2018-05-23 13:14:38 +02:00
|
|
|
const { rtrimSlashes } = require('lib/path-utils');
|
2019-09-12 22:57:23 +02:00
|
|
|
const { urlDecode } = require('lib/string-utils');
|
2018-05-23 13:14:38 +02:00
|
|
|
|
2017-11-08 19:51:55 +02:00
|
|
|
const urlUtils = {};
|
|
|
|
|
|
|
|
urlUtils.hash = function(url) {
|
|
|
|
const s = url.split('#');
|
|
|
|
if (s.length <= 1) return '';
|
|
|
|
return s[s.length - 1];
|
2019-07-29 15:43:53 +02:00
|
|
|
};
|
2017-11-08 19:51:55 +02:00
|
|
|
|
2018-05-23 13:14:38 +02:00
|
|
|
urlUtils.urlWithoutPath = function(url) {
|
|
|
|
const parsed = require('url').parse(url, true);
|
2019-09-19 23:51:18 +02:00
|
|
|
return `${parsed.protocol}//${parsed.host}`;
|
2019-07-29 15:43:53 +02:00
|
|
|
};
|
2018-05-23 13:14:38 +02:00
|
|
|
|
|
|
|
urlUtils.urlProtocol = function(url) {
|
2019-05-10 02:06:06 +02:00
|
|
|
if (!url) return '';
|
2018-05-23 13:14:38 +02:00
|
|
|
const parsed = require('url').parse(url, true);
|
|
|
|
return parsed.protocol;
|
2019-07-29 15:43:53 +02:00
|
|
|
};
|
2018-05-23 13:14:38 +02:00
|
|
|
|
|
|
|
urlUtils.prependBaseUrl = function(url, baseUrl) {
|
|
|
|
baseUrl = rtrimSlashes(baseUrl).trim(); // All the code below assumes that the baseUrl does not end up with a slash
|
|
|
|
url = url.trim();
|
|
|
|
|
|
|
|
if (!url) url = '';
|
|
|
|
if (!baseUrl) return url;
|
2019-06-13 01:26:09 +02:00
|
|
|
if (url.indexOf('#') === 0) return url; // Don't prepend if it's a local anchor
|
2018-05-24 13:44:13 +02:00
|
|
|
if (urlUtils.urlProtocol(url)) return url; // Don't prepend the base URL if the URL already has a scheme
|
2018-05-23 13:14:38 +02:00
|
|
|
|
2019-07-29 15:43:53 +02:00
|
|
|
if (url.length >= 2 && url.indexOf('//') === 0) {
|
|
|
|
// If it starts with // it's a protcol-relative URL
|
2018-05-23 13:14:38 +02:00
|
|
|
return urlUtils.urlProtocol(baseUrl) + url;
|
2019-07-29 15:43:53 +02:00
|
|
|
} else if (url && url[0] === '/') {
|
|
|
|
// If it starts with a slash, it's an absolute URL so it should be relative to the domain (and not to the full baseUrl)
|
2018-05-23 13:14:38 +02:00
|
|
|
return urlUtils.urlWithoutPath(baseUrl) + url;
|
|
|
|
} else {
|
2019-09-19 23:51:18 +02:00
|
|
|
return baseUrl + (url ? `/${url}` : '');
|
2018-05-23 13:14:38 +02:00
|
|
|
}
|
2019-07-29 15:43:53 +02:00
|
|
|
};
|
2018-05-23 13:14:38 +02:00
|
|
|
|
2019-09-12 23:48:10 +02:00
|
|
|
const resourceRegex = /^(joplin:\/\/|:\/)([0-9a-zA-Z]{32})(|#[^\s]*)(|\s".*?")$/;
|
|
|
|
|
2019-09-10 10:25:58 +02:00
|
|
|
urlUtils.isResourceUrl = function(url) {
|
2019-09-12 23:48:10 +02:00
|
|
|
return !!url.match(resourceRegex);
|
2019-09-10 10:25:58 +02:00
|
|
|
};
|
|
|
|
|
2019-09-09 19:16:00 +02:00
|
|
|
urlUtils.parseResourceUrl = function(url) {
|
2019-09-10 10:25:58 +02:00
|
|
|
if (!urlUtils.isResourceUrl(url)) return null;
|
|
|
|
|
2019-09-12 23:48:10 +02:00
|
|
|
const match = url.match(resourceRegex);
|
2019-09-09 19:16:00 +02:00
|
|
|
|
2019-09-12 23:48:10 +02:00
|
|
|
const itemId = match[2];
|
|
|
|
let hash = match[3].trim();
|
2019-09-12 22:57:23 +02:00
|
|
|
|
|
|
|
// In general we want the hash to be decoded so that non-alphabetical languages
|
|
|
|
// appear as-is without being encoded with %.
|
|
|
|
// Fixes https://github.com/laurent22/joplin/issues/1870
|
2019-09-12 23:48:10 +02:00
|
|
|
if (hash) hash = urlDecode(hash.substr(1)); // Remove the first #
|
|
|
|
|
|
|
|
return {
|
|
|
|
itemId: itemId,
|
|
|
|
hash: hash,
|
|
|
|
};
|
|
|
|
};
|
|
|
|
|
|
|
|
urlUtils.extractResourceUrls = function(text) {
|
|
|
|
const markdownLinksRE = /\]\((.*?)\)/g;
|
|
|
|
const output = [];
|
|
|
|
let result = null;
|
|
|
|
|
|
|
|
while ((result = markdownLinksRE.exec(text)) !== null) {
|
|
|
|
const resourceUrlInfo = urlUtils.parseResourceUrl(result[1]);
|
|
|
|
if (resourceUrlInfo) output.push(resourceUrlInfo);
|
|
|
|
}
|
|
|
|
|
|
|
|
const htmlRegexes = [
|
|
|
|
/<img[\s\S]*?src=["']:\/([a-zA-Z0-9]{32})["'][\s\S]*?>/gi,
|
|
|
|
/<a[\s\S]*?href=["']:\/([a-zA-Z0-9]{32})["'][\s\S]*?>/gi,
|
|
|
|
];
|
|
|
|
|
|
|
|
for (const htmlRegex of htmlRegexes) {
|
|
|
|
while (true) {
|
|
|
|
const m = htmlRegex.exec(text);
|
|
|
|
if (!m) break;
|
|
|
|
output.push({ itemId: m[1], hash: '' });
|
|
|
|
}
|
|
|
|
}
|
2019-09-09 19:16:00 +02:00
|
|
|
|
|
|
|
return output;
|
|
|
|
};
|
|
|
|
|
2020-10-28 17:50:34 +02:00
|
|
|
urlUtils.objectToQueryString = function(query) {
|
|
|
|
if (!query) return '';
|
|
|
|
|
|
|
|
let queryString = '';
|
|
|
|
const s = [];
|
|
|
|
for (const k in query) {
|
|
|
|
if (!query.hasOwnProperty(k)) continue;
|
|
|
|
s.push(`${encodeURIComponent(k)}=${encodeURIComponent(query[k])}`);
|
|
|
|
}
|
|
|
|
queryString = s.join('&');
|
|
|
|
|
|
|
|
return queryString;
|
|
|
|
};
|
|
|
|
|
2019-07-29 15:43:53 +02:00
|
|
|
module.exports = urlUtils;
|