pr_comment.c revision 1.154
1 /* $NetBSD: pr_comment.c,v 1.154 2023/06/06 07:14:20 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 1.154 2023/06/06 07:14:20 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_delim(void) 55 { 56 if (opt.star_comment_cont) 57 buf_add_chars(&com, " * ", 3); 58 } 59 60 static bool 61 fits_in_one_line(int com_ind, int max_line_length) 62 { 63 for (const char *start = inp_p, *p = start; *p != '\n'; p++) { 64 if (p[0] == '*' && p[1] == '/') { 65 int len = ind_add(com_ind + 3, 66 start, (size_t)(p - start)); 67 len += p == start || ch_isblank(p[-1]) ? 2 : 3; 68 return len <= max_line_length; 69 } 70 } 71 return false; 72 } 73 74 static void 75 analyze_comment(bool *p_may_wrap, bool *p_delim, 76 int *p_ind, int *p_line_length) 77 { 78 bool may_wrap = true; 79 bool delim = opt.comment_delimiter_on_blankline; 80 int ind; 81 int line_length = opt.max_line_length; 82 83 if (ps.curr_col_1 && !opt.format_col1_comments) { 84 may_wrap = false; 85 delim = false; 86 ind = 0; 87 88 } else { 89 if (inp_p[0] == '-' || inp_p[0] == '*' || 90 token.s[token.len - 1] == '/' || 91 (inp_p[0] == '\n' && !opt.format_block_comments)) { 92 may_wrap = false; 93 delim = false; 94 } 95 if (code.len == 0 && inp_p[strspn(inp_p, "*")] == '\n') 96 out.line_kind = lk_block_comment; 97 98 if (com.len > 0) 99 output_line(); 100 if (lab.len == 0 && code.len == 0) { 101 ind = (ps.ind_level - opt.unindent_displace) 102 * opt.indent_size; 103 if (ind <= 0) 104 ind = opt.format_col1_comments ? 0 : 1; 105 line_length = opt.block_comment_max_line_length; 106 } else { 107 delim = false; 108 109 int target_ind = code.len > 0 110 ? ind_add(compute_code_indent(), code.s, code.len) 111 : ind_add(compute_label_indent(), lab.s, lab.len); 112 113 ind = ps.decl_on_line || ps.ind_level == 0 114 ? opt.decl_comment_column - 1 115 : opt.comment_column - 1; 116 if (ind <= target_ind) 117 ind = next_tab(target_ind); 118 if (ind + 25 > line_length) 119 line_length = ind + 25; 120 } 121 } 122 123 ps.com_ind = ind; 124 125 if (!may_wrap) { 126 /* Find out how much indentation there was originally, because 127 * that much will have to be ignored by output_line. */ 128 size_t len = (size_t)(inp_p - 2 - inp.s); 129 ps.n_comment_delta = -ind_add(0, inp.s, len); 130 } else { 131 ps.n_comment_delta = 0; 132 if (!(inp_p[0] == '\t' && !ch_isblank(inp_p[1]))) 133 while (ch_isblank(inp_p[0])) 134 inp_p++; 135 } 136 137 *p_may_wrap = may_wrap; 138 *p_delim = delim; 139 *p_ind = ind; 140 *p_line_length = line_length; 141 } 142 143 static void 144 copy_comment_start(bool may_wrap, bool *delim, int ind, int line_length) 145 { 146 ps.comment_delta = 0; 147 com_add_char('/'); 148 com_add_char(token.s[token.len - 1]); /* either '*' or '/' */ 149 150 if (may_wrap && !ch_isblank(inp_p[0])) 151 com_add_char(' '); 152 153 if (*delim && fits_in_one_line(ind, line_length)) 154 *delim = false; 155 156 if (*delim) { 157 output_line(); 158 com_add_delim(); 159 } 160 } 161 162 static void 163 copy_comment_wrap_text(int line_length, ssize_t *last_blank) 164 { 165 int now_len = ind_add(ps.com_ind, com.s, com.len); 166 for (;;) { 167 char ch = inp_next(); 168 if (ch_isblank(ch)) 169 *last_blank = (ssize_t)com.len; 170 com_add_char(ch); 171 now_len++; 172 if (memchr("*\n\r\b\t", inp_p[0], 6) != NULL) 173 break; 174 if (now_len >= line_length && *last_blank != -1) 175 break; 176 } 177 178 ps.next_col_1 = false; 179 180 if (now_len <= line_length) 181 return; 182 if (ch_isspace(com.s[com.len - 1])) 183 return; 184 185 if (*last_blank == -1) { 186 /* only a single word in this line */ 187 output_line(); 188 com_add_delim(); 189 return; 190 } 191 192 const char *last_word_s = com.s + *last_blank + 1; 193 size_t last_word_len = com.len - (size_t)(*last_blank + 1); 194 com.len = (size_t)*last_blank; 195 output_line(); 196 com_add_delim(); 197 198 /* Assume that output_line and com_add_delim don't 199 * invalidate the "unused" part of the buffer beyond 200 * com.s + com.len. */ 201 memmove(com.s + com.len, last_word_s, last_word_len); 202 com.len += last_word_len; 203 *last_blank = -1; 204 } 205 206 static bool 207 copy_comment_wrap_newline(ssize_t *last_blank) 208 { 209 *last_blank = -1; 210 if (ps.next_col_1) { 211 if (com.len == 0) 212 com_add_char(' '); /* force empty output line */ 213 if (com.len > 3) { 214 output_line(); 215 com_add_delim(); 216 } 217 output_line(); 218 com_add_delim(); 219 } else { 220 ps.next_col_1 = true; 221 if (!(com.len > 0 && ch_isblank(com.s[com.len - 1]))) 222 com_add_char(' '); 223 *last_blank = (int)com.len - 1; 224 } 225 ++line_no; 226 227 /* flush any blanks and/or tabs at start of next line */ 228 inp_skip(); /* '\n' */ 229 while (ch_isblank(inp_p[0])) 230 inp_p++; 231 if (inp_p[0] == '*' && inp_p[1] == '/') 232 return false; 233 if (inp_p[0] == '*') { 234 inp_p++; 235 while (ch_isblank(inp_p[0])) 236 inp_p++; 237 } 238 239 return true; 240 } 241 242 static void 243 copy_comment_wrap_finish(int line_length, bool delim) 244 { 245 if (delim) { 246 if (com.len > 3) 247 output_line(); 248 else 249 com.len = 0; 250 com_add_char(' '); 251 } else { 252 size_t len = com.len; 253 while (ch_isblank(com.s[len - 1])) 254 len--; 255 int end_ind = ind_add(ps.com_ind, com.s, len); 256 if (end_ind + 3 > line_length) 257 output_line(); 258 } 259 260 inp_p += 2; 261 if (com.len > 0 && ch_isblank(com.s[com.len - 1])) 262 buf_add_chars(&com, "*/", 2); 263 else 264 buf_add_chars(&com, " */", 3); 265 } 266 267 /* 268 * Copy characters from 'inp' to 'com'. Try to keep comments from going over 269 * the maximum line length. To do that, remember where the last blank, tab, or 270 * newline was. When a line is filled, print up to the last blank and continue 271 * copying. 272 */ 273 static void 274 copy_comment_wrap(int line_length, bool delim) 275 { 276 ssize_t last_blank = -1; /* index of the last blank in 'com' */ 277 278 for (;;) { 279 if (inp_p[0] == '\n') { 280 if (had_eof) 281 goto unterminated_comment; 282 if (!copy_comment_wrap_newline(&last_blank)) 283 goto end_of_comment; 284 } else if (inp_p[0] == '*' && inp_p[1] == '/') 285 goto end_of_comment; 286 else 287 copy_comment_wrap_text(line_length, &last_blank); 288 } 289 290 end_of_comment: 291 copy_comment_wrap_finish(line_length, delim); 292 return; 293 294 unterminated_comment: 295 diag(1, "Unterminated comment"); 296 output_line(); 297 } 298 299 static void 300 copy_comment_nowrap(void) 301 { 302 char kind = token.s[token.len - 1]; 303 304 for (;;) { 305 if (inp_p[0] == '\n') { 306 if (kind == '/') 307 return; 308 309 if (had_eof) { 310 diag(1, "Unterminated comment"); 311 output_line(); 312 return; 313 } 314 315 if (com.len == 0) 316 com_add_char(' '); /* force output of an 317 * empty line */ 318 output_line(); 319 ++line_no; 320 inp_skip(); 321 continue; 322 } 323 324 com_add_char(*inp_p++); 325 if (com.len >= 2 326 && com.s[com.len - 2] == '*' 327 && com.s[com.len - 1] == '/' 328 && kind == '*') 329 return; 330 } 331 } 332 333 /* 334 * Scan, reformat and output a single comment, which is either a block comment 335 * starting with '/' '*' or an end-of-line comment starting with '//'. 336 */ 337 void 338 process_comment(void) 339 { 340 bool may_wrap, delim; 341 int ind, line_length; 342 343 analyze_comment(&may_wrap, &delim, &ind, &line_length); 344 copy_comment_start(may_wrap, &delim, ind, line_length); 345 if (may_wrap) 346 copy_comment_wrap(line_length, delim); 347 else 348 copy_comment_nowrap(); 349 } 350
Indexes created Sun Oct 19 16:10:00 GMT 2025