xref: /netbsd-src/usr.bin/indent/io.c (revision d40c8c6b3289d20e46c7cf72f313aecd7a4ca500)
1 /*	$NetBSD: io.c,v 1.237 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: io.c,v 1.237 2025/01/04 10:28:08 rillig Exp $");
42 
43 #include <err.h>
44 #include <stdio.h>
45 
46 #include "indent.h"
47 
48 struct input_state in = {
49 	.token_end_line = 1,
50 };
51 
52 struct output_state out;
53 enum indent_enabled indent_enabled;
54 static int out_ind;		/* width of the line that is being written */
55 static unsigned newlines = 2;	/* the total of written and buffered newlines;
56 				 * 0 in the middle of a line, 1 after a single
57 				 * finished line, anything > 1 are trailing
58 				 * blank lines */
59 static unsigned buffered_newlines;	/* not yet written */
60 static int paren_indent;	/* total indentation when parenthesized */
61 
62 
63 static void
64 inp_read_next_line(void)
65 {
66 	buf_clear(&in.line);
67 
68 	for (;;) {
69 		int ch = getc(in.f);
70 		if (ch == EOF) {
71 			if (indent_enabled == indent_on) {
72 				buf_add_char(&in.line, ' ');
73 				buf_add_char(&in.line, '\n');
74 			}
75 			had_eof = true;
76 			break;
77 		}
78 
79 		if (ch != '\0')
80 			buf_add_char(&in.line, (char)ch);
81 		if (ch == '\n')
82 			break;
83 	}
84 	buf_terminate(&in.line);
85 	in.p = in.line.s;
86 }
87 
88 void
89 inp_read_line(void)
90 {
91 	if (indent_enabled == indent_on)
92 		buf_clear(&out.indent_off_text);
93 	buf_add_chars(&out.indent_off_text, in.line.s, in.line.len);
94 	inp_read_next_line();
95 }
96 
97 void
98 inp_skip(void)
99 {
100 	in.p++;
101 	if ((size_t)(in.p - in.line.s) >= in.line.len)
102 		inp_read_line();
103 }
104 
105 char
106 inp_next(void)
107 {
108 	char ch = in.p[0];
109 	inp_skip();
110 	return ch;
111 }
112 
113 
114 static void
115 add_buffered_newline(void)
116 {
117 	buffered_newlines++;
118 	newlines++;
119 	out_ind = 0;
120 }
121 
122 static void
123 write_buffered_newlines(void)
124 {
125 	for (; buffered_newlines > 0; buffered_newlines--) {
126 		if (fputc('\n', output) == EOF)
127 			err(1, "cannot write output");
128 		debug_println("write_newline");
129 	}
130 }
131 
132 static void
133 write_range(const char *s, size_t len)
134 {
135 	write_buffered_newlines();
136 	if (fwrite(s, 1, len, output) != len)
137 		err(1, "cannot write output");
138 	debug_printf("write_range ");
139 	debug_vis_range(s, len);
140 	debug_println("");
141 	for (size_t i = 0; i < len; i++)
142 		newlines = s[i] == '\n' ? newlines + 1 : 0;
143 	out_ind = ind_add(out_ind, s, len);
144 }
145 
146 static void
147 write_indent(int new_ind)
148 {
149 	write_buffered_newlines();
150 
151 	int ind = out_ind;
152 
153 	if (opt.use_tabs) {
154 		int n = new_ind / opt.tabsize - ind / opt.tabsize;
155 		if (n > 0) {
156 			ind = ind - ind % opt.tabsize + n * opt.tabsize;
157 			while (n-- > 0)
158 				if (fputc('\t', output) == EOF)
159 					err(1, "cannot write output");
160 			newlines = 0;
161 		}
162 	}
163 
164 	for (; ind < new_ind; ind++) {
165 		if (fputc(' ', output) == EOF)
166 			err(1, "cannot write output");
167 		newlines = 0;
168 	}
169 
170 	debug_println("write_indent %d", ind);
171 	out_ind = ind;
172 }
173 
174 static bool
175 want_blank_line(void)
176 {
177 	debug_println("%s: %s -> %s", __func__,
178 	    line_kind_name[out.prev_line_kind], line_kind_name[out.line_kind]);
179 	debug_blank_line();
180 
181 	if (((ps.blank_line_after_decl && ps.declaration == decl_no)
182 	    || ps.badp == badp_yes)
183 	    && (lab.len > 0 || code.len > 0)) {
184 		ps.blank_line_after_decl = false;
185 		ps.badp = badp_none;
186 		return true;
187 	}
188 
189 	if (opt.blank_line_around_conditional_compilation) {
190 		if (out.prev_line_kind != lk_pre_if
191 		    && out.line_kind == lk_pre_if)
192 			return true;
193 		if (out.prev_line_kind == lk_pre_endif
194 		    && out.line_kind != lk_pre_endif)
195 			return true;
196 	}
197 	if (opt.blank_line_after_proc && out.prev_line_kind == lk_func_end
198 	    && out.line_kind != lk_pre_endif && out.line_kind != lk_pre_other)
199 		return true;
200 	if (opt.blank_line_before_block_comment
201 	    && out.line_kind == lk_block_comment)
202 		return true;
203 	return false;
204 }
205 
206 static bool
207 is_blank_line_optional(void)
208 {
209 	if (out.prev_line_kind == lk_stmt_head)
210 		return newlines >= 1;
211 	if (ps.psyms.len >= 3)
212 		return newlines >= 2;
213 	return newlines >= 3;
214 }
215 
216 static int
217 compute_case_label_indent(void)
218 {
219 	size_t i = ps.psyms.len - 1;
220 	while (i > 0 && ps.psyms.sym[i] != psym_switch_expr)
221 		i--;
222 	float case_ind = (float)ps.psyms.ind_level[i] + opt.case_indent;
223 	// TODO: case_ind may become negative here.
224 	return (int)(case_ind * (float)opt.indent_size);
225 }
226 
227 int
228 compute_label_indent(void)
229 {
230 	if (out.line_kind == lk_case_or_default)
231 		return compute_case_label_indent();
232 	if (lab.s[0] == '#')
233 		return 0;
234 	// TODO: the indentation may become negative here.
235 	return opt.indent_size * (ps.ind_level - 2);
236 }
237 
238 static void
239 output_line_label(void)
240 {
241 	write_indent(compute_label_indent());
242 	write_range(lab.s, lab.len);
243 }
244 
245 static int
246 compute_lined_up_code_indent(int base_ind)
247 {
248 	int ind = paren_indent;
249 	int overflow = ind_add(ind, code.s, code.len) - opt.max_line_length;
250 	if (overflow >= 0
251 	    && ind_add(base_ind, code.s, code.len) < opt.max_line_length) {
252 		ind -= 2 + overflow;
253 		if (ind < base_ind)
254 			ind = base_ind;
255 	}
256 
257 	if (ps.extra_expr_indent != eei_no
258 	    && ind == base_ind + opt.indent_size)
259 		ind += opt.continuation_indent;
260 	return ind;
261 }
262 
263 int
264 compute_code_indent(void)
265 {
266 	int base_ind = ps.ind_level * opt.indent_size;
267 
268 	if (ps.ind_paren_level == 0) {
269 		if (ps.line_is_stmt_cont)
270 			return base_ind + opt.continuation_indent;
271 		return base_ind;
272 	}
273 
274 	if (opt.lineup_to_parens) {
275 		if (opt.lineup_to_parens_always)
276 			return paren_indent;
277 		return compute_lined_up_code_indent(base_ind);
278 	}
279 
280 	int rel_ind = opt.continuation_indent * ps.ind_paren_level;
281 	if (ps.extra_expr_indent != eei_no && rel_ind == opt.indent_size)
282 		rel_ind += opt.continuation_indent;
283 	return base_ind + rel_ind;
284 }
285 
286 static void
287 output_line_code(void)
288 {
289 	int target_ind = compute_code_indent();
290 	for (size_t i = 0; i < ps.paren.len; i++) {
291 		int paren_ind = ps.paren.item[i].indent;
292 		if (paren_ind >= 0) {
293 			ps.paren.item[i].indent =
294 			    -1 - (paren_ind + target_ind);
295 			debug_println(
296 			    "setting paren_indents[%zu] from %d to %d "
297 			    "for column %d",
298 			    i, paren_ind,
299 			    ps.paren.item[i].indent, target_ind + 1);
300 		}
301 	}
302 
303 	if (lab.len > 0 && target_ind <= out_ind)
304 		write_range(" ", 1);
305 	write_indent(target_ind);
306 	write_range(code.s, code.len);
307 }
308 
309 static void
310 output_comment(void)
311 {
312 	int target_ind = ps.comment_ind;
313 	const char *p;
314 
315 	if (ps.comment_cont)
316 		target_ind += ps.comment_shift;
317 	ps.comment_cont = true;
318 
319 	/* consider the original indentation in case this is a box comment */
320 	for (p = com.s; *p == '\t'; p++)
321 		target_ind += opt.tabsize;
322 
323 	for (; target_ind < 0; p++) {
324 		if (*p == ' ')
325 			target_ind++;
326 		else if (*p == '\t')
327 			target_ind = next_tab(target_ind);
328 		else {
329 			target_ind = 0;
330 			break;
331 		}
332 	}
333 
334 	if (out_ind > target_ind)
335 		add_buffered_newline();
336 
337 	while (com.s + com.len > p && ch_isspace(com.s[com.len - 1]))
338 		com.len--;
339 	buf_terminate(&com);
340 
341 	write_indent(target_ind);
342 	write_range(p, com.len - (size_t)(p - com.s));
343 }
344 
345 /*
346  * Write a line of formatted source to the output file. The line consists of
347  * the label, the code and the comment.
348  */
349 static void
350 output_indented_line(void)
351 {
352 	if (lab.len == 0 && code.len == 0 && com.len == 0)
353 		out.line_kind = lk_blank;
354 
355 	if (want_blank_line() && newlines < 2 && out.line_kind != lk_blank)
356 		add_buffered_newline();
357 
358 	/* This kludge aligns function definitions correctly. */
359 	if (ps.ind_level == 0)
360 		ps.line_is_stmt_cont = false;
361 
362 	if (opt.blank_line_after_decl && ps.declaration == decl_end
363 	    && ps.psyms.len > 2) {
364 		ps.declaration = decl_no;
365 		ps.blank_line_after_decl = true;
366 	}
367 
368 	if (opt.swallow_optional_blank_lines
369 	    && out.line_kind == lk_blank
370 	    && is_blank_line_optional())
371 		return;
372 
373 	if (lab.len > 0)
374 		output_line_label();
375 	if (code.len > 0)
376 		output_line_code();
377 	if (com.len > 0)
378 		output_comment();
379 	add_buffered_newline();
380 	if (out.line_kind != lk_blank)
381 		write_buffered_newlines();
382 
383 	out.prev_line_kind = out.line_kind;
384 }
385 
386 static bool
387 is_stmt_cont(void)
388 {
389 	if (ps.psyms.len >= 2
390 	    && ps.psyms.sym[ps.psyms.len - 2] == psym_lbrace_enum
391 	    && ps.prev_lsym == lsym_comma
392 	    && ps.paren.len == 0)
393 		return false;
394 	return ps.in_stmt_or_decl
395 	    && (!ps.in_decl || ps.in_init)
396 	    && ps.init_level == 0;
397 }
398 
399 static void
400 prepare_next_line(void)
401 {
402 	ps.line_has_decl = ps.in_decl;
403 	ps.line_has_func_def = false;
404 	ps.line_is_stmt_cont = is_stmt_cont();
405 	ps.decl_indent_done = false;
406 	if (ps.extra_expr_indent == eei_last)
407 		ps.extra_expr_indent = eei_no;
408 	if (!(ps.psyms.sym[ps.psyms.len - 1] == psym_if_expr_stmt_else
409 		&& ps.paren.len > 0))
410 		ps.ind_level = ps.ind_level_follow;
411 	ps.ind_paren_level = (int)ps.paren.len;
412 	ps.want_blank = false;
413 	if ((ps.badp == badp_seen_lbrace || ps.badp == badp_seen_decl)
414 	    && !ps.in_decl)
415 		ps.badp = badp_yes;
416 
417 	if (ps.paren.len > 0) {
418 		/* TODO: explain what negative indentation means */
419 		paren_indent = -1 - ps.paren.item[ps.paren.len - 1].indent;
420 		debug_println("paren_indent is now %d", paren_indent);
421 	}
422 
423 	out.line_kind = lk_other;
424 }
425 
426 void
427 output_line(void)
428 {
429 	debug_blank_line();
430 	debug_buffers(__func__);
431 
432 	if (indent_enabled == indent_on)
433 		output_indented_line();
434 	else if (indent_enabled == indent_last_off_line) {
435 		indent_enabled = indent_on;
436 		write_range(out.indent_off_text.s, out.indent_off_text.len);
437 		buf_clear(&out.indent_off_text);
438 	}
439 
440 	buf_clear(&lab);
441 	buf_clear(&code);
442 	buf_clear(&com);
443 
444 	prepare_next_line();
445 }
446 
447 void
448 finish_output(void)
449 {
450 	output_line();
451 	if (indent_enabled != indent_on) {
452 		indent_enabled = indent_last_off_line;
453 		output_line();
454 	}
455 	if (fflush(output) != 0)
456 		err(1, "output file");
457 }
458