シェブロテインのCST構造

CstElement

export declare type CstElement = IToken | CstNode

IToken

/**
 * Things to note:
 *  - The offset range is inclusive to exclusive.
 *
 * - A lineTerminator as the last character does not effect the Token's line numbering.
 *   In other words a new line only starts **after** a line terminator.
 *
 * - A Token's image is it's **literal** text.
 *   e.g unicode escaping is untouched.
 */
export interface IToken {
  /** The textual representation of the Token as it appeared in the text. */
  image: string
  /** Offset of the first character of the Token. 0-indexed. */
  startOffset: number
  /** Line of the first character of the Token. 1-indexed. */
  startLine?: number
  /**
   * Column of the first character of the Token. 1-indexed.
   *
   * For token foo in the following line, startColumn will be 3 and endColumn will be 5.
   * ```
   * a foo
   * 123456
   * ```
   */
  startColumn?: number
  /**
   * Offset of the last character of the Token. 0-indexed.
   * Note that this points at the last character, not the end of the token, so the original image would be
   * `input.substring(token.startOffset, token.endOffset + 1)`.
   */
  endOffset?: number
  /** Line of the last character of the Token. 1-indexed. Will be the same as startLine for single-line tokens.*/
  endLine?: number
  /** Column of the last character of the Token. 1-indexed. See also startColumn. */
  endColumn?: number
  /** this marks if a Token does not really exist and has been inserted "artificially" during parsing in rule error recovery. */
  isInsertedInRecovery?: boolean
  /** An number index representing the type of the Token use <getTokenConstructor> to get the Token Type from a token "instance"  */
  tokenTypeIdx: number
  /**
   * The actual Token Type of this Token "instance"
   * This is the same Object returned by the "createToken" API.
   * This property is very useful for debugging the Lexing and Parsing phases.
   */
  tokenType: TokenType

  /**
   * Custom Payload value, this is an optional feature of Custom Token Patterns
   * For additional details see the docs:
   * https://chevrotain.io/docs/guide/custom_token_patterns.html#custom-payloads
   */
  payload?: any
}

CstNode

/**
 * A [Concrete Syntax Tree](http://chevrotain.io/docs/guide/concrete_syntax_tree.html) Node.
 * This structure represents the whole parse tree of the grammar
 * This means that information on each and every Token is present.
 * This is unlike an AST (Abstract Syntax Tree) where some of the syntactic information is missing.
 *
 * For example given an ECMAScript grammar, an AST would normally not contain information on the location
 * of Commas, Semi colons, redundant parenthesis ect, however a CST would have that information.
 */
export interface CstNode {
  readonly name: string
  readonly children: CstChildrenDictionary
  /**
   * A flag indicating the whole CSTNode has been recovered during **re-sync** error recovery.
   * This means that some of the node's children may be missing.
   * - Note that single token insertion/deletion recovery would not activate this flag.
   *   This flag would only be activated in **re-sync** recovery when the rule's
   *   grammar cannot be fully parsed.
   * - See: https://chevrotain.io/docs/tutorial/step4_fault_tolerance.html
   *   for more info on error recovery and fault tolerance.
   */
  readonly recoveredNode?: boolean

  /**
   * Will only be present if the {@link IParserConfig.nodeLocationTracking} is
   * **not** set to "none".
   * See: http://chevrotain.io/docs/guide/concrete_syntax_tree.html#cstnode-location
   * For more details.
   */
  readonly location?: CstNodeLocation
}