xref: /netbsd-src/usr.bin/indent/pr_comment.c (revision d40c8c6b3289d20e46c7cf72f313aecd7a4ca500)
1 /*	$NetBSD: pr_comment.c,v 1.174 2025/01/04 10:28:08 rillig Exp $	*/
2 
3 /*-
4  * SPDX-License-Identifier: BSD-4-Clause
5  *
6  * Copyright (c) 1985 Sun Microsystems, Inc.
7  * Copyright (c) 1980, 1993
8  *	The Regents of the University of California.  All rights reserved.
9  * All rights reserved.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  * 3. All advertising materials mentioning features or use of this software
20  *    must display the following acknowledgement:
21  *	This product includes software developed by the University of
22  *	California, Berkeley and its contributors.
23  * 4. Neither the name of the University nor the names of its contributors
24  *    may be used to endorse or promote products derived from this software
25  *    without specific prior written permission.
26  *
27  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
28  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
31  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
32  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
33  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
34  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
35  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
36  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
37  * SUCH DAMAGE.
38  */
39 
40 #include <sys/cdefs.h>
41 __RCSID("$NetBSD: pr_comment.c,v 1.174 2025/01/04 10:28:08 rillig Exp $");
42 
43 #include <string.h>
44 
45 #include "indent.h"
46 
47 static void
48 com_add_char(char ch)
49 {
50 	buf_add_char(&com, ch);
51 }
52 
53 static void
54 com_add_star(void)
55 {
56 	if (opt.star_comment_cont)
57 		buf_add_str(&com, " * ");
58 }
59 
60 static bool
61 fits_in_one_line(int max_line_length)
62 {
63 	for (const char *start = in.p, *p = start; *p != '\n'; p++) {
64 		if (p[0] == '*' && p[1] == '/') {
65 			while (p - in.p >= 2
66 			    && ch_isblank(p[-1])
67 			    && ch_isblank(p[-2]))
68 				p--;
69 			int ind = ind_add(ps.comment_ind + 3,
70 			    start, (size_t)(p - start));
71 			ind += p == start || ch_isblank(p[-1]) ? 2 : 3;
72 			return ind <= max_line_length;
73 		}
74 	}
75 	return false;
76 }
77 
78 static bool
79 is_block_comment(void)
80 {
81 	const char *p = in.p;
82 	while (*p == '*')
83 		p++;
84 	return *p == '\n';
85 }
86 
87 static void
88 analyze_comment(bool *p_may_wrap, bool *p_delim, int *p_line_length)
89 {
90 	bool may_wrap = true;
91 	bool delim = false;	// only relevant if may_wrap
92 	int ind;
93 	int line_length = opt.max_line_length;
94 
95 	if (in.p - in.line.s == 2 && !opt.format_col1_comments) {
96 		may_wrap = false;
97 		ind = 0;
98 	} else {
99 		if (in.p[0] == '-' || in.p[0] == '*' ||
100 		    token.s[token.len - 1] == '/' ||
101 		    (in.p[0] == '\n' && !opt.format_block_comments))
102 			may_wrap = false;
103 
104 		if (com.len > 0)
105 			output_line();
106 		if (lab.len == 0 && code.len == 0) {
107 			if (is_block_comment())
108 				out.line_kind = lk_block_comment;
109 			ind = (ps.ind_level - opt.unindent_displace)
110 			    * opt.indent_size;
111 			if (ind <= 0)
112 				ind = opt.format_col1_comments ? 0 : 1;
113 			line_length = opt.block_comment_max_line_length;
114 			if (may_wrap && in.p[0] == '\n')
115 				delim = true;
116 			if (may_wrap && opt.comment_delimiter_on_blank_line)
117 				delim = true;
118 		} else {
119 			int min_ind = code.len > 0
120 			    ? ind_add(compute_code_indent(), code.s, code.len)
121 			    : ind_add(compute_label_indent(), lab.s, lab.len);
122 
123 			ind = ps.line_has_decl || ps.ind_level == 0
124 			    ? opt.decl_comment_column - 1
125 			    : opt.comment_column - 1;
126 			if (ind <= min_ind)
127 				ind = next_tab(min_ind);
128 			if (ind + 25 > line_length)
129 				line_length = ind + 25;
130 		}
131 	}
132 
133 	if (!may_wrap) {
134 		/* Find out how much indentation there was originally, because
135 		 * that much will have to be ignored by output_line. */
136 		size_t len = (size_t)(in.p - 2 - in.line.s);
137 		ps.comment_shift = -ind_add(0, in.line.s, len);
138 	} else {
139 		ps.comment_shift = 0;
140 		if (!(in.p[0] == '\t' && !ch_isblank(in.p[1])))
141 			while (ch_isblank(in.p[0]))
142 				in.p++;
143 	}
144 
145 	ps.comment_ind = ind;
146 	*p_may_wrap = may_wrap;
147 	*p_delim = delim;
148 	*p_line_length = line_length;
149 }
150 
151 static void
152 copy_comment_start(bool may_wrap, bool *delim, int line_length)
153 {
154 	ps.comment_cont = false;
155 	buf_add_chars(&com, token.s, token.len);	// "/*" or "//"
156 
157 	if (may_wrap) {
158 		if (!ch_isblank(in.p[0]))
159 			com_add_char(' ');
160 
161 		if (*delim && fits_in_one_line(line_length))
162 			*delim = false;
163 		if (*delim) {
164 			output_line();
165 			com_add_star();
166 		}
167 	}
168 }
169 
170 static void
171 copy_comment_wrap_text(int line_length, ssize_t *last_blank)
172 {
173 	int ind = ind_add(ps.comment_ind, com.s, com.len);
174 	for (;;) {
175 		char ch = inp_next();
176 		if (ch_isblank(ch))
177 			*last_blank = (ssize_t)com.len;
178 		com_add_char(ch);
179 		ind++;
180 		if (memchr("*\n\r\t", in.p[0], 5) != NULL)
181 			break;
182 		if (ind >= line_length && *last_blank != -1)
183 			break;
184 	}
185 
186 	if (ind <= line_length)
187 		return;
188 	if (ch_isspace(com.s[com.len - 1]))
189 		return;
190 
191 	if (*last_blank == -1) {	/* only a single word in this line */
192 		output_line();
193 		com_add_star();
194 		return;
195 	}
196 
197 	// Move the overlong word to the next line.
198 	const char *last_word = com.s + *last_blank + 1;
199 	size_t last_word_len = com.len - (size_t)(*last_blank + 1);
200 	com.len = (size_t)*last_blank;
201 	buf_terminate(&com);
202 	output_line();
203 	com_add_star();
204 
205 	/* Assume that output_line and com_add_delim left the "unused" part of
206 	 * the now truncated buffer beyond com.s + com.len as-is. */
207 	memmove(com.s + com.len, last_word, last_word_len);
208 	com.len += last_word_len;
209 	buf_terminate(&com);
210 	*last_blank = -1;
211 }
212 
213 /* In a comment that is re-wrapped, handle a single newline character. */
214 static bool
215 copy_comment_wrap_newline(ssize_t *last_blank, bool seen_newline)
216 {
217 	*last_blank = -1;
218 	if (seen_newline) {
219 		if (com.len > 3) {
220 			output_line();
221 			com_add_star();
222 		}
223 		output_line();
224 		com_add_star();
225 	} else {
226 		if (!(com.len > 0 && ch_isblank(com.s[com.len - 1])))
227 			com_add_char(' ');
228 		*last_blank = (int)com.len - 1;
229 	}
230 	in.token_end_line++;
231 
232 	/* flush any blanks and/or tabs at start of next line */
233 	inp_skip();		/* '\n' */
234 	while (ch_isblank(in.p[0]))
235 		in.p++;
236 	if (in.p[0] == '*' && in.p[1] == '/')
237 		return false;
238 	if (in.p[0] == '*') {
239 		in.p++;
240 		while (ch_isblank(in.p[0]))
241 			in.p++;
242 	}
243 
244 	return true;
245 }
246 
247 static void
248 copy_comment_wrap_finish(int line_length, bool delim)
249 {
250 	if (delim) {
251 		if (com.len > 3)
252 			output_line();
253 		buf_clear(&com);
254 	} else {
255 		size_t len = com.len;
256 		// XXX: This loop differs from the one below.
257 		while (ch_isblank(com.s[len - 1]))
258 			len--;
259 		if (ind_add(ps.comment_ind, com.s, len) + 3 > line_length)
260 			output_line();
261 	}
262 
263 	while (com.len >= 2
264 	    && ch_isblank(com.s[com.len - 1])
265 	    && ch_isblank(com.s[com.len - 2]))
266 		com.len--;
267 	buf_terminate(&com);
268 
269 	in.p += 2;
270 	if (com.len > 0 && ch_isblank(com.s[com.len - 1]))
271 		buf_add_str(&com, "*/");
272 	else
273 		buf_add_str(&com, " */");
274 }
275 
276 static void
277 copy_comment_wrap(int line_length, bool delim)
278 {
279 	ssize_t last_blank = -1;	/* index of the last blank in 'com' */
280 	bool seen_newline = false;
281 
282 	for (;;) {
283 		if (in.p[0] == '\n') {
284 			if (had_eof)
285 				goto unterminated_comment;
286 			if (!copy_comment_wrap_newline(&last_blank,
287 				seen_newline))
288 				break;
289 			seen_newline = true;
290 		} else if (in.p[0] == '*' && in.p[1] == '/')
291 			break;
292 		else {
293 			copy_comment_wrap_text(line_length, &last_blank);
294 			seen_newline = false;
295 		}
296 	}
297 
298 	copy_comment_wrap_finish(line_length, delim);
299 	return;
300 
301 unterminated_comment:
302 	in.token_start_line = in.token_end_line;
303 	diag(1, "Unterminated comment");
304 	output_line();
305 }
306 
307 static void
308 copy_comment_nowrap(void)
309 {
310 	char kind = token.s[token.len - 1];
311 
312 	for (;;) {
313 		if (in.p[0] == '\n') {
314 			if (kind == '/')
315 				return;
316 
317 			if (had_eof) {
318 				in.token_start_line = in.token_end_line;
319 				diag(1, "Unterminated comment");
320 				output_line();
321 				return;
322 			}
323 
324 			output_line();
325 			in.token_end_line++;
326 			inp_skip();
327 			continue;
328 		}
329 
330 		if (kind == '*' && in.p[0] == '*' && in.p[1] == '/') {
331 			com_add_char(*in.p++);
332 			com_add_char(*in.p++);
333 			return;
334 		}
335 
336 		com_add_char(*in.p++);
337 	}
338 }
339 
340 void
341 process_comment(void)
342 {
343 	bool may_wrap, delim;
344 	int line_length;
345 
346 	analyze_comment(&may_wrap, &delim, &line_length);
347 	copy_comment_start(may_wrap, &delim, line_length);
348 	if (may_wrap)
349 		copy_comment_wrap(line_length, delim);
350 	else
351 		copy_comment_nowrap();
352 }
353