leo.abnf

; Copyright (C) 2019-2023 Aleo Systems Inc.
; Copyright (C) 2024 Provable Inc.
; This file is part of the Leo library.

; The Leo library is free software: you can redistribute it and/or modify
; it under the terms of the GNU General Public License as published by
; the Free Software Foundation, either version 3 of the License, or
; (at your option) any later version.

; The Leo library is distributed in the hope that it will be useful,
; but WITHOUT ANY WARRANTY; without even the implied warranty of
; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
; GNU General Public License for more details.

; You should have received a copy of the GNU General Public License
; along with the Leo library. If not, see <https://www.gnu.org/licenses/>.

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

; This file specifies the grammar of Leo in the ABNF notation,
; which is defined in:
; - RFC 5234 (https://www.rfc-editor.org/rfc/rfc5234)
; - RFC 7405 (https://www.rfc-editor.org/rfc/rfc7405)

; Leo files are written in Unicode, encoded in UTF-8.
; The grammar in this file applies to the Unicode scalar values
; that result from the UTF-8 decoding of the files.

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

; Lexical Grammar
; ---------------

ascii = %x00-7F

safe-nonascii = %x80-2029 / %x202F-2065 / %x206A-D7FF / %xE000-10FFFF
                ; excludes bidi embeddings/overrides/isolates
                ; (i.e. %x202A-202E and %x2066-2069)
                ; and excludes high/low surrogates
                ; (i.e. %xD800-DFFF )

character = ascii / safe-nonascii

line-terminator = line-feed / carriage-return line-feed

line-feed = %x0A

carriage-return = %x0D

whitespace = space
           / horizontal-tab
           / form-feed
           / line-terminator
           / carriage-return ; not followed by line feed

space = %x20

horizontal-tab = %x09

form-feed = %x0C

comment = block-comment / line-comment

block-comment = "/*" rest-of-block-comment

rest-of-block-comment =
      "*" rest-of-block-comment-after-star
    / not-star-or-line-feed rest-of-block-comment
    / line-terminator rest-of-block-comment

rest-of-block-comment-after-star =
      "/"
    / "*" rest-of-block-comment-after-star
    / not-star-or-slash-or-line-feed rest-of-block-comment
    / line-terminator rest-of-block-comment

not-star-or-line-feed = %x00-09
                      ; no line feed (%x0A)
                      / %x0B-29
                      ; no * (%x2A)
                      / %x2B-7E
                      / safe-nonascii

not-star-or-slash-or-line-feed = %x00-09
                               ; no line feed (%x0A)
                               / %x0B-29
                               ; no * (%x2A)
                               / %x2B-2E
                               ; no / (%x2F)
                               / %x30-7E
                               / safe-nonascii

line-comment = "//" *not-line-feed line-terminator

not-line-feed = %x00-09
              ; no line feed (%x0A)
              / %x0B-7F
              / safe-nonascii

keyword = %s"address"
        / %s"aleo"
        / %s"as"
        / %s"assert"
        / %s"assert_eq"
        / %s"assert_neq"
        / %s"async"
        / %s"block"
        / %s"bool"
        / %s"const"
        / %s"constant"
        / %s"else"
        / %s"field"
        / %s"Fn"
        / %s"for"
        / %s"function"
        / %s"Future"
        / %s"group"
        / %s"i8"
        / %s"i16"
        / %s"i32"
        / %s"i64"
        / %s"i128"
        / %s"if"
        / %s"import"
        / %s"in"
        / %s"inline"
        / %s"let"
        / %s"mapping"
        / %s"network"
        / %s"private"
        / %s"program"
        / %s"public"
        / %s"record"
        / %s"return"
        / %s"scalar"
        / %s"self"
        / %s"signature"
        / %s"string"
        / %s"struct"
        / %s"transition"
        / %s"u8"
        / %s"u16"
        / %s"u32"
        / %s"u64"
        / %s"u128"

uppercase-letter = %x41-5A ; A-Z

lowercase-letter = %x61-7A ; a-z

letter = uppercase-letter / lowercase-letter

binary-digit = %x30-31 ; 0-1

octal-digit = %x30-37 ; 0-7

decimal-digit = %x30-39 ; 0-9

nonzero-decimal-digit = %x31-39 ; 1-9

hexadecimal-digit = decimal-digit
                  / %x41-46 ; A-F (uppercase)
                  ; no a-f (lowercase)

identifier = letter *( letter / decimal-digit / "_" )
             ; but not a keyword or a boolean literal or aleo1...

binary-numeral = %s"0b" 1*( binary-digit *"_" )

octal-numeral = %s"0o" 1*( octal-digit *"_" )

decimal-numeral = 1*( decimal-digit *"_" )

hexadecimal-numeral = %s"0x" 1*( hexadecimal-digit *"_" )

numeral = binary-numeral
        / octal-numeral
        / decimal-numeral
        / hexadecimal-numeral

tuple-index = "0" / nonzero-decimal-digit *( decimal-digit )
              ; i.e. decimal-numeral without extra leading zeros

unsigned-literal = numeral ( %s"u8" / %s"u16" / %s"u32" / %s"u64" / %s"u128" )

signed-literal = numeral ( %s"i8" / %s"i16" / %s"i32" / %s"i64" / %s"i128" )

integer-literal = unsigned-literal
                / signed-literal

field-literal = decimal-numeral %s"field"

product-group-literal = decimal-numeral %s"group"

scalar-literal = decimal-numeral %s"scalar"

numeric-literal = integer-literal
                / field-literal
                / product-group-literal
                / scalar-literal

boolean-literal = %s"true" / %s"false"

explicit-address-literal = %s"aleo1" 58( lowercase-letter / decimal-digit )

string-literal = double-quote *string-character double-quote

double-quote = %x22

string-character = not-double-quote-or-line-feed
                 / line-terminator

not-double-quote-or-line-feed = %x00-09
                              ; no line feed (%x0A)
                              / %x0B-21
                              ; no double quote (%x22)
                              / %x23-7F
                              / safe-nonascii

atomic-literal = numeric-literal
               / boolean-literal
               / explicit-address-literal
               / string-literal

annotation = "@" identifier

symbol = "!"
       / "&&" / "||"
       / "==" / "!="
       / "<" / "<=" / ">" / ">="
       / "&" / "|" / "^"
       / "<<" / ">>"
       / "+" / "-" / "*" / "/" / "%" / "**"
       / "="
       / "+=" / "-=" / "*=" / "/=" / "%=" / "**="
       / "<<=" / ">>="
       / "&=" / "|=" / "^="
       / "&&=" / "||="
       / "(" / ")"
       / "[" / "]"
       / "{" / "}"
       / "," / "." / ".." / ";" / ":" / "::" / "?"
       / "->" / "=>" / "_"
       / %s")group"

token = keyword
      / identifier
      / atomic-literal
      / decimal-numeral
      / annotation
      / symbol

lexeme = token / comment / whitespace

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

; Syntactic Grammar
; -----------------

program-id = identifier "." network

network = %s"aleo"

locator = program-id "/" identifier

unsigned-type = %s"u8" / %s"u16" / %s"u32" / %s"u64" / %s"u128"

signed-type = %s"i8" / %s"i16" / %s"i32" / %s"i64" / %s"i128"

integer-type = unsigned-type / signed-type

field-type = %s"field"

group-type = %s"group"

scalar-type = %s"scalar"

arithmetic-type = integer-type / field-type / group-type / scalar-type

boolean-type = %s"bool"

address-type = %s"address"

signature-type = %s"signature"

string-type = %s"string"

named-primitive-type = boolean-type
                     / arithmetic-type
                     / address-type
                     / signature-type
                     / string-type

unit-type = "(" ")"

primitive-type = named-primitive-type / unit-type

untyped-future-type = %s"Future"

typed-future-type =
    %s"Future" "<" %s"Fn" "(" [ type 1*( "," type ) [ "," ] ] ")" ">"

future-type = untyped-future-type / typed-future-type

tuple-type = "(" type 1*( "," type ) [ "," ] ")"

array-type = "[" type ";" decimal-numeral "]"

type = primitive-type
     / future-type
     / tuple-type
     / array-type

named-type = named-primitive-type
           / untyped-future-type
           / identifier

group-coordinate = ( [ "-" ] decimal-numeral ) / "+" / "-" / "_"

affine-group-literal = "(" group-coordinate "," group-coordinate %s")group"

program-address-literal = program-id

literal = atomic-literal
        / affine-group-literal
        / program-address-literal

group-literal = product-group-literal / affine-group-literal

address-literal = explicit-address-literal / program-address-literal

primary-expression = literal
                   / variable
                   / locator
                   / associated-constant
                   / "(" expression ")"
                   / free-function-call
                   / associated-function-call
                   / unit-expression
                   / tuple-expression
                   / array-expression
                   / struct-expression
                   / self-expression
                   / block-expression
                   / network-expression

variable = identifier

associated-constant = named-type "::" identifier

free-function-call = identifier function-arguments
                   / locator function-arguments

associated-function-call = named-type "::" identifier function-arguments

function-arguments = "(" [ expression *( "," expression ) [ "," ] ] ")"

unit-expression = "(" ")"

tuple-expression = "(" expression 1*( "," expression ) [ "," ] ")"

array-expression = "[" expression 1*( "," expression ) [ "," ] "]"

struct-expression = identifier "{" struct-component-initializer
                                   *( "," struct-component-initializer )
                                   [ "," ] "}"

struct-component-initializer = identifier
                             / identifier ":" expression

self-expression = %s"self" "." %s"address"
                / %s"self" "." %s"caller"
                / %s"self" "." %s"signer"

block-expression = %s"block" "." %s"height"

network-expression = %s"network" "." %s"id"

postfix-expression = primary-expression
                   / tuple-component-expression
                   / array-component-expression
                   / struct-component-expression
                   / method-call

tuple-component-expression = postfix-expression "." tuple-index

array-component-expression = postfix-expression "[" expression "]"

struct-component-expression = postfix-expression "." identifier

method-call = postfix-expression "." identifier function-arguments

unary-expression = postfix-expression
                 / "!" unary-expression
                 / "-" unary-expression

cast-expression = unary-expression
                / unary-expression %s"as" named-primitive-type

exponential-expression = cast-expression
                       / cast-expression "**" exponential-expression

multiplicative-expression = exponential-expression
                          / multiplicative-expression "*" exponential-expression
                          / multiplicative-expression "/" exponential-expression
                          / multiplicative-expression "%" exponential-expression

additive-expression = multiplicative-expression
                    / additive-expression "+" multiplicative-expression
                    / additive-expression "-" multiplicative-expression

shift-expression = additive-expression
                 / shift-expression "<<" additive-expression
                 / shift-expression ">>" additive-expression

conjunctive-expression = shift-expression
                       / conjunctive-expression "&" shift-expression

inclusive-disjunctive-expression =
      conjunctive-expression
    / inclusive-disjunctive-expression "|" conjunctive-expression

exclusive-disjunctive-expression =
      inclusive-disjunctive-expression
    / exclusive-disjunctive-expression "^" inclusive-disjunctive-expression

ordering-expression =
      exclusive-disjunctive-expression
    / exclusive-disjunctive-expression "<" exclusive-disjunctive-expression
    / exclusive-disjunctive-expression ">" exclusive-disjunctive-expression
    / exclusive-disjunctive-expression "<=" exclusive-disjunctive-expression
    / exclusive-disjunctive-expression ">=" exclusive-disjunctive-expression

equality-expression = ordering-expression
                    / ordering-expression "==" ordering-expression
                    / ordering-expression "!=" ordering-expression

conditional-conjunctive-expression =
      equality-expression
    / conditional-conjunctive-expression "&&" equality-expression

conditional-disjunctive-expression =
      conditional-conjunctive-expression
    / conditional-disjunctive-expression "||" conditional-conjunctive-expression

binary-expression = conditional-disjunctive-expression

conditional-ternary-expression = binary-expression
                               / binary-expression "?" expression ":" expression

expression = conditional-ternary-expression

statement = expression-statement
          / return-statement
          / variable-declaration
          / constant-declaration
          / conditional-statement
          / loop-statement
          / assignment-statement
          / block

expression-statement = expression ";"

block = "{" *statement "}"

return-statement = %s"return" [ expression ] ";"

variable-declaration = %s"let" identifier-or-identifiers ":" type
                       "=" expression ";"

constant-declaration = %s"const" identifier ":" type "=" expression ";"

identifier-or-identifiers = identifier
                          / "(" identifier 1*( "," identifier ) [ "," ] ")"

branch = %s"if" conditional-ternary-expression block

conditional-statement = branch
                      / branch %s"else" block
                      / branch %s"else" conditional-statement

loop-statement = %s"for" identifier ":" type
                 %s"in" expression ".." expression
                 block

assignment-operator = "="
                    / "+="
                    / "-="
                    / "*="
                    / "/="
                    / "%="
                    / "**="
                    / "<<="
                    / ">>="
                    / "&="
                    / "|="
                    / "^="
                    / "&&="
                    / "||="

assignment-statement = expression assignment-operator expression ";"

assert-statement = %s"assert" "(" expression ")" ";"
                 / %s"assert_eq" "(" expression "," expression ")" ";"
                 / %s"assert_neq" "(" expression "," expression ")" ";"

function-declaration = *annotation [ %s"async" ] function-kind identifier
                       "(" [ function-inputs ] ")"
                       [ "->" function-outputs ]
                       ( block / ";" )

function-kind = %s"function" / %s"transition" / %s"inline"

function-inputs = function-input *( "," function-input ) [ "," ]

function-input = [ %s"public" / %s"private" / %s"constant" ]
                 identifier ":" type

function-outputs = function-output
                 / "(" [ function-output *( "," function-output ) [ "," ] ] ")"

function-output = [ %s"public" / %s"private" / %s"constant" ] type

struct-declaration = %s"struct" identifier
                     "{" struct-component-declarations "}"

struct-component-declarations = struct-component-declaration
                                *( "," struct-component-declaration )
                                [ "," ]

struct-component-declaration =
    [ %s"public" / %s"private" / %s"constant" ] identifier ":" type

record-declaration = %s"record" identifier
                     "{" struct-component-declarations "}"

mapping-declaration = %s"mapping" identifier ":" type "=>" type ";"

program-item = function-declaration
             / struct-declaration
             / record-declaration
             / mapping-declaration
             / constant-declaration

program-declaration = %s"program" program-id "{" *program-item "}"

import-declaration = %s"import" program-id ";"

file = *import-declaration program-declaration