gcfixer/src/utils.ts

Normalize Geocaching GPX file for Garmin devices

import * as I from './interfaces';
import cleaner from 'clean-html';
import he from 'he';
import unidecode from 'unidecode-plus';
import htmlToText from 'html-to-text';

const HTML_CLEAN = {
  'add-break-around-tags': ['b', 'i', 'strong', 'em'],
  'add-remove-attributes': ['style', 'class'],
  'add-remove-tags': ['img', 'picture', 'a'],
  'remove-comments': true,
  'remove-empty-tags': ['p', 'div']
};
// chars < and > outside of tags can screw Garmin html parser, so we replace them with
// visually similar from ASCII range. Possible are also unicode triangles &#9665; &#9655;
const FAKE_LT = '&#171;'; // «
const FAKE_GT = '&#187;'; // »

/**
 * Helps he.decode to process some oddities
 */
export function unescape(txt: string): string {
  let out = txt;

  out = out.replace(/&amp;lt;/g, '#@AMP@LT@#');
  out = out.replace(/&amp;gt;/g, '#@AMP@GT@#');
  out = out.replace(/&amp;amp;/g, '#@AMP@#');
  out = out.replace(/&amp;/g, '&');
  out = out.replace(/ ,,/g, ' "');
  out = out.replace(/„/g, '"');
  out = out.replace(/“/g, '"');

  return out;
}

export function escape(txt: string, escapeTags = false): string {
  let out = txt;
  out = out.replace(/&/g, '&amp;');
  out = out.replace(/#@AMP@#/g, '&amp;amp;');
  out = out.replace(/#@AMP@LT@#/g, FAKE_LT);
  out = out.replace(/#@AMP@GT@#/g, FAKE_GT);
  out = out.replace(/ ,,/g, '"');

  if (escapeTags) {
    out = out.replace(/</g, '&lt;');
    out = out.replace(/>/g, '&gt;');
    out = out.replace(/°/g, '&#176;');
  }

  return out;
}

export function untag(txt: string): string {
  let out = txt;

  out = out.replace(/</g, FAKE_LT);
  out = out.replace(/>/g, FAKE_GT);
  out = out.replace(/&amp;lt;/g, FAKE_LT);
  out = out.replace(/&amp;gt;/g, FAKE_GT);
  out = out.replace(/#@AMP@LT@#/g, FAKE_LT);
  out = out.replace(/#@AMP@GT@#/g, FAKE_GT);

  return out;
}

export function removeUrl(txt: string): string {
  return txt.replace(/\(https?:\/\/(.*)\)/gi, '');
}

export function processText(text: string, flag: I.Flag) {
  let out = text;

  if (flag.desc) {
    // clean wrong html entities
    out = unescape(out);
    // replace html entities
    out = he.decode(out);
    // sanitize html
    cleaner.clean(out, HTML_CLEAN, (html) => {
      out = html;
    });
    // remove html if needed
    if (flag.stripHtml) {
      out = htmlToText.fromString(out, {
        decodeOptions: { strict: false },
        ignoreHref: true,
        ignoreImage: true,
        uppercaseHeadings: false
      });
    }
    // utf8 to ascii
    out = unidecode(out, { skipRanges: [[0xb0, 0xb0]] });
    // escape
    out = escape(out, !flag.stripHtml);
    if (flag.stripHtml) {
      out = untag(out);
    }
  }

  if (flag.text) {
    out = unidecode(out, { skipRanges: [[0xb0, 0xb0]] });
    out = escape(out);
    out = untag(out);
  }

  if (flag.removeUrl) {
    out = removeUrl(out);
  }

  return out;
}

export function createTag(options: I.TagOptions) {
  const howManyTabs = options.closing && options.name === options.lastParent ? 0 : options.level;
  let levelMinus = false;
  let singleTag = false;
  let tabs = '';
  let tag = '';

  for (let i = 0; i < howManyTabs; i++) {
    tabs = `${tabs}  `;
  }

  if (options.closing) {
    tag = options.name === options.lastParent ? '</' : `\n${tabs}  </`;
  } else {
    tag = `\n${tabs}<`;
  }

  tag = tag + options.name;

  if (options.attrs !== undefined) {
    Object.keys(options.attrs).forEach((attr) => {
      let val: string = options.attrs![attr];
      if (val.substr(-2) === ' /') {
        singleTag = true;
        val = val.substr(0, val.length - 2);
      }
      tag = `${tag} ${attr}="${val}"`;
    });
  }

  if (options.name.includes('/') || singleTag) {
    levelMinus = true;
  }

  tag = `${tag}${singleTag ? ' />' : '>'}`;

  return {
    levelMinus,
    tag
  };
}