Update extractIdentifiers #595

src/js/common/identifiers.js

@@ -1,15 +1,67 @@
-const cleanDOI = text => {
-	const doi = text.match(/10(?:\.[0-9]{4,})?\/[^\s]*[^\s\.,]/);
-	return doi ? doi[0] : null;
+/* eslint-disable no-useless-escape */
+/* eslint-disable no-cond-assign */
+// `cleanISBN`, `cleanDOI` and `extractIdentifiers` adapted from https://github.com/zotero/utilities/blob/43142236a282e5e1a3190694628f329aa2e0ba8e/utilities.js
+
+/**
+ * Strip info:doi prefix and any suffixes from a DOI
+ * @type String
+ */
+function cleanDOI(/**String**/ x) {
+	if (typeof (x) != "string") {
+		throw new Error("cleanDOI: argument must be a string");
+	}
+	// If it's a URL, try to decode it
+	if (/^https?:/.test(x)) {
+		try {
+			x = decodeURIComponent(x);
+		}
+		catch (e) {
+			// URI contains an invalid escape sequence
+			console.warn("Not decoding URL-like DOI because of invalid escape sequence: " + 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\.,]/);
+	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;
 }
 
-const cleanISBN = (isbnStr, dontValidate) => {
+/**
+ * Clean and validate ISBN.
+ * Return isbn if valid, otherwise return false
+ * @param {String} isbn
+ * @param {Boolean} [dontValidate=false] Do not validate check digit
+ * @return {String|Boolean} Valid ISBN or false
+ */
+function cleanISBN(isbnStr, dontValidate) {
 	isbnStr = isbnStr.toUpperCase()
 		.replace(/[\x2D\xAD\u2010-\u2015\u2043\u2212]+/g, ''); // Ignore dashes
 	var isbnRE = /\b(?:97[89]\s*(?:\d\s*){9}\d|(?:\d\s*){9}[\dX])\b/g,
 		isbnMatch;
-
-	// eslint-disable-next-line no-cond-assign
 	while (isbnMatch = isbnRE.exec(isbnStr)) {
 		var isbn = isbnMatch[0].replace(/\s+/g, '');
 
@@ -43,10 +95,9 @@ const cleanISBN = (isbnStr, dontValidate) => {
 	return false;
 }
 
-// https://github.com/zotero/zotero/blob/57989260935703f0c7d570a39bcf6516b8c61df6/chrome/content/zotero/xpcom/utilities_internal.js#L1409
-const extractIdentifiers = text => {
-	const identifiers = [];
-	const foundIDs = new Set(); // keep track of identifiers to avoid duplicates
+function extractIdentifiers(text) {
+	var identifiers = [];
+	var foundIDs = new Set(); // keep track of identifiers to avoid duplicates
 
 	// First look for DOIs
 	var ids = text.split(/[\s\u00A0]+/); // whitespace + non-breaking space
@@ -67,8 +118,6 @@ const extractIdentifiers = text => {
 			.toUpperCase();
 		let ISBN_RE = /(?:\D|^)(97[89]\d{10}|\d{9}[\dX])(?!\d)/g;
 		let isbn;
-
-		// eslint-disable-next-line no-cond-assign
 		while (isbn = ISBN_RE.exec(ids)) {
 			isbn = cleanISBN(isbn[1]);
 			if (isbn && !foundIDs.has(isbn)) {
@@ -82,8 +131,6 @@ const extractIdentifiers = text => {
 		// Next try spaces
 		if (!identifiers.length) {
 			ids = ids.replace(/[ \u00A0]+/g, ""); // space + non-breaking space
-
-			// eslint-disable-next-line no-cond-assign
 			while (isbn = ISBN_RE.exec(ids)) {
 				isbn = cleanISBN(isbn[1]);
 				if (isbn && !foundIDs.has(isbn)) {
@@ -112,7 +159,20 @@ const extractIdentifiers = text => {
 		}
 	}
 
-	// Finally try for PMID
+	// Next, try ADS Bibcodes
+	if (!identifiers.length) {
+		// regex as in the ADS Bibcode translator
+		let adsBibcode_RE = /\b(\d{4}\D\S{13}[A-Z.:])\b/g;
+		let adsBibcode;
+		while ((adsBibcode = adsBibcode_RE.exec(text)) && !foundIDs.has(adsBibcode)) {
+			identifiers.push({
+				adsBibcode: adsBibcode[1]
+			});
+			foundIDs.add(adsBibcode);
+		}
+	}
+
+	// Finally, try PMID
 	if (!identifiers.length) {
 		// PMID; right now, the longest PMIDs are 8 digits, so it doesn't seem like we'll
 		// need to discriminate for a fairly long time

src/js/component/item/actions/add-by-identifier.jsx

@@ -194,7 +194,7 @@ const AddByIdentifier = props => {
 				<div className="popover-inner" role="tooltip">
 					<h3 className="popover-header">
 						<label htmlFor={ `${id.current}-input` }>
-							Enter a URL, ISBNs, DOIs, PMIDs, or arXiv IDs
+							Enter a URL, ISBNs, DOIs, PMIDs, arXiv IDs, or ADS Bibcodes to add to your library:
 						</label>
 					</h3>
 					<div className="popover-body">

src/js/component/modal/add-by-identifier.jsx

@@ -149,7 +149,7 @@ const AddByIdentifierModal = () => {
 							onChange={ handleInputChange }
 							onCommit={ handleInputCommit }
 							onPaste={ handlePaste }
-							placeholder="URL, ISBNs, DOIs, PMIDs, or arXiv IDs"
+							placeholder="URL, ISBNs, DOIs, PMIDs, arXiv IDs, or ADS Bibcodes"
 							ref={ inputEl }
 							tabIndex={ 0 }
 							value={ identifier }

src/js/component/modal/create-parent-item.jsx

@@ -200,7 +200,7 @@ const CreateParentItemModal = () => {
 			>
 				<h3>
 					<label htmlFor={`${id}-input`}>
-						Enter a DOI, ISBN, PMID, or arXiv IDs to identify this file
+						Enter a DOI, ISBN, PMID, arXiv ID, or ADS Bibcode to identify this file:
 					</label>
 				</h3>
 				<div className="form">

test/items.test.jsx

@@ -139,7 +139,7 @@ describe('Items', () => {
 		);
 
 		const input = screen.getByRole('textbox',
-			{ name: 'Enter a URL, ISBNs, DOIs, PMIDs, or arXiv IDs' }
+			{ name: 'Enter a URL, ISBNs, DOIs, PMIDs, arXiv IDs, or ADS Bibcodes to add to your library:' }
 		);
 
 		await userEvent.type(

test/parent.test.jsx

@@ -101,8 +101,8 @@ describe('Create Parent Item', () => {
 		await waitForPosition();
 		await user.click(screen.getByRole('menuitem', { name: 'Create Parent Item' }));
 		const dialog = await screen.findByRole('dialog', { name: 'Create Parent Item' });
-		await waitFor(() => expect(getByRole(dialog, 'textbox', { name: 'Enter a DOI, ISBN, PMID, or arXiv IDs to identify this file' })).toHaveFocus());
-		const input = getByRole(dialog, 'textbox', { name: 'Enter a DOI, ISBN, PMID, or arXiv IDs to identify this file' });
+		await waitFor(() => expect(getByRole(dialog, 'textbox', { name: 'Enter a DOI, ISBN, PMID, arXiv ID, or ADS Bibcode to identify this file:' })).toHaveFocus());
+		const input = getByRole(dialog, 'textbox', { name: 'Enter a DOI, ISBN, PMID, arXiv ID, or ADS Bibcode to identify this file:' });
 		await user.type(input, '1706.03762{enter}', { skipClick: true });
 
 		expect(await screen.findByRole('row', { name: 'Attention Is All You Need' })).toBeInTheDocument();