grammarkdown
Summary
grammarkdown is a markdown-style parser for syntactic grammars, designed to make it easily to rapidly prototype a grammar and statically verify its consistency.
The grammar supported by grammarkdown is based on the parametric grammar used by ECMA-262 (the JavaScript language standard).
Usage
Syntax: grammarkdown [options] [...files]
Examples: grammarkdown es6.grammar
grammarkdown --out es6.md --format markdown es6.grammar
Options:
-f, --format FORMAT The output format.
-h, --help Prints this message.
--noChecks Does not perform static checking of the grammar.
--noEmit Does not emit output.
--noEmitOnError Does not emit output if there are errors.
-o, --out FILE Specify the output file.
-v, --version Prints the version.
Syntax
A grammarkdown grammar file uses significant whitespace in the form of line terminators and indentation. Tab (ASCII 0x9) characters are preferred,
however using multiple spaces for indentation is supported as long as all nested elements have the same amount of leading whitespace.
Productions
A Production consists of a left-hand-side Nonterminal followed by a colon (:) separator and one or more right-hand-side sentences consisting of
various forms of terminal and nonterminal symbols. For example:
NameSpaceImport : `*` `as` ImportedBinding
It is recommended that Productions should follow pascal-case naming conventions, to avoid collision with reserved keywords.
You may specify multiple productions for a Nonterminal on multiple lines, as follows:
NamedImports : `{` `}`
NamedImports : `{` ImportList `}`
NamedImports : `{` ImportList `,` `}`
You may also specify multiple right-hand-side sentences for a single production by indenting them:
NamedImports :
`{` `}`
`{` ImportList `}`
`{` ImportList `,` `}`
A Production may specify one or more parameters that can be used to reuse a Nonterminal in various circumstances:
IdentifierReference[Yield] :
Identifier
[~Yield] `yield`
A Production may also specify a limited set of terminals, by using the one of keyphrase:
Keyword :: one of
`break` `do` `in` `typeof`
`case` `else` `instanceof` `var`
`catch` `export` `new` `void`
`class` `extends` `return` `while`
`const` `finally` `super` `with`
`continue` `for` `switch` `yield`
`debugger` `function` `this`
`default` `if` `throw`
`delete` `import` `try`
Parameters
If a Nonterminal on the right-hand-side of a production needs to set a parameter, they supply it in an argument list. Supplying the name of the argument sets the parameter, prefixing the name with a question mark ('?) passes the current value of the parameter, and eliding the argument clears the parameter:
Declaration[Yield] :
HoistableDeclaration[?Yield]
ClassDeclaration[?Yield]
LexicalDeclaration[In, ?Yield]
The right-hand-side of a Production consists of one or more Terminal or Nonterminal symbols, a sentence of Prose, or an Assertion.
Terminals
A Terminal symbol can be one of the following:
- A literal string of one or more characters enclosed in backticks ('`'). For example:
`function` - A sequence of three backtick characters, which denotes a backtick token. For example:
``` - A unicode character literal enclosed in a leading less-than ('<') character and a trailing greater-than ('>') character. For example:
<TAB>
Nonterminals
A Nonterminal symbol is an identifier, followed by an optional argument list, and an optional question mark ('?'). The question mark changes the cardinality of the Nonterminal from "exactly one" to "zero or one".
The identifier may optionally be enclosed in | characters, if it happens to collide with a keyword.
Character Literals and Ranges
Character literals may be specified using one of the following forms:
- A Unicode code point, of the form
U+followed by four to six non-lowercase hexadecimal digits with no leading zeros other than those necessary for padding to a minimum of four digits, in accordance with The Unicode Standard, Version 15.0.0, Appendix A, Notational Conventions (i.e., matching Unicode extended BNF pattern"U+" ( [1-9 A-F] | "10" )? H H H Hor regular expression pattern^U[+]([1-9A-F]|10)?[0-9A-F]{4}$as inU+00A0orU+1D306). - The preceding representation followed by a space and a printable ASCII prose explanation (such as a character name) free of
<and>and line terminators, all wrapped in<and>(i.e., matching Unicode extended BNF pattern"<" "U+" ( [1-9 A-F] | "10" )? H H H H " " [\u0020-\u007E -- [<>]]+ ">"or regular expression pattern^<U[+]([1-9A-F]|10)?[0-9A-F]{4} [\x20-\x3b\x3d\x3f-\x7e]+>$as in<U+2212 MINUS SIGN>) - An abbreviation defined somewhere outside the grammar as an ASCII identifier name, wrapped in
<and>(i.e., matching Unicode extended BNF pattern"<" [A-Z a-z _] [A-Z a-z _ 0-9]* ">"or regular expression pattern^<[A-Za-z_][A-Za-z_0-9]*$>as in<NBSP>).
Character ranges may be specified using the through keyword:
SourceCharacter but not one of `"` or `\` or U+0000 through U+001F
Prose
A sentence of Prose is a single line with a leading greater-than ('>') character. For example: > any Unicode code point
The but not Condition
The but not condition allows you to reference a production, excluding some part of that production. For example:
MultiLineNotAsteriskChar ::
SourceCharacter but not `*`
Here, MultiLineNotAsteriskChar may contain any alternative from SourceCharacter, except the terminal `*`.
The one of Condition
You can exclude multiple alternatives by including a list of symbols to exclude through the use of the one of keyphrase.
Each entry in the list is separated by or:
MultiLineNotForwardSlashOrAsteriskChar ::
SourceCharacter but not one of `/` or `*`
Assertions
An Assertion is a zero-width test that must evaluate successfully for the Production to be considered. Assertions are enclosed in a leading open bracket ('[') character and a trailing close-bracket (']') character.
The possible assertions include:
- The empty assertion, which matches exactly zero tokens:
[empty] - The lookahead assertion, which verifies the next tokens in the stream:
[lookahead != `function`] - The no-symbol-here assertion, which verifies the next token is not the provided symbol:
[no LineTerminator here] - The lexical-goal assertion, which states that the current lexical goal is the supplied Nonterminal:
[lexical goal InputElementRegExp] - The parameter assertion, which states the supplied parameter to the current production is either set (using the plus ('+') character), or cleared (using the tilde ('~') character):
[~Yield] `yield` - The prose assertion, which allows for arbitrary prose, mixed with terminals and nonterminals:
[> prose text `terminal` prose text |NonTerminal| prose text]
A lookahead assertion has the following operators:
- The
==operator states the lookahead phrase is matched:[lookahead == `class`] - The
!=operator states the lookahead phrase is not matched:[lookahead != `function`] - The
<-operator states that any matching phrase in the provided set is matched:[lookahead <- { `public`, `private` }] - The
<!operator states that any matching phrase in the provided set is not matched:[lookahead <! { `{`, `function` }]
Linking
During emit, grammarkdown implicitly adds a generated name for each Production and Right-hand side that can be used to
link directly to the production using a URI fragment. You can explicitly set the name for a production by tagging it with a custom link name:
Declaration[Yield] :
HoistableDeclaration[?Yield] #declaration-hoistable
ClassDeclaration[?Yield] #declaration-class
LexicalDeclaration[In, ?Yield] #declaration-lexical
Comments
You can also annotate your grammar with C-style single-line and multi-line comments.
Examples
For comprehensive examples of grammarkdown syntax and output, you can review the following samples:
- ECMA-262 version 2015 (ES6) Grammar
- TypeScript 1.5 Supplemental Grammar
API
grammarkdown has an API that can be consumed:
var grammarkdown = require("grammarkdown")
, Grammar = grammarkdown.Grammar
, EmitFormat = grammarkdown.EmitFormat
var filename = "...";
var source = "...";
var output;
// parse
var grammar = new Grammar(
[filename],
{ format: EmitFormat.markdown },
function () { return source; });
// bind (optional, bind happens automatically during check)
grammar.bind();
// check (optional, check happens automatically during emit)
grammar.check();
// emit
grammar.emit(undefined, function (file, text) { output = text; });
console.log(output);