Parse query string in using URLSearchParams

[calixteman]

• I just noticed in reading the code that we parse that stuff when something exists in the web api;
• see https://developer.mozilla.org/en-US/docs/Web/API/URLSearchParams/URLSearchParams.

[Snuffleupagus]

I just noticed in reading the code that we parse that stuff when something exists in the web api;

For reference: Note that this PDF.js code actually pre-dates the existence of the URLSearchParams functionality, which is why this code looks the way it does. In particular Safari were quite slow, comparatively, to implement URLSearchParams support.

If we're making these changes, shouldn't the following code be updated similar to the web/ui_utils.js code?

pdf.js/test/resources/reftest-analyzer.js

Lines 53 to 63 in 85c6dd5

	function hashParameters() {
	const query = window.location.hash.substring(1);
	const params = new Map();
	for (const part of query.split(/[&;]/)) {
	const param = part.split("="),
	key = param[0].toLowerCase(),
	value = param.length > 1 ? param[1] : "";
	params.set(decodeURIComponent(key), decodeURIComponent(value));
	}
	return params;
	}

[calixteman]

If we're making these changes, shouldn't the following code be updated similar to the web/ui_utils.js code?

pdf.js/test/resources/reftest-analyzer.js

Lines 53 to 63 in 85c6dd5

	function hashParameters() {
	const query = window.location.hash.substring(1);
	const params = new Map();
	for (const part of query.split(/[&;]/)) {
	const param = part.split("="),
	key = param[0].toLowerCase(),
	value = param.length > 1 ? param[1] : "";
	params.set(decodeURIComponent(key), decodeURIComponent(value));
	}
	return params;
	}

In this code we split on & and ; when URLSearchParams handles only & as separator, it's why I didn't change it.
Is this ; here because it has been forgotten or do we really have some urls containing it ?

[Snuffleupagus]

Is this ; here because it has been forgotten or do we really have some urls containing it ?

As mentioned at the top of that file, that code was originally imported from Firefox and the "special" separator pattern thus came from the original file.
As far as I can find, we only have a single call-site that invoke the reftest-analyzer code and there we're only passing one hash-parameter; see

pdf.js/test/test.js

Line 231 in de7cfed

const startUrl = `http://${host}:${server.port}/test/resources/reftest-analyzer.html#web=/test/eq.log`;

Hence, as long as all tests and the reftest-analyzer itself works, I believe that it's should be safe to change that one as well.

[calixteman]

/botio test

[pdfjsbot]

From: Bot.io (Windows)

---

Received

Command cmd_test from @calixteman received. Current queue size: 0

Live output at: http://54.193.163.58:8877/6a3431c663788e8/output.txt

[pdfjsbot]

From: Bot.io (Linux m4)

---

Received

Command cmd_test from @calixteman received. Current queue size: 0

Live output at: http://54.241.84.105:8877/03c9f802892c5dd/output.txt

[pdfjsbot]

From: Bot.io (Linux m4)

---

Failed

Full output at http://54.241.84.105:8877/03c9f802892c5dd/output.txt

Total script time: 24.07 mins

• Font tests: Passed
• Unit tests: Passed
• Integration Tests: Passed
• Regression tests: FAILED

  different ref/snapshot: 13
  different first/second rendering: 1

Image differences available at: http://54.241.84.105:8877/03c9f802892c5dd/reftest-analyzer.html#web=eq.log

[pdfjsbot]

From: Bot.io (Windows)

---

Failed

Full output at http://54.193.163.58:8877/6a3431c663788e8/output.txt

Total script time: 41.52 mins

• Font tests: Passed
• Unit tests: Passed
• Integration Tests: Passed
• Regression tests: FAILED

  different ref/snapshot: 8
  different first/second rendering: 1

Image differences available at: http://54.193.163.58:8877/6a3431c663788e8/reftest-analyzer.html#web=eq.log