OLD | NEW |
---|---|
1 // Copyright 2014 The Chromium Authors. All rights reserved. | 1 // Copyright 2014 The Chromium Authors. All rights reserved. |
2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
4 | 4 |
5 #include "config.h" | 5 #include "config.h" |
6 #include "core/css/parser/CSSTokenizer.h" | 6 #include "core/css/parser/CSSTokenizer.h" |
7 | 7 |
8 namespace blink { | 8 namespace blink { |
9 #include "core/CSSTokenizerCodepoints.cpp" | 9 #include "core/CSSTokenizerCodepoints.cpp" |
10 } | 10 } |
11 | 11 |
12 #include "core/css/parser/CSSParserObserverWrapper.h" | 12 #include "core/css/parser/CSSParserObserverWrapper.h" |
13 #include "core/css/parser/CSSParserTokenRange.h" | 13 #include "core/css/parser/CSSParserTokenRange.h" |
14 #include "core/css/parser/CSSTokenizerInputStream.h" | 14 #include "core/css/parser/CSSTokenizerInputStream.h" |
15 #include "core/html/parser/HTMLParserIdioms.h" | 15 #include "core/html/parser/HTMLParserIdioms.h" |
16 #include "wtf/text/CharacterNames.h" | 16 #include "wtf/text/CharacterNames.h" |
17 | 17 |
18 namespace blink { | 18 namespace blink { |
19 | 19 |
20 CSSTokenizer::Scope::Scope(const String& string) | 20 CSSTokenizer::Scope::Scope(const String& string) |
21 : m_string(string) | 21 : m_string(string) |
22 { | 22 { |
23 // According to the spec, we should perform preprocessing here. | 23 // According to the spec, we should perform preprocessing here. |
24 // See: http://dev.w3.org/csswg/css-syntax/#input-preprocessing | 24 // See: http://dev.w3.org/csswg/css-syntax/#input-preprocessing |
25 // | 25 // |
26 // However, we can skip this step since: | 26 // However, we can skip this step since: |
27 // * We're using HTML spaces (which accept \r and \f as a valid white space) | 27 // * We're using HTML spaces (which accept \r and \f as a valid white space) |
28 // * Do not count white spaces | 28 // * Do not count white spaces |
29 // * CSSTokenizerInputStream::peek replaces NULLs for replacement characters | 29 // * CSSTokenizerInputStream::peek replaces NULLs for replacement characters |
30 | 30 |
31 if (string.isEmpty()) | 31 if (string.isEmpty()) |
32 return; | 32 return; |
33 | 33 |
34 // To avoid resizing we err on the side of reserving too much space. | 34 // To avoid resizing we err on the side of reserving too much space. |
35 // Most strings we tokenize have about 3.5 to 5 characters per token. | 35 // Most strings we tokenize have about 3.5 to 5 characters per token. |
36 m_tokens.reserveInitialCapacity(string.length() / 3); | 36 m_tokens.reserveInitialCapacity(string.length() / 3); |
37 | 37 |
38 CSSTokenizerInputStream input(string); | 38 CSSTokenizerInputStream input(string); |
39 CSSTokenizer tokenizer(input, *this); | 39 CSSTokenizer tokenizer(input, *this); |
40 while (true) { | 40 while (true) { |
41 CSSParserToken token = tokenizer.nextToken(); | 41 CSSParserToken token = tokenizer.nextToken(); |
42 if (token.type() == CommentToken) | 42 if (token.type() == CommentToken) |
43 continue; | 43 continue; |
44 if (token.type() == EOFToken) | 44 if (token.type() == EOFToken) |
45 return; | 45 return; |
46 m_tokens.append(token); | 46 m_tokens.append(token); |
47 } | 47 } |
48 } | 48 } |
49 | 49 |
50 CSSTokenizer::Scope::Scope(const String& string, CSSParserObserverWrapper& wrapp er) | 50 CSSTokenizer::Scope::Scope(const String& string, CSSParserObserverWrapper& wrapp er) |
51 : m_string(string) | 51 : m_string(string) |
52 { | 52 { |
53 if (string.isEmpty()) | 53 if (string.isEmpty()) |
54 return; | 54 return; |
55 | 55 |
56 CSSTokenizerInputStream input(string); | 56 CSSTokenizerInputStream input(string); |
57 CSSTokenizer tokenizer(input, *this); | 57 CSSTokenizer tokenizer(input, *this); |
58 | 58 |
59 unsigned offset = 0; | 59 unsigned offset = 0; |
60 while (true) { | 60 while (true) { |
61 CSSParserToken token = tokenizer.nextToken(); | 61 CSSParserToken token = tokenizer.nextToken(); |
(...skipping 197 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
259 return CSSParserToken(DelimiterToken, cc); | 259 return CSSParserToken(DelimiterToken, cc); |
260 } | 260 } |
261 | 261 |
262 CSSParserToken CSSTokenizer::colon(UChar cc) | 262 CSSParserToken CSSTokenizer::colon(UChar cc) |
263 { | 263 { |
264 return CSSParserToken(ColonToken); | 264 return CSSParserToken(ColonToken); |
265 } | 265 } |
266 | 266 |
267 CSSParserToken CSSTokenizer::semiColon(UChar cc) | 267 CSSParserToken CSSTokenizer::semiColon(UChar cc) |
268 { | 268 { |
269 return CSSParserToken(SemicolonToken); | 269 return CSSParserToken(SemicolonToken, lastConsumedCharacter()); |
alancutter (OOO until 2018)
2015/08/05 08:01:43
It's not clear to me why we need to store the semi
| |
270 } | 270 } |
271 | 271 |
272 CSSParserToken CSSTokenizer::hash(UChar cc) | 272 CSSParserToken CSSTokenizer::hash(UChar cc) |
273 { | 273 { |
274 UChar nextChar = m_input.nextInputChar(); | 274 UChar nextChar = m_input.nextInputChar(); |
275 if (isNameChar(nextChar) || twoCharsAreValidEscape(nextChar, m_input.peek(1) )) { | 275 if (isNameChar(nextChar) || twoCharsAreValidEscape(nextChar, m_input.peek(1) )) { |
276 HashTokenType type = nextCharsAreIdentifier() ? HashTokenId : HashTokenU nrestricted; | 276 HashTokenType type = nextCharsAreIdentifier() ? HashTokenId : HashTokenU nrestricted; |
277 return CSSParserToken(type, consumeName()); | 277 return CSSParserToken(type, consumeName()); |
278 } | 278 } |
279 | 279 |
(...skipping 348 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
628 { | 628 { |
629 while (true) { | 629 while (true) { |
630 UChar cc = consume(); | 630 UChar cc = consume(); |
631 if (cc == ')' || cc == kEndOfFileMarker) | 631 if (cc == ')' || cc == kEndOfFileMarker) |
632 return; | 632 return; |
633 if (twoCharsAreValidEscape(cc, m_input.nextInputChar())) | 633 if (twoCharsAreValidEscape(cc, m_input.nextInputChar())) |
634 consumeEscape(); | 634 consumeEscape(); |
635 } | 635 } |
636 } | 636 } |
637 | 637 |
638 void CSSTokenizer::consumeUntilNonWhitespace() | 638 unsigned CSSTokenizer::consumeUntilNonWhitespace() |
639 { | 639 { |
640 unsigned count = 0; | |
640 // Using HTML space here rather than CSS space since we don't do preprocessi ng | 641 // Using HTML space here rather than CSS space since we don't do preprocessi ng |
641 while (isHTMLSpace<UChar>(m_input.nextInputChar())) | 642 while (isHTMLSpace<UChar>(m_input.nextInputChar())) { |
643 ++count; | |
642 consume(); | 644 consume(); |
645 } | |
646 return count; | |
643 } | 647 } |
644 | 648 |
645 void CSSTokenizer::consumeSingleWhitespaceIfNext() | 649 void CSSTokenizer::consumeSingleWhitespaceIfNext() |
646 { | 650 { |
647 // We check for \r\n and HTML spaces since we don't do preprocessing | 651 // We check for \r\n and HTML spaces since we don't do preprocessing |
648 UChar c = m_input.nextInputChar(); | 652 UChar c = m_input.nextInputChar(); |
649 if (c == '\r' && m_input.peek(1) == '\n') | 653 if (c == '\r' && m_input.peek(1) == '\n') |
650 consume(2); | 654 consume(2); |
651 else if (isHTMLSpace(c)) | 655 else if (isHTMLSpace(c)) |
652 consume(); | 656 consume(); |
(...skipping 48 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
701 } | 705 } |
702 if (twoCharsAreValidEscape(cc, m_input.nextInputChar())) { | 706 if (twoCharsAreValidEscape(cc, m_input.nextInputChar())) { |
703 result.append(consumeEscape()); | 707 result.append(consumeEscape()); |
704 continue; | 708 continue; |
705 } | 709 } |
706 reconsume(cc); | 710 reconsume(cc); |
707 return registerString(result.toString()); | 711 return registerString(result.toString()); |
708 } | 712 } |
709 } | 713 } |
710 | 714 |
715 CSSParserString CSSTokenizer::lastConsumedCharacter() | |
716 { | |
717 return m_input.rangeAsCSSParserString(m_input.offset() - 1, 1); | |
718 } | |
719 | |
711 // http://dev.w3.org/csswg/css-syntax/#consume-an-escaped-code-point | 720 // http://dev.w3.org/csswg/css-syntax/#consume-an-escaped-code-point |
712 UChar32 CSSTokenizer::consumeEscape() | 721 UChar32 CSSTokenizer::consumeEscape() |
713 { | 722 { |
714 UChar cc = consume(); | 723 UChar cc = consume(); |
715 ASSERT(!isNewLine(cc)); | 724 ASSERT(!isNewLine(cc)); |
716 if (isASCIIHexDigit(cc)) { | 725 if (isASCIIHexDigit(cc)) { |
717 unsigned consumedHexDigits = 1; | 726 unsigned consumedHexDigits = 1; |
718 StringBuilder hexChars; | 727 StringBuilder hexChars; |
719 hexChars.append(cc); | 728 hexChars.append(cc); |
720 while (consumedHexDigits < 6 && isASCIIHexDigit(m_input.nextInputChar()) ) { | 729 while (consumedHexDigits < 6 && isASCIIHexDigit(m_input.nextInputChar()) ) { |
(...skipping 64 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
785 | 794 |
786 CSSParserString CSSTokenizer::registerString(const String& string) | 795 CSSParserString CSSTokenizer::registerString(const String& string) |
787 { | 796 { |
788 m_scope.storeString(string); | 797 m_scope.storeString(string); |
789 CSSParserString result; | 798 CSSParserString result; |
790 result.init(string); | 799 result.init(string); |
791 return result; | 800 return result; |
792 } | 801 } |
793 | 802 |
794 } // namespace blink | 803 } // namespace blink |
OLD | NEW |