summaryrefslogtreecommitdiffstats
path: root/contrib/snowball/compiler/tokeniser.c
diff options
context:
space:
mode:
Diffstat (limited to 'contrib/snowball/compiler/tokeniser.c')
-rw-r--r--contrib/snowball/compiler/tokeniser.c470
1 files changed, 470 insertions, 0 deletions
diff --git a/contrib/snowball/compiler/tokeniser.c b/contrib/snowball/compiler/tokeniser.c
new file mode 100644
index 000000000..3dae5f744
--- /dev/null
+++ b/contrib/snowball/compiler/tokeniser.c
@@ -0,0 +1,470 @@
+
+#include <stdio.h> /* stderr etc */
+#include <stdlib.h> /* malloc free */
+#include <string.h> /* strlen */
+#include <ctype.h> /* isalpha etc */
+#include "header.h"
+
+struct system_word {
+ int s_size; /* size of system word */
+ const byte * s; /* pointer to the system word */
+ int code; /* its internal code */
+};
+
+
+/* ASCII collating assumed in syswords.c */
+
+#include "syswords.h"
+
+static int smaller(int a, int b) { return a < b ? a : b; }
+
+extern symbol * get_input(symbol * p, char ** p_file) {
+
+ char * s = b_to_s(p);
+ {
+ FILE * input = fopen(s, "r");
+ if (input == 0) { free(s); return 0; }
+ *p_file = s;
+ {
+ symbol * u = create_b(STARTSIZE);
+ int size = 0;
+ repeat
+ { int ch = getc(input);
+ if (ch == EOF) break;
+ if (size >= CAPACITY(u)) u = increase_capacity(u, size/2);
+ u[size++] = ch;
+ }
+ fclose(input);
+ SIZE(u) = size; return u;
+ }
+ }
+}
+
+static void error(struct tokeniser * t, char * s1, int n, symbol * p, char * s2) {
+ if (t->error_count == 20) { fprintf(stderr, "... etc\n"); exit(1); }
+ fprintf(stderr, "%s:%d: ", t->file, t->line_number);
+ unless (s1 == 0) fprintf(stderr, "%s", s1);
+ unless (p == 0) {
+ int i;
+ for (i = 0; i < n; i++) fprintf(stderr, "%c", p[i]);
+ }
+ unless (s2 == 0) fprintf(stderr, "%s", s2);
+ fprintf(stderr, "\n");
+ t->error_count++;
+}
+
+static void error1(struct tokeniser * t, char * s) {
+ error(t, s, 0,0, 0);
+}
+
+static void error2(struct tokeniser * t, char * s) {
+ error(t, "unexpected end of text after ", 0,0, s);
+}
+
+static int compare_words(int m, symbol * p, int n, const byte * q) {
+ unless (m == n) return m - n;
+ {
+ int i; for (i = 0; i < n; i++) {
+ int diff = p[i] - q[i];
+ unless (diff == 0) return diff;
+ }
+ }
+ return 0;
+}
+
+static int find_word(int n, symbol * p) {
+ int i = 0; int j = vocab->code;
+ repeat {
+ int k = i + (j - i)/2;
+ const struct system_word * w = vocab + k;
+ int diff = compare_words(n, p, w->s_size, w->s);
+ if (diff == 0) return w->code;
+ if (diff < 0) j = k; else i = k;
+ if (j - i == 1) break;
+ }
+ return -1;
+}
+
+static int get_number(int n, symbol * p) {
+ int x = 0;
+ int i; for (i = 0; i < n; i++) x = 10*x + p[i] - '0';
+ return x;
+}
+
+static int eq_s(struct tokeniser * t, char * s) {
+ int l = strlen(s);
+ if (SIZE(t->p) - t->c < l) return false;
+ {
+ int i;
+ for (i = 0; i < l; i++) if (t->p[t->c + i] != s[i]) return false;
+ }
+ t->c += l; return true;
+}
+
+static int white_space(struct tokeniser * t, int ch) {
+ switch (ch) {
+ case '\n': t->line_number++;
+ case '\r':
+ case '\t':
+ case ' ': return true;
+ }
+ return false;
+}
+
+static symbol * find_in_m(struct tokeniser * t, int n, symbol * p) {
+ struct m_pair * q = t->m_pairs;
+ repeat {
+ if (q == 0) return 0;
+ {
+ symbol * name = q->name;
+ if (n == SIZE(name) && memcmp(name, p, n * sizeof(symbol)) == 0) return q->value;
+ }
+ q = q->next;
+ }
+}
+
+static int read_literal_string(struct tokeniser * t, int c) {
+ symbol * p = t->p;
+ int ch;
+ SIZE(t->b) = 0;
+ repeat {
+ if (c >= SIZE(p)) { error2(t, "'"); return c; }
+ ch = p[c];
+ if (ch == '\n') { error1(t, "string not terminated"); return c; }
+ c++;
+ if (ch == t->m_start) {
+ int c0 = c;
+ int newlines = false; /* no newlines as yet */
+ int black_found = false; /* no printing chars as yet */
+ repeat {
+ if (c >= SIZE(p)) { error2(t, "'"); return c; }
+ ch = p[c]; c++;
+ if (ch == t->m_end) break;
+ unless (white_space(t, ch)) black_found = true;
+ if (ch == '\n') newlines = true;
+ if (newlines && black_found) {
+ error1(t, "string not terminated");
+ return c;
+ }
+ }
+ unless (newlines) {
+ int n = c - c0 - 1; /* macro size */
+ int firstch = p[c0];
+ symbol * q = find_in_m(t, n, p + c0);
+ if (q == 0) {
+ if (n == 1 && (firstch == '\'' || firstch == t->m_start))
+ t->b = add_to_b(t->b, 1, p + c0);
+ else
+ error(t, "string macro '", n, p + c0, "' undeclared");
+ } else
+ t->b = add_to_b(t->b, SIZE(q), q);
+ }
+ } else {
+ if (ch == '\'') return c;
+ t->b = add_to_b(t->b, 1, p + c - 1);
+ }
+ }
+}
+
+static int next_token(struct tokeniser * t) {
+ symbol * p = t->p;
+ int c = t->c;
+ int ch;
+ int code = -1;
+ repeat {
+ if (c >= SIZE(p)) { t->c = c; return -1; }
+ ch = p[c];
+ if (white_space(t, ch)) { c++; continue; }
+ if (isalpha(ch)) {
+ int c0 = c;
+ while (c < SIZE(p) && (isalnum(p[c]) || p[c] == '_')) c++;
+ code = find_word(c - c0, p + c0);
+ if (code < 0) {
+ t->b = move_to_b(t->b, c - c0, p + c0);
+ code = c_name;
+ }
+ } else
+ if (isdigit(ch)) {
+ int c0 = c;
+ while (c < SIZE(p) && isdigit(p[c])) c++;
+ t->number = get_number(c - c0, p + c0);
+ code = c_number;
+ } else
+ if (ch == '\'') {
+ c = read_literal_string(t, c + 1);
+ code = c_literalstring;
+ } else
+ {
+ int lim = smaller(2, SIZE(p) - c);
+ int i;
+ for (i = lim; i > 0; i--) {
+ code = find_word(i, p + c);
+ if (code >= 0) { c += i; break; }
+ }
+ }
+ if (code >= 0) {
+ t->c = c;
+ return code;
+ }
+ error(t, "'", 1, p + c, "' unknown");
+ c++;
+ continue;
+ }
+}
+
+static int next_char(struct tokeniser * t) {
+ if (t->c >= SIZE(t->p)) return -1;
+ return t->p[t->c++];
+}
+
+static int next_real_char(struct tokeniser * t) {
+ repeat {
+ int ch = next_char(t);
+ if (white_space(t, ch)) continue;
+ return ch;
+ }
+}
+
+static void read_chars(struct tokeniser * t) {
+ int ch = next_real_char(t);
+ if (ch < 0) { error2(t, "stringdef"); return; }
+ {
+ int c0 = t->c-1;
+ repeat {
+ ch = next_char(t);
+ if (white_space(t, ch) || ch < 0) break;
+ }
+ t->b2 = move_to_b(t->b2, t->c - c0 - 1, t->p + c0);
+ }
+}
+
+static int decimal_to_num(int ch) {
+ if ('0' <= ch && ch <= '9') return ch - '0';
+ return -1;
+}
+
+static int hex_to_num(int ch) {
+ if ('0' <= ch && ch <= '9') return ch - '0';
+ if ('a' <= ch && ch <= 'f') return ch - 'a' + 10;
+ return -1;
+}
+
+static void convert_numeric_string(struct tokeniser * t, symbol * p, int base) {
+ int c = 0; int d = 0;
+ repeat {
+ while (c < SIZE(p) && p[c] == ' ') c++;
+ if (c == SIZE(p)) break;
+ {
+ int number = 0;
+ repeat {
+ int ch = p[c];
+ if (c == SIZE(p) || ch == ' ') break;
+ if (base == 10) {
+ ch = decimal_to_num(ch);
+ if (ch < 0) {
+ error1(t, "decimal string contains non-digits");
+ return;
+ }
+ } else {
+ ch = hex_to_num(tolower(ch));
+ if (ch < 0) {
+ error1(t, "hex string contains non-hex characters");
+ return;
+ }
+ }
+ number = base * number + ch;
+ c++;
+ }
+ if (t->widechars || t->utf8) {
+ unless (0 <= number && number <= 0xffff) {
+ error1(t, "character values exceed 64K");
+ return;
+ }
+ } else {
+ unless (0 <= number && number <= 0xff) {
+ error1(t, "character values exceed 256");
+ return;
+ }
+ }
+ if (t->utf8)
+ d += put_utf8(number, p + d);
+ else
+ p[d++] = number;
+ }
+ }
+ SIZE(p) = d;
+}
+
+extern int read_token(struct tokeniser * t) {
+ symbol * p = t->p;
+ int held = t->token_held;
+ t->token_held = false;
+ if (held) return t->token;
+ repeat {
+ int code = next_token(t);
+ switch (code) {
+ case c_comment1: /* slash-slash comment */
+ while (t->c < SIZE(p) && p[t->c] != '\n') t->c++;
+ continue;
+ case c_comment2: /* slash-star comment */
+ repeat {
+ if (t->c >= SIZE(p)) {
+ error1(t, "/* comment not terminated");
+ t->token = -1;
+ return -1;
+ }
+ if (p[t->c] == '\n') t->line_number++;
+ if (eq_s(t, "*/")) break;
+ t->c++;
+ }
+ continue;
+ case c_stringescapes:
+ {
+ int ch1 = next_real_char(t);
+ int ch2 = next_real_char(t);
+ if (ch2 < 0)
+ { error2(t, "stringescapes"); continue; }
+ if (ch1 == '\'')
+ { error1(t, "first stringescape cannot be '"); continue; }
+ t->m_start = ch1;
+ t->m_end = ch2;
+ }
+ continue;
+ case c_stringdef:
+ {
+ int base = 0;
+ read_chars(t);
+ code = read_token(t);
+ if (code == c_hex) { base = 16; code = read_token(t); } else
+ if (code == c_decimal) { base = 10; code = read_token(t); }
+ unless (code == c_literalstring)
+ { error1(t, "string omitted after stringdef"); continue; }
+ if (base > 0) convert_numeric_string(t, t->b, base);
+ { NEW(m_pair, q);
+ q->next = t->m_pairs;
+ q->name = copy_b(t->b2);
+ q->value = copy_b(t->b);
+ t->m_pairs = q;
+ }
+ }
+ continue;
+ case c_get:
+ code = read_token(t);
+ unless (code == c_literalstring) {
+ error1(t, "string omitted after get"); continue;
+ }
+ t->get_depth++;
+ if (t->get_depth > 10) {
+ fprintf(stderr, "get directives go 10 deep. Looping?\n");
+ exit(1);
+ }
+ {
+ char * file;
+ NEW(input, q);
+ symbol * u = get_input(t->b, &file);
+ if (u == 0) {
+ struct include * r = t->includes;
+ until (r == 0) {
+ symbol * b = copy_b(r->b);
+ b = add_to_b(b, SIZE(t->b), t->b);
+ u = get_input(b, &file);
+ lose_b(b);
+ unless (u == 0) break;
+ r = r->next;
+ }
+ }
+ if (u == 0) {
+ error(t, "Can't get '", SIZE(t->b), t->b, "'");
+ exit(1);
+ }
+ memmove(q, t, sizeof(struct input));
+ t->next = q;
+ t->p = u;
+ t->c = 0;
+ t->file = file;
+ t->line_number = 1;
+ }
+ p = t->p;
+ continue;
+ case -1:
+ unless (t->next == 0) {
+ lose_b(p);
+ {
+ struct input * q = t->next;
+ memmove(t, q, sizeof(struct input)); p = t->p;
+ FREE(q);
+ }
+ t->get_depth--;
+ continue;
+ }
+ /* drop through */
+ default:
+ t->previous_token = t->token;
+ t->token = code;
+ return code;
+ }
+ }
+}
+
+extern const char * name_of_token(int code) {
+ int i;
+ for (i = 1; i < vocab->code; i++)
+ if ((vocab + i)->code == code) return (const char *)(vocab + i)->s;
+ switch (code) {
+ case c_mathassign: return "=";
+ case c_name: return "name";
+ case c_number: return "number";
+ case c_literalstring:return "literal";
+ case c_neg: return "neg";
+ case c_grouping: return "grouping";
+ case c_call: return "call";
+ case c_booltest: return "Boolean test";
+ case -2: return "start of text";
+ case -1: return "end of text";
+ default: return "?";
+ }
+}
+
+extern struct tokeniser * create_tokeniser(symbol * p, char * file) {
+ NEW(tokeniser, t);
+ t->next = 0;
+ t->p = p;
+ t->c = 0;
+ t->file = file;
+ t->line_number = 1;
+ t->b = create_b(0);
+ t->b2 = create_b(0);
+ t->m_start = -1;
+ t->m_pairs = 0;
+ t->get_depth = 0;
+ t->error_count = 0;
+ t->token_held = false;
+ t->token = -2;
+ t->previous_token = -2;
+ return t;
+}
+
+extern void close_tokeniser(struct tokeniser * t) {
+ lose_b(t->b);
+ lose_b(t->b2);
+ {
+ struct m_pair * q = t->m_pairs;
+ until (q == 0) {
+ struct m_pair * q_next = q->next;
+ lose_b(q->name);
+ lose_b(q->value);
+ FREE(q);
+ q = q_next;
+ }
+ }
+ {
+ struct input * q = t->next;
+ until (q == 0) {
+ struct input * q_next = q->next;
+ FREE(q);
+ q = q_next;
+ }
+ }
+ free(t->file);
+ FREE(t);
+}
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-07 01:25:50 +0000