chromium/third_party/polymer/v3_0/components-chromium/polymer/lib/mixins/strict-binding-parser.d.ts

// tslint:disable:variable-name Describing an API that's defined elsewhere.
// tslint:disable:no-any describes the API as best we are able today

import {isPath} from '../utils/path.js';

import {dedupingMixin} from '../utils/mixin.js';

import {PropertyEffects} from './property-effects.js';


/**
 * Mixin that parses binding expressions and generates corresponding metadata.
 * The implementation is different than in `property-effects`, as it uses a
 * state machine instead of a regex. As such, this implementation is able to
 * handle more cases, with the potential performance hit.
 */
declare function StrictBindingParser<T extends new (...args: any[]) => {}>(base: T): T & StrictBindingParserConstructor & PropertyEffectsConstructor & TemplateStampConstructor & PropertyAccessorsConstructor & PropertiesChangedConstructor;

import {PropertyEffectsConstructor} from './property-effects.js';

import {TemplateStampConstructor, TemplateStamp} from './template-stamp.js';

import {PropertyAccessorsConstructor, PropertyAccessors} from './property-accessors.js';

import {PropertiesChangedConstructor, PropertiesChanged} from './properties-changed.js';

interface StrictBindingParserConstructor {
  new(...args: any[]): StrictBindingParser;

  /**
   * Called to parse text in a template (either attribute values or
   * textContent) into binding metadata.
   *
   * Any overrides of this method should return an array of binding part
   * metadata  representing one or more bindings found in the provided text
   * and any "literal" text in between.  Any non-literal parts will be passed
   * to `_evaluateBinding` when any dependencies change.  The only required
   * fields of each "part" in the returned array are as follows:
   *
   * - `dependencies` - Array containing trigger metadata for each property
   *   that should trigger the binding to update
   * - `literal` - String containing text if the part represents a literal;
   *   in this case no `dependencies` are needed
   *
   * Additional metadata for use by `_evaluateBinding` may be provided in
   * each part object as needed.
   *
   * The default implementation handles the following types of bindings
   * (one or more may be intermixed with literal strings):
   * - Property binding: `[[prop]]`
   * - Path binding: `[[object.prop]]`
   * - Negated property or path bindings: `[[!prop]]` or `[[!object.prop]]`
   * - Two-way property or path bindings (supports negation):
   *   `{{prop}}`, `{{object.prop}}`, `{{!prop}}` or `{{!object.prop}}`
   * - Inline computed method (supports negation):
   *   `[[compute(a, 'literal', b)]]`, `[[!compute(a, 'literal', b)]]`
   *
   * @param text Text to parse from attribute or textContent
   * @param templateInfo Current template metadata
   * @returns Array of binding part metadata
   */
  _parseBindings(text: string, templateInfo: object|null): BindingPart[]|null;
}

export {StrictBindingParserConstructor};

interface StrictBindingParser extends PropertyEffects, TemplateStamp, PropertyAccessors, PropertiesChanged {
}

export {StrictBindingParser};

import {BindingPart} from '../../interfaces';