@briancullen/aws-textract-parser/lib/factory/DocumentParser.js

Library for converting AWS Textract responses into a more usable structure.

"use strict";
var __importDefault = (this && this.__importDefault) || function (mod) {
    return (mod && mod.__esModule) ? mod : { "default": mod };
};
Object.defineProperty(exports, "__esModule", { value: true });
var DocumentNode_1 = __importDefault(require("../model/DocumentNode"));
var DocumentParser = /** @class */ (function () {
    function DocumentParser(pageParserFactory) {
        this.pageParserFactory = pageParserFactory;
    }
    DocumentParser.prototype.process = function (input) {
        var _a, _b;
        if (((_a = input.DocumentMetadata) === null || _a === void 0 ? void 0 : _a.Pages) === undefined) {
            return new DocumentNode_1.default([]);
        }
        var blocks = (_b = input.Blocks, (_b !== null && _b !== void 0 ? _b : []));
        var blocksById = blocks.reduce(function (map, block) { return map.set(block.Id, block); }, new Map());
        var parser = this.pageParserFactory(blocksById);
        var pages = blocks.filter(function (block) { return block.BlockType === 'PAGE'; })
            .map(function (block) { return parser.process(block); });
        return new DocumentNode_1.default(pages);
    };
    return DocumentParser;
}());
exports.DocumentParser = DocumentParser;