xref: /netbsd-src/usr.bin/indent/io.c (revision 901e7e84758515fbf39dfc064cb0b45ab146d8b0)
1 /*	$NetBSD: io.c,v 1.232 2023/06/27 04:41:23 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 1.232 2023/06/27 04:41:23 rillig Exp $");
42 
43 #include <stdio.h>
44 
45 #include "indent.h"
46 
47 struct buffer inp;
48 const char *inp_p;
49 
50 struct output_state out;
51 enum indent_enabled indent_enabled;
52 static int out_ind;		/* width of the line that is being written */
53 static unsigned newlines = 2;	/* the total of written and buffered newlines;
54 				 * 0 in the middle of a line, 1 after a single
55 				 * finished line, anything > 1 are trailing
56 				 * blank lines */
57 static unsigned buffered_newlines;	/* not yet written */
58 static int paren_indent;	/* total indentation when parenthesized */
59 
60 
61 static void
62 inp_read_next_line(void)
63 {
64 	buf_clear(&inp);
65 
66 	for (;;) {
67 		int ch = getc(input);
68 		if (ch == EOF) {
69 			if (indent_enabled == indent_on) {
70 				buf_add_char(&inp, ' ');
71 				buf_add_char(&inp, '\n');
72 			}
73 			had_eof = true;
74 			break;
75 		}
76 
77 		if (ch != '\0')
78 			buf_add_char(&inp, (char)ch);
79 		if (ch == '\n')
80 			break;
81 	}
82 	buf_terminate(&inp);
83 	inp_p = inp.s;
84 }
85 
86 void
87 inp_read_line(void)
88 {
89 	if (indent_enabled == indent_on)
90 		buf_clear(&out.indent_off_text);
91 	buf_add_chars(&out.indent_off_text, inp.s, inp.len);
92 	inp_read_next_line();
93 }
94 
95 void
96 inp_skip(void)
97 {
98 	inp_p++;
99 	if ((size_t)(inp_p - inp.s) >= inp.len)
100 		inp_read_line();
101 }
102 
103 char
104 inp_next(void)
105 {
106 	char ch = inp_p[0];
107 	inp_skip();
108 	return ch;
109 }
110 
111 
112 static void
113 add_buffered_newline(void)
114 {
115 	buffered_newlines++;
116 	newlines++;
117 	out_ind = 0;
118 }
119 
120 static void
121 write_buffered_newlines(void)
122 {
123 	for (; buffered_newlines > 0; buffered_newlines--) {
124 		fputc('\n', output);
125 		debug_println("write_newline");
126 	}
127 }
128 
129 static void
130 write_range(const char *s, size_t len)
131 {
132 	write_buffered_newlines();
133 	fwrite(s, 1, len, output);
134 	debug_printf("write_range ");
135 	debug_vis_range(s, len);
136 	debug_println("");
137 	for (size_t i = 0; i < len; i++)
138 		newlines = s[i] == '\n' ? newlines + 1 : 0;
139 	out_ind = ind_add(out_ind, s, len);
140 }
141 
142 static void
143 write_indent(int new_ind)
144 {
145 	write_buffered_newlines();
146 
147 	int ind = out_ind;
148 
149 	if (opt.use_tabs) {
150 		int n = new_ind / opt.tabsize - ind / opt.tabsize;
151 		if (n > 0) {
152 			ind = ind - ind % opt.tabsize + n * opt.tabsize;
153 			while (n-- > 0)
154 				fputc('\t', output);
155 			newlines = 0;
156 		}
157 	}
158 
159 	for (; ind < new_ind; ind++) {
160 		fputc(' ', output);
161 		newlines = 0;
162 	}
163 
164 	debug_println("write_indent %d", ind);
165 	out_ind = ind;
166 }
167 
168 static bool
169 want_blank_line(void)
170 {
171 	debug_println("%s: %s -> %s", __func__,
172 	    line_kind_name[out.prev_line_kind], line_kind_name[out.line_kind]);
173 
174 	if (((ps.blank_line_after_decl && ps.declaration == decl_no)
175 	    || ps.badp == badp_yes)
176 	    && (lab.len > 0 || code.len > 0)) {
177 		ps.blank_line_after_decl = false;
178 		ps.badp = badp_none;
179 		return true;
180 	}
181 
182 	if (opt.blank_line_around_conditional_compilation) {
183 		if (out.prev_line_kind != lk_pre_if
184 		    && out.line_kind == lk_pre_if)
185 			return true;
186 		if (out.prev_line_kind == lk_pre_endif
187 		    && out.line_kind != lk_pre_endif)
188 			return true;
189 	}
190 	if (opt.blank_line_after_proc && out.prev_line_kind == lk_func_end
191 	    && out.line_kind != lk_pre_endif && out.line_kind != lk_pre_other)
192 		return true;
193 	if (opt.blank_line_before_block_comment
194 	    && out.line_kind == lk_block_comment)
195 		return true;
196 	return false;
197 }
198 
199 static bool
200 is_blank_line_optional(void)
201 {
202 	if (out.prev_line_kind == lk_stmt_head)
203 		return newlines >= 1;
204 	if (ps.psyms.len >= 3)
205 		return newlines >= 2;
206 	return newlines >= 3;
207 }
208 
209 static int
210 compute_case_label_indent(void)
211 {
212 	size_t i = ps.psyms.len - 1;
213 	while (i > 0 && ps.psyms.sym[i] != psym_switch_expr)
214 		i--;
215 	float case_ind = (float)ps.psyms.ind_level[i] + opt.case_indent;
216 	// TODO: case_ind may become negative here.
217 	return (int)(case_ind * (float)opt.indent_size);
218 }
219 
220 int
221 compute_label_indent(void)
222 {
223 	if (out.line_kind == lk_case_or_default)
224 		return compute_case_label_indent();
225 	if (lab.s[0] == '#')
226 		return 0;
227 	// TODO: the indentation may become negative here.
228 	return opt.indent_size * (ps.ind_level - 2);
229 }
230 
231 static void
232 output_line_label(void)
233 {
234 	write_indent(compute_label_indent());
235 	write_range(lab.s, lab.len);
236 }
237 
238 static int
239 compute_lined_up_code_indent(int base_ind)
240 {
241 	int ind = paren_indent;
242 	int overflow = ind_add(ind, code.s, code.len) - opt.max_line_length;
243 	if (overflow >= 0
244 	    && ind_add(base_ind, code.s, code.len) < opt.max_line_length) {
245 		ind -= 2 + overflow;
246 		if (ind < base_ind)
247 			ind = base_ind;
248 	}
249 
250 	if (ps.extra_expr_indent != eei_no
251 	    && ind == base_ind + opt.indent_size)
252 		ind += opt.continuation_indent;
253 	return ind;
254 }
255 
256 int
257 compute_code_indent(void)
258 {
259 	int base_ind = ps.ind_level * opt.indent_size;
260 
261 	if (ps.ind_paren_level == 0) {
262 		if (ps.line_is_stmt_cont)
263 			return base_ind + opt.continuation_indent;
264 		return base_ind;
265 	}
266 
267 	if (opt.lineup_to_parens) {
268 		if (opt.lineup_to_parens_always)
269 			return paren_indent;
270 		return compute_lined_up_code_indent(base_ind);
271 	}
272 
273 	int rel_ind = opt.continuation_indent * ps.ind_paren_level;
274 	if (ps.extra_expr_indent != eei_no && rel_ind == opt.indent_size)
275 		rel_ind += opt.continuation_indent;
276 	return base_ind + rel_ind;
277 }
278 
279 static void
280 output_line_code(void)
281 {
282 	int target_ind = compute_code_indent();
283 	for (size_t i = 0; i < ps.paren.len; i++) {
284 		int paren_ind = ps.paren.item[i].indent;
285 		if (paren_ind >= 0) {
286 			ps.paren.item[i].indent =
287 			    -1 - (paren_ind + target_ind);
288 			debug_println(
289 			    "setting paren_indents[%zu] from %d to %d "
290 			    "for column %d",
291 			    i, paren_ind,
292 			    ps.paren.item[i].indent, target_ind + 1);
293 		}
294 	}
295 
296 	if (lab.len > 0 && target_ind <= out_ind)
297 		write_range(" ", 1);
298 	write_indent(target_ind);
299 	write_range(code.s, code.len);
300 }
301 
302 static void
303 output_comment(void)
304 {
305 	int target_ind = ps.comment_ind;
306 	const char *p;
307 
308 	if (ps.comment_cont)
309 		target_ind += ps.comment_shift;
310 	ps.comment_cont = true;
311 
312 	/* consider the original indentation in case this is a box comment */
313 	for (p = com.s; *p == '\t'; p++)
314 		target_ind += opt.tabsize;
315 
316 	for (; target_ind < 0; p++) {
317 		if (*p == ' ')
318 			target_ind++;
319 		else if (*p == '\t')
320 			target_ind = next_tab(target_ind);
321 		else {
322 			target_ind = 0;
323 			break;
324 		}
325 	}
326 
327 	if (out_ind > target_ind)
328 		add_buffered_newline();
329 
330 	while (com.s + com.len > p && ch_isspace(com.s[com.len - 1]))
331 		com.len--;
332 	buf_terminate(&com);
333 
334 	write_indent(target_ind);
335 	write_range(p, com.len - (size_t)(p - com.s));
336 }
337 
338 /*
339  * Write a line of formatted source to the output file. The line consists of
340  * the label, the code and the comment.
341  */
342 static void
343 output_indented_line(void)
344 {
345 	if (lab.len == 0 && code.len == 0 && com.len == 0)
346 		out.line_kind = lk_blank;
347 
348 	if (want_blank_line() && newlines < 2 && out.line_kind != lk_blank)
349 		add_buffered_newline();
350 
351 	/* This kludge aligns function definitions correctly. */
352 	if (ps.ind_level == 0)
353 		ps.line_is_stmt_cont = false;
354 
355 	if (opt.blank_line_after_decl && ps.declaration == decl_end
356 	    && ps.psyms.len > 2) {
357 		ps.declaration = decl_no;
358 		ps.blank_line_after_decl = true;
359 	}
360 
361 	if (opt.swallow_optional_blank_lines
362 	    && out.line_kind == lk_blank
363 	    && is_blank_line_optional())
364 		return;
365 
366 	if (lab.len > 0)
367 		output_line_label();
368 	if (code.len > 0)
369 		output_line_code();
370 	if (com.len > 0)
371 		output_comment();
372 	add_buffered_newline();
373 	if (out.line_kind != lk_blank)
374 		write_buffered_newlines();
375 
376 	out.prev_line_kind = out.line_kind;
377 }
378 
379 static bool
380 is_stmt_cont(void)
381 {
382 	if (ps.psyms.len >= 2
383 	    && ps.psyms.sym[ps.psyms.len - 2] == psym_lbrace_enum
384 	    && ps.prev_lsym == lsym_comma
385 	    && ps.paren.len == 0)
386 		return false;
387 	return ps.in_stmt_or_decl
388 	    && (!ps.in_decl || ps.in_init)
389 	    && ps.init_level == 0;
390 }
391 
392 static void
393 prepare_next_line(void)
394 {
395 	ps.line_has_decl = ps.in_decl;
396 	ps.line_has_func_def = false;
397 	ps.line_is_stmt_cont = is_stmt_cont();
398 	ps.decl_indent_done = false;
399 	if (ps.extra_expr_indent == eei_last)
400 		ps.extra_expr_indent = eei_no;
401 	if (!(ps.psyms.sym[ps.psyms.len - 1] == psym_if_expr_stmt_else
402 		&& ps.paren.len > 0))
403 		ps.ind_level = ps.ind_level_follow;
404 	ps.ind_paren_level = (int)ps.paren.len;
405 	ps.want_blank = false;
406 	if ((ps.badp == badp_seen_lbrace || ps.badp == badp_seen_decl)
407 	    && !ps.in_decl)
408 		ps.badp = badp_yes;
409 
410 	if (ps.paren.len > 0) {
411 		/* TODO: explain what negative indentation means */
412 		paren_indent = -1 - ps.paren.item[ps.paren.len - 1].indent;
413 		debug_println("paren_indent is now %d", paren_indent);
414 	}
415 
416 	out.line_kind = lk_other;
417 }
418 
419 void
420 output_line(void)
421 {
422 	debug_blank_line();
423 	debug_printf("%s", __func__);
424 	debug_buffers();
425 
426 	if (indent_enabled == indent_on)
427 		output_indented_line();
428 	else if (indent_enabled == indent_last_off_line) {
429 		indent_enabled = indent_on;
430 		write_range(out.indent_off_text.s, out.indent_off_text.len);
431 		buf_clear(&out.indent_off_text);
432 	}
433 
434 	buf_clear(&lab);
435 	buf_clear(&code);
436 	buf_clear(&com);
437 
438 	prepare_next_line();
439 }
440 
441 void
442 finish_output(void)
443 {
444 	output_line();
445 	if (indent_enabled != indent_on) {
446 		indent_enabled = indent_last_off_line;
447 		output_line();
448 	}
449 	fflush(output);
450 }
451