Disallow different characters after certain operators

corpus/expressions.txt

@@ -682,12 +682,16 @@ SignalProducer<(), CarthageError>.empty
 Suppressing the syntactic significance of a newline
 ===
 
-let _ = (1 + 2)
+let _ = (1+2)
     / 3
 
 let additionIsSane = (2 + 2)
     == 4
 
+let two = 2
+    + 2
+    - 2
+
 ---
 
 (source_file
@@ -710,4 +714,13 @@ let additionIsSane = (2 + 2)
         (additive_expression
           (integer_literal)
           (integer_literal)))
+          (integer_literal)))
+      (property_declaration
+        (value_binding_pattern
+          (non_binding_pattern
+            (simple_identifier)))
+        (additive_expression
+          (additive_expression
+            (integer_literal)
+            (integer_literal))
         (integer_literal))))

grammar.js

@@ -152,12 +152,15 @@ module.exports = grammar({
     $._nil_coalescing_operator,
     $._equal_sign,
     $._eq_eq,
+    $._plus_then_ws, // + symbol with whitespace after it
+    $._minus_then_ws, // - symbol with whitespace after it
     $._throws_keyword,
     $._rethrows_keyword,
     $.default_keyword,
     $._where_keyword,
     $._else,
     $._catch,
+    $._as,
     $._as_quest,
     $._as_bang,
   ],
@@ -834,11 +837,12 @@ module.exports = grammar({
 
     _is_operator: ($) => "is",
 
-    _additive_operator: ($) => choice("+", "-"),
+    _additive_operator: ($) =>
+      choice($._plus_then_ws, $._minus_then_ws, "+", "-"),
 
     _multiplicative_operator: ($) => choice("*", "/", "%"),
 
-    _as_operator: ($) => choice("as", $._as_quest, $._as_bang),
+    _as_operator: ($) => choice($._as, $._as_quest, $._as_bang),
 
     _prefix_unary_operator: ($) =>
       prec.right(
@@ -1634,7 +1638,11 @@ function generate_case_pattern($, allows_binding, force) {
 function generate_type_casting_pattern($, allows_binding) {
   return choice(
     seq("is", $._type),
-    seq(generate_pattern_matching_rule($, allows_binding, false), "as", $._type)
+    seq(
+      generate_pattern_matching_rule($, allows_binding, false),
+      $._as,
+      $._type
+    )
   );
 }
 

script-data/known_failures.txt

@@ -2,5 +2,3 @@ firefox-ios/Shared/Functions.swift
 RxSwift/RxExample/RxExample/Examples/GitHubSearchRepositories/GitHubSearchRepositories.swift
 RxSwift/Rx.playground/Pages/Transforming_Operators.xcplaygroundpage/Contents.swift
 SwiftLint/Source/SwiftLintFramework/Extensions/String+SwiftLint.swift
-SwiftLint/Source/SwiftLintFramework/Rules/Metrics/NestingRuleExamples.swift
-SwiftLint/Source/SwiftLintFramework/Rules/Style/StatementPositionRule.swift

scripts/top-repos.sh

@@ -99,10 +99,16 @@ while read line ; do
     cd $tmpdir
     checkout $line
     cd $parser_dir
-    validate $line &
-    pids+=($!)
+    if [ -z "$1" ]; then
+        validate $line &
+        pids+=($!)
+    else
+        validate $line
+    fi
 done <<<"$repos"
 
-for pid in "${pids[@]}" ; do
-    wait $pid
-done
+if [ -z "$1" ]; then
+    for pid in "${pids[@]}" ; do
+        wait $pid
+    done
+fi

src/scanner.c

@@ -16,17 +16,20 @@ enum TokenType {
     NIL_COALESCING_OPERATOR,
     EQUAL_SIGN,
     EQ_EQ,
+    PLUS_THEN_WS,
+    MINUS_THEN_WS,
     THROWS_KEYWORD,
     RETHROWS_KEYWORD,
     DEFAULT_KEYWORD,
     WHERE_KEYWORD,
     ELSE_KEYWORD,
     CATCH_KEYWORD,
+    AS_KEYWORD,
     AS_QUEST,
     AS_BANG
 };
 
-#define CROSS_SEMI_OPERATOR_COUNT 17
+#define CROSS_SEMI_OPERATOR_COUNT 20
 
 const char* CROSS_SEMI_OPERATORS[CROSS_SEMI_OPERATOR_COUNT] = {
     "->",
@@ -38,16 +41,49 @@ const char* CROSS_SEMI_OPERATORS[CROSS_SEMI_OPERATOR_COUNT] = {
     "??",
     "=",
     "==",
+    "+",
+    "-",
     "throws",
     "rethrows",
     "default",
     "where",
     "else",
     "catch",
+    "as",
     "as?",
     "as!"
 };
 
+enum IllegalTerminatorGroup {
+    ALPHANUMERIC,
+    OPERATOR_SYMBOLS,
+    OPERATOR_OR_DOT,
+    NON_WHITESPACE
+};
+
+const enum IllegalTerminatorGroup CROSS_SEMI_ILLEGAL_TERMINATORS[CROSS_SEMI_OPERATOR_COUNT] = {
+    OPERATOR_SYMBOLS, // ->
+    OPERATOR_OR_DOT,  // .
+    OPERATOR_OR_DOT,  // ...
+    OPERATOR_OR_DOT,  // ..<
+    OPERATOR_SYMBOLS, // &&
+    OPERATOR_SYMBOLS, // ||
+    OPERATOR_SYMBOLS, // ??
+    OPERATOR_SYMBOLS, // =
+    OPERATOR_SYMBOLS, // ==
+    NON_WHITESPACE,   // +
+    NON_WHITESPACE,   // -
+    ALPHANUMERIC,     // throws
+    ALPHANUMERIC,     // rethrows
+    ALPHANUMERIC,     // default
+    ALPHANUMERIC,     // where
+    ALPHANUMERIC,     // else
+    ALPHANUMERIC,     // catch
+    ALPHANUMERIC,     // as
+    OPERATOR_SYMBOLS, // as?
+    OPERATOR_SYMBOLS  // as!
+};
+
 const enum TokenType CROSS_SEMI_SYMBOLS[CROSS_SEMI_OPERATOR_COUNT] = {
     ARROW_OPERATOR,
     DOT_OPERATOR,
@@ -58,18 +94,21 @@ const enum TokenType CROSS_SEMI_SYMBOLS[CROSS_SEMI_OPERATOR_COUNT] = {
     NIL_COALESCING_OPERATOR,
     EQUAL_SIGN,
     EQ_EQ,
+    PLUS_THEN_WS,
+    MINUS_THEN_WS,
     THROWS_KEYWORD,
     RETHROWS_KEYWORD,
     DEFAULT_KEYWORD,
     WHERE_KEYWORD,
     ELSE_KEYWORD,
     CATCH_KEYWORD,
+    AS_KEYWORD,
     AS_QUEST,
     AS_BANG
 };
 
 #define NON_CONSUMING_CROSS_SEMI_CHAR_COUNT 3
-const char NON_CONSUMING_CROSS_SEMI_CHARS[CROSS_SEMI_OPERATOR_COUNT] = { '?', ':', '{' };
+const char NON_CONSUMING_CROSS_SEMI_CHARS[NON_CONSUMING_CROSS_SEMI_CHAR_COUNT] = { '?', ':', '{' };
 
 /**
  * All possible results of having performed some sort of parsing.
@@ -175,6 +214,8 @@ static bool eat_operators(
             }
 
             if (CROSS_SEMI_OPERATORS[op_idx][str_idx] == '\0') {
+                // Make sure that the operator is allowed to have the next character as its lookahead.
+                enum IllegalTerminatorGroup illegal_terminators = CROSS_SEMI_ILLEGAL_TERMINATORS[op_idx];
                 switch (lexer->lookahead) {
                 // See "Operators":
                 // https://docs.swift.org/swift-book/ReferenceManual/LexicalStructure.html#ID418
@@ -192,10 +233,22 @@ static bool eat_operators(
                 case '^':
                 case '?':
                 case '~':
-                    break;
+                    if (illegal_terminators == OPERATOR_SYMBOLS) {
+                        break;
+                    } // Otherwise, intentionally fall through to the OPERATOR_OR_DOT case
+                case '.':
+                    if (illegal_terminators == OPERATOR_OR_DOT) {
+                        break;
+                    } // Otherwise, fall through to DEFAULT which checks its groups directly
                 default:
-                    // Only match if this is the _end_ of an operator. If it's possible for a custom
-                    // operator to continue from here, don't treat it as a full match.
+                    if (iswalnum(lexer->lookahead) && illegal_terminators == ALPHANUMERIC) {
+                        break;
+                    }
+
+                    if (!iswspace(lexer->lookahead) && illegal_terminators == NON_WHITESPACE) {
+                        break;
+                    }
+
                     full_match = op_idx;
                     if (mark_end) {
                         lexer->mark_end(lexer);