const PREC = { // See https://github.com/elixir-lang/elixir/blob/master/lib/elixir/src/elixir_parser.yrl IN_MATCH_OPS: 10, WHEN_OP: 20, TYPE_OP: 30, BAR_OP: 40, ASSOC_OP: 50, CAPTURE_OP: 60, MATCH_OP: 70, OR_OPS: 80, AND_OPS: 90, COMP_OPS: 100, REL_OPS: 110, ARROW_OPS: 120, IN_OPS: 130, XOR_OP: 140, TERNARY_OP: 150, CONCAT_OPS: 160, RANGE_OP: 160, ADD_OPS: 170, MULT_OPS: 180, POWER_OP: 190, UNARY_OPS: 200, ACCESS: 205, DOT_OP: 210, AT_OP: 220, CAPTURE_OPERAND: 235, }; const IN_MATCH_OPS = ["<-", "\\\\"]; const OR_OPS = ["||", "|||", "or"]; const AND_OPS = ["&&", "&&&", "and"]; const COMP_OPS = ["==", "!=", "=~", "===", "!=="]; const REL_OPS = ["<", ">", "<=", ">="]; const ARROW_OPS = ["|>", "<<<", ">>>", "<<~", "~>>", "<~", "~>", "<~>", "<|>"]; const IN_OPS = ["in", "not in"]; const CONCAT_OPS = ["++", "--", "+++", "---", "<>"]; const ADD_OPS = ["+", "-"]; const MULT_OPS = ["*", "/"]; const UNARY_OPS = ["+", "-", "!", "^", "~~~", "not"]; const ALL_OPS = [ ["->", "when", "::", "|", "=>", "&", "=", "^^^", "//", "..", "**", ".", "@"], IN_MATCH_OPS, OR_OPS, AND_OPS, COMP_OPS, REL_OPS, ARROW_OPS, IN_OPS, CONCAT_OPS, ADD_OPS, MULT_OPS, UNARY_OPS, ].flat(); // Ignore word literals and "=>" which is not a valid atom const ATOM_OPERATOR_LITERALS = ALL_OPS.filter( (operator) => !/[a-z]/.test(operator) && operator !== "=>" ); const ATOM_SPECIAL_LITERALS = ["...", "%{}", "{}", "%", "<<>>", "..//"]; // See Ref 6. in the docs const ATOM_WORD_LITERAL = /[\p{ID_Start}_][\p{ID_Continue}@]*[?!]?/u; // Word tokens used directly in the grammar const RESERVED_WORD_TOKENS = [ // Operators ["and", "in", "not", "or", "when"], // Literals ["true", "false", "nil"], // Other ["after", "catch", "do", "else", "end", "fn", "rescue"], ].flat(); const DIGITS = /[0-9]+/; const BIN_DIGITS = /[0-1]+/; const OCT_DIGITS = /[0-7]+/; const HEX_DIGITS = /[0-9a-fA-F]+/; const NUMBER_DEC = sep1(DIGITS, "_"); const NUMBER_BIN = seq("0b", sep1(BIN_DIGITS, "_")); const NUMBER_OCT = seq("0o", sep1(OCT_DIGITS, "_")); const NUMBER_HEX = seq("0x", sep1(HEX_DIGITS, "_")); const INTEGER = choice(NUMBER_DEC, NUMBER_BIN, NUMBER_OCT, NUMBER_HEX); const FLOAT_SCIENTIFIC_PART = seq(/[eE]/, optional(choice("-", "+")), INTEGER); const FLOAT = seq(NUMBER_DEC, ".", NUMBER_DEC, optional(FLOAT_SCIENTIFIC_PART)); const NEWLINE = /\r?\n/; module.exports = grammar({ name: "elixir", externals: ($) => [ // See Ref 1. in the docs $._quoted_content_i_single, $._quoted_content_i_double, $._quoted_content_i_heredoc_single, $._quoted_content_i_heredoc_double, $._quoted_content_i_parenthesis, $._quoted_content_i_curly, $._quoted_content_i_square, $._quoted_content_i_angle, $._quoted_content_i_bar, $._quoted_content_i_slash, $._quoted_content_single, $._quoted_content_double, $._quoted_content_heredoc_single, $._quoted_content_heredoc_double, $._quoted_content_parenthesis, $._quoted_content_curly, $._quoted_content_square, $._quoted_content_angle, $._quoted_content_bar, $._quoted_content_slash, // See Ref 2. in the docs $._newline_before_do, $._newline_before_binary_operator, $._newline_before_comment, // See Ref 3. in the docs $._before_unary_op, // See Ref 4. in the docs $._not_in, // See Ref 5. in the docs $._quoted_atom_start, ], extras: ($) => [ NEWLINE, /[ \t]|\r?\n|\\\r?\n/, $.comment, $._newline_before_comment, // Placing this directly in the binary operator rule leads // to conflicts, but we can place it here without any drawbacks. // If we detect binary operator and the previous line is not a // valid expression, it's a syntax error either way $._newline_before_binary_operator, ], conflicts: ($) => [ // Given `left • *`, `left` identifier can be either: // * expression in `left * right` // * call identifier in `left * / 2` [$._expression, $._local_call_without_parentheses], // Given `left • when`, `left` expression can be either: // * binary operator operand in `left when right` // * stab arguments item in `left when right ->` // // Given `arg1, left • when`, `left` expression can be either: // * binary operator operand in `arg1, left when right, arg3` // * stab arguments item in `arg1, left when right ->` [$.binary_operator, $._stab_clause_arguments_without_parentheses], // Given `((arg1, arg2 • ,`, `arg3` expression can be either: // * stab parenthesised arguments item in `((arg1, arg2, arg3) ->)` // * stab non-parenthesised arguments item in `((arg1, arg2, arg3 ->))` [ $._stab_clause_arguments_without_parentheses, $._stab_clause_arguments_with_parentheses, ], // Given `(-> • /`, stab can be either: // * stab clause operator in `(-> / / 2)` // * operator identifier in `(-> / 2)` [$.operator_identifier, $.stab_clause], // Given `& • /`, ampersand can be either: // * capture operator in `& / / 2` // * operator identifier in `& / 1` [$.unary_operator, $.operator_identifier], // Given `(arg -> expression • \n`, the newline could be either: // * terminator separating expressions in `(arg -> expression \n expression)` // * terminator separating clauses in `(arg -> expression \n arg -> expression)` [$.body], ], rules: { source: ($) => seq( optional($._terminator), optional( seq(sep1($._expression, $._terminator), optional($._terminator)) ) ), _terminator: ($) => // Right precedence, because we want to consume `;` after newlines if present prec.right(choice(seq(repeat(NEWLINE), ";"), repeat1(NEWLINE))), _expression: ($) => choice( $.block, $.identifier, $.alias, $.integer, $.float, $.char, $.boolean, $.nil, $._atom, $.string, $.charlist, $.sigil, $.list, $.tuple, $.bitstring, $.map, $._nullary_operator, $.unary_operator, $.binary_operator, $.dot, $.call, $.access_call, $.anonymous_function ), block: ($) => seq( "(", optional($._terminator), optional( choice( sep1(choice($.stab_clause), $._terminator), seq( sep1(choice($._expression), $._terminator), optional($._terminator) ) ) ), ")" ), identifier: ($) => choice( // See Ref 6. in the docs /[_\p{Ll}\p{Lm}\p{Lo}\p{Nl}\u1885\u1886\u2118\u212E\u309B\u309C][\p{ID_Continue}]*[?!]?/u, "..." ), alias: ($) => token(sep1(/[A-Z][_a-zA-Z0-9]*/, /\s*\.\s*/)), integer: ($) => token(INTEGER), float: ($) => token(FLOAT), char: ($) => /\?(.|\\.)/, boolean: ($) => choice("true", "false"), nil: ($) => "nil", _atom: ($) => choice($.atom, $.quoted_atom), atom: ($) => token( seq( ":", choice( ATOM_WORD_LITERAL, ...ATOM_OPERATOR_LITERALS, ...ATOM_SPECIAL_LITERALS ) ) ), quoted_atom: ($) => seq( alias($._quoted_atom_start, ":"), choice($._quoted_i_double, $._quoted_i_single) ), // Defines $._quoted_content_i_{name} and $._quoted_content_{name} rules, // content with and without interpolation respectively ...defineQuoted(`"`, `"`, "double"), ...defineQuoted(`'`, `'`, "single"), ...defineQuoted(`'''`, `'''`, "heredoc_single"), ...defineQuoted(`"""`, `"""`, "heredoc_double"), ...defineQuoted(`(`, `)`, "parenthesis"), ...defineQuoted(`{`, `}`, "curly"), ...defineQuoted(`[`, `]`, "square"), ...defineQuoted(`<`, `>`, "angle"), ...defineQuoted(`|`, `|`, "bar"), ...defineQuoted(`/`, `/`, "slash"), string: ($) => choice($._quoted_i_double, $._quoted_i_heredoc_double), charlist: ($) => choice($._quoted_i_single, $._quoted_i_heredoc_single), interpolation: ($) => seq("#{", optional($._expression), "}"), escape_sequence: ($) => token( seq( "\\", choice( // Single escaped character /[^ux]/, // Hex byte /x[0-9a-fA-F]{1,2}/, /x\{[0-9a-fA-F]+\}/, // Unicode code point /u\{[0-9a-fA-F]+\}/, /u[0-9a-fA-F]{4}/ ) ) ), sigil: ($) => seq( "~", choice( seq( alias(token.immediate(/[a-z]/), $.sigil_name), choice( $._quoted_i_double, $._quoted_i_single, $._quoted_i_heredoc_single, $._quoted_i_heredoc_double, $._quoted_i_parenthesis, $._quoted_i_curly, $._quoted_i_square, $._quoted_i_angle, $._quoted_i_bar, $._quoted_i_slash ) ), seq( alias(token.immediate(/[A-Z][A-Z0-9]*/), $.sigil_name), choice( $._quoted_double, $._quoted_single, $._quoted_heredoc_single, $._quoted_heredoc_double, $._quoted_parenthesis, $._quoted_curly, $._quoted_square, $._quoted_angle, $._quoted_bar, $._quoted_slash ) ) ), optional(alias(token.immediate(/[a-zA-Z0-9]+/), $.sigil_modifiers)) ), keywords: ($) => // Right precedence, because we want to consume next items as long // as there is a comma ahead prec.right(sep1($.pair, ",")), _keywords_with_trailing_separator: ($) => seq(sep1($.pair, ","), optional(",")), pair: ($) => seq(field("key", $._keyword), field("value", $._expression)), _keyword: ($) => choice($.keyword, $.quoted_keyword), keyword: ($) => // See Ref 7. in the docs token( seq( choice( ATOM_WORD_LITERAL, ...ATOM_OPERATOR_LITERALS.filter((op) => op !== "::"), ...ATOM_SPECIAL_LITERALS ), /:\s/ ) ), quoted_keyword: ($) => seq( choice($._quoted_i_double, $._quoted_i_single), token.immediate(/:\s/) ), list: ($) => seq("[", optional($._items_with_trailing_separator), "]"), tuple: ($) => seq("{", optional($._items_with_trailing_separator), "}"), bitstring: ($) => seq("<<", optional($._items_with_trailing_separator), ">>"), map: ($) => // Precedence over tuple prec( 1, seq( "%", optional($.struct), "{", optional(alias($._items_with_trailing_separator, $.map_content)), "}" ) ), struct: ($) => // Left precedence, because if there is a conflict involving `{}`, // we want to treat it as map continuation rather than tuple prec.left( choice( $.alias, $._atom, $.identifier, $.unary_operator, $.dot, alias($._call_with_parentheses, $.call) ) ), _items_with_trailing_separator: ($) => seq( choice( seq(sep1($._expression, ","), optional(",")), seq( optional(seq(sep1($._expression, ","), ",")), alias($._keywords_with_trailing_separator, $.keywords) ) ) ), _nullary_operator: ($) => // Nullary operators don't have any child nodes, so we reuse the // operator_identifier node alias(prec(PREC.RANGE_OP, ".."), $.operator_identifier), unary_operator: ($) => choice( unaryOp($, prec, PREC.CAPTURE_OP, "&", $._capture_expression), unaryOp($, prec, PREC.UNARY_OPS, choice(...UNARY_OPS)), unaryOp($, prec, PREC.AT_OP, "@"), // Capture operand like &1 is a special case with higher precedence unaryOp($, prec, PREC.CAPTURE_OPERAND, "&", $.integer) ), _capture_expression: ($) => choice( // Note that block expression is not allowed as capture operand, // so we have an explicit sequence with the parentheses and higher // precedence prec(1, seq("(", $._expression, ")")), $._expression ), binary_operator: ($) => choice( binaryOp($, prec.left, PREC.IN_MATCH_OPS, choice(...IN_MATCH_OPS)), binaryOp( $, prec.right, PREC.WHEN_OP, "when", $._expression, choice($._expression, $.keywords) ), binaryOp($, prec.right, PREC.TYPE_OP, "::"), binaryOp( $, prec.right, PREC.BAR_OP, "|", $._expression, choice($._expression, $.keywords) ), binaryOp($, prec.right, PREC.ASSOC_OP, "=>"), binaryOp($, prec.right, PREC.MATCH_OP, "="), binaryOp($, prec.left, PREC.OR_OPS, choice(...OR_OPS)), binaryOp($, prec.left, PREC.AND_OPS, choice(...AND_OPS)), binaryOp($, prec.left, PREC.COMP_OPS, choice(...COMP_OPS)), binaryOp($, prec.left, PREC.REL_OPS, choice(...REL_OPS)), binaryOp($, prec.left, PREC.ARROW_OPS, choice(...ARROW_OPS)), binaryOp( $, prec.left, PREC.IN_OPS, choice("in", alias($._not_in, "not in")) ), binaryOp($, prec.left, PREC.XOR_OP, "^^^"), binaryOp($, prec.right, PREC.TERNARY_OP, "//"), binaryOp($, prec.right, PREC.CONCAT_OPS, choice(...CONCAT_OPS)), binaryOp($, prec.right, PREC.RANGE_OP, ".."), binaryOp($, prec.left, PREC.ADD_OPS, choice(...ADD_OPS)), binaryOp($, prec.left, PREC.MULT_OPS, choice(...MULT_OPS)), binaryOp($, prec.left, PREC.POWER_OP, "**"), // Operator with arity binaryOp( $, prec.left, PREC.MULT_OPS, "/", $.operator_identifier, $.integer ) ), operator_identifier: ($) => // Operators with the following changes: // // * exclude "=>" since it's not a valid operator identifier // * exclude // since it's only valid after .. // * exclude binary "-" and "+" as they are handled as unary below // // For unary operator identifiers we use the same precedence as // operators, so that we get conflicts and resolve them dynamically // (see grammar.conflicts for more details) choice( // Unary operators prec(PREC.CAPTURE_OP, "&"), prec(PREC.UNARY_OPS, choice(...UNARY_OPS)), prec(PREC.AT_OP, "@"), // Binary operators ...IN_MATCH_OPS, "when", "::", "|", "=", ...OR_OPS, ...AND_OPS, ...COMP_OPS, ...REL_OPS, ...ARROW_OPS, "in", alias($._not_in, "not in"), "^^^", ...CONCAT_OPS, // The range operator has both a binary and a nullary version. // The nullary version is already parsed as operator_identifier, // so it covers this case // ".." ...MULT_OPS, "**", "->" ), dot: ($) => prec( PREC.DOT_OP, seq( field("left", $._expression), field("operator", "."), field("right", choice($.alias, $.tuple)) ) ), call: ($) => choice($._call_without_parentheses, $._call_with_parentheses), _call_without_parentheses: ($) => choice( $._local_call_without_parentheses, $._local_call_just_do_block, $._remote_call_without_parentheses ), _call_with_parentheses: ($) => choice( $._local_call_with_parentheses, $._remote_call_with_parentheses, $._anonymous_call, $._double_call ), // Note, calls have left precedence, so that `do end` block sticks to // the outermost call _local_call_without_parentheses: ($) => prec.left( seq( field("target", $.identifier), alias($._call_arguments_without_parentheses, $.arguments), optional(seq(optional($._newline_before_do), $.do_block)) ) ), _local_call_with_parentheses: ($) => prec.left( seq( field("target", $.identifier), alias($._call_arguments_with_parentheses_immediate, $.arguments), optional(seq(optional($._newline_before_do), $.do_block)) ) ), _local_call_just_do_block: ($) => // Lower precedence than identifier, because `foo bar do` is `foo(bar) do end` prec(-1, seq(field("target", $.identifier), $.do_block)), _remote_call_without_parentheses: ($) => prec.left( seq( field("target", alias($._remote_dot, $.dot)), optional(alias($._call_arguments_without_parentheses, $.arguments)), optional(seq(optional($._newline_before_do), $.do_block)) ) ), _remote_call_with_parentheses: ($) => prec.left( seq( field("target", alias($._remote_dot, $.dot)), alias($._call_arguments_with_parentheses_immediate, $.arguments), optional(seq(optional($._newline_before_do), $.do_block)) ) ), _remote_dot: ($) => prec( PREC.DOT_OP, seq( field("left", $._expression), field("operator", "."), field( "right", choice( $.identifier, alias(choice(...RESERVED_WORD_TOKENS), $.identifier), $.operator_identifier, alias($._quoted_i_double, $.string), alias($._quoted_i_single, $.charlist) ) ) ) ), _anonymous_call: ($) => seq( field("target", alias($._anonymous_dot, $.dot)), alias($._call_arguments_with_parentheses, $.arguments) ), _anonymous_dot: ($) => prec( PREC.DOT_OP, seq(field("left", $._expression), field("operator", ".")) ), _double_call: ($) => prec.left( seq( field( "target", alias( choice( $._local_call_with_parentheses, $._remote_call_with_parentheses, $._anonymous_call ), $.call ) ), alias($._call_arguments_with_parentheses, $.arguments), optional(seq(optional($._newline_before_do), $.do_block)) ) ), _call_arguments_with_parentheses: ($) => seq("(", optional($._call_arguments_with_trailing_separator), ")"), _call_arguments_with_parentheses_immediate: ($) => seq( token.immediate("("), optional($._call_arguments_with_trailing_separator), ")" ), _call_arguments_with_trailing_separator: ($) => choice( seq( sep1($._expression, ","), optional( seq(",", alias($._keywords_with_trailing_separator, $.keywords)) ) ), alias($._keywords_with_trailing_separator, $.keywords) ), _call_arguments_without_parentheses: ($) => // In stab clauses a newline can either separate multiple body expressions // or multiple stab clauses, this falls under the $.body conflict. Given a // multiline stab clause with trailing identifier like `1 -> 1 \n x \n 2 -> x`, // there are two matching interpretations: // * `x` as identifier and `2` as stab argument // * `x 2` call as stab argument // Similarly for `Mod.fun` or `mod.fun` the newline should terminate the call. // Consequently, we reject the second interpretation using dynamic precedence prec.dynamic( -1, // Right precedence, because `fun1 fun2 x, y` is `fun1(fun2(x, y))` prec.right( choice( seq(sep1($._expression, ","), optional(seq(",", $.keywords))), $.keywords ) ) ), do_block: ($) => seq( callKeywordBlock($, "do"), repeat( choice($.after_block, $.rescue_block, $.catch_block, $.else_block) ), "end" ), after_block: ($) => callKeywordBlock($, "after"), rescue_block: ($) => callKeywordBlock($, "rescue"), catch_block: ($) => callKeywordBlock($, "catch"), else_block: ($) => callKeywordBlock($, "else"), access_call: ($) => prec( PREC.ACCESS, seq( field("target", $._expression), token.immediate("["), field("key", $._expression), "]" ) ), stab_clause: ($) => // Right precedence, because we want to consume body if any prec.right( seq( optional(field("left", $._stab_clause_left)), field("operator", "->"), optional(field("right", $.body)) ) ), _stab_clause_left: ($) => choice( alias($._stab_clause_arguments_with_parentheses, $.arguments), alias( $._stab_clause_arguments_with_parentheses_with_guard, $.binary_operator ), alias($._stab_clause_arguments_without_parentheses, $.arguments), alias( $._stab_clause_arguments_without_parentheses_with_guard, $.binary_operator ) ), _stab_clause_arguments_with_parentheses: ($) => // Precedence over block expression prec( 1, seq( "(", optional( choice( seq( // We need the same expression precedence as below, so that we don't // discard this rule in favour of the one below. We use right precedence, // because in this case we can consume expression until the next comma sep1(prec.right(PREC.WHEN_OP, $._expression), ","), optional(seq(",", $.keywords)) ), $.keywords ) ), ")" ) ), _stab_clause_arguments_without_parentheses: ($) => // We give the arguments and expression the same precedence as "when" // binary operator, so that we get conflicts and resolve them dynamically // (see the grammar.conflicts for more details) prec( PREC.WHEN_OP, choice( seq( sep1(prec(PREC.WHEN_OP, $._expression), ","), optional(seq(",", $.keywords)) ), $.keywords ) ), _stab_clause_arguments_with_parentheses_with_guard: ($) => seq( field( "left", alias($._stab_clause_arguments_with_parentheses, $.arguments) ), field("operator", "when"), field("right", $._expression) ), _stab_clause_arguments_without_parentheses_with_guard: ($) => // Given `a when b ->`, the left stab operand can be interpreted either // as a single argument item, or as binary operator with arguments on // the left and guard expression on the right. Using dynamic precedence // we favour the latter interpretation during dynamic conflict resolution prec.dynamic( 1, seq( field( "left", alias($._stab_clause_arguments_without_parentheses, $.arguments) ), field("operator", "when"), field("right", $._expression) ) ), body: ($) => choice( $._terminator, seq( optional($._terminator), sep1($._expression, $._terminator), optional($._terminator) ) ), anonymous_function: ($) => seq( "fn", optional($._terminator), sep1($.stab_clause, $._terminator), "end" ), // A comment may be anywhere, we give it a lower precedence, // so it doesn't intercept interpolation comment: ($) => token(prec(-1, seq("#", /.*/))), }, }); function sep1(rule, separator) { return seq(rule, repeat(seq(separator, rule))); } function unaryOp($, assoc, precedence, operator, right = null) { // Expression such as `x + y` falls under the "expression vs local call" // conflict that we already have. By using dynamic precedence we penalize // unary operator, so `x + y` is interpreted as binary operator (unless // _before_unary_op is tokenized and forces unary operator interpretation) return prec.dynamic( -1, assoc( precedence, seq( optional($._before_unary_op), field("operator", operator), field("operand", right || $._expression) ) ) ); } function binaryOp($, assoc, precedence, operator, left = null, right = null) { return assoc( precedence, seq( field("left", left || $._expression), field("operator", operator), field("right", right || $._expression) ) ); } function callKeywordBlock($, start) { return seq( start, optional($._terminator), optional( choice( sep1(choice($.stab_clause), $._terminator), seq(sep1(choice($._expression), $._terminator), optional($._terminator)) ) ) ); } function defineQuoted(start, end, name) { return { [`_quoted_i_${name}`]: ($) => seq( field("quoted_start", start), optional(alias($[`_quoted_content_i_${name}`], $.quoted_content)), repeat( seq( choice($.interpolation, $.escape_sequence), optional(alias($[`_quoted_content_i_${name}`], $.quoted_content)) ) ), field("quoted_end", end) ), [`_quoted_${name}`]: ($) => seq( field("quoted_start", start), optional(alias($[`_quoted_content_${name}`], $.quoted_content)), repeat( seq( // The end delimiter may be escaped in non-interpolating strings too $.escape_sequence, optional(alias($[`_quoted_content_${name}`], $.quoted_content)) ) ), field("quoted_end", end) ), }; }