Thu, 31 Aug 2017 15:17:03 +0800
merge
aoqi@0 | 1 | /* |
aoqi@0 | 2 | * Copyright (c) 2012, 2014, Oracle and/or its affiliates. All rights reserved. |
aoqi@0 | 3 | * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. |
aoqi@0 | 4 | * |
aoqi@0 | 5 | * This code is free software; you can redistribute it and/or modify it |
aoqi@0 | 6 | * under the terms of the GNU General Public License version 2 only, as |
aoqi@0 | 7 | * published by the Free Software Foundation. Oracle designates this |
aoqi@0 | 8 | * particular file as subject to the "Classpath" exception as provided |
aoqi@0 | 9 | * by Oracle in the LICENSE file that accompanied this code. |
aoqi@0 | 10 | * |
aoqi@0 | 11 | * This code is distributed in the hope that it will be useful, but WITHOUT |
aoqi@0 | 12 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
aoqi@0 | 13 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
aoqi@0 | 14 | * version 2 for more details (a copy is included in the LICENSE file that |
aoqi@0 | 15 | * accompanied this code). |
aoqi@0 | 16 | * |
aoqi@0 | 17 | * You should have received a copy of the GNU General Public License version |
aoqi@0 | 18 | * 2 along with this work; if not, write to the Free Software Foundation, |
aoqi@0 | 19 | * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. |
aoqi@0 | 20 | * |
aoqi@0 | 21 | * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA |
aoqi@0 | 22 | * or visit www.oracle.com if you need additional information or have any |
aoqi@0 | 23 | * questions. |
aoqi@0 | 24 | */ |
aoqi@0 | 25 | |
aoqi@0 | 26 | package com.sun.tools.javac.parser; |
aoqi@0 | 27 | |
aoqi@0 | 28 | import java.text.BreakIterator; |
aoqi@0 | 29 | import java.util.Arrays; |
aoqi@0 | 30 | import java.util.HashMap; |
aoqi@0 | 31 | import java.util.HashSet; |
aoqi@0 | 32 | import java.util.Locale; |
aoqi@0 | 33 | import java.util.Map; |
aoqi@0 | 34 | import java.util.Set; |
aoqi@0 | 35 | |
aoqi@0 | 36 | import com.sun.source.doctree.AttributeTree.ValueKind; |
aoqi@0 | 37 | import com.sun.tools.javac.parser.DocCommentParser.TagParser.Kind; |
aoqi@0 | 38 | import com.sun.tools.javac.parser.Tokens.Comment; |
aoqi@0 | 39 | import com.sun.tools.javac.parser.Tokens.TokenKind; |
aoqi@0 | 40 | import com.sun.tools.javac.tree.DCTree; |
aoqi@0 | 41 | import com.sun.tools.javac.tree.DCTree.DCAttribute; |
aoqi@0 | 42 | import com.sun.tools.javac.tree.DCTree.DCDocComment; |
aoqi@0 | 43 | import com.sun.tools.javac.tree.DCTree.DCEndElement; |
aoqi@0 | 44 | import com.sun.tools.javac.tree.DCTree.DCEndPosTree; |
aoqi@0 | 45 | import com.sun.tools.javac.tree.DCTree.DCErroneous; |
aoqi@0 | 46 | import com.sun.tools.javac.tree.DCTree.DCIdentifier; |
aoqi@0 | 47 | import com.sun.tools.javac.tree.DCTree.DCReference; |
aoqi@0 | 48 | import com.sun.tools.javac.tree.DCTree.DCStartElement; |
aoqi@0 | 49 | import com.sun.tools.javac.tree.DCTree.DCText; |
aoqi@0 | 50 | import com.sun.tools.javac.tree.DocTreeMaker; |
aoqi@0 | 51 | import com.sun.tools.javac.tree.JCTree; |
aoqi@0 | 52 | import com.sun.tools.javac.util.DiagnosticSource; |
aoqi@0 | 53 | import com.sun.tools.javac.util.List; |
aoqi@0 | 54 | import com.sun.tools.javac.util.ListBuffer; |
aoqi@0 | 55 | import com.sun.tools.javac.util.Log; |
aoqi@0 | 56 | import com.sun.tools.javac.util.Name; |
aoqi@0 | 57 | import com.sun.tools.javac.util.Names; |
aoqi@0 | 58 | import com.sun.tools.javac.util.Options; |
aoqi@0 | 59 | import com.sun.tools.javac.util.Position; |
aoqi@0 | 60 | import com.sun.tools.javac.util.StringUtils; |
aoqi@0 | 61 | import static com.sun.tools.javac.util.LayoutCharacters.*; |
aoqi@0 | 62 | |
aoqi@0 | 63 | /** |
aoqi@0 | 64 | * |
aoqi@0 | 65 | * <p><b>This is NOT part of any supported API. |
aoqi@0 | 66 | * If you write code that depends on this, you do so at your own risk. |
aoqi@0 | 67 | * This code and its internal interfaces are subject to change or |
aoqi@0 | 68 | * deletion without notice.</b> |
aoqi@0 | 69 | */ |
aoqi@0 | 70 | public class DocCommentParser { |
aoqi@0 | 71 | static class ParseException extends Exception { |
aoqi@0 | 72 | private static final long serialVersionUID = 0; |
aoqi@0 | 73 | ParseException(String key) { |
aoqi@0 | 74 | super(key); |
aoqi@0 | 75 | } |
aoqi@0 | 76 | } |
aoqi@0 | 77 | |
aoqi@0 | 78 | final ParserFactory fac; |
aoqi@0 | 79 | final DiagnosticSource diagSource; |
aoqi@0 | 80 | final Comment comment; |
aoqi@0 | 81 | final DocTreeMaker m; |
aoqi@0 | 82 | final Names names; |
aoqi@0 | 83 | |
aoqi@0 | 84 | BreakIterator sentenceBreaker; |
aoqi@0 | 85 | |
aoqi@0 | 86 | /** The input buffer, index of most recent character read, |
aoqi@0 | 87 | * index of one past last character in buffer. |
aoqi@0 | 88 | */ |
aoqi@0 | 89 | protected char[] buf; |
aoqi@0 | 90 | protected int bp; |
aoqi@0 | 91 | protected int buflen; |
aoqi@0 | 92 | |
aoqi@0 | 93 | /** The current character. |
aoqi@0 | 94 | */ |
aoqi@0 | 95 | protected char ch; |
aoqi@0 | 96 | |
aoqi@0 | 97 | int textStart = -1; |
aoqi@0 | 98 | int lastNonWhite = -1; |
aoqi@0 | 99 | boolean newline = true; |
aoqi@0 | 100 | |
aoqi@0 | 101 | Map<Name, TagParser> tagParsers; |
aoqi@0 | 102 | |
aoqi@0 | 103 | DocCommentParser(ParserFactory fac, DiagnosticSource diagSource, Comment comment) { |
aoqi@0 | 104 | this.fac = fac; |
aoqi@0 | 105 | this.diagSource = diagSource; |
aoqi@0 | 106 | this.comment = comment; |
aoqi@0 | 107 | names = fac.names; |
aoqi@0 | 108 | m = fac.docTreeMaker; |
aoqi@0 | 109 | |
aoqi@0 | 110 | Locale locale = (fac.locale == null) ? Locale.getDefault() : fac.locale; |
aoqi@0 | 111 | |
aoqi@0 | 112 | Options options = fac.options; |
aoqi@0 | 113 | boolean useBreakIterator = options.isSet("breakIterator"); |
aoqi@0 | 114 | if (useBreakIterator || !locale.getLanguage().equals(Locale.ENGLISH.getLanguage())) |
aoqi@0 | 115 | sentenceBreaker = BreakIterator.getSentenceInstance(locale); |
aoqi@0 | 116 | |
aoqi@0 | 117 | initTagParsers(); |
aoqi@0 | 118 | } |
aoqi@0 | 119 | |
aoqi@0 | 120 | DCDocComment parse() { |
aoqi@0 | 121 | String c = comment.getText(); |
aoqi@0 | 122 | buf = new char[c.length() + 1]; |
aoqi@0 | 123 | c.getChars(0, c.length(), buf, 0); |
aoqi@0 | 124 | buf[buf.length - 1] = EOI; |
aoqi@0 | 125 | buflen = buf.length - 1; |
aoqi@0 | 126 | bp = -1; |
aoqi@0 | 127 | nextChar(); |
aoqi@0 | 128 | |
aoqi@0 | 129 | List<DCTree> body = blockContent(); |
aoqi@0 | 130 | List<DCTree> tags = blockTags(); |
aoqi@0 | 131 | |
aoqi@0 | 132 | // split body into first sentence and body |
aoqi@0 | 133 | ListBuffer<DCTree> fs = new ListBuffer<DCTree>(); |
aoqi@0 | 134 | loop: |
aoqi@0 | 135 | for (; body.nonEmpty(); body = body.tail) { |
aoqi@0 | 136 | DCTree t = body.head; |
aoqi@0 | 137 | switch (t.getKind()) { |
aoqi@0 | 138 | case TEXT: |
aoqi@0 | 139 | String s = ((DCText) t).getBody(); |
aoqi@0 | 140 | int i = getSentenceBreak(s); |
aoqi@0 | 141 | if (i > 0) { |
aoqi@0 | 142 | int i0 = i; |
aoqi@0 | 143 | while (i0 > 0 && isWhitespace(s.charAt(i0 - 1))) |
aoqi@0 | 144 | i0--; |
aoqi@0 | 145 | fs.add(m.at(t.pos).Text(s.substring(0, i0))); |
aoqi@0 | 146 | int i1 = i; |
aoqi@0 | 147 | while (i1 < s.length() && isWhitespace(s.charAt(i1))) |
aoqi@0 | 148 | i1++; |
aoqi@0 | 149 | body = body.tail; |
aoqi@0 | 150 | if (i1 < s.length()) |
aoqi@0 | 151 | body = body.prepend(m.at(t.pos + i1).Text(s.substring(i1))); |
aoqi@0 | 152 | break loop; |
aoqi@0 | 153 | } else if (body.tail.nonEmpty()) { |
aoqi@0 | 154 | if (isSentenceBreak(body.tail.head)) { |
aoqi@0 | 155 | int i0 = s.length() - 1; |
aoqi@0 | 156 | while (i0 > 0 && isWhitespace(s.charAt(i0))) |
aoqi@0 | 157 | i0--; |
aoqi@0 | 158 | fs.add(m.at(t.pos).Text(s.substring(0, i0 + 1))); |
aoqi@0 | 159 | body = body.tail; |
aoqi@0 | 160 | break loop; |
aoqi@0 | 161 | } |
aoqi@0 | 162 | } |
aoqi@0 | 163 | break; |
aoqi@0 | 164 | |
aoqi@0 | 165 | case START_ELEMENT: |
aoqi@0 | 166 | case END_ELEMENT: |
aoqi@0 | 167 | if (isSentenceBreak(t)) |
aoqi@0 | 168 | break loop; |
aoqi@0 | 169 | break; |
aoqi@0 | 170 | } |
aoqi@0 | 171 | fs.add(t); |
aoqi@0 | 172 | } |
aoqi@0 | 173 | |
aoqi@0 | 174 | @SuppressWarnings("unchecked") |
aoqi@0 | 175 | DCTree first = getFirst(fs.toList(), body, tags); |
aoqi@0 | 176 | int pos = (first == null) ? Position.NOPOS : first.pos; |
aoqi@0 | 177 | |
aoqi@0 | 178 | DCDocComment dc = m.at(pos).DocComment(comment, fs.toList(), body, tags); |
aoqi@0 | 179 | return dc; |
aoqi@0 | 180 | } |
aoqi@0 | 181 | |
aoqi@0 | 182 | void nextChar() { |
aoqi@0 | 183 | ch = buf[bp < buflen ? ++bp : buflen]; |
aoqi@0 | 184 | switch (ch) { |
aoqi@0 | 185 | case '\f': case '\n': case '\r': |
aoqi@0 | 186 | newline = true; |
aoqi@0 | 187 | } |
aoqi@0 | 188 | } |
aoqi@0 | 189 | |
aoqi@0 | 190 | /** |
aoqi@0 | 191 | * Read block content, consisting of text, html and inline tags. |
aoqi@0 | 192 | * Terminated by the end of input, or the beginning of the next block tag: |
aoqi@0 | 193 | * i.e. @ as the first non-whitespace character on a line. |
aoqi@0 | 194 | */ |
aoqi@0 | 195 | @SuppressWarnings("fallthrough") |
aoqi@0 | 196 | protected List<DCTree> blockContent() { |
aoqi@0 | 197 | ListBuffer<DCTree> trees = new ListBuffer<DCTree>(); |
aoqi@0 | 198 | textStart = -1; |
aoqi@0 | 199 | |
aoqi@0 | 200 | loop: |
aoqi@0 | 201 | while (bp < buflen) { |
aoqi@0 | 202 | switch (ch) { |
aoqi@0 | 203 | case '\n': case '\r': case '\f': |
aoqi@0 | 204 | newline = true; |
aoqi@0 | 205 | // fallthrough |
aoqi@0 | 206 | |
aoqi@0 | 207 | case ' ': case '\t': |
aoqi@0 | 208 | nextChar(); |
aoqi@0 | 209 | break; |
aoqi@0 | 210 | |
aoqi@0 | 211 | case '&': |
aoqi@0 | 212 | entity(trees); |
aoqi@0 | 213 | break; |
aoqi@0 | 214 | |
aoqi@0 | 215 | case '<': |
aoqi@0 | 216 | newline = false; |
aoqi@0 | 217 | addPendingText(trees, bp - 1); |
aoqi@0 | 218 | trees.add(html()); |
aoqi@0 | 219 | if (textStart == -1) { |
aoqi@0 | 220 | textStart = bp; |
aoqi@0 | 221 | lastNonWhite = -1; |
aoqi@0 | 222 | } |
aoqi@0 | 223 | break; |
aoqi@0 | 224 | |
aoqi@0 | 225 | case '>': |
aoqi@0 | 226 | newline = false; |
aoqi@0 | 227 | addPendingText(trees, bp - 1); |
aoqi@0 | 228 | trees.add(m.at(bp).Erroneous(newString(bp, bp+1), diagSource, "dc.bad.gt")); |
aoqi@0 | 229 | nextChar(); |
aoqi@0 | 230 | if (textStart == -1) { |
aoqi@0 | 231 | textStart = bp; |
aoqi@0 | 232 | lastNonWhite = -1; |
aoqi@0 | 233 | } |
aoqi@0 | 234 | break; |
aoqi@0 | 235 | |
aoqi@0 | 236 | case '{': |
aoqi@0 | 237 | inlineTag(trees); |
aoqi@0 | 238 | break; |
aoqi@0 | 239 | |
aoqi@0 | 240 | case '@': |
aoqi@0 | 241 | if (newline) { |
aoqi@0 | 242 | addPendingText(trees, lastNonWhite); |
aoqi@0 | 243 | break loop; |
aoqi@0 | 244 | } |
aoqi@0 | 245 | // fallthrough |
aoqi@0 | 246 | |
aoqi@0 | 247 | default: |
aoqi@0 | 248 | newline = false; |
aoqi@0 | 249 | if (textStart == -1) |
aoqi@0 | 250 | textStart = bp; |
aoqi@0 | 251 | lastNonWhite = bp; |
aoqi@0 | 252 | nextChar(); |
aoqi@0 | 253 | } |
aoqi@0 | 254 | } |
aoqi@0 | 255 | |
aoqi@0 | 256 | if (lastNonWhite != -1) |
aoqi@0 | 257 | addPendingText(trees, lastNonWhite); |
aoqi@0 | 258 | |
aoqi@0 | 259 | return trees.toList(); |
aoqi@0 | 260 | } |
aoqi@0 | 261 | |
aoqi@0 | 262 | /** |
aoqi@0 | 263 | * Read a series of block tags, including their content. |
aoqi@0 | 264 | * Standard tags parse their content appropriately. |
aoqi@0 | 265 | * Non-standard tags are represented by {@link UnknownBlockTag}. |
aoqi@0 | 266 | */ |
aoqi@0 | 267 | protected List<DCTree> blockTags() { |
aoqi@0 | 268 | ListBuffer<DCTree> tags = new ListBuffer<DCTree>(); |
aoqi@0 | 269 | while (ch == '@') |
aoqi@0 | 270 | tags.add(blockTag()); |
aoqi@0 | 271 | return tags.toList(); |
aoqi@0 | 272 | } |
aoqi@0 | 273 | |
aoqi@0 | 274 | /** |
aoqi@0 | 275 | * Read a single block tag, including its content. |
aoqi@0 | 276 | * Standard tags parse their content appropriately. |
aoqi@0 | 277 | * Non-standard tags are represented by {@link UnknownBlockTag}. |
aoqi@0 | 278 | */ |
aoqi@0 | 279 | protected DCTree blockTag() { |
aoqi@0 | 280 | int p = bp; |
aoqi@0 | 281 | try { |
aoqi@0 | 282 | nextChar(); |
aoqi@0 | 283 | if (isIdentifierStart(ch)) { |
aoqi@0 | 284 | Name name = readTagName(); |
aoqi@0 | 285 | TagParser tp = tagParsers.get(name); |
aoqi@0 | 286 | if (tp == null) { |
aoqi@0 | 287 | List<DCTree> content = blockContent(); |
aoqi@0 | 288 | return m.at(p).UnknownBlockTag(name, content); |
aoqi@0 | 289 | } else { |
aoqi@0 | 290 | switch (tp.getKind()) { |
aoqi@0 | 291 | case BLOCK: |
aoqi@0 | 292 | return tp.parse(p); |
aoqi@0 | 293 | case INLINE: |
aoqi@0 | 294 | return erroneous("dc.bad.inline.tag", p); |
aoqi@0 | 295 | } |
aoqi@0 | 296 | } |
aoqi@0 | 297 | } |
aoqi@0 | 298 | blockContent(); |
aoqi@0 | 299 | |
aoqi@0 | 300 | return erroneous("dc.no.tag.name", p); |
aoqi@0 | 301 | } catch (ParseException e) { |
aoqi@0 | 302 | blockContent(); |
aoqi@0 | 303 | return erroneous(e.getMessage(), p); |
aoqi@0 | 304 | } |
aoqi@0 | 305 | } |
aoqi@0 | 306 | |
aoqi@0 | 307 | protected void inlineTag(ListBuffer<DCTree> list) { |
aoqi@0 | 308 | newline = false; |
aoqi@0 | 309 | nextChar(); |
aoqi@0 | 310 | if (ch == '@') { |
aoqi@0 | 311 | addPendingText(list, bp - 2); |
aoqi@0 | 312 | list.add(inlineTag()); |
aoqi@0 | 313 | textStart = bp; |
aoqi@0 | 314 | lastNonWhite = -1; |
aoqi@0 | 315 | } else { |
aoqi@0 | 316 | if (textStart == -1) |
aoqi@0 | 317 | textStart = bp - 1; |
aoqi@0 | 318 | lastNonWhite = bp; |
aoqi@0 | 319 | } |
aoqi@0 | 320 | } |
aoqi@0 | 321 | |
aoqi@0 | 322 | /** |
aoqi@0 | 323 | * Read a single inline tag, including its content. |
aoqi@0 | 324 | * Standard tags parse their content appropriately. |
aoqi@0 | 325 | * Non-standard tags are represented by {@link UnknownBlockTag}. |
aoqi@0 | 326 | * Malformed tags may be returned as {@link Erroneous}. |
aoqi@0 | 327 | */ |
aoqi@0 | 328 | protected DCTree inlineTag() { |
aoqi@0 | 329 | int p = bp - 1; |
aoqi@0 | 330 | try { |
aoqi@0 | 331 | nextChar(); |
aoqi@0 | 332 | if (isIdentifierStart(ch)) { |
aoqi@0 | 333 | Name name = readTagName(); |
aoqi@0 | 334 | skipWhitespace(); |
aoqi@0 | 335 | |
aoqi@0 | 336 | TagParser tp = tagParsers.get(name); |
aoqi@0 | 337 | if (tp == null) { |
aoqi@0 | 338 | DCTree text = inlineText(); |
aoqi@0 | 339 | if (text != null) { |
aoqi@0 | 340 | nextChar(); |
aoqi@0 | 341 | return m.at(p).UnknownInlineTag(name, List.of(text)).setEndPos(bp); |
aoqi@0 | 342 | } |
aoqi@0 | 343 | } else if (tp.getKind() == TagParser.Kind.INLINE) { |
aoqi@0 | 344 | DCEndPosTree<?> tree = (DCEndPosTree<?>) tp.parse(p); |
aoqi@0 | 345 | if (tree != null) { |
aoqi@0 | 346 | return tree.setEndPos(bp); |
aoqi@0 | 347 | } |
aoqi@0 | 348 | } else { |
aoqi@0 | 349 | inlineText(); // skip content |
aoqi@0 | 350 | nextChar(); |
aoqi@0 | 351 | } |
aoqi@0 | 352 | } |
aoqi@0 | 353 | return erroneous("dc.no.tag.name", p); |
aoqi@0 | 354 | } catch (ParseException e) { |
aoqi@0 | 355 | return erroneous(e.getMessage(), p); |
aoqi@0 | 356 | } |
aoqi@0 | 357 | } |
aoqi@0 | 358 | |
aoqi@0 | 359 | /** |
aoqi@0 | 360 | * Read plain text content of an inline tag. |
aoqi@0 | 361 | * Matching pairs of { } are skipped; the text is terminated by the first |
aoqi@0 | 362 | * unmatched }. It is an error if the beginning of the next tag is detected. |
aoqi@0 | 363 | */ |
aoqi@0 | 364 | protected DCTree inlineText() throws ParseException { |
aoqi@0 | 365 | skipWhitespace(); |
aoqi@0 | 366 | int pos = bp; |
aoqi@0 | 367 | int depth = 1; |
aoqi@0 | 368 | |
aoqi@0 | 369 | loop: |
aoqi@0 | 370 | while (bp < buflen) { |
aoqi@0 | 371 | switch (ch) { |
aoqi@0 | 372 | case '\n': case '\r': case '\f': |
aoqi@0 | 373 | newline = true; |
aoqi@0 | 374 | break; |
aoqi@0 | 375 | |
aoqi@0 | 376 | case ' ': case '\t': |
aoqi@0 | 377 | break; |
aoqi@0 | 378 | |
aoqi@0 | 379 | case '{': |
aoqi@0 | 380 | newline = false; |
aoqi@0 | 381 | lastNonWhite = bp; |
aoqi@0 | 382 | depth++; |
aoqi@0 | 383 | break; |
aoqi@0 | 384 | |
aoqi@0 | 385 | case '}': |
aoqi@0 | 386 | if (--depth == 0) { |
aoqi@0 | 387 | return m.at(pos).Text(newString(pos, bp)); |
aoqi@0 | 388 | } |
aoqi@0 | 389 | newline = false; |
aoqi@0 | 390 | lastNonWhite = bp; |
aoqi@0 | 391 | break; |
aoqi@0 | 392 | |
aoqi@0 | 393 | case '@': |
aoqi@0 | 394 | if (newline) |
aoqi@0 | 395 | break loop; |
aoqi@0 | 396 | newline = false; |
aoqi@0 | 397 | lastNonWhite = bp; |
aoqi@0 | 398 | break; |
aoqi@0 | 399 | |
aoqi@0 | 400 | default: |
aoqi@0 | 401 | newline = false; |
aoqi@0 | 402 | lastNonWhite = bp; |
aoqi@0 | 403 | break; |
aoqi@0 | 404 | } |
aoqi@0 | 405 | nextChar(); |
aoqi@0 | 406 | } |
aoqi@0 | 407 | throw new ParseException("dc.unterminated.inline.tag"); |
aoqi@0 | 408 | } |
aoqi@0 | 409 | |
aoqi@0 | 410 | /** |
aoqi@0 | 411 | * Read Java class name, possibly followed by member |
aoqi@0 | 412 | * Matching pairs of < > are skipped. The text is terminated by the first |
aoqi@0 | 413 | * unmatched }. It is an error if the beginning of the next tag is detected. |
aoqi@0 | 414 | */ |
aoqi@0 | 415 | // TODO: boolean allowMember should be enum FORBID, ALLOW, REQUIRE |
aoqi@0 | 416 | // TODO: improve quality of parse to forbid bad constructions. |
aoqi@0 | 417 | @SuppressWarnings("fallthrough") |
aoqi@0 | 418 | protected DCReference reference(boolean allowMember) throws ParseException { |
aoqi@0 | 419 | int pos = bp; |
aoqi@0 | 420 | int depth = 0; |
aoqi@0 | 421 | |
aoqi@0 | 422 | // scan to find the end of the signature, by looking for the first |
aoqi@0 | 423 | // whitespace not enclosed in () or <>, or the end of the tag |
aoqi@0 | 424 | loop: |
aoqi@0 | 425 | while (bp < buflen) { |
aoqi@0 | 426 | switch (ch) { |
aoqi@0 | 427 | case '\n': case '\r': case '\f': |
aoqi@0 | 428 | newline = true; |
aoqi@0 | 429 | // fallthrough |
aoqi@0 | 430 | |
aoqi@0 | 431 | case ' ': case '\t': |
aoqi@0 | 432 | if (depth == 0) |
aoqi@0 | 433 | break loop; |
aoqi@0 | 434 | break; |
aoqi@0 | 435 | |
aoqi@0 | 436 | case '(': |
aoqi@0 | 437 | case '<': |
aoqi@0 | 438 | newline = false; |
aoqi@0 | 439 | depth++; |
aoqi@0 | 440 | break; |
aoqi@0 | 441 | |
aoqi@0 | 442 | case ')': |
aoqi@0 | 443 | case '>': |
aoqi@0 | 444 | newline = false; |
aoqi@0 | 445 | --depth; |
aoqi@0 | 446 | break; |
aoqi@0 | 447 | |
aoqi@0 | 448 | case '}': |
aoqi@0 | 449 | if (bp == pos) |
aoqi@0 | 450 | return null; |
aoqi@0 | 451 | newline = false; |
aoqi@0 | 452 | break loop; |
aoqi@0 | 453 | |
aoqi@0 | 454 | case '@': |
aoqi@0 | 455 | if (newline) |
aoqi@0 | 456 | break loop; |
aoqi@0 | 457 | // fallthrough |
aoqi@0 | 458 | |
aoqi@0 | 459 | default: |
aoqi@0 | 460 | newline = false; |
aoqi@0 | 461 | |
aoqi@0 | 462 | } |
aoqi@0 | 463 | nextChar(); |
aoqi@0 | 464 | } |
aoqi@0 | 465 | |
aoqi@0 | 466 | if (depth != 0) |
aoqi@0 | 467 | throw new ParseException("dc.unterminated.signature"); |
aoqi@0 | 468 | |
aoqi@0 | 469 | String sig = newString(pos, bp); |
aoqi@0 | 470 | |
aoqi@0 | 471 | // Break sig apart into qualifiedExpr member paramTypes. |
aoqi@0 | 472 | JCTree qualExpr; |
aoqi@0 | 473 | Name member; |
aoqi@0 | 474 | List<JCTree> paramTypes; |
aoqi@0 | 475 | |
aoqi@0 | 476 | Log.DeferredDiagnosticHandler deferredDiagnosticHandler |
aoqi@0 | 477 | = new Log.DeferredDiagnosticHandler(fac.log); |
aoqi@0 | 478 | |
aoqi@0 | 479 | try { |
aoqi@0 | 480 | int hash = sig.indexOf("#"); |
aoqi@0 | 481 | int lparen = sig.indexOf("(", hash + 1); |
aoqi@0 | 482 | if (hash == -1) { |
aoqi@0 | 483 | if (lparen == -1) { |
aoqi@0 | 484 | qualExpr = parseType(sig); |
aoqi@0 | 485 | member = null; |
aoqi@0 | 486 | } else { |
aoqi@0 | 487 | qualExpr = null; |
aoqi@0 | 488 | member = parseMember(sig.substring(0, lparen)); |
aoqi@0 | 489 | } |
aoqi@0 | 490 | } else { |
aoqi@0 | 491 | qualExpr = (hash == 0) ? null : parseType(sig.substring(0, hash)); |
aoqi@0 | 492 | if (lparen == -1) |
aoqi@0 | 493 | member = parseMember(sig.substring(hash + 1)); |
aoqi@0 | 494 | else |
aoqi@0 | 495 | member = parseMember(sig.substring(hash + 1, lparen)); |
aoqi@0 | 496 | } |
aoqi@0 | 497 | |
aoqi@0 | 498 | if (lparen < 0) { |
aoqi@0 | 499 | paramTypes = null; |
aoqi@0 | 500 | } else { |
aoqi@0 | 501 | int rparen = sig.indexOf(")", lparen); |
aoqi@0 | 502 | if (rparen != sig.length() - 1) |
aoqi@0 | 503 | throw new ParseException("dc.ref.bad.parens"); |
aoqi@0 | 504 | paramTypes = parseParams(sig.substring(lparen + 1, rparen)); |
aoqi@0 | 505 | } |
aoqi@0 | 506 | |
aoqi@0 | 507 | if (!deferredDiagnosticHandler.getDiagnostics().isEmpty()) |
aoqi@0 | 508 | throw new ParseException("dc.ref.syntax.error"); |
aoqi@0 | 509 | |
aoqi@0 | 510 | } finally { |
aoqi@0 | 511 | fac.log.popDiagnosticHandler(deferredDiagnosticHandler); |
aoqi@0 | 512 | } |
aoqi@0 | 513 | |
aoqi@0 | 514 | return m.at(pos).Reference(sig, qualExpr, member, paramTypes).setEndPos(bp); |
aoqi@0 | 515 | } |
aoqi@0 | 516 | |
aoqi@0 | 517 | JCTree parseType(String s) throws ParseException { |
aoqi@0 | 518 | JavacParser p = fac.newParser(s, false, false, false); |
aoqi@0 | 519 | JCTree tree = p.parseType(); |
aoqi@0 | 520 | if (p.token().kind != TokenKind.EOF) |
aoqi@0 | 521 | throw new ParseException("dc.ref.unexpected.input"); |
aoqi@0 | 522 | return tree; |
aoqi@0 | 523 | } |
aoqi@0 | 524 | |
aoqi@0 | 525 | Name parseMember(String s) throws ParseException { |
aoqi@0 | 526 | JavacParser p = fac.newParser(s, false, false, false); |
aoqi@0 | 527 | Name name = p.ident(); |
aoqi@0 | 528 | if (p.token().kind != TokenKind.EOF) |
aoqi@0 | 529 | throw new ParseException("dc.ref.unexpected.input"); |
aoqi@0 | 530 | return name; |
aoqi@0 | 531 | } |
aoqi@0 | 532 | |
aoqi@0 | 533 | List<JCTree> parseParams(String s) throws ParseException { |
aoqi@0 | 534 | if (s.trim().isEmpty()) |
aoqi@0 | 535 | return List.nil(); |
aoqi@0 | 536 | |
aoqi@0 | 537 | JavacParser p = fac.newParser(s.replace("...", "[]"), false, false, false); |
aoqi@0 | 538 | ListBuffer<JCTree> paramTypes = new ListBuffer<JCTree>(); |
aoqi@0 | 539 | paramTypes.add(p.parseType()); |
aoqi@0 | 540 | |
aoqi@0 | 541 | if (p.token().kind == TokenKind.IDENTIFIER) |
aoqi@0 | 542 | p.nextToken(); |
aoqi@0 | 543 | |
aoqi@0 | 544 | while (p.token().kind == TokenKind.COMMA) { |
aoqi@0 | 545 | p.nextToken(); |
aoqi@0 | 546 | paramTypes.add(p.parseType()); |
aoqi@0 | 547 | |
aoqi@0 | 548 | if (p.token().kind == TokenKind.IDENTIFIER) |
aoqi@0 | 549 | p.nextToken(); |
aoqi@0 | 550 | } |
aoqi@0 | 551 | |
aoqi@0 | 552 | if (p.token().kind != TokenKind.EOF) |
aoqi@0 | 553 | throw new ParseException("dc.ref.unexpected.input"); |
aoqi@0 | 554 | |
aoqi@0 | 555 | return paramTypes.toList(); |
aoqi@0 | 556 | } |
aoqi@0 | 557 | |
aoqi@0 | 558 | /** |
aoqi@0 | 559 | * Read Java identifier |
aoqi@0 | 560 | * Matching pairs of { } are skipped; the text is terminated by the first |
aoqi@0 | 561 | * unmatched }. It is an error if the beginning of the next tag is detected. |
aoqi@0 | 562 | */ |
aoqi@0 | 563 | @SuppressWarnings("fallthrough") |
aoqi@0 | 564 | protected DCIdentifier identifier() throws ParseException { |
aoqi@0 | 565 | skipWhitespace(); |
aoqi@0 | 566 | int pos = bp; |
aoqi@0 | 567 | |
aoqi@0 | 568 | if (isJavaIdentifierStart(ch)) { |
aoqi@0 | 569 | Name name = readJavaIdentifier(); |
aoqi@0 | 570 | return m.at(pos).Identifier(name); |
aoqi@0 | 571 | } |
aoqi@0 | 572 | |
aoqi@0 | 573 | throw new ParseException("dc.identifier.expected"); |
aoqi@0 | 574 | } |
aoqi@0 | 575 | |
aoqi@0 | 576 | /** |
aoqi@0 | 577 | * Read a quoted string. |
aoqi@0 | 578 | * It is an error if the beginning of the next tag is detected. |
aoqi@0 | 579 | */ |
aoqi@0 | 580 | @SuppressWarnings("fallthrough") |
aoqi@0 | 581 | protected DCText quotedString() { |
aoqi@0 | 582 | int pos = bp; |
aoqi@0 | 583 | nextChar(); |
aoqi@0 | 584 | |
aoqi@0 | 585 | loop: |
aoqi@0 | 586 | while (bp < buflen) { |
aoqi@0 | 587 | switch (ch) { |
aoqi@0 | 588 | case '\n': case '\r': case '\f': |
aoqi@0 | 589 | newline = true; |
aoqi@0 | 590 | break; |
aoqi@0 | 591 | |
aoqi@0 | 592 | case ' ': case '\t': |
aoqi@0 | 593 | break; |
aoqi@0 | 594 | |
aoqi@0 | 595 | case '"': |
aoqi@0 | 596 | nextChar(); |
aoqi@0 | 597 | // trim trailing white-space? |
aoqi@0 | 598 | return m.at(pos).Text(newString(pos, bp)); |
aoqi@0 | 599 | |
aoqi@0 | 600 | case '@': |
aoqi@0 | 601 | if (newline) |
aoqi@0 | 602 | break loop; |
aoqi@0 | 603 | |
aoqi@0 | 604 | } |
aoqi@0 | 605 | nextChar(); |
aoqi@0 | 606 | } |
aoqi@0 | 607 | return null; |
aoqi@0 | 608 | } |
aoqi@0 | 609 | |
aoqi@0 | 610 | /** |
aoqi@0 | 611 | * Read general text content of an inline tag, including HTML entities and elements. |
aoqi@0 | 612 | * Matching pairs of { } are skipped; the text is terminated by the first |
aoqi@0 | 613 | * unmatched }. It is an error if the beginning of the next tag is detected. |
aoqi@0 | 614 | */ |
aoqi@0 | 615 | @SuppressWarnings("fallthrough") |
aoqi@0 | 616 | protected List<DCTree> inlineContent() { |
aoqi@0 | 617 | ListBuffer<DCTree> trees = new ListBuffer<DCTree>(); |
aoqi@0 | 618 | |
aoqi@0 | 619 | skipWhitespace(); |
aoqi@0 | 620 | int pos = bp; |
aoqi@0 | 621 | int depth = 1; |
aoqi@0 | 622 | textStart = -1; |
aoqi@0 | 623 | |
aoqi@0 | 624 | loop: |
aoqi@0 | 625 | while (bp < buflen) { |
aoqi@0 | 626 | |
aoqi@0 | 627 | switch (ch) { |
aoqi@0 | 628 | case '\n': case '\r': case '\f': |
aoqi@0 | 629 | newline = true; |
aoqi@0 | 630 | // fall through |
aoqi@0 | 631 | |
aoqi@0 | 632 | case ' ': case '\t': |
aoqi@0 | 633 | nextChar(); |
aoqi@0 | 634 | break; |
aoqi@0 | 635 | |
aoqi@0 | 636 | case '&': |
aoqi@0 | 637 | entity(trees); |
aoqi@0 | 638 | break; |
aoqi@0 | 639 | |
aoqi@0 | 640 | case '<': |
aoqi@0 | 641 | newline = false; |
aoqi@0 | 642 | addPendingText(trees, bp - 1); |
aoqi@0 | 643 | trees.add(html()); |
aoqi@0 | 644 | break; |
aoqi@0 | 645 | |
aoqi@0 | 646 | case '{': |
aoqi@0 | 647 | newline = false; |
aoqi@0 | 648 | depth++; |
aoqi@0 | 649 | nextChar(); |
aoqi@0 | 650 | break; |
aoqi@0 | 651 | |
aoqi@0 | 652 | case '}': |
aoqi@0 | 653 | newline = false; |
aoqi@0 | 654 | if (--depth == 0) { |
aoqi@0 | 655 | addPendingText(trees, bp - 1); |
aoqi@0 | 656 | nextChar(); |
aoqi@0 | 657 | return trees.toList(); |
aoqi@0 | 658 | } |
aoqi@0 | 659 | nextChar(); |
aoqi@0 | 660 | break; |
aoqi@0 | 661 | |
aoqi@0 | 662 | case '@': |
aoqi@0 | 663 | if (newline) |
aoqi@0 | 664 | break loop; |
aoqi@0 | 665 | // fallthrough |
aoqi@0 | 666 | |
aoqi@0 | 667 | default: |
aoqi@0 | 668 | if (textStart == -1) |
aoqi@0 | 669 | textStart = bp; |
aoqi@0 | 670 | nextChar(); |
aoqi@0 | 671 | break; |
aoqi@0 | 672 | } |
aoqi@0 | 673 | } |
aoqi@0 | 674 | |
aoqi@0 | 675 | return List.<DCTree>of(erroneous("dc.unterminated.inline.tag", pos)); |
aoqi@0 | 676 | } |
aoqi@0 | 677 | |
aoqi@0 | 678 | protected void entity(ListBuffer<DCTree> list) { |
aoqi@0 | 679 | newline = false; |
aoqi@0 | 680 | addPendingText(list, bp - 1); |
aoqi@0 | 681 | list.add(entity()); |
aoqi@0 | 682 | if (textStart == -1) { |
aoqi@0 | 683 | textStart = bp; |
aoqi@0 | 684 | lastNonWhite = -1; |
aoqi@0 | 685 | } |
aoqi@0 | 686 | } |
aoqi@0 | 687 | |
aoqi@0 | 688 | /** |
aoqi@0 | 689 | * Read an HTML entity. |
aoqi@0 | 690 | * {@literal &identifier; } or {@literal &#digits; } or {@literal &#xhex-digits; } |
aoqi@0 | 691 | */ |
aoqi@0 | 692 | protected DCTree entity() { |
aoqi@0 | 693 | int p = bp; |
aoqi@0 | 694 | nextChar(); |
aoqi@0 | 695 | Name name = null; |
aoqi@0 | 696 | boolean checkSemi = false; |
aoqi@0 | 697 | if (ch == '#') { |
aoqi@0 | 698 | int namep = bp; |
aoqi@0 | 699 | nextChar(); |
aoqi@0 | 700 | if (isDecimalDigit(ch)) { |
aoqi@0 | 701 | nextChar(); |
aoqi@0 | 702 | while (isDecimalDigit(ch)) |
aoqi@0 | 703 | nextChar(); |
aoqi@0 | 704 | name = names.fromChars(buf, namep, bp - namep); |
aoqi@0 | 705 | } else if (ch == 'x' || ch == 'X') { |
aoqi@0 | 706 | nextChar(); |
aoqi@0 | 707 | if (isHexDigit(ch)) { |
aoqi@0 | 708 | nextChar(); |
aoqi@0 | 709 | while (isHexDigit(ch)) |
aoqi@0 | 710 | nextChar(); |
aoqi@0 | 711 | name = names.fromChars(buf, namep, bp - namep); |
aoqi@0 | 712 | } |
aoqi@0 | 713 | } |
aoqi@0 | 714 | } else if (isIdentifierStart(ch)) { |
aoqi@0 | 715 | name = readIdentifier(); |
aoqi@0 | 716 | } |
aoqi@0 | 717 | |
aoqi@0 | 718 | if (name == null) |
aoqi@0 | 719 | return erroneous("dc.bad.entity", p); |
aoqi@0 | 720 | else { |
aoqi@0 | 721 | if (ch != ';') |
aoqi@0 | 722 | return erroneous("dc.missing.semicolon", p); |
aoqi@0 | 723 | nextChar(); |
aoqi@0 | 724 | return m.at(p).Entity(name); |
aoqi@0 | 725 | } |
aoqi@0 | 726 | } |
aoqi@0 | 727 | |
aoqi@0 | 728 | /** |
aoqi@0 | 729 | * Read the start or end of an HTML tag, or an HTML comment |
aoqi@0 | 730 | * {@literal <identifier attrs> } or {@literal </identifier> } |
aoqi@0 | 731 | */ |
aoqi@0 | 732 | protected DCTree html() { |
aoqi@0 | 733 | int p = bp; |
aoqi@0 | 734 | nextChar(); |
aoqi@0 | 735 | if (isIdentifierStart(ch)) { |
aoqi@0 | 736 | Name name = readIdentifier(); |
aoqi@0 | 737 | List<DCTree> attrs = htmlAttrs(); |
aoqi@0 | 738 | if (attrs != null) { |
aoqi@0 | 739 | boolean selfClosing = false; |
aoqi@0 | 740 | if (ch == '/') { |
aoqi@0 | 741 | nextChar(); |
aoqi@0 | 742 | selfClosing = true; |
aoqi@0 | 743 | } |
aoqi@0 | 744 | if (ch == '>') { |
aoqi@0 | 745 | nextChar(); |
aoqi@0 | 746 | return m.at(p).StartElement(name, attrs, selfClosing).setEndPos(bp); |
aoqi@0 | 747 | } |
aoqi@0 | 748 | } |
aoqi@0 | 749 | } else if (ch == '/') { |
aoqi@0 | 750 | nextChar(); |
aoqi@0 | 751 | if (isIdentifierStart(ch)) { |
aoqi@0 | 752 | Name name = readIdentifier(); |
aoqi@0 | 753 | skipWhitespace(); |
aoqi@0 | 754 | if (ch == '>') { |
aoqi@0 | 755 | nextChar(); |
aoqi@0 | 756 | return m.at(p).EndElement(name); |
aoqi@0 | 757 | } |
aoqi@0 | 758 | } |
aoqi@0 | 759 | } else if (ch == '!') { |
aoqi@0 | 760 | nextChar(); |
aoqi@0 | 761 | if (ch == '-') { |
aoqi@0 | 762 | nextChar(); |
aoqi@0 | 763 | if (ch == '-') { |
aoqi@0 | 764 | nextChar(); |
aoqi@0 | 765 | while (bp < buflen) { |
aoqi@0 | 766 | int dash = 0; |
aoqi@0 | 767 | while (ch == '-') { |
aoqi@0 | 768 | dash++; |
aoqi@0 | 769 | nextChar(); |
aoqi@0 | 770 | } |
aoqi@0 | 771 | // strictly speaking, a comment should not contain "--" |
aoqi@0 | 772 | // so dash > 2 is an error, dash == 2 implies ch == '>' |
aoqi@0 | 773 | if (dash >= 2 && ch == '>') { |
aoqi@0 | 774 | nextChar(); |
aoqi@0 | 775 | return m.at(p).Comment(newString(p, bp)); |
aoqi@0 | 776 | } |
aoqi@0 | 777 | |
aoqi@0 | 778 | nextChar(); |
aoqi@0 | 779 | } |
aoqi@0 | 780 | } |
aoqi@0 | 781 | } |
aoqi@0 | 782 | } |
aoqi@0 | 783 | |
aoqi@0 | 784 | bp = p + 1; |
aoqi@0 | 785 | ch = buf[bp]; |
aoqi@0 | 786 | return erroneous("dc.malformed.html", p); |
aoqi@0 | 787 | } |
aoqi@0 | 788 | |
aoqi@0 | 789 | /** |
aoqi@0 | 790 | * Read a series of HTML attributes, terminated by {@literal > }. |
aoqi@0 | 791 | * Each attribute is of the form {@literal identifier[=value] }. |
aoqi@0 | 792 | * "value" may be unquoted, single-quoted, or double-quoted. |
aoqi@0 | 793 | */ |
aoqi@0 | 794 | protected List<DCTree> htmlAttrs() { |
aoqi@0 | 795 | ListBuffer<DCTree> attrs = new ListBuffer<DCTree>(); |
aoqi@0 | 796 | skipWhitespace(); |
aoqi@0 | 797 | |
aoqi@0 | 798 | loop: |
aoqi@0 | 799 | while (isIdentifierStart(ch)) { |
aoqi@0 | 800 | int namePos = bp; |
aoqi@0 | 801 | Name name = readIdentifier(); |
aoqi@0 | 802 | skipWhitespace(); |
aoqi@0 | 803 | List<DCTree> value = null; |
aoqi@0 | 804 | ValueKind vkind = ValueKind.EMPTY; |
aoqi@0 | 805 | if (ch == '=') { |
aoqi@0 | 806 | ListBuffer<DCTree> v = new ListBuffer<DCTree>(); |
aoqi@0 | 807 | nextChar(); |
aoqi@0 | 808 | skipWhitespace(); |
aoqi@0 | 809 | if (ch == '\'' || ch == '"') { |
aoqi@0 | 810 | vkind = (ch == '\'') ? ValueKind.SINGLE : ValueKind.DOUBLE; |
aoqi@0 | 811 | char quote = ch; |
aoqi@0 | 812 | nextChar(); |
aoqi@0 | 813 | textStart = bp; |
aoqi@0 | 814 | while (bp < buflen && ch != quote) { |
aoqi@0 | 815 | if (newline && ch == '@') { |
aoqi@0 | 816 | attrs.add(erroneous("dc.unterminated.string", namePos)); |
aoqi@0 | 817 | // No point trying to read more. |
aoqi@0 | 818 | // In fact, all attrs get discarded by the caller |
aoqi@0 | 819 | // and superseded by a malformed.html node because |
aoqi@0 | 820 | // the html tag itself is not terminated correctly. |
aoqi@0 | 821 | break loop; |
aoqi@0 | 822 | } |
aoqi@0 | 823 | attrValueChar(v); |
aoqi@0 | 824 | } |
aoqi@0 | 825 | addPendingText(v, bp - 1); |
aoqi@0 | 826 | nextChar(); |
aoqi@0 | 827 | } else { |
aoqi@0 | 828 | vkind = ValueKind.UNQUOTED; |
aoqi@0 | 829 | textStart = bp; |
aoqi@0 | 830 | while (bp < buflen && !isUnquotedAttrValueTerminator(ch)) { |
aoqi@0 | 831 | attrValueChar(v); |
aoqi@0 | 832 | } |
aoqi@0 | 833 | addPendingText(v, bp - 1); |
aoqi@0 | 834 | } |
aoqi@0 | 835 | skipWhitespace(); |
aoqi@0 | 836 | value = v.toList(); |
aoqi@0 | 837 | } |
aoqi@0 | 838 | DCAttribute attr = m.at(namePos).Attribute(name, vkind, value); |
aoqi@0 | 839 | attrs.add(attr); |
aoqi@0 | 840 | } |
aoqi@0 | 841 | |
aoqi@0 | 842 | return attrs.toList(); |
aoqi@0 | 843 | } |
aoqi@0 | 844 | |
aoqi@0 | 845 | protected void attrValueChar(ListBuffer<DCTree> list) { |
aoqi@0 | 846 | switch (ch) { |
aoqi@0 | 847 | case '&': |
aoqi@0 | 848 | entity(list); |
aoqi@0 | 849 | break; |
aoqi@0 | 850 | |
aoqi@0 | 851 | case '{': |
aoqi@0 | 852 | inlineTag(list); |
aoqi@0 | 853 | break; |
aoqi@0 | 854 | |
aoqi@0 | 855 | default: |
aoqi@0 | 856 | nextChar(); |
aoqi@0 | 857 | } |
aoqi@0 | 858 | } |
aoqi@0 | 859 | |
aoqi@0 | 860 | protected void addPendingText(ListBuffer<DCTree> list, int textEnd) { |
aoqi@0 | 861 | if (textStart != -1) { |
aoqi@0 | 862 | if (textStart <= textEnd) { |
aoqi@0 | 863 | list.add(m.at(textStart).Text(newString(textStart, textEnd + 1))); |
aoqi@0 | 864 | } |
aoqi@0 | 865 | textStart = -1; |
aoqi@0 | 866 | } |
aoqi@0 | 867 | } |
aoqi@0 | 868 | |
aoqi@0 | 869 | protected DCErroneous erroneous(String code, int pos) { |
aoqi@0 | 870 | int i = bp - 1; |
aoqi@0 | 871 | loop: |
aoqi@0 | 872 | while (i > pos) { |
aoqi@0 | 873 | switch (buf[i]) { |
aoqi@0 | 874 | case '\f': case '\n': case '\r': |
aoqi@0 | 875 | newline = true; |
aoqi@0 | 876 | break; |
aoqi@0 | 877 | case '\t': case ' ': |
aoqi@0 | 878 | break; |
aoqi@0 | 879 | default: |
aoqi@0 | 880 | break loop; |
aoqi@0 | 881 | } |
aoqi@0 | 882 | i--; |
aoqi@0 | 883 | } |
aoqi@0 | 884 | textStart = -1; |
aoqi@0 | 885 | return m.at(pos).Erroneous(newString(pos, i + 1), diagSource, code); |
aoqi@0 | 886 | } |
aoqi@0 | 887 | |
aoqi@0 | 888 | @SuppressWarnings("unchecked") |
aoqi@0 | 889 | <T> T getFirst(List<T>... lists) { |
aoqi@0 | 890 | for (List<T> list: lists) { |
aoqi@0 | 891 | if (list.nonEmpty()) |
aoqi@0 | 892 | return list.head; |
aoqi@0 | 893 | } |
aoqi@0 | 894 | return null; |
aoqi@0 | 895 | } |
aoqi@0 | 896 | |
aoqi@0 | 897 | protected boolean isIdentifierStart(char ch) { |
aoqi@0 | 898 | return Character.isUnicodeIdentifierStart(ch); |
aoqi@0 | 899 | } |
aoqi@0 | 900 | |
aoqi@0 | 901 | protected Name readIdentifier() { |
aoqi@0 | 902 | int start = bp; |
aoqi@0 | 903 | nextChar(); |
aoqi@0 | 904 | while (bp < buflen && Character.isUnicodeIdentifierPart(ch)) |
aoqi@0 | 905 | nextChar(); |
aoqi@0 | 906 | return names.fromChars(buf, start, bp - start); |
aoqi@0 | 907 | } |
aoqi@0 | 908 | |
aoqi@0 | 909 | protected Name readTagName() { |
aoqi@0 | 910 | int start = bp; |
aoqi@0 | 911 | nextChar(); |
aoqi@0 | 912 | while (bp < buflen && (Character.isUnicodeIdentifierPart(ch) || ch == '.')) |
aoqi@0 | 913 | nextChar(); |
aoqi@0 | 914 | return names.fromChars(buf, start, bp - start); |
aoqi@0 | 915 | } |
aoqi@0 | 916 | |
aoqi@0 | 917 | protected boolean isJavaIdentifierStart(char ch) { |
aoqi@0 | 918 | return Character.isJavaIdentifierStart(ch); |
aoqi@0 | 919 | } |
aoqi@0 | 920 | |
aoqi@0 | 921 | protected Name readJavaIdentifier() { |
aoqi@0 | 922 | int start = bp; |
aoqi@0 | 923 | nextChar(); |
aoqi@0 | 924 | while (bp < buflen && Character.isJavaIdentifierPart(ch)) |
aoqi@0 | 925 | nextChar(); |
aoqi@0 | 926 | return names.fromChars(buf, start, bp - start); |
aoqi@0 | 927 | } |
aoqi@0 | 928 | |
aoqi@0 | 929 | protected boolean isDecimalDigit(char ch) { |
aoqi@0 | 930 | return ('0' <= ch && ch <= '9'); |
aoqi@0 | 931 | } |
aoqi@0 | 932 | |
aoqi@0 | 933 | protected boolean isHexDigit(char ch) { |
aoqi@0 | 934 | return ('0' <= ch && ch <= '9') |
aoqi@0 | 935 | || ('a' <= ch && ch <= 'f') |
aoqi@0 | 936 | || ('A' <= ch && ch <= 'F'); |
aoqi@0 | 937 | } |
aoqi@0 | 938 | |
aoqi@0 | 939 | protected boolean isUnquotedAttrValueTerminator(char ch) { |
aoqi@0 | 940 | switch (ch) { |
aoqi@0 | 941 | case '\f': case '\n': case '\r': case '\t': |
aoqi@0 | 942 | case ' ': |
aoqi@0 | 943 | case '"': case '\'': case '`': |
aoqi@0 | 944 | case '=': case '<': case '>': |
aoqi@0 | 945 | return true; |
aoqi@0 | 946 | default: |
aoqi@0 | 947 | return false; |
aoqi@0 | 948 | } |
aoqi@0 | 949 | } |
aoqi@0 | 950 | |
aoqi@0 | 951 | protected boolean isWhitespace(char ch) { |
aoqi@0 | 952 | return Character.isWhitespace(ch); |
aoqi@0 | 953 | } |
aoqi@0 | 954 | |
aoqi@0 | 955 | protected void skipWhitespace() { |
aoqi@0 | 956 | while (isWhitespace(ch)) |
aoqi@0 | 957 | nextChar(); |
aoqi@0 | 958 | } |
aoqi@0 | 959 | |
aoqi@0 | 960 | protected int getSentenceBreak(String s) { |
aoqi@0 | 961 | if (sentenceBreaker != null) { |
aoqi@0 | 962 | sentenceBreaker.setText(s); |
aoqi@0 | 963 | int i = sentenceBreaker.next(); |
aoqi@0 | 964 | return (i == s.length()) ? -1 : i; |
aoqi@0 | 965 | } |
aoqi@0 | 966 | |
aoqi@0 | 967 | // scan for period followed by whitespace |
aoqi@0 | 968 | boolean period = false; |
aoqi@0 | 969 | for (int i = 0; i < s.length(); i++) { |
aoqi@0 | 970 | switch (s.charAt(i)) { |
aoqi@0 | 971 | case '.': |
aoqi@0 | 972 | period = true; |
aoqi@0 | 973 | break; |
aoqi@0 | 974 | |
aoqi@0 | 975 | case ' ': |
aoqi@0 | 976 | case '\f': |
aoqi@0 | 977 | case '\n': |
aoqi@0 | 978 | case '\r': |
aoqi@0 | 979 | case '\t': |
aoqi@0 | 980 | if (period) |
aoqi@0 | 981 | return i; |
aoqi@0 | 982 | break; |
aoqi@0 | 983 | |
aoqi@0 | 984 | default: |
aoqi@0 | 985 | period = false; |
aoqi@0 | 986 | break; |
aoqi@0 | 987 | } |
aoqi@0 | 988 | } |
aoqi@0 | 989 | return -1; |
aoqi@0 | 990 | } |
aoqi@0 | 991 | |
aoqi@0 | 992 | |
aoqi@0 | 993 | Set<String> htmlBlockTags = new HashSet<String>(Arrays.asList( |
aoqi@0 | 994 | "h1", "h2", "h3", "h4", "h5", "h6", "p", "pre")); |
aoqi@0 | 995 | |
aoqi@0 | 996 | protected boolean isSentenceBreak(Name n) { |
aoqi@0 | 997 | return htmlBlockTags.contains(StringUtils.toLowerCase(n.toString())); |
aoqi@0 | 998 | } |
aoqi@0 | 999 | |
aoqi@0 | 1000 | protected boolean isSentenceBreak(DCTree t) { |
aoqi@0 | 1001 | switch (t.getKind()) { |
aoqi@0 | 1002 | case START_ELEMENT: |
aoqi@0 | 1003 | return isSentenceBreak(((DCStartElement) t).getName()); |
aoqi@0 | 1004 | |
aoqi@0 | 1005 | case END_ELEMENT: |
aoqi@0 | 1006 | return isSentenceBreak(((DCEndElement) t).getName()); |
aoqi@0 | 1007 | } |
aoqi@0 | 1008 | return false; |
aoqi@0 | 1009 | } |
aoqi@0 | 1010 | |
aoqi@0 | 1011 | /** |
aoqi@0 | 1012 | * @param start position of first character of string |
aoqi@0 | 1013 | * @param end position of character beyond last character to be included |
aoqi@0 | 1014 | */ |
aoqi@0 | 1015 | String newString(int start, int end) { |
aoqi@0 | 1016 | return new String(buf, start, end - start); |
aoqi@0 | 1017 | } |
aoqi@0 | 1018 | |
aoqi@0 | 1019 | static abstract class TagParser { |
aoqi@0 | 1020 | enum Kind { INLINE, BLOCK } |
aoqi@0 | 1021 | |
aoqi@0 | 1022 | Kind kind; |
aoqi@0 | 1023 | DCTree.Kind treeKind; |
aoqi@0 | 1024 | |
aoqi@0 | 1025 | TagParser(Kind k, DCTree.Kind tk) { |
aoqi@0 | 1026 | kind = k; |
aoqi@0 | 1027 | treeKind = tk; |
aoqi@0 | 1028 | } |
aoqi@0 | 1029 | |
aoqi@0 | 1030 | Kind getKind() { |
aoqi@0 | 1031 | return kind; |
aoqi@0 | 1032 | } |
aoqi@0 | 1033 | |
aoqi@0 | 1034 | DCTree.Kind getTreeKind() { |
aoqi@0 | 1035 | return treeKind; |
aoqi@0 | 1036 | } |
aoqi@0 | 1037 | |
aoqi@0 | 1038 | abstract DCTree parse(int pos) throws ParseException; |
aoqi@0 | 1039 | } |
aoqi@0 | 1040 | |
aoqi@0 | 1041 | /** |
aoqi@0 | 1042 | * @see <a href="http://docs.oracle.com/javase/7/docs/technotes/tools/solaris/javadoc.html#javadoctags">Javadoc Tags</a> |
aoqi@0 | 1043 | */ |
aoqi@0 | 1044 | private void initTagParsers() { |
aoqi@0 | 1045 | TagParser[] parsers = { |
aoqi@0 | 1046 | // @author name-text |
aoqi@0 | 1047 | new TagParser(Kind.BLOCK, DCTree.Kind.AUTHOR) { |
aoqi@0 | 1048 | public DCTree parse(int pos) { |
aoqi@0 | 1049 | List<DCTree> name = blockContent(); |
aoqi@0 | 1050 | return m.at(pos).Author(name); |
aoqi@0 | 1051 | } |
aoqi@0 | 1052 | }, |
aoqi@0 | 1053 | |
aoqi@0 | 1054 | // {@code text} |
aoqi@0 | 1055 | new TagParser(Kind.INLINE, DCTree.Kind.CODE) { |
aoqi@0 | 1056 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1057 | DCTree text = inlineText(); |
aoqi@0 | 1058 | nextChar(); |
aoqi@0 | 1059 | return m.at(pos).Code((DCText) text); |
aoqi@0 | 1060 | } |
aoqi@0 | 1061 | }, |
aoqi@0 | 1062 | |
aoqi@0 | 1063 | // @deprecated deprecated-text |
aoqi@0 | 1064 | new TagParser(Kind.BLOCK, DCTree.Kind.DEPRECATED) { |
aoqi@0 | 1065 | public DCTree parse(int pos) { |
aoqi@0 | 1066 | List<DCTree> reason = blockContent(); |
aoqi@0 | 1067 | return m.at(pos).Deprecated(reason); |
aoqi@0 | 1068 | } |
aoqi@0 | 1069 | }, |
aoqi@0 | 1070 | |
aoqi@0 | 1071 | // {@docRoot} |
aoqi@0 | 1072 | new TagParser(Kind.INLINE, DCTree.Kind.DOC_ROOT) { |
aoqi@0 | 1073 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1074 | if (ch == '}') { |
aoqi@0 | 1075 | nextChar(); |
aoqi@0 | 1076 | return m.at(pos).DocRoot(); |
aoqi@0 | 1077 | } |
aoqi@0 | 1078 | inlineText(); // skip unexpected content |
aoqi@0 | 1079 | nextChar(); |
aoqi@0 | 1080 | throw new ParseException("dc.unexpected.content"); |
aoqi@0 | 1081 | } |
aoqi@0 | 1082 | }, |
aoqi@0 | 1083 | |
aoqi@0 | 1084 | // @exception class-name description |
aoqi@0 | 1085 | new TagParser(Kind.BLOCK, DCTree.Kind.EXCEPTION) { |
aoqi@0 | 1086 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1087 | skipWhitespace(); |
aoqi@0 | 1088 | DCReference ref = reference(false); |
aoqi@0 | 1089 | List<DCTree> description = blockContent(); |
aoqi@0 | 1090 | return m.at(pos).Exception(ref, description); |
aoqi@0 | 1091 | } |
aoqi@0 | 1092 | }, |
aoqi@0 | 1093 | |
aoqi@0 | 1094 | // {@inheritDoc} |
aoqi@0 | 1095 | new TagParser(Kind.INLINE, DCTree.Kind.INHERIT_DOC) { |
aoqi@0 | 1096 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1097 | if (ch == '}') { |
aoqi@0 | 1098 | nextChar(); |
aoqi@0 | 1099 | return m.at(pos).InheritDoc(); |
aoqi@0 | 1100 | } |
aoqi@0 | 1101 | inlineText(); // skip unexpected content |
aoqi@0 | 1102 | nextChar(); |
aoqi@0 | 1103 | throw new ParseException("dc.unexpected.content"); |
aoqi@0 | 1104 | } |
aoqi@0 | 1105 | }, |
aoqi@0 | 1106 | |
aoqi@0 | 1107 | // {@link package.class#member label} |
aoqi@0 | 1108 | new TagParser(Kind.INLINE, DCTree.Kind.LINK) { |
aoqi@0 | 1109 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1110 | DCReference ref = reference(true); |
aoqi@0 | 1111 | List<DCTree> label = inlineContent(); |
aoqi@0 | 1112 | return m.at(pos).Link(ref, label); |
aoqi@0 | 1113 | } |
aoqi@0 | 1114 | }, |
aoqi@0 | 1115 | |
aoqi@0 | 1116 | // {@linkplain package.class#member label} |
aoqi@0 | 1117 | new TagParser(Kind.INLINE, DCTree.Kind.LINK_PLAIN) { |
aoqi@0 | 1118 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1119 | DCReference ref = reference(true); |
aoqi@0 | 1120 | List<DCTree> label = inlineContent(); |
aoqi@0 | 1121 | return m.at(pos).LinkPlain(ref, label); |
aoqi@0 | 1122 | } |
aoqi@0 | 1123 | }, |
aoqi@0 | 1124 | |
aoqi@0 | 1125 | // {@literal text} |
aoqi@0 | 1126 | new TagParser(Kind.INLINE, DCTree.Kind.LITERAL) { |
aoqi@0 | 1127 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1128 | DCTree text = inlineText(); |
aoqi@0 | 1129 | nextChar(); |
aoqi@0 | 1130 | return m.at(pos).Literal((DCText) text); |
aoqi@0 | 1131 | } |
aoqi@0 | 1132 | }, |
aoqi@0 | 1133 | |
aoqi@0 | 1134 | // @param parameter-name description |
aoqi@0 | 1135 | new TagParser(Kind.BLOCK, DCTree.Kind.PARAM) { |
aoqi@0 | 1136 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1137 | skipWhitespace(); |
aoqi@0 | 1138 | |
aoqi@0 | 1139 | boolean typaram = false; |
aoqi@0 | 1140 | if (ch == '<') { |
aoqi@0 | 1141 | typaram = true; |
aoqi@0 | 1142 | nextChar(); |
aoqi@0 | 1143 | } |
aoqi@0 | 1144 | |
aoqi@0 | 1145 | DCIdentifier id = identifier(); |
aoqi@0 | 1146 | |
aoqi@0 | 1147 | if (typaram) { |
aoqi@0 | 1148 | if (ch != '>') |
aoqi@0 | 1149 | throw new ParseException("dc.gt.expected"); |
aoqi@0 | 1150 | nextChar(); |
aoqi@0 | 1151 | } |
aoqi@0 | 1152 | |
aoqi@0 | 1153 | skipWhitespace(); |
aoqi@0 | 1154 | List<DCTree> desc = blockContent(); |
aoqi@0 | 1155 | return m.at(pos).Param(typaram, id, desc); |
aoqi@0 | 1156 | } |
aoqi@0 | 1157 | }, |
aoqi@0 | 1158 | |
aoqi@0 | 1159 | // @return description |
aoqi@0 | 1160 | new TagParser(Kind.BLOCK, DCTree.Kind.RETURN) { |
aoqi@0 | 1161 | public DCTree parse(int pos) { |
aoqi@0 | 1162 | List<DCTree> description = blockContent(); |
aoqi@0 | 1163 | return m.at(pos).Return(description); |
aoqi@0 | 1164 | } |
aoqi@0 | 1165 | }, |
aoqi@0 | 1166 | |
aoqi@0 | 1167 | // @see reference | quoted-string | HTML |
aoqi@0 | 1168 | new TagParser(Kind.BLOCK, DCTree.Kind.SEE) { |
aoqi@0 | 1169 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1170 | skipWhitespace(); |
aoqi@0 | 1171 | switch (ch) { |
aoqi@0 | 1172 | case '"': |
aoqi@0 | 1173 | DCText string = quotedString(); |
aoqi@0 | 1174 | if (string != null) { |
aoqi@0 | 1175 | skipWhitespace(); |
aoqi@0 | 1176 | if (ch == '@' |
aoqi@0 | 1177 | || ch == EOI && bp == buf.length - 1) { |
aoqi@0 | 1178 | return m.at(pos).See(List.<DCTree>of(string)); |
aoqi@0 | 1179 | } |
aoqi@0 | 1180 | } |
aoqi@0 | 1181 | break; |
aoqi@0 | 1182 | |
aoqi@0 | 1183 | case '<': |
aoqi@0 | 1184 | List<DCTree> html = blockContent(); |
aoqi@0 | 1185 | if (html != null) |
aoqi@0 | 1186 | return m.at(pos).See(html); |
aoqi@0 | 1187 | break; |
aoqi@0 | 1188 | |
aoqi@0 | 1189 | case '@': |
aoqi@0 | 1190 | if (newline) |
aoqi@0 | 1191 | throw new ParseException("dc.no.content"); |
aoqi@0 | 1192 | break; |
aoqi@0 | 1193 | |
aoqi@0 | 1194 | case EOI: |
aoqi@0 | 1195 | if (bp == buf.length - 1) |
aoqi@0 | 1196 | throw new ParseException("dc.no.content"); |
aoqi@0 | 1197 | break; |
aoqi@0 | 1198 | |
aoqi@0 | 1199 | default: |
aoqi@0 | 1200 | if (isJavaIdentifierStart(ch) || ch == '#') { |
aoqi@0 | 1201 | DCReference ref = reference(true); |
aoqi@0 | 1202 | List<DCTree> description = blockContent(); |
aoqi@0 | 1203 | return m.at(pos).See(description.prepend(ref)); |
aoqi@0 | 1204 | } |
aoqi@0 | 1205 | } |
aoqi@0 | 1206 | throw new ParseException("dc.unexpected.content"); |
aoqi@0 | 1207 | } |
aoqi@0 | 1208 | }, |
aoqi@0 | 1209 | |
aoqi@0 | 1210 | // @serialData data-description |
aoqi@0 | 1211 | new TagParser(Kind.BLOCK, DCTree.Kind.SERIAL_DATA) { |
aoqi@0 | 1212 | public DCTree parse(int pos) { |
aoqi@0 | 1213 | List<DCTree> description = blockContent(); |
aoqi@0 | 1214 | return m.at(pos).SerialData(description); |
aoqi@0 | 1215 | } |
aoqi@0 | 1216 | }, |
aoqi@0 | 1217 | |
aoqi@0 | 1218 | // @serialField field-name field-type description |
aoqi@0 | 1219 | new TagParser(Kind.BLOCK, DCTree.Kind.SERIAL_FIELD) { |
aoqi@0 | 1220 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1221 | skipWhitespace(); |
aoqi@0 | 1222 | DCIdentifier name = identifier(); |
aoqi@0 | 1223 | skipWhitespace(); |
aoqi@0 | 1224 | DCReference type = reference(false); |
aoqi@0 | 1225 | List<DCTree> description = null; |
aoqi@0 | 1226 | if (isWhitespace(ch)) { |
aoqi@0 | 1227 | skipWhitespace(); |
aoqi@0 | 1228 | description = blockContent(); |
aoqi@0 | 1229 | } |
aoqi@0 | 1230 | return m.at(pos).SerialField(name, type, description); |
aoqi@0 | 1231 | } |
aoqi@0 | 1232 | }, |
aoqi@0 | 1233 | |
aoqi@0 | 1234 | // @serial field-description | include | exclude |
aoqi@0 | 1235 | new TagParser(Kind.BLOCK, DCTree.Kind.SERIAL) { |
aoqi@0 | 1236 | public DCTree parse(int pos) { |
aoqi@0 | 1237 | List<DCTree> description = blockContent(); |
aoqi@0 | 1238 | return m.at(pos).Serial(description); |
aoqi@0 | 1239 | } |
aoqi@0 | 1240 | }, |
aoqi@0 | 1241 | |
aoqi@0 | 1242 | // @since since-text |
aoqi@0 | 1243 | new TagParser(Kind.BLOCK, DCTree.Kind.SINCE) { |
aoqi@0 | 1244 | public DCTree parse(int pos) { |
aoqi@0 | 1245 | List<DCTree> description = blockContent(); |
aoqi@0 | 1246 | return m.at(pos).Since(description); |
aoqi@0 | 1247 | } |
aoqi@0 | 1248 | }, |
aoqi@0 | 1249 | |
aoqi@0 | 1250 | // @throws class-name description |
aoqi@0 | 1251 | new TagParser(Kind.BLOCK, DCTree.Kind.THROWS) { |
aoqi@0 | 1252 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1253 | skipWhitespace(); |
aoqi@0 | 1254 | DCReference ref = reference(false); |
aoqi@0 | 1255 | List<DCTree> description = blockContent(); |
aoqi@0 | 1256 | return m.at(pos).Throws(ref, description); |
aoqi@0 | 1257 | } |
aoqi@0 | 1258 | }, |
aoqi@0 | 1259 | |
aoqi@0 | 1260 | // {@value package.class#field} |
aoqi@0 | 1261 | new TagParser(Kind.INLINE, DCTree.Kind.VALUE) { |
aoqi@0 | 1262 | public DCTree parse(int pos) throws ParseException { |
aoqi@0 | 1263 | DCReference ref = reference(true); |
aoqi@0 | 1264 | skipWhitespace(); |
aoqi@0 | 1265 | if (ch == '}') { |
aoqi@0 | 1266 | nextChar(); |
aoqi@0 | 1267 | return m.at(pos).Value(ref); |
aoqi@0 | 1268 | } |
aoqi@0 | 1269 | nextChar(); |
aoqi@0 | 1270 | throw new ParseException("dc.unexpected.content"); |
aoqi@0 | 1271 | } |
aoqi@0 | 1272 | }, |
aoqi@0 | 1273 | |
aoqi@0 | 1274 | // @version version-text |
aoqi@0 | 1275 | new TagParser(Kind.BLOCK, DCTree.Kind.VERSION) { |
aoqi@0 | 1276 | public DCTree parse(int pos) { |
aoqi@0 | 1277 | List<DCTree> description = blockContent(); |
aoqi@0 | 1278 | return m.at(pos).Version(description); |
aoqi@0 | 1279 | } |
aoqi@0 | 1280 | }, |
aoqi@0 | 1281 | }; |
aoqi@0 | 1282 | |
aoqi@0 | 1283 | tagParsers = new HashMap<Name,TagParser>(); |
aoqi@0 | 1284 | for (TagParser p: parsers) |
aoqi@0 | 1285 | tagParsers.put(names.fromString(p.getTreeKind().tagName), p); |
aoqi@0 | 1286 | |
aoqi@0 | 1287 | } |
aoqi@0 | 1288 | } |