xref: /freebsd/contrib/bc/src/dc_lex.c (revision d4eeb02986980bf33dd56c41ceb9fc5f180c0d47)
1 /*
2  * *****************************************************************************
3  *
4  * SPDX-License-Identifier: BSD-2-Clause
5  *
6  * Copyright (c) 2018-2021 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
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
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
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->is_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->is_stdin || l->is_exprs))
145 			{
146 				got_more = bc_lex_readLine(l);
147 			}
148 			if (got_more) bc_vec_popAll(&l->str);
149 		}
150 	}
151 	while (got_more && depth);
152 
153 	// Obviously, if we didn't balance, that's an error.
154 	if (BC_ERR(c == '\0' && depth))
155 	{
156 		l->i = i;
157 		bc_lex_err(l, BC_ERR_PARSE_STRING);
158 	}
159 
160 	bc_vec_pushByte(&l->str, '\0');
161 
162 	l->i = i;
163 	l->line += nls;
164 }
165 
166 /**
167  * Lexes a dc token. This is the dc implementation of BcLexNext.
168  * @param l  The lexer.
169  */
170 void
171 dc_lex_token(BcLex* l)
172 {
173 	char c = l->buf[l->i++], c2;
174 	size_t i;
175 
176 	BC_SIG_ASSERT_LOCKED;
177 
178 	// If the last token was a command that needs a register, we need to parse a
179 	// register, so do so.
180 	for (i = 0; i < dc_lex_regs_len; ++i)
181 	{
182 		// If the token is a register token, take care of it and return.
183 		if (l->last == dc_lex_regs[i])
184 		{
185 			dc_lex_register(l);
186 			return;
187 		}
188 	}
189 
190 	// These lines are for tokens that easily correspond to one character. We
191 	// just set the token.
192 	if (c >= '"' && c <= '~' &&
193 	    (l->t = dc_lex_tokens[(c - '"')]) != BC_LEX_INVALID)
194 	{
195 		return;
196 	}
197 
198 	// This is the workhorse of the lexer when more complicated things are
199 	// needed.
200 	switch (c)
201 	{
202 		case '\0':
203 		case '\n':
204 		case '\t':
205 		case '\v':
206 		case '\f':
207 		case '\r':
208 		case ' ':
209 		{
210 			bc_lex_commonTokens(l, c);
211 			break;
212 		}
213 
214 		// We don't have the ! command, so we always expect certain things
215 		// after the exclamation point.
216 		case '!':
217 		{
218 			c2 = l->buf[l->i];
219 
220 			if (c2 == '=') l->t = BC_LEX_OP_REL_NE;
221 			else if (c2 == '<') l->t = BC_LEX_OP_REL_LE;
222 			else if (c2 == '>') l->t = BC_LEX_OP_REL_GE;
223 			else bc_lex_invalidChar(l, c);
224 
225 			l->i += 1;
226 
227 			break;
228 		}
229 
230 		case '#':
231 		{
232 			bc_lex_lineComment(l);
233 			break;
234 		}
235 
236 		case '.':
237 		{
238 			c2 = l->buf[l->i];
239 
240 			// If the character after is a number, this dot is part of a number.
241 			// Otherwise, it's the BSD dot (equivalent to last).
242 			if (BC_NO_ERR(BC_LEX_NUM_CHAR(c2, true, false)))
243 			{
244 				bc_lex_number(l, c);
245 			}
246 			else bc_lex_invalidChar(l, c);
247 
248 			break;
249 		}
250 
251 		case '0':
252 		case '1':
253 		case '2':
254 		case '3':
255 		case '4':
256 		case '5':
257 		case '6':
258 		case '7':
259 		case '8':
260 		case '9':
261 		case 'A':
262 		case 'B':
263 		case 'C':
264 		case 'D':
265 		case 'E':
266 		case 'F':
267 		{
268 			bc_lex_number(l, c);
269 			break;
270 		}
271 
272 		case 'g':
273 		{
274 			c2 = l->buf[l->i];
275 
276 			if (c2 == 'l') l->t = BC_LEX_KW_LINE_LENGTH;
277 			else if (c2 == 'z') l->t = BC_LEX_KW_LEADING_ZERO;
278 			else bc_lex_invalidChar(l, c2);
279 
280 			l->i += 1;
281 
282 			break;
283 		}
284 
285 		case '[':
286 		{
287 			dc_lex_string(l);
288 			break;
289 		}
290 
291 		default:
292 		{
293 			bc_lex_invalidChar(l, c);
294 		}
295 	}
296 }
297 #endif // DC_ENABLED
298