From 35cc4391595553cb9695af8b03e7d622b20c5130 Mon Sep 17 00:00:00 2001
From: Amaan Qureshi <amaanq12@gmail.com>
Date: Wed, 31 Jan 2024 17:20:25 -0500
Subject: [PATCH] fix: format scanner and don't bother parsing during error
 recovery

---
 common/define-grammar.js |   1 +
 common/scanner.h         | 391 +++++++++++++++++++++------------------
 2 files changed, 215 insertions(+), 177 deletions(-)

diff --git a/common/define-grammar.js b/common/define-grammar.js
index ade7dd88..f1bd7220 100644
--- a/common/define-grammar.js
+++ b/common/define-grammar.js
@@ -6,6 +6,7 @@ module.exports = function defineGrammar(dialect) {
 
     externals: ($, previous) => previous.concat([
       $._function_signature_automatic_semicolon,
+      $.__error_recovery,
     ]),
 
     supertypes: ($, previous) => previous.concat([
diff --git a/common/scanner.h b/common/scanner.h
index 37ebc976..0662dd64 100644
--- a/common/scanner.h
+++ b/common/scanner.h
@@ -3,200 +3,234 @@
 #include <wctype.h>
 
 enum TokenType {
-  AUTOMATIC_SEMICOLON,
-  TEMPLATE_CHARS,
-  TERNARY_QMARK,
-  HTML_COMMENT,
-  LOGICAL_OR,
-  ESCAPE_SEQUENCE,
-  FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON,
+    AUTOMATIC_SEMICOLON,
+    TEMPLATE_CHARS,
+    TERNARY_QMARK,
+    HTML_COMMENT,
+    LOGICAL_OR,
+    ESCAPE_SEQUENCE,
+    FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON,
+    ERROR_RECOVERY,
 };
 
 static void advance(TSLexer *lexer) { lexer->advance(lexer, false); }
+
 static void skip(TSLexer *lexer) { lexer->advance(lexer, true); }
 
 static bool scan_template_chars(TSLexer *lexer) {
-  lexer->result_symbol = TEMPLATE_CHARS;
-  for (bool has_content = false;; has_content = true) {
-    lexer->mark_end(lexer);
-    switch (lexer->lookahead) {
-    case '`':
-      return has_content;
-    case '\0':
-      return false;
-    case '$':
-      advance(lexer);
-      if (lexer->lookahead == '{') return has_content;
-      break;
-    case '\\':
-      return has_content;
-    default:
-      advance(lexer);
+    lexer->result_symbol = TEMPLATE_CHARS;
+    for (bool has_content = false;; has_content = true) {
+        lexer->mark_end(lexer);
+        switch (lexer->lookahead) {
+            case '`':
+                return has_content;
+            case '\0':
+                return false;
+            case '$':
+                advance(lexer);
+                if (lexer->lookahead == '{') {
+                    return has_content;
+                }
+                break;
+            case '\\':
+                return has_content;
+            default:
+                advance(lexer);
+        }
     }
-  }
 }
 
 static bool scan_whitespace_and_comments(TSLexer *lexer, bool *scanned_comment) {
-  for (;;) {
-    while (iswspace(lexer->lookahead)) {
-      skip(lexer);
-    }
-
-    if (lexer->lookahead == '/') {
-      skip(lexer);
-
-      if (lexer->lookahead == '/') {
-        skip(lexer);
-        while (lexer->lookahead != 0 && lexer->lookahead != '\n') {
-          skip(lexer);
+    for (;;) {
+        while (iswspace(lexer->lookahead)) {
+            skip(lexer);
         }
-        *scanned_comment = true;
-      } else if (lexer->lookahead == '*') {
-        skip(lexer);
-        while (lexer->lookahead != 0) {
-          if (lexer->lookahead == '*') {
+
+        if (lexer->lookahead == '/') {
             skip(lexer);
+
             if (lexer->lookahead == '/') {
-              skip(lexer);
-              break;
+                skip(lexer);
+                while (lexer->lookahead != 0 && lexer->lookahead != '\n') {
+                    skip(lexer);
+                }
+                *scanned_comment = true;
+            } else if (lexer->lookahead == '*') {
+                skip(lexer);
+                while (lexer->lookahead != 0) {
+                    if (lexer->lookahead == '*') {
+                        skip(lexer);
+                        if (lexer->lookahead == '/') {
+                            skip(lexer);
+                            break;
+                        }
+                    } else {
+                        skip(lexer);
+                    }
+                }
+            } else {
+                return false;
             }
-          } else {
-            skip(lexer);
-          }
+        } else {
+            return true;
         }
-      } else {
-        return false;
-      }
-    } else {
-      return true;
     }
-  }
 }
 
 static bool scan_automatic_semicolon(TSLexer *lexer, const bool *valid_symbols, bool *scanned_comment) {
-  lexer->result_symbol = AUTOMATIC_SEMICOLON;
-  lexer->mark_end(lexer);
-
-  for (;;) {
-    if (lexer->lookahead == 0) return true;
-    if (lexer->lookahead == '}') {
-      // Automatic semicolon insertion breaks detection of object patterns
-      // in a typed context:
-      //   type F = ({a}: {a: number}) => number;
-      // Therefore, disable automatic semicolons when followed by typing
-      do {
+    lexer->result_symbol = AUTOMATIC_SEMICOLON;
+    lexer->mark_end(lexer);
+
+    for (;;) {
+        if (lexer->lookahead == 0) {
+            return true;
+        }
+        if (lexer->lookahead == '}') {
+            // Automatic semicolon insertion breaks detection of object patterns
+            // in a typed context:
+            //   type F = ({a}: {a: number}) => number;
+            // Therefore, disable automatic semicolons when followed by typing
+            do {
+                skip(lexer);
+            } while (iswspace(lexer->lookahead));
+            if (lexer->lookahead == ':') {
+                return false;
+            }
+            return true;
+        }
+        if (!iswspace(lexer->lookahead)) {
+            return false;
+        }
+        if (lexer->lookahead == '\n') {
+            break;
+        }
         skip(lexer);
-      } while (iswspace(lexer->lookahead));
-      if (lexer->lookahead == ':') return false;
-      return true;
     }
-    if (!iswspace(lexer->lookahead)) return false;
-    if (lexer->lookahead == '\n') break;
+
     skip(lexer);
-  }
-
-  skip(lexer);
-
-  if (!scan_whitespace_and_comments(lexer, scanned_comment)) return false;
-
-  switch (lexer->lookahead) {
-  case ',':
-  case '.':
-  case ';':
-  case '*':
-  case '%':
-  case '>':
-  case '<':
-  case '=':
-  case '?':
-  case '^':
-  case '|':
-  case '&':
-  case '/':
-  case ':':
-    return false;
 
-  case '{':
-    if (valid_symbols[FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON]) return false;
-    break;
+    if (!scan_whitespace_and_comments(lexer, scanned_comment)) {
+        return false;
+    }
 
-    // Don't insert a semicolon before a '[' or '(', unless we're parsing
-    // a type. Detect whether we're parsing a type or an expression using
-    // the validity of a binary operator token.
-  case '(':
-  case '[':
-    if (valid_symbols[LOGICAL_OR]) return false;
-    break;
+    switch (lexer->lookahead) {
+        case ',':
+        case '.':
+        case ';':
+        case '*':
+        case '%':
+        case '>':
+        case '<':
+        case '=':
+        case '?':
+        case '^':
+        case '|':
+        case '&':
+        case '/':
+        case ':':
+            return false;
+
+        case '{':
+            if (valid_symbols[FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON]) {
+                return false;
+            }
+            break;
+
+            // Don't insert a semicolon before a '[' or '(', unless we're parsing
+            // a type. Detect whether we're parsing a type or an expression using
+            // the validity of a binary operator token.
+        case '(':
+        case '[':
+            if (valid_symbols[LOGICAL_OR]) {
+                return false;
+            }
+            break;
 
-    // Insert a semicolon before `--` and `++`, but not before binary `+` or `-`.
-  case '+':
-    skip(lexer);
-    return lexer->lookahead == '+';
-  case '-':
-    skip(lexer);
-    return lexer->lookahead == '-';
+            // Insert a semicolon before `--` and `++`, but not before binary `+` or `-`.
+        case '+':
+            skip(lexer);
+            return lexer->lookahead == '+';
+        case '-':
+            skip(lexer);
+            return lexer->lookahead == '-';
 
-    // Don't insert a semicolon before `!=`, but do insert one before a unary `!`.
-  case '!':
-    skip(lexer);
-    return lexer->lookahead != '=';
+            // Don't insert a semicolon before `!=`, but do insert one before a unary `!`.
+        case '!':
+            skip(lexer);
+            return lexer->lookahead != '=';
 
-    // Don't insert a semicolon before `in` or `instanceof`, but do insert one
-    // before an identifier.
-  case 'i':
-    skip(lexer);
+            // Don't insert a semicolon before `in` or `instanceof`, but do insert one
+            // before an identifier.
+        case 'i':
+            skip(lexer);
 
-    if (lexer->lookahead != 'n') return true;
-    skip(lexer);
+            if (lexer->lookahead != 'n') {
+                return true;
+            }
+            skip(lexer);
 
-    if (!iswalpha(lexer->lookahead)) return false;
+            if (!iswalpha(lexer->lookahead)) {
+                return false;
+            }
 
-    for (unsigned i = 0; i < 8; i++) {
-      if (lexer->lookahead != "stanceof"[i]) return true;
-      skip(lexer);
-    }
+            for (unsigned i = 0; i < 8; i++) {
+                if (lexer->lookahead != "stanceof"[i]) {
+                    return true;
+                }
+                skip(lexer);
+            }
 
-    if (!iswalpha(lexer->lookahead)) return false;
-    break;
-  }
+            if (!iswalpha(lexer->lookahead)) {
+                return false;
+            }
+            break;
+    }
 
-  return true;
+    return true;
 }
 
 static bool scan_ternary_qmark(TSLexer *lexer) {
-  for(;;) {
-    if (!iswspace(lexer->lookahead)) break;
-    skip(lexer);
-  }
+    for (;;) {
+        if (!iswspace(lexer->lookahead)) {
+            break;
+        }
+        skip(lexer);
+    }
 
-  if (lexer->lookahead == '?') {
-    advance(lexer);
+    if (lexer->lookahead == '?') {
+        advance(lexer);
 
-    if (lexer->lookahead == '?') return false;
-    /* Optional chaining. */
-    if (lexer->lookahead == '.') return false;
+        /* Optional chaining. */
+        if (lexer->lookahead == '?' || lexer->lookahead == '.') {
+            return false;
+        }
 
-    lexer->mark_end(lexer);
-    lexer->result_symbol = TERNARY_QMARK;
+        lexer->mark_end(lexer);
+        lexer->result_symbol = TERNARY_QMARK;
 
-    /* TypeScript optional arguments contain the ?: sequence, possibly
-       with whitespace. */
-    for(;;) {
-      if (!iswspace(lexer->lookahead)) break;
-      advance(lexer);
-    }
-    if (lexer->lookahead == ':') return false;
-    if (lexer->lookahead == ')') return false;
-    if (lexer->lookahead == ',') return false;
-
-    if (lexer->lookahead == '.') {
-      advance(lexer);
-      if (iswdigit(lexer->lookahead)) return true;
-      return false;
+        /* TypeScript optional arguments contain the ?: sequence, possibly
+           with whitespace. */
+        for (;;) {
+            if (!iswspace(lexer->lookahead)) {
+                break;
+            }
+            advance(lexer);
+        }
+
+        if (lexer->lookahead == ':' || lexer->lookahead == ')' || lexer->lookahead == ',') {
+            return false;
+        }
+
+        if (lexer->lookahead == '.') {
+            advance(lexer);
+            if (iswdigit(lexer->lookahead)) {
+                return true;
+            }
+            return false;
+        }
+        return true;
     }
-    return true;
-  }
-  return false;
+    return false;
 }
 
 static bool scan_closing_comment(TSLexer *lexer) {
@@ -236,29 +270,32 @@ static bool scan_closing_comment(TSLexer *lexer) {
     return true;
 }
 
-
 static inline bool external_scanner_scan(void *payload, TSLexer *lexer, const bool *valid_symbols) {
-  if (valid_symbols[TEMPLATE_CHARS]) {
-    if (valid_symbols[AUTOMATIC_SEMICOLON]) return false;
-    return scan_template_chars(lexer);
-  } else if (
-    valid_symbols[AUTOMATIC_SEMICOLON] ||
-    valid_symbols[FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON]
-             ) {
-    bool scanned_comment = false;
-    bool ret = scan_automatic_semicolon(lexer, valid_symbols, &scanned_comment);
-    if (!ret && !scanned_comment && valid_symbols[TERNARY_QMARK] && lexer->lookahead == '?')
-      return scan_ternary_qmark(lexer);
-    return ret;
-  }
-  if (valid_symbols[TERNARY_QMARK]) {
-    return scan_ternary_qmark(lexer);
-  }
-
-  if (valid_symbols[HTML_COMMENT] && !valid_symbols[LOGICAL_OR] && !valid_symbols[ESCAPE_SEQUENCE]) {
-    return scan_closing_comment(lexer);
-  }
-
-  return false;
+    if (valid_symbols[ERROR_RECOVERY]) {
+        return false;
+    }
+
+    if (valid_symbols[TEMPLATE_CHARS]) {
+        if (valid_symbols[AUTOMATIC_SEMICOLON]) {
+            return false;
+        }
+        return scan_template_chars(lexer);
+    }
+    if (valid_symbols[AUTOMATIC_SEMICOLON] || valid_symbols[FUNCTION_SIGNATURE_AUTOMATIC_SEMICOLON]) {
+        bool scanned_comment = false;
+        bool ret = scan_automatic_semicolon(lexer, valid_symbols, &scanned_comment);
+        if (!ret && !scanned_comment && valid_symbols[TERNARY_QMARK] && lexer->lookahead == '?') {
+            return scan_ternary_qmark(lexer);
+        }
+        return ret;
+    }
+    if (valid_symbols[TERNARY_QMARK]) {
+        return scan_ternary_qmark(lexer);
+    }
 
+    if (valid_symbols[HTML_COMMENT] && !valid_symbols[LOGICAL_OR] && !valid_symbols[ESCAPE_SEQUENCE]) {
+        return scan_closing_comment(lexer);
+    }
+
+    return false;
 }