct dice_;
 typedef struct dice_ * dice_t;
 
+struct dice_expression_;
+typedef struct dice_expression_ * dice_expression_t;
+
 typedef struct {
     int64_t result;
 } dice_result_t;
@@ -16,6 +19,7 @@ typedef struct {
 typedef enum {
     DICEOPTION_AMOUNT = 0,
     DICEOPTION_SIDES,
+    DICEOPTION_ERROR,
 } dice_option_t;
 
 void dice_result_free(dice_result_t *r);
@@ -31,4 +35,9 @@ bool dice_get(dice_t d, dice_option_t opt, ...);
 int64_t dice_roll(dice_t d);
 bool dice_evaluate(dice_t d, dice_result_t **res, size_t *reslen);
 
+dice_expression_t dice_expression_parse(char const *s, int *error);
+void dice_expression_free(dice_expression_t e);
+bool dice_expression_evaluate(dice_expression_t e, int64_t *result);
+bool dice_expression_print(dice_expression_t e);
+
 #endif

lib/dice_lexer.l

@@ -5,15 +5,21 @@
 #include "dice.h"
 #include "dice_parse.h"
 
+extern void dice_update_consumed(dice_t d, int off);
+
 %}
 
 %%
 [0-9]+ {
+    dice_t d = yyget_extra(yyscanner);
+    dice_update_consumed(d, yyleng);
     yylval->integer = atoi(yytext);
     return TOK_INTEGER;
 }
 
 [d] {
+    dice_t d = yyget_extra(yyscanner);
+    dice_update_consumed(d, yyleng);
     return TOK_DICESEP;
 }
 %%

lib/dice_parse.y

@@ -10,6 +10,7 @@ extern int yylex(void *lval, void *scanner);
 
 void yyerror(void *scanner, dice_t dice, char const *err)
 {
+    dice_set(dice, DICEOPTION_ERROR, err);
 }
 
 int yywrap(void)
@@ -33,10 +34,14 @@ dice:           TOK_INTEGER TOK_DICESEP TOK_INTEGER
                 {
                     dice_set(dice, DICEOPTION_AMOUNT, $1);
                     dice_set(dice, DICEOPTION_SIDES, $3);
+
+                    YYACCEPT;
                 }
         |       TOK_DICESEP TOK_INTEGER
                 {
                     dice_set(dice, DICEOPTION_AMOUNT, 1);
                     dice_set(dice, DICEOPTION_SIDES, $2);
+
+                    YYACCEPT;
                 }
         ;

lib/diceexpr.c

@@ -0,0 +1,771 @@
+/*
+ * TINYEXPR - Tiny recursive descent parser and evaluation engine in C
+ *
+ * Copyright (c) 2015, 2016 Lewis Van Winkle
+ *
+ * http://CodePlea.com
+ *
+ * This software is provided 'as-is', without any express or implied
+ * warranty. In no event will the authors be held liable for any damages
+ * arising from the use of this software.
+ *
+ * Permission is granted to anyone to use this software for any purpose,
+ * including commercial applications, and to alter it and redistribute it
+ * freely, subject to the following restrictions:
+ *
+ * 1. The origin of this software must not be misrepresented; you must not
+ * claim that you wrote the original software. If you use this software
+ * in a product, an acknowledgement in the product documentation would be
+ * appreciated but is not required.
+ * 2. Altered source versions must be plainly marked as such, and must not be
+ * misrepresented as being the original software.
+ * 3. This notice may not be removed or altered from any source distribution.
+ */
+
+/* COMPILE TIME OPTIONS */
+
+/* Exponentiation associativity:
+For a^b^c = (a^b)^c and -a^b = (-a)^b do nothing.
+For a^b^c = a^(b^c) and -a^b = -(a^b) uncomment the next line.*/
+#define TE_POW_FROM_RIGHT
+
+/* Logarithms
+For log = base 10 log do nothing
+For log = natural log uncomment the next line. */
+/* #define TE_NAT_LOG */
+
+#include <stdlib.h>
+#include <math.h>
+#include <string.h>
+#include <stdio.h>
+#include <limits.h>
+#include <ctype.h>
+
+#include "dice.h"
+
+#ifndef NAN
+#define NAN (0.0/0.0)
+#endif
+
+#ifndef INFINITY
+#define INFINITY (1.0/0.0)
+#endif
+
+extern int dice_consumed(dice_t d);
+
+typedef struct te_expr {
+    int type;
+    union {
+        double value;
+        const double *bound;
+        const void *function;
+        dice_t dice;
+    };
+    void *parameters[1];
+} te_expr;
+
+enum {
+    TE_VARIABLE = 0, TE_DICE = 2,
+
+    TE_FUNCTION0 = 8, TE_FUNCTION1, TE_FUNCTION2, TE_FUNCTION3,
+    TE_FUNCTION4, TE_FUNCTION5, TE_FUNCTION6, TE_FUNCTION7,
+
+    TE_CLOSURE0 = 16, TE_CLOSURE1, TE_CLOSURE2, TE_CLOSURE3,
+    TE_CLOSURE4, TE_CLOSURE5, TE_CLOSURE6, TE_CLOSURE7,
+
+    TE_FLAG_PURE = 64,
+};
+
+typedef struct te_variable {
+    const char *name;
+    const void *address;
+    int type;
+    void *context;
+} te_variable;
+
+
+typedef double (*te_fun2)(double, double);
+
+enum {
+    TOK_NULL = TE_CLOSURE7+1, TOK_ERROR, TOK_END, TOK_SEP,
+    TOK_OPEN, TOK_CLOSE, TOK_NUMBER, TOK_VARIABLE, TOK_INFIX, TOK_DICE,
+};
+
+
+enum {TE_CONSTANT = 1};
+
+
+typedef struct state {
+    const char *start;
+    const char *next;
+    int type;
+    union {
+        double value;
+        const double *bound;
+        const void *function;
+        dice_t dice;
+    };
+    void *context;
+
+    const te_variable *lookup;
+    int lookup_len;
+} state;
+
+static void te_free(te_expr *);
+
+#define TYPE_MASK(TYPE) ((TYPE)&0x0000003F)
+
+#define IS_PURE(TYPE) (((TYPE) & TE_FLAG_PURE) != 0)
+#define IS_FUNCTION(TYPE) (((TYPE) & TE_FUNCTION0) != 0)
+#define IS_CLOSURE(TYPE) (((TYPE) & TE_CLOSURE0) != 0)
+#define ARITY(TYPE) ( ((TYPE) & (TE_FUNCTION0 | TE_CLOSURE0)) ? ((TYPE) & 0x00000007) : 0 )
+#define NEW_EXPR(type, ...) new_expr((type), (const te_expr*[]){__VA_ARGS__})
+
+static te_expr *new_expr(const int type, const te_expr *parameters[]) {
+    const int arity = ARITY(type);
+    const int psize = sizeof(void*) * arity;
+    const int size = (sizeof(te_expr) - sizeof(void*)) + psize + (IS_CLOSURE(type) ? sizeof(void*) : 0);
+    te_expr *ret = malloc(size);
+    memset(ret, 0, size);
+    if (arity && parameters) {
+        memcpy(ret->parameters, parameters, psize);
+    }
+    ret->type = type;
+    ret->bound = 0;
+    return ret;
+}
+
+static void te_free_parameters(te_expr *n) {
+    if (!n) return;
+    switch (TYPE_MASK(n->type)) {
+        case TE_FUNCTION7: case TE_CLOSURE7: te_free(n->parameters[6]);
+        case TE_FUNCTION6: case TE_CLOSURE6: te_free(n->parameters[5]);
+        case TE_FUNCTION5: case TE_CLOSURE5: te_free(n->parameters[4]);
+        case TE_FUNCTION4: case TE_CLOSURE4: te_free(n->parameters[3]);
+        case TE_FUNCTION3: case TE_CLOSURE3: te_free(n->parameters[2]);
+        case TE_FUNCTION2: case TE_CLOSURE2: te_free(n->parameters[1]);
+        case TE_FUNCTION1: case TE_CLOSURE1: te_free(n->parameters[0]);
+    }
+    if (TYPE_MASK(n->type) == TE_DICE) {
+        dice_free(n->dice);
+    }
+}
+
+static void te_free(te_expr *n) {
+    if (!n) return;
+    te_free_parameters(n);
+    free(n);
+}
+
+static double pi() {return 3.14159265358979323846;}
+static double e() {return 2.71828182845904523536;}
+static double fac(double a) {/* simplest version of fac */
+    if (a < 0.0)
+        return NAN;
+    if (a > UINT_MAX)
+        return INFINITY;
+    unsigned int ua = (unsigned int)(a);
+    unsigned long int result = 1, i;
+    for (i = 1; i <= ua; i++) {
+        if (i > ULONG_MAX / result)
+            return INFINITY;
+        result *= i;
+    }
+    return (double)result;
+}
+static double ncr(double n, double r) {
+    if (n < 0.0 || r < 0.0 || n < r) return NAN;
+    if (n > UINT_MAX || r > UINT_MAX) return INFINITY;
+    unsigned long int un = (unsigned int)(n), ur = (unsigned int)(r), i;
+    unsigned long int result = 1;
+    if (ur > un / 2) ur = un - ur;
+    for (i = 1; i <= ur; i++) {
+        if (result > ULONG_MAX / (un - ur + i))
+            return INFINITY;
+        result *= un - ur + i;
+        result /= i;
+    }
+    return result;
+}
+static double npr(double n, double r) {return ncr(n, r) * fac(r);}
+
+static const te_variable functions[] = {
+    /* must be in alphabetical order */
+    {"abs", fabs,     TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"acos", acos,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"asin", asin,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"atan", atan,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"atan2", atan2,  TE_FUNCTION2 | TE_FLAG_PURE, 0},
+    {"ceil", ceil,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"cos", cos,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"cosh", cosh,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"e", e,          TE_FUNCTION0 | TE_FLAG_PURE, 0},
+    {"exp", exp,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"fac", fac,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"floor", floor,  TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"ln", log,       TE_FUNCTION1 | TE_FLAG_PURE, 0},
+#ifdef TE_NAT_LOG
+    {"log", log,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+#else
+    {"log", log10,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+#endif
+    {"log10", log10,  TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"ncr", ncr,      TE_FUNCTION2 | TE_FLAG_PURE, 0},
+    {"npr", npr,      TE_FUNCTION2 | TE_FLAG_PURE, 0},
+    {"pi", pi,        TE_FUNCTION0 | TE_FLAG_PURE, 0},
+    {"pow", pow,      TE_FUNCTION2 | TE_FLAG_PURE, 0},
+    {"sin", sin,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"sinh", sinh,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"sqrt", sqrt,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"tan", tan,      TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {"tanh", tanh,    TE_FUNCTION1 | TE_FLAG_PURE, 0},
+    {0, 0, 0, 0}
+};
+
+static const te_variable *find_builtin(const char *name, int len) {
+    int imin = 0;
+    int imax = sizeof(functions) / sizeof(te_variable) - 2;
+
+    /*Binary search.*/
+    while (imax >= imin) {
+        const int i = (imin + ((imax-imin)/2));
+        int c = strncmp(name, functions[i].name, len);
+        if (!c) c = '\0' - functions[i].name[len];
+        if (c == 0) {
+            return functions + i;
+        } else if (c > 0) {
+            imin = i + 1;
+        } else {
+            imax = i - 1;
+        }
+    }
+
+    return 0;
+}
+
+static const te_variable *find_lookup(const state *s, const char *name, int len) {
+    int iters;
+    const te_variable *var;
+    if (!s->lookup) return 0;
+
+    for (var = s->lookup, iters = s->lookup_len; iters; ++var, --iters) {
+        if (strncmp(name, var->name, len) == 0 && var->name[len] == '\0') {
+            return var;
+        }
+    }
+    return 0;
+}
+
+
+
+static double add(double a, double b) {return a + b;}
+static double sub(double a, double b) {return a - b;}
+static double mul(double a, double b) {return a * b;}
+static double divide(double a, double b) {return a / b;}
+static double negate(double a) {return -a;}
+static double comma(double a, double b) {(void)a; return b;}
+
+
+void next_token(state *s) {
+    s->type = TOK_NULL;
+
+    do {
+
+        if (!*s->next){
+            s->type = TOK_END;
+            return;
+        }
+
+        if (isdigit(s->next[0]) || s->next[0] == '.' || s->next[0] == 'd') {
+            /* first try reading a dice expression, if that fails, we go
+               back to trying a number instead */
+            dice_t d = dice_parse(s->next);
+            if (d != NULL) {
+                int consumed = dice_consumed(d);
+                s->type = TOK_DICE;
+                s->dice = d;
+                s->next += consumed;
+            } else if (d == NULL) {
+                /* Try reading a number. */
+                s->value = strtod(s->next, (char**)&s->next);
+                s->type = TOK_NUMBER;
+            }
+        } else {
+            /* Look for a variable or builtin function call. */
+            if (s->next[0] >= 'a' && s->next[0] <= 'z') {
+                const char *start;
+                start = s->next;
+                while ((s->next[0] >= 'a' && s->next[0] <= 'z') || (s->next[0] >= '0' && s->next[0] <= '9') || (s->next[0] == '_')) s->next++;
+
+                const te_variable *var = find_lookup(s, start, s->next - start);
+                if (!var) var = find_builtin(start, s->next - start);
+
+                if (!var) {
+                    s->type = TOK_ERROR;
+                } else {
+                    switch(TYPE_MASK(var->type))
+                    {
+                        case TE_VARIABLE:
+                            s->type = TOK_VARIABLE;
+                            s->bound = var->address;
+                            break;
+
+                        case TE_CLOSURE0: case TE_CLOSURE1: case TE_CLOSURE2: case TE_CLOSURE3:
+                        case TE_CLOSURE4: case TE_CLOSURE5: case TE_CLOSURE6: case TE_CLOSURE7:
+                            s->context = var->context;
+
+                        case TE_FUNCTION0: case TE_FUNCTION1: case TE_FUNCTION2: case TE_FUNCTION3:
+                        case TE_FUNCTION4: case TE_FUNCTION5: case TE_FUNCTION6: case TE_FUNCTION7:
+                            s->type = var->type;
+                            s->function = var->address;
+                            break;
+                    }
+                }
+
+            } else {
+                /* Look for an operator or special character. */
+                switch (s->next++[0]) {
+                    case '+': s->type = TOK_INFIX; s->function = add; break;
+                    case '-': s->type = TOK_INFIX; s->function = sub; break;
+                    case '*': s->type = TOK_INFIX; s->function = mul; break;
+                    case '/': s->type = TOK_INFIX; s->function = divide; break;
+                    case '^': s->type = TOK_INFIX; s->function = pow; break;
+                    case '%': s->type = TOK_INFIX; s->function = fmod; break;
+                    case '(': s->type = TOK_OPEN; break;
+                    case ')': s->type = TOK_CLOSE; break;
+                    case ',': s->type = TOK_SEP; break;
+                    case ' ': case '\t': case '\n': case '\r': break;
+                    default: s->type = TOK_ERROR; break;
+                }
+            }
+        }
+    } while (s->type == TOK_NULL);
+}
+
+
+static te_expr *list(state *s);
+static te_expr *expr(state *s);
+static te_expr *power(state *s);
+
+static te_expr *base(state *s) {
+    /* <base>      =    <constant> | <variable> | <function-0> {"(" ")"} | <function-1> <power> | <function-X> "(" <expr> {"," <expr>} ")" | "(" <list> ")" */
+    te_expr *ret;
+    int arity;
+
+    switch (TYPE_MASK(s->type)) {
+        case TOK_NUMBER:
+            ret = new_expr(TE_CONSTANT, 0);
+            ret->value = s->value;
+            next_token(s);
+            break;
+
+        case TOK_VARIABLE:
+            ret = new_expr(TE_VARIABLE, 0);
+            ret->bound = s->bound;
+            next_token(s);
+            break;
+
+        case TOK_DICE:
+            ret = new_expr(TE_DICE, 0);
+            ret->dice = s->dice;
+            next_token(s);
+            break;
+
+        case TE_FUNCTION0:
+        case TE_CLOSURE0:
+            ret = new_expr(s->type, 0);
+            ret->function = s->function;
+            if (IS_CLOSURE(s->type)) ret->parameters[0] = s->context;
+            next_token(s);
+            if (s->type == TOK_OPEN) {
+                next_token(s);
+                if (s->type != TOK_CLOSE) {
+                    s->type = TOK_ERROR;
+                } else {
+                    next_token(s);
+                }
+            }
+            break;
+
+        case TE_FUNCTION1:
+        case TE_CLOSURE1:
+            ret = new_expr(s->type, 0);
+            ret->function = s->function;
+            if (IS_CLOSURE(s->type)) ret->parameters[1] = s->context;
+            next_token(s);
+            ret->parameters[0] = power(s);
+            break;
+
+        case TE_FUNCTION2: case TE_FUNCTION3: case TE_FUNCTION4:
+        case TE_FUNCTION5: case TE_FUNCTION6: case TE_FUNCTION7:
+        case TE_CLOSURE2: case TE_CLOSURE3: case TE_CLOSURE4:
+        case TE_CLOSURE5: case TE_CLOSURE6: case TE_CLOSURE7:
+            arity = ARITY(s->type);
+
+            ret = new_expr(s->type, 0);
+            ret->function = s->function;
+            if (IS_CLOSURE(s->type)) ret->parameters[arity] = s->context;
+            next_token(s);
+
+            if (s->type != TOK_OPEN) {
+                s->type = TOK_ERROR;
+            } else {
+                int i;
+                for(i = 0; i < arity; i++) {
+                    next_token(s);
+                    ret->parameters[i] = expr(s);
+                    if(s->type != TOK_SEP) {
+                        break;
+                    }
+                }
+                if(s->type != TOK_CLOSE || i != arity - 1) {
+                    s->type = TOK_ERROR;
+                } else {
+                    next_token(s);
+                }
+            }
+
+            break;
+
+        case TOK_OPEN:
+            next_token(s);
+            ret = list(s);
+            if (s->type != TOK_CLOSE) {
+                s->type = TOK_ERROR;
+            } else {
+                next_token(s);
+            }
+            break;
+
+        default:
+            ret = new_expr(0, 0);
+            s->type = TOK_ERROR;
+            ret->value = NAN;
+            break;
+    }
+
+    return ret;
+}
+
+
+static te_expr *power(state *s) {
+    /* <power>     =    {("-" | "+")} <base> */
+    int sign = 1;
+    while (s->type == TOK_INFIX && (s->function == add || s->function == sub)) {
+        if (s->function == sub) sign = -sign;
+        next_token(s);
+    }
+
+    te_expr *ret;
+
+    if (sign == 1) {
+        ret = base(s);
+    } else {
+        ret = NEW_EXPR(TE_FUNCTION1 | TE_FLAG_PURE, base(s));
+        ret->function = negate;
+    }
+
+    return ret;
+}
+
+#ifdef TE_POW_FROM_RIGHT
+static te_expr *factor(state *s) {
+    /* <factor>    =    <power> {"^" <power>} */
+    te_expr *ret = power(s);
+
+    int neg = 0;
+    te_expr *insertion = 0;
+
+    if (ret->type == (TE_FUNCTION1 | TE_FLAG_PURE) && ret->function == negate) {
+        te_expr *se = ret->parameters[0];
+        free(ret);
+        ret = se;
+        neg = 1;
+    }
+
+    while (s->type == TOK_INFIX && (s->function == pow)) {
+        te_fun2 t = s->function;
+        next_token(s);
+
+        if (insertion) {
+            /* Make exponentiation go right-to-left. */
+            te_expr *insert = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, insertion->parameters[1], power(s));
+            insert->function = t;
+            insertion->parameters[1] = insert;
+            insertion = insert;
+        } else {
+            ret = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, ret, power(s));
+            ret->function = t;
+            insertion = ret;
+        }
+    }
+
+    if (neg) {
+        ret = NEW_EXPR(TE_FUNCTION1 | TE_FLAG_PURE, ret);
+        ret->function = negate;
+    }
+
+    return ret;
+}
+#else
+static te_expr *factor(state *s) {
+    /* <factor>    =    <power> {"^" <power>} */
+    te_expr *ret = power(s);
+
+    while (s->type == TOK_INFIX && (s->function == pow)) {
+        te_fun2 t = s->function;
+        next_token(s);
+        ret = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, ret, power(s));
+        ret->function = t;
+    }
+
+    return ret;
+}
+#endif
+
+
+
+static te_expr *term(state *s) {
+    /* <term>      =    <factor> {("*" | "/" | "%") <factor>} */
+    te_expr *ret = factor(s);
+
+    while (s->type == TOK_INFIX && (s->function == mul || s->function == divide || s->function == fmod)) {
+        te_fun2 t = s->function;
+        next_token(s);
+        ret = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, ret, factor(s));
+        ret->function = t;
+    }
+
+    return ret;
+}
+
+
+static te_expr *expr(state *s) {
+    /* <expr>      =    <term> {("+" | "-") <term>} */
+    te_expr *ret = term(s);
+
+    while (s->type == TOK_INFIX && (s->function == add || s->function == sub)) {
+        te_fun2 t = s->function;
+        next_token(s);
+        ret = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, ret, term(s));
+        ret->function = t;
+    }
+
+    return ret;
+}
+
+
+static te_expr *list(state *s) {
+    /* <list>      =    <expr> {"," <expr>} */
+    te_expr *ret = expr(s);
+
+    while (s->type == TOK_SEP) {
+        next_token(s);
+        ret = NEW_EXPR(TE_FUNCTION2 | TE_FLAG_PURE, ret, expr(s));
+        ret->function = comma;
+    }
+
+    return ret;
+}
+
+
+#define TE_FUN(...) ((double(*)(__VA_ARGS__))n->function)
+#define M(e) te_eval(n->parameters[e])
+
+
+double te_eval(const te_expr *n) {
+    if (!n) return NAN;
+
+    switch(TYPE_MASK(n->type)) {
+        case TE_CONSTANT: return n->value;
+        case TE_VARIABLE: return *n->bound;
+        case TE_DICE: return (double)(dice_roll(n->dice) * 1.0);
+
+        case TE_FUNCTION0: case TE_FUNCTION1: case TE_FUNCTION2: case TE_FUNCTION3:
+        case TE_FUNCTION4: case TE_FUNCTION5: case TE_FUNCTION6: case TE_FUNCTION7:
+            switch(ARITY(n->type)) {
+                case 0: return TE_FUN(void)();
+                case 1: return TE_FUN(double)(M(0));
+                case 2: return TE_FUN(double, double)(M(0), M(1));
+                case 3: return TE_FUN(double, double, double)(M(0), M(1), M(2));
+                case 4: return TE_FUN(double, double, double, double)(M(0), M(1), M(2), M(3));
+                case 5: return TE_FUN(double, double, double, double, double)(M(0), M(1), M(2), M(3), M(4));
+                case 6: return TE_FUN(double, double, double, double, double, double)(M(0), M(1), M(2), M(3), M(4), M(5));
+                case 7: return TE_FUN(double, double, double, double, double, double, double)(M(0), M(1), M(2), M(3), M(4), M(5), M(6));
+                default: return NAN;
+            }
+
+        case TE_CLOSURE0: case TE_CLOSURE1: case TE_CLOSURE2: case TE_CLOSURE3:
+        case TE_CLOSURE4: case TE_CLOSURE5: case TE_CLOSURE6: case TE_CLOSURE7:
+            switch(ARITY(n->type)) {
+                case 0: return TE_FUN(void*)(n->parameters[0]);
+                case 1: return TE_FUN(void*, double)(n->parameters[1], M(0));
+                case 2: return TE_FUN(void*, double, double)(n->parameters[2], M(0), M(1));
+                case 3: return TE_FUN(void*, double, double, double)(n->parameters[3], M(0), M(1), M(2));
+                case 4: return TE_FUN(void*, double, double, double, double)(n->parameters[4], M(0), M(1), M(2), M(3));
+                case 5: return TE_FUN(void*, double, double, double, double, double)(n->parameters[5], M(0), M(1), M(2), M(3), M(4));
+                case 6: return TE_FUN(void*, double, double, double, double, double, double)(n->parameters[6], M(0), M(1), M(2), M(3), M(4), M(5));
+                case 7: return TE_FUN(void*, double, double, double, double, double, double, double)(n->parameters[7], M(0), M(1), M(2), M(3), M(4), M(5), M(6));
+                default: return NAN;
+            }
+
+        default: return NAN;
+    }
+
+}
+
+#undef TE_FUN
+#undef M
+
+static void optimize(te_expr *n) {
+    /* Evaluates as much as possible. */
+    if (n->type == TE_CONSTANT) return;
+    if (n->type == TE_VARIABLE) return;
+
+    /* Only optimize out functions flagged as pure. */
+    if (IS_PURE(n->type)) {
+        const int arity = ARITY(n->type);
+        int known = 1;
+        int i;
+        for (i = 0; i < arity; ++i) {
+            optimize(n->parameters[i]);
+            if (((te_expr*)(n->parameters[i]))->type != TE_CONSTANT) {
+                known = 0;
+            }
+        }
+        if (known) {
+            const double value = te_eval(n);
+            te_free_parameters(n);
+            n->type = TE_CONSTANT;
+            n->value = value;
+        }
+    }
+}
+
+
+static te_expr *te_compile(const char *expression, const te_variable *variables,
+                           int var_count, int *error) {
+    state s;
+    s.start = s.next = expression;
+    s.lookup = variables;
+    s.lookup_len = var_count;
+
+    next_token(&s);
+    te_expr *root = list(&s);
+
+    if (s.type != TOK_END) {
+        te_free(root);
+        if (error) {
+            *error = (s.next - s.start);
+            if (*error == 0) *error = 1;
+        }
+        return 0;
+    } else {
+        optimize(root);
+        if (error) *error = 0;
+        return root;
+    }
+}
+
+
+static double te_interp(const char *expression, int *error) {
+    te_expr *n = te_compile(expression, 0, 0, error);
+    double ret;
+    if (n) {
+        ret = te_eval(n);
+        te_free(n);
+    } else {
+        ret = NAN;
+    }
+    return ret;
+}
+
+static void pn (const te_expr *n, int depth) {
+    int i, arity;
+    printf("%*s", depth, "");
+
+    switch(TYPE_MASK(n->type)) {
+    case TE_CONSTANT: printf("%f\n", n->value); break;
+    case TE_VARIABLE: printf("bound %p\n", n->bound); break;
+
+    case TE_FUNCTION0: case TE_FUNCTION1: case TE_FUNCTION2: case TE_FUNCTION3:
+    case TE_FUNCTION4: case TE_FUNCTION5: case TE_FUNCTION6: case TE_FUNCTION7:
+    case TE_CLOSURE0: case TE_CLOSURE1: case TE_CLOSURE2: case TE_CLOSURE3:
+    case TE_CLOSURE4: case TE_CLOSURE5: case TE_CLOSURE6: case TE_CLOSURE7:
+         arity = ARITY(n->type);
+         printf("f%d", arity);
+         for(i = 0; i < arity; i++) {
+             printf(" %p", n->parameters[i]);
+         }
+         printf("\n");
+         for(i = 0; i < arity; i++) {
+             pn(n->parameters[i], depth + 1);
+         }
+         break;
+    }
+}
+
+
+static void te_print(const te_expr *n) {
+    pn(n, 0);
+}
+
+struct dice_expression_
+{
+    te_expr *expr;
+};
+
+void dice_expression_free(dice_expression_t e)
+{
+    if (e == NULL) {
+        return;
+    }
+
+    te_free(e->expr);
+    free(e);
+}
+
+dice_expression_t dice_expression_parse(char const *n, int *error)
+{
+    dice_expression_t e = NULL;
+
+    e = calloc(1, sizeof(struct dice_expression_));
+    if (e == NULL) {
+        return NULL;
+    }
+
+    e->expr = te_compile(n, 0, 0, error);
+    if (e->expr == NULL) {
+        free(e);
+        return NULL;
+    }
+
+    return e;
+}
+
+bool dice_expression_evaluate(dice_expression_t e, int64_t *result)
+{
+    double val;
+    if (e == NULL || e->expr == NULL) {
+        return false;
+    }
+
+    val = te_eval(e->expr);
+    if (val == NAN) {
+        return false;
+    }
+
+    *result = (int64_t)trunc(val);
+
+    return true;
+}
+
+bool dice_expression_print(dice_expression_t e)
+{
+    if (e == NULL || e->expr == NULL) {
+        return false;
+    }
+
+    te_print(e->expr);
+    return true;
+}

tests/Makefile.am

@@ -1,6 +1,7 @@
 bin_PROGRAMS =	test_dice_simple_roll \
 		test_dice_parse \
-		test_dice_evaluate
+		test_dice_evaluate \
+		test_expr_parse
 
 AM_CFLAGS =	-I../lib ${CMOCKA_CFLAGS}
 AM_LDFLAGS =	${CMOCKA_LIBS} ../libdice.la

tests/test_dice_parse.c

@@ -6,6 +6,10 @@
 #include <string.h>
 #include <stdlib.h>
 
+/* private library function
+ */
+int dice_consumed(dice_t d);
+
 static void test_dice_parse_none(void **data)
 {
     dice_t d = dice_parse("");
@@ -31,6 +35,8 @@ static void test_dice_parse_amount_sides(void **data)
     assert_true(dice_get(d, DICEOPTION_SIDES, &i));
     assert_int_equal(i, 10);
 
+    assert_int_equal(dice_consumed(d), 4);
+
     dice_free(d);
 }
 
@@ -44,6 +50,27 @@ static void test_dice_parse_sides(void **data)
     assert_true(dice_get(d, DICEOPTION_SIDES, &i));
     assert_int_equal(i, 12);
 
+    assert_int_equal(dice_consumed(d), 3);
+
+    dice_free(d);
+}
+
+static void test_dice_parse_big(void **data)
+{
+    char const *dice_str = "1000d120000";
+    dice_t d = dice_parse(dice_str);
+    int i = 0;
+
+    assert_non_null(d);
+
+    assert_true(dice_get(d, DICEOPTION_SIDES, &i));
+    assert_int_equal(i, 120000);
+
+    assert_true(dice_get(d, DICEOPTION_AMOUNT, &i));
+    assert_int_equal(i, 1000);
+
+    assert_int_equal(dice_consumed(d), strlen(dice_str));
+
     dice_free(d);
 }
 
@@ -54,6 +81,7 @@ int main(int ac, char **av)
         cmocka_unit_test(test_dice_parse_amount),
         cmocka_unit_test(test_dice_parse_amount_sides),
         cmocka_unit_test(test_dice_parse_sides),
+        cmocka_unit_test(test_dice_parse_big),
     };
 
     return cmocka_run_group_tests(tests, NULL, NULL);