reStructuredText的ANTLR语法(规则优先级)

时间:2011-05-30 16:16:04

标签: antlr antlr3 restructuredtext

第一个问题流

大家好,

这可以是对此问题的跟进:Antlr rule priorities

我正在尝试为reStructuredText markup language写一个ANTLR语法。

我面临的主要问题是:“如何匹配任何字符序列(常规文本)而不屏蔽其他语法规则?”

让我们举一个带内联标记的段落的例子:

In `Figure 17-6`_, we have positioned ``before_ptr`` so that it points to the element 
*before* the insert point. The variable ``after_ptr`` points to the element *after* the 
insert. In other words, we are going to put our new element **in between** ``before_ptr`` 
and ``after_ptr``.

我认为编写内联标记文本的规则很容易。所以我写了一个简单的语法:

grammar Rst;

options {
    output=AST;
    language=Java;
    backtrack=true;
    //memoize=true;
}

@members {
boolean inInlineMarkup = false;
}

// PARSER

text
    : inline_markup (WS? inline_markup)* WS? EOF
    ;


inline_markup
@after {
inInlineMarkup = false;
}
    : {!inInlineMarkup}? (emphasis|strong|litteral|link)
    ;

emphasis
@init {
inInlineMarkup = true;
}
    : '*' (~'*')+ '*' {System.out.println("emphasis: " + $text);}
    ;

strong
@init {
inInlineMarkup = true;
}
    : '**' (~'*')+ '**' {System.out.println("bold: " + $text);}
    ;

litteral
@init {
inInlineMarkup = true;
}
    : '``' (~'`')+ '``' {System.out.println("litteral: " + $text);}
    ;

link
@init {
inInlineMarkup = true;
}
    : inline_internal_target
    | footnote_reference
    | hyperlink_reference
    ;

inline_internal_target
    : '_`' (~'`')+ '`' {System.out.println("inline_internal_target: " + $text);}
    ;

footnote_reference
    : '[' (~']')+ ']_' {System.out.println("footnote_reference: " + $text);}
    ;


hyperlink_reference
    : ~(' '|'\t'|'\u000C'|'_')+ '_' {System.out.println("hyperlink_reference: " + $text);}
    |   '`' (~'`')+ '`_' {System.out.println("hyperlink_reference (long): " + $text);}
    ;

// LEXER

WS  
  : (' '|'\t'|'\u000C')+
  ; 

NEWLINE
  : '\r'? '\n'
  ;

这个简单的语法不起作用。我甚至没有尝试匹配常规文字......

我的问题:

  • 有人可以指出我的错误,也许可以给我一个如何匹配常规文字的提示吗?
  • 有没有办法设定语法规则的优先级?也许这可能是一个领先。

先谢谢你的帮助: - )

罗宾


第二个问题流

非常感谢你的帮助!我本来很难搞清楚我的错误...我不是在编写那种语法(仅)来学习ANTLR,我正在尝试编写一个用于eclipse的IDE插件。为此,我需要一个语法;)

我设法进一步使用语法并编写了text规则:

grammar Rst;

options {
    output=AST;
    language=Java;
}



@members {
boolean inInlineMarkup = false;
}

//////////////////
// PARSER RULES //
//////////////////

file
  : line* EOF
  ;


line
  : text* NEWLINE
  ;

text
    : inline_markup
    | normal_text
    ;

inline_markup
@after {
inInlineMarkup = false;
}
    : {!inInlineMarkup}? {inInlineMarkup = true;} 
  (
  | STRONG
  | EMPHASIS
  | LITTERAL
  | INTERPRETED_TEXT
  | SUBSTITUTION_REFERENCE
  | link
  )
    ;


link
    : INLINE_INTERNAL_TARGET
    | FOOTNOTE_REFERENCE
    | HYPERLINK_REFERENCE
    ;

normal_text
  : {!inInlineMarkup}? 
   ~(EMPHASIS
      |SUBSTITUTION_REFERENCE
      |STRONG
      |LITTERAL
      |INTERPRETED_TEXT
      |INLINE_INTERNAL_TARGET
      |FOOTNOTE_REFERENCE
      |HYPERLINK_REFERENCE
      |NEWLINE
      )
  ;
//////////////////
// LEXER TOKENS //
//////////////////

EMPHASIS
    : STAR ANY_BUT_STAR+ STAR {System.out.println("EMPHASIS: " + $text);}
    ;

SUBSTITUTION_REFERENCE
  : PIPE ANY_BUT_PIPE+ PIPE  {System.out.println("SUBST_REF: " + $text);}
  ;

STRONG
    : STAR STAR ANY_BUT_STAR+ STAR STAR {System.out.println("STRONG: " + $text);}
    ;

LITTERAL
    : BACKTICK BACKTICK ANY_BUT_BACKTICK+ BACKTICK BACKTICK {System.out.println("LITTERAL: " + $text);}
    ;
INTERPRETED_TEXT
  : BACKTICK ANY_BUT_BACKTICK+ BACKTICK {System.out.println("LITTERAL: " + $text);}
  ;

INLINE_INTERNAL_TARGET
    : UNDERSCORE BACKTICK ANY_BUT_BACKTICK+ BACKTICK {System.out.println("INLINE_INTERNAL_TARGET: " + $text);}
    ;

FOOTNOTE_REFERENCE
    : L_BRACKET ANY_BUT_BRACKET+ R_BRACKET UNDERSCORE {System.out.println("FOOTNOTE_REFERENCE: " + $text);}
    ;


HYPERLINK_REFERENCE
  : BACKTICK ANY_BUT_BACKTICK+ BACKTICK UNDERSCORE {System.out.println("HYPERLINK_REFERENCE (long): " + $text);}
  | ANY_BUT_ENDLINK+ UNDERSCORE {System.out.println("HYPERLINK_REFERENCE (short): " + $text);}
  ;

WS  
  : (' '|'\t')+ {$channel=HIDDEN;}
  ; 

NEWLINE
  : '\r'? '\n' {$channel=HIDDEN;}
  ;


///////////////
// FRAGMENTS //
///////////////

fragment ANY_BUT_PIPE
  : ESC PIPE
  | ~(PIPE|'\n'|'\r')
  ;
fragment ANY_BUT_BRACKET
  : ESC R_BRACKET
  | ~(R_BRACKET|'\n'|'\r')
  ;
fragment ANY_BUT_STAR
  : ESC STAR
  | ~(STAR|'\n'|'\r')
  ;
fragment ANY_BUT_BACKTICK
  : ESC BACKTICK
  | ~(BACKTICK|'\n'|'\r')
  ;
fragment ANY_BUT_ENDLINK
  : ~(UNDERSCORE|' '|'\t'|'\n'|'\r')
  ;



fragment ESC
  : '\\'
  ;
fragment STAR
  : '*'
  ;
fragment BACKTICK
  : '`'
  ;
fragment PIPE
  : '|'
  ;
fragment L_BRACKET
  : '['
  ;
fragment R_BRACKET
  : ']'
  ;
fragment UNDERSCORE
  : '_'
  ;

语法对于inline_markup工作正常但是normal_text不匹配。

这是我的测试类:

import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.Reader;

import org.antlr.runtime.ANTLRStringStream;
import org.antlr.runtime.CommonTokenStream;
import org.antlr.runtime.RecognitionException;
import org.antlr.runtime.tree.Tree;

public class Test {

    public static void main(String[] args) throws RecognitionException, IOException {

        InputStream is = Test.class.getResourceAsStream("test.rst");
        Reader r = new InputStreamReader(is);
        StringBuilder source = new StringBuilder();
        char[] buffer = new char[1024];
        int readLenght = 0;
        while ((readLenght = r.read(buffer)) > 0) {
            if (readLenght < buffer.length) {
                source.append(buffer, 0, readLenght);
            } else {
                source.append(buffer);
            }
        }
        r.close();
        System.out.println(source.toString());

        ANTLRStringStream in = new ANTLRStringStream(source.toString());
        RstLexer lexer = new RstLexer(in);
        CommonTokenStream tokens = new CommonTokenStream(lexer);
        RstParser parser = new RstParser(tokens);
        RstParser.file_return out = parser.file();
        System.out.println(((Tree)out.getTree()).toStringTree());
    }
}

我使用的输入文件:

In `Figure 17-6`_, we have positioned ``before_ptr`` so that it points to the element 
*before* the insert point. The variable ``after_ptr`` points to the |element| *after* the 
insert. In other words, `we are going`_ to put_ our new element **in between** ``before_ptr`` 
and ``after_ptr``.

我得到了这个输出:

HYPERLINK_REFERENCE (short): 7-6`_
line 1:2 mismatched character ' ' expecting '_'
line 1:10 mismatched character ' ' expecting '_'
line 1:18 mismatched character ' ' expecting '_'
line 1:21 mismatched character ' ' expecting '_'
line 1:26 mismatched character ' ' expecting '_'
line 1:37 mismatched character ' ' expecting '_'
LITTERAL: `before_ptr`
line 1:86 no viable alternative at character '\r'
line 1:55 mismatched character ' ' expecting '_'
line 1:60 mismatched character ' ' expecting '_'
line 1:63 mismatched character ' ' expecting '_'
line 1:70 mismatched character ' ' expecting '_'
line 1:73 mismatched character ' ' expecting '_'
line 1:77 mismatched character ' ' expecting '_'
line 1:85 mismatched character ' ' expecting '_'
EMPHASIS: *before*
line 2:12 mismatched character ' ' expecting '_'
line 2:19 mismatched character ' ' expecting '_'
line 2:26 mismatched character ' ' expecting '_'
LITTERAL: `after_ptr`
line 2:30 mismatched character ' ' expecting '_'
line 2:39 mismatched character ' ' expecting '_'
line 2:90 no viable alternative at character '\r'
line 2:60 mismatched character ' ' expecting '_'
line 2:63 mismatched character ' ' expecting '_'
line 2:67 mismatched character ' ' expecting '_'
line 2:77 mismatched character ' ' expecting '_'
line 2:85 mismatched character ' ' expecting '_'
line 2:89 mismatched character ' ' expecting '_'
line 3:7 mismatched character ' ' expecting '_'
line 3:10 mismatched character ' ' expecting '_'
line 3:16 mismatched character ' ' expecting '_'
line 3:23 mismatched character ' ' expecting '_'
line 3:27 mismatched character ' ' expecting '_'
line 3:31 mismatched character ' ' expecting '_'
line 3:42 mismatched character ' ' expecting '_'
line 3:51 mismatched character ' ' expecting '_'
line 3:55 mismatched character ' ' expecting '_'
line 3:63 mismatched character ' ' expecting '_'
line 3:94 mismatched character '\r' expecting '*'
line 4:3 mismatched character ' ' expecting '_'
line 4:18 no viable alternative at character '\r'
line 4:18 mismatched character '\r' expecting '_'
HYPERLINK_REFERENCE (short): oing`_
HYPERLINK_REFERENCE (short): ut_
EMPHASIS: *in between*
LITTERAL: `after_ptr`
BR.recoverFromMismatchedToken
line 0:-1 mismatched input '<EOF>' expecting NEWLINE
null

你能指出我的错误吗? (当我添加filter = true;语法选项时,解析器适用于内联标记而没有错误)

罗宾

2 个答案:

答案 0 :(得分:6)

以下是 解析此reStructeredText的快速演示。请注意,它只处理所有可用标记语法的一小部分,并且通过向其添加更多内容,影响现有的解析器/词法分析器规则:所以有很多,很多还有更多工作要做!

演示

grammar RST;

options {
  output=AST;
  backtrack=true;
  memoize=true;
}

tokens {
  ROOT;
  PARAGRAPH;
  INDENTATION;
  LINE;
  WORD;
  BOLD;
  ITALIC;
  INTERPRETED_TEXT;
  INLINE_LITERAL;
  REFERENCE;
}

parse
  :  paragraph+ EOF -> ^(ROOT paragraph+)
  ;

paragraph
  :  line+ -> ^(PARAGRAPH line+)
  |  Space* LineBreak -> /* omit line-breaks between paragraphs from AST */
  ;

line
  :  indentation text+ LineBreak -> ^(LINE text+)
  ;

indentation
  :  Space* -> ^(INDENTATION Space*)
  ;

text
  :  styledText
  |  interpretedText
  |  inlineLiteral
  |  reference
  |  Space
  |  Star
  |  EscapeSequence
  |  Any
  ;

styledText
  :  bold
  |  italic
  ;

bold
  :  Star Star boldAtom+ Star Star -> ^(BOLD boldAtom+)
  ;  

italic
  :  Star italicAtom+ Star -> ^(ITALIC italicAtom+)
  ;

boldAtom
  :  ~(Star | LineBreak)
  |  italic
  ;

italicAtom
  :  ~(Star | LineBreak)
  |  bold
  ;

interpretedText
  :  BackTick interpretedTextAtoms BackTick -> ^(INTERPRETED_TEXT interpretedTextAtoms)
  ;

interpretedTextAtoms
  :  ~BackTick+
  ;

inlineLiteral
  :  BackTick BackTick inlineLiteralAtoms BackTick BackTick -> ^(INLINE_LITERAL inlineLiteralAtoms)
  ;

inlineLiteralAtoms
  :  inlineLiteralAtom+
  ;

inlineLiteralAtom
  :  ~BackTick
  |  BackTick ~BackTick
  ;

reference
  :  Any+ UnderScore -> ^(REFERENCE Any+)
  ;

UnderScore
  :  '_'
  ;

BackTick
  :  '`'
  ;

Star
  :  '*'
  ;

Space
  :  ' ' 
  |  '\t'
  ;

EscapeSequence
  :  '\\' ('\\' | '*')
  ;

LineBreak
  :  '\r'? '\n'
  |  '\r'
  ;

Any
  :  .
  ;

当您从上面生成解析器和词法分析器时,让它解析以下输入文件:

***x*** **yyy** *zz* *
a b c

P2 ``*a*`b`` `q`
Python_

(注意尾随换行!)

解析器将生成以下AST:

enter image description here

修改

可以通过运行此类来创建图表:

import org.antlr.runtime.*;
import org.antlr.runtime.tree.*;
import org.antlr.stringtemplate.*;

public class Main {
  public static void main(String[] args) throws Exception {
    String source =
        "***x*** **yyy** *zz* *\n" +
        "a b c\n" +
        "\n" +
        "P2 ``*a*`b`` `q`\n" +
        "Python_\n";
    RSTLexer lexer = new RSTLexer(new ANTLRStringStream(source));
    RSTParser parser = new RSTParser(new CommonTokenStream(lexer));
    CommonTree tree = (CommonTree)parser.parse().getTree();
    DOTTreeGenerator gen = new DOTTreeGenerator();
    StringTemplate st = gen.toDOT(tree);
    System.out.println(st);
  }
}

或者如果您的来源来自文件,请执行以下操作:

RSTLexer lexer = new RSTLexer(new ANTLRFileStream("test.rst"));

RSTLexer lexer = new RSTLexer(new ANTLRFileStream("test.rst", "???"));

其中"???"是您文件的编码。

上面的类会将AST作为DOT文件打印到控制台。您可以使用DOT查看器显示AST。在这种情况下,我发布了由kgraphviewer创建的图像。但有many more viewers around。一个不错的在线版本是this one,它似乎在“引擎盖”下使用kgraphviewer。祝你好运!

答案 1 :(得分:4)

  

罗宾写道:

     

我认为编写内联标记文本的规则很容易

我必须承认我不熟悉这种标记语言,但它似乎类似于BB-Code或Wiki标记,它们不易翻译成(ANTLR)语法!这些语言不容易被标记化,因为它取决于这些令牌发生的位置。空格有时具有特殊含义(带有定义列表)。所以不,这一点都不容易,IMO。因此,如果这只是让您熟悉ANTLR(或一般的解析器生成器)的练习,我高度建议选择其他内容进行解析。

  

罗宾写道:

     

有人可以指出我的错误,也许可以给我一个如何匹配常规文本的提示吗?

您必须首先意识到ANTLR会创建词法分析器(tokenizer)和解析器。 Lexer规则以大写字母开头,解析器规则以小写字母开头。解析器只能对令牌(词法分析器规则生成的对象)进行操作。为了保持秩序, 不应在解析器规则中使用标记文字(请参阅下面语法中的规则q)。此外,~(否定)元字符具有不同的含义,具体取决于它的使用位置(在解析器或词法分析器规则中)。

采用以下语法:

p : T;
q : ~'z';

T : ~'x';
U : 'y';

ANTLR会首先将'z'文字“移动”到词法分析器规则,如下所示:

p : T;
q : ~RANDOM_NAME;

T : ~'x';
U : 'y';
RANDOM_NAME : 'z';

(名称RANDOM_NAME未使用,但无关紧要)。现在,解析器规则q 匹配'z'以外的任何字符!解析器规则中的否定否定了令牌(或词法分析器规则)。因此,~RANDOM_NAME将匹配词法分析器T或词法分析器U

在lexer规则中,~否定(单个!)字符。因此,词法分析器规则T将匹配范围\u0000 .. \uFFFF中的任何字符,但'x'除外。请注意以下内容:~'ab'在词法分析器规则中无效:您只能否定单个字符集。

因此,解析器规则中的所有这些~'???'都是错误的(错误,因为:它们的行为与您期望的不同)。

  

罗宾写道:

     

有没有办法为语法规则设置优先级?也许这可能是一个领先。

是的,在词法分析器和解析器规则中,顺序是从上到下(顶部具有最高优先级)。假设parse是你语法的入口点:

parse
  :  p
  |  q
  ;

然后首先尝试p,如果失败,则尝试匹配q

对于词法分析器规则,例如关键字规则在可能与所述关键字匹配的规则之前匹配:

// first keywords:
WHILE : 'while';
IF    : 'if'
ELSE  : 'else';

// and only then, the identifier rule: 
ID    : ('a'..'z' | 'A'..'Z' | '_') ('a'..'z' | 'A'..'Z' | '_' | '0'..'9')*;