EBNF Grammar for AWK

README.md

Note: if you wish to understand these notations, please read this: https://gist.github.com/Chubek/52884d1fa766fa16ae8d8f226ba105ad

So, again, why did I write the EBNF grammar for AWK?

Basically, I have two ongoing projects where AWK is involved. Firs is Squawk, and implementation of AWK and second is AWK2c, which obviously translates AWK to C.

Plus, I am thinking of making a Github page called 'The Internet Grammar Database' where I would post EBNF, Yacc, PEG, Lex, definitions of languages. However, I don't have much experience in web development, so if you can help me, let me know (chubakbidpaa [at] riseup [dot] net).

So anyways, awk.ebnf contains the EBNF grammar for AWK. Some considerations:

• The grammar is based on POSIX specs.
• It may not be complete.
• It may not be the best representation of AWK as EBNF.
• The file, ad verbatim, is best viewed on NeoVim with my ebnf.vim which I linked in the Gist above.
• Some people asked if this gramamr can be fed to any parser generator? No parser generator exists that can feeds off of EBNF. There's BNFC, which uses LBNF but that's a whole other beast. This was true until last year. Today, with ChatGPT, you can generate almost anything you want with a well-specified EBNF grammar!

Thanks; ~ Chubak

awk.ebnf

awk-program		::= { top-level-element | comment }

top-level-element	::= function-definition
			| pattern-action-pair

function-definition	::= "function" identifier '(' [ identifier-list ] ')' compound-statements

pattern-action-pair	::= [ pattern ] action

action			::= compound-statements

pattern			::= expression-range
			| "BEGIN"
			| "END"

expression-range	::= expression [ ',' expression ]

compound-labeled	::= '{' labeled-statement-list '}'

compound-statements	::= '{' statement-list '}'

labeled-statement-list  ::= labeled-statement { terminate labeled-statement } 

labeld-statement	::= label ':' statement-list

label			::= "case" identifier
			| "default"

statement-list		::= statement { terminate statement }

statement		::= expression-statement
			| print-statement
			| printf-statement
			| for-statement
			| do-while-statement
			| while-statement
			| if-statement
			| switch-statement
			| controlflow-statement
			| memory-statement

memory-statement	::= "delete" variable

controlflow-statement	::= "break"
			| "continue"
			| "next"
			| "return" expression

switch-statment		::= "switch" '(' expression ')' compound-labeled

if-statement            ::= "if" '(' expression ')' compound-statements [ "else" ( if-statement | compound-statements ) ]

while-statement		::= "while" '(' expression ')'compound-statements

do-while-statement	::= "do" compound-statements "while" '(' expression ')'

for-statment 		::= "for" '(' [ expr-list [ ';' expr-list ';' expr-list ] ] ')' compound-statements

printf-statement	::= "printf" format-string ',' expr-list [ output-redirection ] terminate

print-statement		::= "print" expr-list [ output-redirection ] terminate

expression-statement	::= expression terminate

output-redirection	::= '>' string-const
			| ">>" string-const
			| '|' string-const

expr-list		::= expression { ',' expression }

expression		::= primary
			| unary-expr
			| binary-expr
			| ternary-expr
			| assign-expr

assign-expr		::= variable "^=" expression
			|  variable "%=" expression
			|  variable "*=" expression
			|  variable "/=" expression
			|  variable "+=" expression
			|  variable "-=" expression
			|  variable '=' expression

ternary-expr		::= binary-expr   '?' binary-expr ':' binary-expr 

binary-expr		::= logical-or-expr

logical-or-expr		::= logical-and-expr "||" logical-and-expr

logical-and-expr	::= in-expr "&&" logical-and-expr

in-expr			::= match-xpr "in" variable

match-expr		::= relop-expr'~' match-expr
			| relop-expr "!~" match-expr

relop-expr		::= concat-expr '<' relop-expr
			| concat-expr "<=" relop-expr
			| concat-expr '!=' relop-expr
			| concat-expr "==" relop-expr
			| concat-expr '>' relop-expr
			| concat-expr ">=" relop-expr

concat-expr		::= additive-expr concat-expr

additive-expr		::= multiplicative-expr '+' additive-expr
			| multiplicative-expr '-' additive-expr

multiplicative-expr	::= exponential-expr '*' multiplicative-expr
			| exponential-expr '/' multiplicative-expr
			| exponential-expr '%' multiplicative-expr

exponential-expr	::= unary-expr '^' exponential-expr

unary-expr		::= '+' primary
			| '-' primary
			| '!' primary
			| '~' primary
			| '$' primary
			| "++" primary
			| "--" primary
			| postfix-expr

postfix-expr		::= primary "++"
			| primary "--"

primary			::= variable
			| const-literal
			| getline
			| identifier '(' [ expr-list ] ')'
			| '(' expr ')'

getline			::= [ string-const '|' ] "getline" [ variable ] [ '<' string-const ]

variable		::= identifier [ index ]

index			::= '[' { alphanumeric } ']'

identifier-list		::= identifier { ',' identifier }

identifier		::= alphabetic { alphanumeric | '_' }

alphanumeric		::= letter | int-digit

alphabetic		::= upper-case | lower-case

upper-case		::= 'A' | 'B' | 'C' | ... | 'X' | 'Y' | 'Z'

lower-case		::= 'a' | 'b' | 'c' | ... | 'x' | 'y' | 'z'

format-string		::= '"' { format-glyph } '"'
			| "'" { format-glyph } "'"

const-literal		::= regex-const
			| string-const
			| float-const
			| integer-const

regex-const		::= '/' { vis-glyph } '/'

string-const		::= '"' { vis-glyph } '"'
			| "'" { vis-glyph } "'"

float-const		::= [ dec-integer ] '.' dec-integer

integer-const		::= dec-integer
			| hex-integer
			| oct-integer
			| bin-integer

dec-integer		::= int-digit { int-digit }

hex-integer		::= ( "0x" | "0X" ) hex-digit { hex-digit }

oct-integer		::= ( "0o" | "0O" ) oct-digit { oct-digit }

bin-integer		::= ( "0b" | "0B" ) bin-digit { bin-digit }

hex-digit		::= int-digit | 'a' | 'b' | 'c' | 'd' | 'e' | 'f'
				     | 'A' | 'B' | 'C' | 'D' | 'E' | 'F'

int-digit		::= oct-digit | '8' | '9'

oct-digit		::= bin-digit | '3' | '4' | '5' | '6' | '7'

bin-digit		::= '0' | '1' 

format-glyph		::= vis-glyph | format-tag

format-tag		::= '%' no-percentage

vis-glyph		::= printable | c-escape

printable		::= ' ' | '!' | ... | '}' | '~'

c-escape		::= '\' ( "'" | '"' | 'n' | 't' | 'r' | 'b' | 'f' | 'v' | 'a' | '?' | '0' | '\' )
			| '\' oct-digit oct-digit oct-digit
			| "\x" hex-digit hex-digit
			| "\U" hex-digit hex-digit hex-digit hex-digit

comment			::= "# " { no-newline } sys-newline

no-percentage		::= ? any character except percentage sign ?

no-newline		::= ? any character except sys-newline ?

terminate		::= [ ';' ] sys-newline

whitespace		::= ? space and horizontal tab ?

sys-newline		::= ? system newline character ?

Thanks for the EBNF!

A small typo: line with non-terminal oct-digit (https://gist.github.com/Chubek/30e20238ab198c73644512e6ccf41eed#file-awk-ebnf-L185 in the current revision) missed the "2" digit.