1 /*
2 * *****************************************************************************
3 *
4 * SPDX-License-Identifier: BSD-2-Clause
5 *
6 * Copyright (c) 2018-2023 Gavin D. Howard and contributors.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions are met:
10 *
11 * * Redistributions of source code must retain the above copyright notice, this
12 * list of conditions and the following disclaimer.
13 *
14 * * Redistributions in binary form must reproduce the above copyright notice,
15 * this list of conditions and the following disclaimer in the documentation
16 * and/or other materials provided with the distribution.
17 *
18 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
19 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
22 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
23 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
24 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
25 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
26 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
27 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
28 * POSSIBILITY OF SUCH DAMAGE.
29 *
30 * *****************************************************************************
31 *
32 * The lexer for dc.
33 *
34 */
35
36 #if DC_ENABLED
37
38 #include <ctype.h>
39
40 #include <dc.h>
41 #include <vm.h>
42
43 bool
dc_lex_negCommand(BcLex * l)44 dc_lex_negCommand(BcLex* l)
45 {
46 char c = l->buf[l->i];
47 return !BC_LEX_NUM_CHAR(c, false, false);
48 }
49
50 /**
51 * Processes a dc command that needs a register. This is where the
52 * extended-register extension is implemented.
53 * @param l The lexer.
54 */
55 static void
dc_lex_register(BcLex * l)56 dc_lex_register(BcLex* l)
57 {
58 // If extended register is enabled and the character is whitespace...
59 if (DC_X && isspace(l->buf[l->i - 1]))
60 {
61 char c;
62
63 // Eat the whitespace.
64 bc_lex_whitespace(l);
65 c = l->buf[l->i];
66
67 // Check for a letter or underscore.
68 if (BC_ERR(!isalpha(c) && c != '_'))
69 {
70 bc_lex_verr(l, BC_ERR_PARSE_CHAR, c);
71 }
72
73 // Parse a normal identifier.
74 l->i += 1;
75 bc_lex_name(l);
76 }
77 else
78 {
79 // I don't allow newlines because newlines are used for controlling when
80 // execution happens, and allowing newlines would just be complex.
81 if (BC_ERR(l->buf[l->i - 1] == '\n'))
82 {
83 bc_lex_verr(l, BC_ERR_PARSE_CHAR, l->buf[l->i - 1]);
84 }
85
86 // Set the lexer string and token.
87 bc_vec_popAll(&l->str);
88 bc_vec_pushByte(&l->str, (uchar) l->buf[l->i - 1]);
89 bc_vec_pushByte(&l->str, '\0');
90 l->t = BC_LEX_NAME;
91 }
92 }
93
94 /**
95 * Parses a dc string. Since dc's strings need to check for balanced brackets,
96 * we can't just parse bc and dc strings with different start and end
97 * characters. Oh, and dc strings need to check for escaped brackets.
98 * @param l The lexer.
99 */
100 static void
dc_lex_string(BcLex * l)101 dc_lex_string(BcLex* l)
102 {
103 size_t depth, nls, i;
104 char c;
105 bool got_more;
106
107 // Set the token and clear the string.
108 l->t = BC_LEX_STR;
109 bc_vec_popAll(&l->str);
110
111 do
112 {
113 depth = 1;
114 nls = 0;
115 got_more = false;
116
117 assert(l->mode != BC_MODE_STDIN || l->buf == vm->buffer.v);
118
119 // This is the meat. As long as we don't run into the NUL byte, and we
120 // have "depth", which means we haven't completely balanced brackets
121 // yet, we continue eating the string.
122 for (i = l->i; (c = l->buf[i]) && depth; ++i)
123 {
124 // Check for escaped brackets and set the depths as appropriate.
125 if (c == '\\')
126 {
127 c = l->buf[++i];
128 if (!c) break;
129 }
130 else
131 {
132 depth += (c == '[');
133 depth -= (c == ']');
134 }
135
136 // We want to adjust the line in the lexer as necessary.
137 nls += (c == '\n');
138
139 if (depth) bc_vec_push(&l->str, &c);
140 }
141
142 if (BC_ERR(c == '\0' && depth))
143 {
144 if (!vm->eof && l->mode != BC_MODE_FILE)
145 {
146 got_more = bc_lex_readLine(l);
147 }
148
149 if (got_more)
150 {
151 bc_vec_popAll(&l->str);
152 }
153 }
154 }
155 while (got_more && depth);
156
157 // Obviously, if we didn't balance, that's an error.
158 if (BC_ERR(c == '\0' && depth))
159 {
160 l->i = i;
161 bc_lex_err(l, BC_ERR_PARSE_STRING);
162 }
163
164 bc_vec_pushByte(&l->str, '\0');
165
166 l->i = i;
167 l->line += nls;
168 }
169
170 /**
171 * Lexes a dc token. This is the dc implementation of BcLexNext.
172 * @param l The lexer.
173 */
174 void
dc_lex_token(BcLex * l)175 dc_lex_token(BcLex* l)
176 {
177 char c = l->buf[l->i++], c2;
178 size_t i;
179
180 BC_SIG_ASSERT_LOCKED;
181
182 // If the last token was a command that needs a register, we need to parse a
183 // register, so do so.
184 for (i = 0; i < dc_lex_regs_len; ++i)
185 {
186 // If the token is a register token, take care of it and return.
187 if (l->last == dc_lex_regs[i])
188 {
189 dc_lex_register(l);
190 return;
191 }
192 }
193
194 // These lines are for tokens that easily correspond to one character. We
195 // just set the token.
196 if (c >= '"' && c <= '~' &&
197 (l->t = dc_lex_tokens[(c - '"')]) != BC_LEX_INVALID)
198 {
199 return;
200 }
201
202 // This is the workhorse of the lexer when more complicated things are
203 // needed.
204 switch (c)
205 {
206 case '\0':
207 case '\n':
208 case '\t':
209 case '\v':
210 case '\f':
211 case '\r':
212 case ' ':
213 {
214 bc_lex_commonTokens(l, c);
215 break;
216 }
217
218 // We don't have the ! command, so we always expect certain things
219 // after the exclamation point.
220 case '!':
221 {
222 c2 = l->buf[l->i];
223
224 if (c2 == '=') l->t = BC_LEX_OP_REL_NE;
225 else if (c2 == '<') l->t = BC_LEX_OP_REL_LE;
226 else if (c2 == '>') l->t = BC_LEX_OP_REL_GE;
227 else bc_lex_invalidChar(l, c);
228
229 l->i += 1;
230
231 break;
232 }
233
234 case '#':
235 {
236 bc_lex_lineComment(l);
237 break;
238 }
239
240 case '.':
241 {
242 c2 = l->buf[l->i];
243
244 // If the character after is a number, this dot is part of a number.
245 // Otherwise, it's the BSD dot (equivalent to last).
246 if (BC_NO_ERR(BC_LEX_NUM_CHAR(c2, true, false)))
247 {
248 bc_lex_number(l, c);
249 }
250 else bc_lex_invalidChar(l, c);
251
252 break;
253 }
254
255 case '0':
256 case '1':
257 case '2':
258 case '3':
259 case '4':
260 case '5':
261 case '6':
262 case '7':
263 case '8':
264 case '9':
265 case 'A':
266 case 'B':
267 case 'C':
268 case 'D':
269 case 'E':
270 case 'F':
271 {
272 bc_lex_number(l, c);
273 break;
274 }
275
276 case 'g':
277 {
278 c2 = l->buf[l->i];
279
280 if (c2 == 'l') l->t = BC_LEX_KW_LINE_LENGTH;
281 else if (c2 == 'x') l->t = BC_LEX_EXTENDED_REGISTERS;
282 else if (c2 == 'z') l->t = BC_LEX_KW_LEADING_ZERO;
283 else bc_lex_invalidChar(l, c2);
284
285 l->i += 1;
286
287 break;
288 }
289
290 case '[':
291 {
292 dc_lex_string(l);
293 break;
294 }
295
296 default:
297 {
298 bc_lex_invalidChar(l, c);
299 }
300 }
301 }
302 #endif // DC_ENABLED
303