Parse EBNF: Difference between revisions

{{clarified-review}}{{draft task}}[[Category:Recursion]]

<langsyntaxhighlight lang="go">package main

}</langsyntaxhighlight>

<langsyntaxhighlight lang="haskell">import Control.Applicative

ws = many $ oneOf " \n\t"</langsyntaxhighlight>

<langsyntaxhighlight lang="modula2">MODULE EBNF;

END EBNF.</langsyntaxhighlight>

<langsyntaxhighlight lang="modula2">IMPLEMENTATION MODULE EBNFScanner;

END EBNFScanner.</langsyntaxhighlight>

<langsyntaxhighlight lang="perl">#!/usr/bin/perl

----------------------------------------------------------------------</langsyntaxhighlight>

-- identifiers are sdxstrictly +=a-z only, not A-Z or 1-9</span>

<span style="color: #000000;">get_token</span><span style="color: #0000FF;">()</span>

-- {"repeatterminal", <e1>"a1"} -- as per "{}" inliteral ebnfconstants

-- {"optionalor", <&lt;e1>} &gt;, &lt;e2&gt;, ...} -- as(any) per "[]"one inof ebnfn

-- {"identrepeat", <name>, idx&lt;e1&gt;} -- applyas theper sub-rule"{}" in ebnf

} "za" """,{

factor expr = numberterm |{ '('plus exprterm ')'} .

plus factor = "+"number | "-"'(' expr ')' .

number =times digit= {"*" digit| }"/" .

<span style="color: #008000;">"a1 a4a2a6"</span><span a5style="color: a6#0000FF;">,</span>

<span style="color: #008000;">"a1 a2a3 a4 a5a6"</span><span a5style="color: a6#0000FF;">,</span>

<span style="color: #008000;">"a1 a2 a4 a5 a6"</span><span style="color: a7#0000FF;">,</span>

<span style="yourcolor: ad#008000;">"a1 a2 a4 a5 a5 a6"</span><span style="color: here#0000FF;"}>,</span>

<span style="2*3color: +#008000;">"your 4ad here"</23span><span -style="color: 7#0000FF;">},</span>

<span style="color: #008000;">"-2*3 + 4/23 - 7"</span><span style="color: #0000FF;">,</span>

<span style="color: #008000;">"(3 + 4) * 6-2+(4*(4))"</span><span style="color: #0000FF;">,</span>

<span style="(4color: +#008000;">"-2"</span><span style="color: 3#0000FF;"}}>,</span>

<langsyntaxhighlight PicoLisplang="picolisp">(de EBNF

(put (car @S) 'ebnf @E) ) )</langsyntaxhighlight>

<langsyntaxhighlight PicoLisplang="picolisp">(de matchEbnf (Pat)

(parseLst (get 'expr 'ebnf)) ) ) )</langsyntaxhighlight>

This parses the EBNF rule set using a perl 6Raku grammar, then if it parses as valid EBNF, constructs a grammar and parses the test strings with that. EBNF rule sets that are naively syntactically correct but missing rules will parse as valid but will give a runtime failure warning about missing methods.

<syntaxhighlight lang="raku" perl6line># A perl 6Raku grammar to parse EBNF

rule TOP { ^ <title>? '{' [ <production> ]+ '}' <comment>? $ }

rule production { <name> '=' <expression> <[.;]> }

rule expression { <term> +% "|" }

rule term { <factor>+ }

rule factor { <group> | <repeat> | <optional> | <identifier> | <literal> }

rule group { '(' <expression> ')' }

rule repeat { '{' <expression> '}' }

rule optional { '[' <expression> ']' }

token identifier { <-[\|\(\)\{\}\[\]\.\;\"\'\s]>+ } #"

token literal { ["'" <-[']>+ "'" | '"' <-["]>+ '"'] } #"

token title { <literal> }

token comment { <literal> }

token name { <identifier> <?before \h* '='> }

method production($/) {

method expression($/) { make join '|', $<term>>>.ast }

method term($/) { make join '\h*', $<factor>>>.ast }

method factor($/) {

method repeat($/) { make $<expression>.ast }

method optional($/) { make $<expression>.ast }

method group($/) { make $<expression>.ast }

say qqx/perl6raku $fn/;

}</langsyntaxhighlight>

<langsyntaxhighlight lang="ruby">#--

<langsyntaxhighlight lang="tcl">package require Tcl 8.6

}</langsyntaxhighlight>

<langsyntaxhighlight lang="tcl"># Demonstration code

puts [parse "1 - 2 - -3 * (4 + 5)"]</langsyntaxhighlight>