#include "ao_lisp.h"
#include "ao_lisp_read.h"
+#include <math.h>
+#include <stdlib.h>
static const uint16_t lex_classes[128] = {
IGNORE, /* ^@ */
PRINTABLE|SPECIAL, /* ) */
PRINTABLE, /* * */
PRINTABLE|SIGN, /* + */
- PRINTABLE, /* , */
+ PRINTABLE|SPECIAL, /* , */
PRINTABLE|SIGN, /* - */
- PRINTABLE|SPECIAL, /* . */
+ PRINTABLE|DOTC|FLOATC, /* . */
PRINTABLE, /* / */
PRINTABLE|DIGIT, /* 0 */
PRINTABLE|DIGIT, /* 1 */
PRINTABLE, /* B */
PRINTABLE, /* C */
PRINTABLE, /* D */
- PRINTABLE, /* E */
+ PRINTABLE|FLOATC, /* E */
PRINTABLE, /* F */
PRINTABLE, /* G */
PRINTABLE, /* H */
PRINTABLE, /* ] */
PRINTABLE, /* ^ */
PRINTABLE, /* _ */
- PRINTABLE, /* ` */
+ PRINTABLE|SPECIAL, /* ` */
PRINTABLE, /* a */
PRINTABLE, /* b */
PRINTABLE, /* c */
PRINTABLE, /* d */
- PRINTABLE, /* e */
+ PRINTABLE|FLOATC, /* e */
PRINTABLE, /* f */
PRINTABLE, /* g */
PRINTABLE, /* h */
PRINTABLE, /* y */
PRINTABLE, /* z */
PRINTABLE, /* { */
- PRINTABLE|VBAR, /* | */
+ PRINTABLE, /* | */
PRINTABLE, /* } */
- PRINTABLE|TWIDDLE, /* ~ */
+ PRINTABLE, /* ~ */
IGNORE, /* ^? */
};
lex_unget_c = c;
}
+static uint16_t lex_class;
+
static int
-lex_quoted (void)
+lexc(void)
+{
+ int c;
+ do {
+ c = lex_get();
+ if (c == EOF) {
+ c = 0;
+ lex_class = ENDOFFILE;
+ } else {
+ c &= 0x7f;
+ lex_class = lex_classes[c];
+ }
+ } while (lex_class & IGNORE);
+ return c;
+}
+
+static int
+lex_quoted(void)
{
int c;
int v;
int count;
c = lex_get();
- if (c == EOF)
- return EOF;
+ if (c == EOF) {
+ lex_class = ENDOFFILE;
+ return 0;
+ }
+ lex_class = 0;
c &= 0x7f;
switch (c) {
case 'n':
}
}
-static uint16_t lex_class;
-
-static int
-lexc(void)
-{
- int c;
- do {
- c = lex_get();
- if (c == EOF) {
- lex_class = ENDOFFILE;
- c = 0;
- } else {
- c &= 0x7f;
- lex_class = lex_classes[c];
- if (lex_class & BACKSLASH) {
- c = lex_quoted();
- if (c == EOF)
- lex_class = ENDOFFILE;
- else
- lex_class = PRINTABLE;
- }
- }
- } while (lex_class & IGNORE);
- return c;
-}
-
#define AO_LISP_TOKEN_MAX 32
static char token_string[AO_LISP_TOKEN_MAX];
-static int token_int;
+static int32_t token_int;
static int token_len;
+static float token_float;
static inline void add_token(int c) {
if (c && token_len < AO_LISP_TOKEN_MAX - 1)
token_string[token_len++] = c;
}
+static inline void del_token(void) {
+ if (token_len > 0)
+ token_len--;
+}
+
static inline void end_token(void) {
token_string[token_len] = '\0';
}
+struct namedfloat {
+ const char *name;
+ float value;
+};
+
+static const struct namedfloat namedfloats[] = {
+ { .name = "+inf.0", .value = INFINITY },
+ { .name = "-inf.0", .value = -INFINITY },
+ { .name = "+nan.0", .value = NAN },
+ { .name = "-nan.0", .value = NAN },
+};
+
+#define NUM_NAMED_FLOATS (sizeof namedfloats / sizeof namedfloats[0])
+
static int
_lex(void)
{
continue;
}
- if (lex_class & SPECIAL) {
+ if (lex_class & (SPECIAL|DOTC)) {
add_token(c);
end_token();
switch (c) {
return QUOTE;
case '.':
return DOT;
+ case '`':
+ return QUASIQUOTE;
+ case ',':
+ c = lexc();
+ if (c == '@') {
+ add_token(c);
+ end_token();
+ return UNQUOTE_SPLICING;
+ } else {
+ lex_unget(c);
+ return UNQUOTE;
+ }
}
}
- if (lex_class & TWIDDLE) {
- token_int = lexc();
- return NUM;
- }
if (lex_class & POUND) {
- for (;;) {
- c = lexc();
+ c = lexc();
+ switch (c) {
+ case 't':
+ add_token(c);
+ end_token();
+ return BOOL;
+ case 'f':
add_token(c);
- switch (c) {
- case 't':
- return BOOL;
- case 'f':
- return BOOL;
+ end_token();
+ return BOOL;
+ case '\\':
+ for (;;) {
+ int alphabetic;
+ c = lexc();
+ alphabetic = (('a' <= c && c <= 'z') || ('A' <= c && c <= 'Z'));
+ if (token_len == 0) {
+ add_token(c);
+ if (!alphabetic)
+ break;
+ } else {
+ if (alphabetic)
+ add_token(c);
+ else {
+ lex_unget(c);
+ break;
+ }
+ }
+ }
+ end_token();
+ if (token_len == 1)
+ token_int = token_string[0];
+ else if (!strcmp(token_string, "space"))
+ token_int = ' ';
+ else if (!strcmp(token_string, "newline"))
+ token_int = '\n';
+ else if (!strcmp(token_string, "tab"))
+ token_int = '\t';
+ else if (!strcmp(token_string, "return"))
+ token_int = '\r';
+ else if (!strcmp(token_string, "formfeed"))
+ token_int = '\f';
+ else {
+ ao_lisp_error(AO_LISP_INVALID, "invalid character token #\\%s", token_string);
+ continue;
}
+ return NUM;
}
}
if (lex_class & STRINGC) {
for (;;) {
c = lexc();
+ if (lex_class & BACKSLASH)
+ c = lex_quoted();
if (lex_class & (STRINGC|ENDOFFILE)) {
end_token();
return STRING;
}
}
if (lex_class & PRINTABLE) {
- int isnum;
+ int isfloat;
int hasdigit;
int isneg;
+ int isint;
+ int epos;
- isnum = 1;
+ isfloat = 1;
+ isint = 1;
hasdigit = 0;
token_int = 0;
isneg = 0;
+ epos = 0;
for (;;) {
if (!(lex_class & NUMBER)) {
- isnum = 0;
+ isint = 0;
+ isfloat = 0;
} else {
- if (token_len != 0 &&
+ if (!(lex_class & INTEGER))
+ isint = 0;
+ if (token_len != epos &&
(lex_class & SIGN))
{
- isnum = 0;
+ isint = 0;
+ isfloat = 0;
}
if (c == '-')
isneg = 1;
+ if (c == '.' && epos != 0)
+ isfloat = 0;
+ if (c == 'e' || c == 'E') {
+ if (token_len == 0)
+ isfloat = 0;
+ else
+ epos = token_len + 1;
+ }
if (lex_class & DIGIT) {
hasdigit = 1;
- if (isnum)
+ if (isint)
token_int = token_int * 10 + c - '0';
}
}
add_token (c);
c = lexc ();
- if (lex_class & (NOTNAME)) {
+ if ((lex_class & (NOTNAME)) && (c != '.' || !isfloat)) {
+ unsigned int u;
// if (lex_class & ENDOFFILE)
// clearerr (f);
lex_unget(c);
end_token ();
- if (isnum && hasdigit) {
+ if (isint && hasdigit) {
if (isneg)
token_int = -token_int;
return NUM;
}
+ if (isfloat && hasdigit) {
+ token_float = strtof(token_string, NULL);
+ return FLOAT;
+ }
+ for (u = 0; u < NUM_NAMED_FLOATS; u++)
+ if (!strcmp(namedfloats[u].name, token_string)) {
+ token_float = namedfloats[u].value;
+ return FLOAT;
+ }
return NAME;
}
}
-
}
}
}
static inline int lex(void)
{
int parse_token = _lex();
- DBGI("token %d (%s)\n", parse_token, token_string);
+ RDBGI("token %d (%s)\n", parse_token, token_string);
return parse_token;
}
static int
push_read_stack(int cons, int read_state)
{
- DBGI("push read stack %p 0x%x\n", ao_lisp_read_cons, read_state);
- DBG_IN();
+ RDBGI("push read stack %p 0x%x\n", ao_lisp_read_cons, read_state);
+ RDBG_IN();
if (cons) {
ao_lisp_read_stack = ao_lisp_cons_cons(ao_lisp_cons_poly(ao_lisp_read_cons),
ao_lisp__cons(ao_lisp_int_poly(read_state),
ao_lisp_read_cons_tail = 0;
ao_lisp_read_stack = 0;
}
- DBG_OUT();
- DBGI("pop read stack %p %d\n", ao_lisp_read_cons, read_state);
+ RDBG_OUT();
+ RDBGI("pop read stack %p %d\n", ao_lisp_read_cons, read_state);
return read_state;
}
char *string;
int cons;
int read_state;
- ao_poly v;
-
+ ao_poly v = AO_LISP_NIL;
cons = 0;
read_state = 0;
v = AO_LISP_NIL;
break;
case NUM:
- v = ao_lisp_int_poly(token_int);
+ v = ao_lisp_integer_poly(token_int);
+ break;
+ case FLOAT:
+ v = ao_lisp_float_get(token_float);
break;
case BOOL:
if (token_string[0] == 't')
v = AO_LISP_NIL;
break;
case QUOTE:
+ case QUASIQUOTE:
+ case UNQUOTE:
+ case UNQUOTE_SPLICING:
if (!push_read_stack(cons, read_state))
return AO_LISP_NIL;
cons++;
- read_state |= READ_IN_QUOTE;
- v = _ao_lisp_atom_quote;
+ read_state = READ_IN_QUOTE;
+ switch (parse_token) {
+ case QUOTE:
+ v = _ao_lisp_atom_quote;
+ break;
+ case QUASIQUOTE:
+ v = _ao_lisp_atom_quasiquote;
+ break;
+ case UNQUOTE:
+ v = _ao_lisp_atom_unquote;
+ break;
+ case UNQUOTE_SPLICING:
+ v = _ao_lisp_atom_unquote2dsplicing;
+ break;
+ }
break;
case CLOSE:
if (!cons) {
projects / fw / altos / blobdiff