@lvce-editor/shared-process 0.11.12 → 0.12.1

package/extensions/builtin.language-basics-jsx/src/tokenizeJsx.js

@@ -0,0 +1,648 @@
+/**
+ * @enum number
+ */
+const State = {
+  None: 0,
+  TopLevelContent: 1,
+  AfterOpeningAngleBracket: 2,
+  InsideOpeningTag: 3,
+  InsideOpeningTagAndHasSeenWhitespace: 4,
+  AfterClosingTagAngleBrackets: 5,
+  AfterClosingTagName: 6,
+  AfterAttributeName: 7,
+  AfterAttributeEqualSign: 8,
+  InsideAttributeDoubleQuote: 9,
+  InsideBlockComment: 10,
+  InsideString: 11,
+  AfterKeyword: 12,
+  Keyword: 13,
+  InsideLineComment: 14,
+  InsideSingleQuoteString: 15,
+  InsideDoubleQuoteString: 16,
+  InsideBacktickString: 17,
+  AfterPropertyDot: 18,
+  InsideTag: 19,
+}
+
+/**
+ * @enum number
+ */
+export const TokenType = {
+  None: 99999999,
+  Numeric: 30,
+  String: 50,
+  Whitespace: 777,
+  Comment: 60,
+  Text: 117,
+  PunctuationTag: 228,
+  TagName: 118,
+  AttributeName: 119,
+  Punctuation: 10,
+  Error: 141,
+  PunctuationString: 11,
+  NewLine: 771,
+  Keyword: 951,
+  VariableName: 952,
+  LanguageConstant: 13,
+  Regex: 14,
+  KeywordImport: 215,
+  KeywordControl: 881,
+  KeywordModifier: 882,
+  KeywordReturn: 883,
+  KeywordNew: 884,
+  FunctionName: 885,
+  KeywordThis: 886,
+  KeywordOperator: 887,
+  KeywordFunction: 889,
+}
+
+export const TokenMap = {
+  [TokenType.None]: 'None',
+  [TokenType.Numeric]: 'Numeric',
+  [TokenType.String]: 'String',
+  [TokenType.Whitespace]: 'Whitespace',
+  [TokenType.Comment]: 'Comment',
+  [TokenType.Text]: 'Text',
+  [TokenType.PunctuationTag]: 'PunctuationTag',
+  [TokenType.TagName]: 'TagName',
+  [TokenType.AttributeName]: 'AttributeName',
+  [TokenType.Punctuation]: 'Punctuation',
+  [TokenType.Error]: 'Error',
+  [TokenType.PunctuationString]: 'PunctuationString',
+  [TokenType.NewLine]: 'NewLine',
+  [TokenType.Keyword]: 'Keyword',
+  [TokenType.VariableName]: 'VariableName',
+  [TokenType.LanguageConstant]: 'LanguageConstant',
+  [TokenType.Regex]: 'Regex',
+  [TokenType.KeywordImport]: 'KeywordImport',
+  [TokenType.KeywordControl]: 'KeywordControl',
+  [TokenType.KeywordModifier]: 'KeywordModifier',
+  [TokenType.KeywordReturn]: 'KeywordReturn',
+  [TokenType.KeywordNew]: 'KeywordNew',
+  [TokenType.FunctionName]: 'Function',
+  [TokenType.KeywordThis]: 'KeywordThis',
+  [TokenType.KeywordOperator]: 'KeywordOperator',
+  [TokenType.KeywordFunction]: 'KeywordFunction',
+}
+
+const RE_KEYWORD =
+  /^(?:as|async|await|break|case|catch|class|const|continue|debugger|default|delete|do|else|enum|export|extends|false|finally|for|from|function|if|implements|import|in|Infinity|instanceof|interface|let|new|null|of|package|private|protected|public|return|super|switch|static|this|throw|try|true|typeof|undefined|var|void|while|with|yield)\b/
+const RE_CURLY_OPEN = /^\{/
+const RE_CURLY_CLOSE = /^\}/
+const RE_SQUARE_OPEN = /^\[/
+const RE_SQUARE_CLOSE = /^\]/
+const RE_COMMA = /^,/
+const RE_COLON = /^:/
+const RE_NUMERIC =
+  /^((0(x|X)[0-9a-fA-F]*)|(([0-9]+\.?[0-9]*)|(\.[0-9]+))((e|E)(\+|-)?[0-9]+)?)\b/
+const RE_NUMERIC_OCTAL = /0(?:o|O)?[0-7][0-7_]*(n)?\b/
+const RE_LINE_COMMENT_START = /^\/\//
+const RE_LINE_COMMENT_CONTENT = /^[^\n]+/
+const RE_NEWLINE_WHITESPACE = /^\n\s*/
+const RE_BLOCK_COMMENT_START = /^\/\*/
+const RE_BLOCK_COMMENT_CONTENT = /^.+?(?=\*\/)/
+const RE_BLOCK_COMMENT_END = /^\*\//
+const RE_UNKNOWN_VALUE = /^[^\}\{\s,"]+/
+const RE_IMPORT = /^[a-zA-Z\.]+/
+const RE_SEMICOLON = /^;/
+const RE_VARIABLE_NAME = /^[a-zA-Z_$][a-zA-Z\d\_]*/
+const RE_LINE_COMMENT = /^\/\/[^\n]*/
+const RE_ROUND_OPEN = /^\(/
+const RE_ROUND_CLOSE = /^\)/
+const RE_DOT = /^\./
+const RE_EQUAL_SIGN = /^=/
+const RE_SINGLE_QUOTE = /^'/
+const RE_PUNCTUATION = /^[\(\)=\+\-><\.:;\{\}\[\]!,&\|\^\?\*%~]/
+const RE_SLASH = /^\//
+const RE_ANYTHING_UNTIL_END = /^.+/s
+const RE_WHITESPACE = /^\s+/
+const RE_QUOTE_SINGLE = /^'/
+const RE_QUOTE_DOUBLE = /^"/
+const RE_QUOTE_BACKTICK = /^`/
+const RE_STRING_SINGLE_QUOTE_CONTENT = /^[^'\\]+/
+const RE_STRING_DOUBLE_QUOTE_CONTENT = /^[^"\\]+/
+const RE_STRING_BACKTICK_QUOTE_CONTENT = /^[^`\\\$]+/
+const RE_STRING_ESCAPE = /^\\./
+const RE_SHEBANG = /^#!.*/
+const RE_FUNCTION_CALL_NAME = /^[\w]+(?=\s*(\(|\=\s*function|\=\s*\())/
+const RE_DECORATOR = /^@\w+/
+const RE_BACKSLASH = /^\\/
+const RE_DOLLAR_CURLY_OPEN = /^\$\{/
+const RE_DOLLAR = /^\$/
+const RE_ANYTHING = /^.+/u
+const RE_ANGLE_BRACKET_OPEN_TAG = /^<(?![\s!<=])/
+const RE_ANGLE_BRACKET_CLOSE_TAG = /^<\/(?![\s!])/
+const RE_TAGNAME = /^[!\w\:\.]+/
+const RE_ANGLE_BRACKET_CLOSE = /^>/
+const RE_ANGLE_BRACKET_ONLY = /^</
+const RE_ANGLE_BRACKET_OPEN = /^</
+const RE_ANY_TEXT = /^[^\n]+/
+const RE_PUNCTUATION_SELF_CLOSING = /^\/>/
+const RE_TAG_TEXT = /^[^\s>]+/
+const RE_ATTRIBUTE_NAME = /^[a-zA-Z\d\-\:\_]+/
+const RE_DOUBLE_QUOTE = /^"/
+const RE_ATTRIBUTE_VALUE_UNQUOTED = /^[^<>\s]+/
+
+// copied from https://github.com/PrismJS/prism/blob/master/components/prism-jsx.js#L57
+const RE_REGEX =
+  /^((?:^|[^$\w\xA0-\uFFFF."'\])\s]|\b(?:return|yield))\s*)\/(?:\[(?:[^\]\\\r\n]|\\.)*\]|\\.|[^/\\\[\r\n])+\/[dgimyus]{0,7}(?=(?:\s|\/\*(?:[^*]|\*(?!\/))*\*\/)*(?:$|[\r\n,.;:})\]]|\/\/))/
+
+const RE_VARIABLE_NAME_SPECIAL = /\p{L}/u
+const RE_VARIABLE_NAME_SPECIAL_2 = /./u
+const RE_SELF_CLOSING = /^\/>/
+const RE_TEXT = /^[^<>\n\{\}\)]+/
+const RE_EMPTY_FRAGMENT = /^<\s*\>/
+const RE_VAR_CONST_LET_EXPORT =
+  /^\s*(?=(?:if|class|export|var|const|let)(?:\s+|$))/
+
+export const initialLineState = {
+  state: State.TopLevelContent,
+  /**
+   * @type {any[]}
+   */
+  stack: [],
+  tag: '',
+}
+
+export const hasArrayReturn = true
+
+/**
+ *
+ * @param {string} line
+ * @param {*} lineState
+ * @returns
+ */
+export const tokenizeLine = (line, lineState) => {
+  let next = null
+  let index = 0
+  let tokens = []
+  let token = TokenType.None
+  let state = lineState.state
+  let stack = [...lineState.stack]
+  let tag = lineState.tag
+
+  while (index < line.length) {
+    const part = line.slice(index)
+    switch (state) {
+      case State.TopLevelContent:
+        if ((next = part.match(RE_WHITESPACE))) {
+          token = TokenType.Whitespace
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_KEYWORD))) {
+          switch (next[0]) {
+            case 'true':
+            case 'false':
+            case 'null':
+            case 'undefined':
+              token = TokenType.LanguageConstant
+              break
+            case 'import':
+            case 'export':
+            case 'from':
+              token = TokenType.KeywordImport
+              break
+            case 'as':
+            case 'switch':
+            case 'default':
+            case 'case':
+            case 'else':
+            case 'if':
+            case 'break':
+            case 'throw':
+            case 'for':
+            case 'try':
+            case 'catch':
+            case 'finally':
+            case 'continue':
+            case 'while':
+              token = TokenType.KeywordControl
+              break
+            case 'async':
+            case 'await':
+              token = TokenType.KeywordModifier
+              break
+            case 'return':
+              token = TokenType.KeywordReturn
+              break
+            case 'new':
+              token = TokenType.KeywordNew
+              break
+            case 'this':
+              token = TokenType.KeywordThis
+              break
+            case 'delete':
+            case 'typeof':
+            case 'in':
+              token = TokenType.KeywordOperator
+              break
+            case 'function':
+              token = TokenType.KeywordFunction
+              break
+            case 'Infinity':
+              token = TokenType.Numeric
+              state = State.TopLevelContent
+              break
+            case 'of':
+              token = TokenType.KeywordOperator
+              break
+            default:
+              token = TokenType.Keyword
+              break
+          }
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_FUNCTION_CALL_NAME))) {
+          token = TokenType.FunctionName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_VARIABLE_NAME))) {
+          token = TokenType.VariableName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_NUMERIC))) {
+          token = TokenType.Numeric
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ANGLE_BRACKET_OPEN_TAG))) {
+          token = TokenType.PunctuationTag
+          state = State.AfterOpeningAngleBracket
+        } else if ((next = part.match(RE_CURLY_CLOSE))) {
+          token = TokenType.Punctuation
+          state = stack.pop() || State.TopLevelContent
+        } else if ((next = part.match(RE_CURLY_OPEN))) {
+          token = TokenType.Punctuation
+          stack.push(State.TopLevelContent)
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_PUNCTUATION))) {
+          token = TokenType.Punctuation
+          if (next[0] === '.') {
+            state = State.AfterPropertyDot
+          } else if (next[0] === '}') {
+            state = stack.pop() || State.TopLevelContent
+          } else {
+            state = State.TopLevelContent
+          }
+        } else if ((next = part.match(RE_SLASH))) {
+          if ((next = part.match(RE_BLOCK_COMMENT_START))) {
+            token = TokenType.Comment
+            state = State.InsideBlockComment
+          } else if ((next = part.match(RE_LINE_COMMENT_START))) {
+            token = TokenType.Comment
+            state = State.InsideLineComment
+          } else if ((next = part.match(RE_REGEX))) {
+            token = TokenType.Regex
+            state = State.TopLevelContent
+          } else {
+            next = part.match(RE_SLASH)
+            token = TokenType.Punctuation
+            state = State.TopLevelContent
+          }
+        } else if ((next = part.match(RE_QUOTE_SINGLE))) {
+          token = TokenType.Punctuation
+          state = State.InsideSingleQuoteString
+        } else if ((next = part.match(RE_QUOTE_DOUBLE))) {
+          token = TokenType.Punctuation
+          stack.push(state)
+          state = State.InsideDoubleQuoteString
+        } else if ((next = part.match(RE_QUOTE_BACKTICK))) {
+          token = TokenType.Punctuation
+          state = State.InsideBacktickString
+        } else if ((next = part.match(RE_NUMERIC_OCTAL))) {
+          token = TokenType.Numeric
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_SHEBANG))) {
+          token = TokenType.Comment
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_DECORATOR))) {
+          token = TokenType.VariableName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_VARIABLE_NAME_SPECIAL))) {
+          token = TokenType.VariableName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ANYTHING))) {
+          token = TokenType.Text
+          state = State.TopLevelContent
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.Keyword:
+        const keyword = next[0]
+        switch (keyword) {
+          case 'true':
+          case 'false':
+            break
+          default:
+            token = TokenType.Keyword
+            state = State.TopLevelContent
+            break
+        }
+        break
+      case State.InsideLineComment:
+        if ((next = part.match(RE_ANYTHING_UNTIL_END))) {
+          token = TokenType.Comment
+          state = State.TopLevelContent
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.InsideBlockComment:
+        if ((next = part.match(RE_BLOCK_COMMENT_END))) {
+          token = TokenType.Comment
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_BLOCK_COMMENT_CONTENT))) {
+          token = TokenType.Comment
+          state = State.InsideBlockComment
+        } else if ((next = part.match(RE_ANYTHING_UNTIL_END))) {
+          token = TokenType.Comment
+          state = State.InsideBlockComment
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.InsideSingleQuoteString:
+        if ((next = part.match(RE_QUOTE_SINGLE))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_STRING_SINGLE_QUOTE_CONTENT))) {
+          token = TokenType.String
+          state = State.InsideSingleQuoteString
+        } else if ((next = part.match(RE_STRING_ESCAPE))) {
+          token = TokenType.String
+          state = State.InsideSingleQuoteString
+        } else if ((next = part.match(RE_BACKSLASH))) {
+          token = TokenType.String
+          state = State.InsideSingleQuoteString
+        } else {
+          part
+          throw new Error('no')
+        }
+        break
+      case State.InsideDoubleQuoteString:
+        if ((next = part.match(RE_QUOTE_DOUBLE))) {
+          token = TokenType.Punctuation
+          state = stack.pop() || State.TopLevelContent
+        } else if ((next = part.match(RE_STRING_DOUBLE_QUOTE_CONTENT))) {
+          token = TokenType.String
+          state = State.InsideDoubleQuoteString
+        } else if ((next = part.match(RE_STRING_ESCAPE))) {
+          token = TokenType.String
+          state = State.InsideDoubleQuoteString
+        } else if ((next = part.match(RE_BACKSLASH))) {
+          token = TokenType.String
+          state = State.InsideDoubleQuoteString
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.InsideBacktickString:
+        if ((next = part.match(RE_QUOTE_BACKTICK))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_STRING_BACKTICK_QUOTE_CONTENT))) {
+          token = TokenType.String
+          state = State.InsideBacktickString
+        } else if ((next = part.match(RE_DOLLAR_CURLY_OPEN))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+          stack.push(State.InsideBacktickString)
+        } else if ((next = part.match(RE_STRING_ESCAPE))) {
+          token = TokenType.String
+          state = State.InsideBacktickString
+        } else if ((next = part.match(RE_BACKSLASH))) {
+          token = TokenType.String
+          state = State.InsideDoubleQuoteString
+        } else if ((next = part.match(RE_DOLLAR))) {
+          token = TokenType.String
+          state = State.InsideBacktickString
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.AfterPropertyDot:
+        if ((next = part.match(RE_WHITESPACE))) {
+          token = TokenType.Whitespace
+          state = State.AfterPropertyDot
+        } else if ((next = part.match(RE_FUNCTION_CALL_NAME))) {
+          token = TokenType.FunctionName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_VARIABLE_NAME))) {
+          token = TokenType.VariableName
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_LINE_COMMENT))) {
+          token = TokenType.Comment
+          state = State.InsideLineComment
+        } else if ((next = part.match(RE_BLOCK_COMMENT_START))) {
+          token = TokenType.Comment
+          state = State.InsideBlockComment
+        } else if ((next = part.match(RE_PUNCTUATION))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_VARIABLE_NAME_SPECIAL_2))) {
+          token = TokenType.VariableName
+          state = State.TopLevelContent
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.AfterOpeningAngleBracket:
+        if ((next = part.match(RE_TAGNAME))) {
+          token = TokenType.TagName
+          state = State.InsideOpeningTag
+          tag = next[0]
+        } else if ((next = part.match(RE_SLASH))) {
+          token = TokenType.PunctuationTag
+          state = State.AfterClosingTagAngleBrackets
+        } else if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ANGLE_BRACKET_OPEN))) {
+          token = TokenType.PunctuationTag
+          state = State.AfterOpeningAngleBracket
+        } else if ((next = part.match(RE_ANY_TEXT))) {
+          token = TokenType.Text
+          state = State.TopLevelContent
+        } else {
+          part //?
+          throw new Error('no')
+        }
+        break
+      case State.InsideOpeningTag:
+        if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else if ((next = part.match(RE_WHITESPACE))) {
+          token = TokenType.Whitespace
+          state = State.InsideOpeningTagAndHasSeenWhitespace
+        } else if ((next = part.match(RE_PUNCTUATION_SELF_CLOSING))) {
+          token = TokenType.PunctuationTag
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_TAG_TEXT))) {
+          token = TokenType.Text
+          state = State.InsideOpeningTag
+        } else {
+          part //?
+          throw new Error('no')
+        }
+        break
+      case State.InsideOpeningTagAndHasSeenWhitespace:
+        if ((next = part.match(RE_ATTRIBUTE_NAME))) {
+          token = TokenType.AttributeName
+          state = State.AfterAttributeName
+        } else if ((next = part.match(RE_SELF_CLOSING))) {
+          token = TokenType.PunctuationTag
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else if ((next = part.match(RE_CURLY_OPEN))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+          stack.push(State.InsideOpeningTag)
+        } else if ((next = part.match(RE_TAG_TEXT))) {
+          token = TokenType.Text
+          state = State.TopLevelContent
+        } else {
+          part //?
+          throw new Error('no')
+        }
+        break
+      case State.AfterClosingTagAngleBrackets:
+        if ((next = part.match(RE_TAGNAME))) {
+          token = TokenType.TagName
+          state = State.AfterClosingTagName
+        } else if ((next = part.match(RE_WHITESPACE))) {
+          token = TokenType.Whitespace
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else if ((next = part.match(RE_ANY_TEXT))) {
+          token = TokenType.Text
+          state = State.TopLevelContent
+        } else {
+          part //?
+          throw new Error('no')
+        }
+        break
+      case State.AfterClosingTagName:
+        if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else if ((next = part.match(RE_TEXT))) {
+          token = TokenType.Text
+          state = State.TopLevelContent
+        } else {
+          throw new Error('no')
+        }
+        break
+      case State.InsideTag:
+        if ((next = part.match(RE_TEXT))) {
+          token = TokenType.Text
+          state = State.InsideTag
+          if (index === 0) {
+            const match = part.match(RE_VAR_CONST_LET_EXPORT)
+            if (match) {
+              next = match
+              token = TokenType.Whitespace
+              state = State.TopLevelContent
+            }
+          }
+        } else if ((next = part.match(RE_ANGLE_BRACKET_OPEN_TAG))) {
+          token = TokenType.PunctuationTag
+          state = State.AfterOpeningAngleBracket
+        } else if ((next = part.match(RE_ANGLE_BRACKET_CLOSE_TAG))) {
+          token = TokenType.PunctuationTag
+          state = State.AfterClosingTagAngleBrackets
+        } else if ((next = part.match(RE_CURLY_OPEN))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+          stack.push(State.InsideTag)
+        } else if ((next = part.match(RE_CURLY_CLOSE))) {
+          token = TokenType.Punctuation
+          state = stack.pop() || State.TopLevelContent
+        } else if ((next = part.match(RE_ROUND_CLOSE))) {
+          if (stack.length === 0) {
+            token = TokenType.Punctuation
+            state = State.TopLevelContent
+          } else {
+            token = TokenType.Text
+            state = State.InsideTag
+          }
+        } else if ((next = part.match(RE_EMPTY_FRAGMENT))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else {
+          part
+          throw new Error('no')
+        }
+        break
+      case State.AfterAttributeName:
+        if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.InsideTag
+        } else if ((next = part.match(RE_EQUAL_SIGN))) {
+          token = TokenType.Punctuation
+          state = State.AfterAttributeEqualSign
+        } else if ((next = part.match(RE_DOUBLE_QUOTE))) {
+          token = TokenType.PunctuationString
+          state = State.InsideDoubleQuoteString
+        } else if ((next = part.match(RE_SINGLE_QUOTE))) {
+          token = TokenType.PunctuationString
+          state = State.InsideSingleQuoteString
+        } else if ((next = part.match(RE_TAG_TEXT))) {
+          token = TokenType.Text
+          state = State.InsideOpeningTag
+        } else if ((next = part.match(RE_WHITESPACE))) {
+          token = TokenType.Whitespace
+          state = State.InsideOpeningTagAndHasSeenWhitespace
+        } else {
+          // TODO check quotation mark
+          part //?
+          throw new Error('no')
+        }
+        break
+      case State.AfterAttributeEqualSign:
+        if ((next = part.match(RE_DOUBLE_QUOTE))) {
+          token = TokenType.PunctuationString
+          state = State.InsideDoubleQuoteString
+          stack.push(State.InsideOpeningTag)
+        } else if ((next = part.match(RE_SINGLE_QUOTE))) {
+          token = TokenType.PunctuationString
+          state = State.InsideSingleQuoteString
+          stack.push(State.InsideOpeningTag)
+        } else if ((next = part.match(RE_CURLY_OPEN))) {
+          token = TokenType.Punctuation
+          state = State.TopLevelContent
+          stack.push(State.InsideOpeningTag)
+        } else if ((next = part.match(RE_ANGLE_BRACKET_CLOSE))) {
+          token = TokenType.PunctuationTag
+          state = State.TopLevelContent
+        } else if ((next = part.match(RE_ATTRIBUTE_VALUE_UNQUOTED))) {
+          token = TokenType.String
+          state = State.InsideOpeningTag
+        } else {
+          part
+          throw new Error('no')
+        }
+        break
+      default:
+        state
+        throw new Error('no')
+    }
+    const tokenLength = next[0].length
+    index += tokenLength
+    tokens.push(token, tokenLength)
+  }
+  if (state === State.InsideLineComment) {
+    state = State.TopLevelContent
+  }
+  if (state === State.InsideSingleQuoteString && !line.endsWith('\\')) {
+    state = State.TopLevelContent
+  }
+  if (state === State.InsideTag && (line.endsWith(',') || line.endsWith(';'))) {
+    state = stack.pop() || State.TopLevelContent
+  }
+  return {
+    state,
+    tokens,
+    stack,
+    tag,
+  }
+}