From 98fd1540f3840be85e2ed0b4bf0fa29426db1afd Mon Sep 17 00:00:00 2001
From: abaevbog <bogdan@zotero.org>
Date: Sun, 4 Aug 2024 21:27:33 -0700
Subject: [PATCH] decode uri components in doi when possible (#28)

Also, attempt to drop last bracket or paren if there
is an opened one before DOI to not skip and pass existing tests.

Fixes: zotero/zotero#3218
---
 test/tests/utilitiesTest.js | 18 ++++++++++++++----
 utilities.js                | 33 +++++++++++++++++++++++++++++++--
 2 files changed, 45 insertions(+), 6 deletions(-)

diff --git a/test/tests/utilitiesTest.js b/test/tests/utilitiesTest.js
index 17ba91f..60b1bc4 100644
--- a/test/tests/utilitiesTest.js
+++ b/test/tests/utilitiesTest.js
@@ -93,13 +93,23 @@ describe("Zotero.Utilities", function() {
 			assert.equal(cleanDOI(`Foo bar ${doi}. Foo bar`), doi);
 		});
 		
-		// FIXME
-		it.skip("should parse a DOI in parentheses", function () {
+		it("should parse a DOI with URL-encoded < and >", function () {
+			const encodedUri = "10.1002/1096-9128(200005)12:6%3C375::AID-CPE480%3E3.0.CO;2-M";
+			const expected = "10.1002/1096-9128(200005)12:6<375::AID-CPE480>3.0.CO;2-M";
+			assert.equal(cleanDOI(`Foo bar ${encodedUri}. Foo bar`), expected);
+		});
+
+		it("should parse a DOI URL with encoded characters", function () {
+			const encodedUri = "https://doi.org/10.1002/1096-9128(200005)12:6%3C375::AID-CPE480%3E3.0.CO;2-M";
+			const expected = "10.1002/1096-9128(200005)12:6<375::AID-CPE480>3.0.CO;2-M";
+			assert.equal(cleanDOI(`Foo bar ${encodedUri}. Foo bar`), expected);
+		});
+		
+		it("should parse a DOI in parentheses", function () {
 			assert.equal(cleanDOI(`Foo bar (${doi}) foo bar`), doi);
 		});
 		
-		// FIXME
-		it.skip("should parse a DOI in brackets", function () {
+		it("should parse a DOI in brackets", function () {
 			assert.equal(cleanDOI(`Foo bar [${doi}] foo bar`), doi);
 		});
 	});
diff --git a/utilities.js b/utilities.js
index 6808118..7c3581f 100644
--- a/utilities.js
+++ b/utilities.js
@@ -482,9 +482,38 @@ var Utilities = {
 		if(typeof(x) != "string") {
 			throw new Error("cleanDOI: argument must be a string");
 		}
-
+		// If it's a URL, decode it
+		if (x.match(/^https?:/)) {
+			x = decodeURIComponent(x);
+		}
+		// Even if it's not a URL, decode %3C followed by %3E as < >
+		var openingPos = x.indexOf("%3C");
+		if (openingPos != -1 && openingPos < x.indexOf("%3E")) {
+			x = x.replace(/%3C/g, "<");
+			x = x.replace(/%3E/g, ">");
+		}
 		var doi = x.match(/10(?:\.[0-9]{4,})?\/[^\s]*[^\s\.,]/);
-		return doi ? doi[0] : null;
+		if (!doi) {
+			return null;
+		}
+		var result = doi[0];
+
+		// Check if the DOI ends with a bracket
+		var trailingBracket = result.slice(-1);
+		if ([']', ')', '}'].includes(trailingBracket)) {
+			// Check the portion of the string before the matched DOI for an unclosed bracket
+			let beforeDOI = x.slice(0, doi.index);
+			let openingBracket = {
+				']': '[',
+				')': '(',
+				'}': '{'
+			}[trailingBracket];
+			if (beforeDOI.lastIndexOf(openingBracket) > beforeDOI.lastIndexOf(trailingBracket)) {
+				// Remove the trailing bracket from the DOI
+				result = result.slice(0, -1);
+			}
+		}
+		return result;
 	},
 
 	/**