Decode content in XP tags

src/tag-names-0th-ifd.js

@@ -230,11 +230,26 @@ export default {
     0x87b0: 'GeoTiffDoubleParams',
     0x87b1: 'GeoTiffAsciiParams',
     0x8825: 'GPS Info IFD Pointer',
-    0x9c9b: 'XPTitle',
-    0x9c9c: 'XPComment',
-    0x9c9d: 'XPAuthor',
-    0x9c9e: 'XPKeywords',
-    0x9c9f: 'XPSubject',
+    0x9c9b: {
+        name: 'XPTitle',
+        description: decodeXPValue,
+    },
+    0x9c9c: {
+        name: 'XPComment',
+        description: decodeXPValue
+    },
+    0x9c9d: {
+        name: 'XPAuthor',
+        description: decodeXPValue,
+    },
+    0x9c9e: {
+        name: 'XPKeywords',
+        description: decodeXPValue,
+    },
+    0x9c9f: {
+        name: 'XPSubject',
+        description: decodeXPValue,
+    },
     0xa480: 'GDALMetadata',
     0xa481: 'GDALNoData',
     0xc4a5: 'PrintIM',
@@ -355,3 +370,12 @@ export default {
     0xc7a7: 'NewRawImageDigest',
     0xc7a8: 'RawToPreviewGain'
 };
+
+function decodeXPValue(value) {
+    // The XP tags are encoded as UCS-2 which uses two bytes per character but
+    // it's close to UTF-16 so we can use that to decode them.
+    // https://www.loc.gov/preservation/digital/formats/content/tiff_tags.shtml
+    const decodedValue = new TextDecoder('utf-16').decode(new Uint8Array(value));
+    // Some softwares pad the string with null characters so we remove them.
+    return decodedValue.replace(/\u0000+$/, ''); // eslint-disable-line no-control-regex
+}

test/unit/tag-names-0th-ifd-spec.js

@@ -359,23 +359,28 @@ describe('tag-names-0th-ifd', () => {
     });
 
     it('should have tag XPTitle', () => {
-        expect(TagNames0thIfd[0x9c9b]).to.equal('XPTitle');
+        expect(TagNames0thIfd[0x9c9b].name).to.equal('XPTitle');
+        expect(TagNames0thIfd[0x9c9b].description([116, 0, 105, 0, 116, 0, 108, 0, 101, 0])).to.equal('title');
     });
 
     it('should have tag XPComment', () => {
-        expect(TagNames0thIfd[0x9c9c]).to.equal('XPComment');
+        expect(TagNames0thIfd[0x9c9c].name).to.equal('XPComment');
+        expect(TagNames0thIfd[0x9c9c].description([99, 0, 111, 0, 109, 0, 109, 0, 101, 0, 110, 0, 116, 0])).to.equal('comment');
     });
 
     it('should have tag XPAuthor', () => {
-        expect(TagNames0thIfd[0x9c9d]).to.equal('XPAuthor');
+        expect(TagNames0thIfd[0x9c9d].name).to.equal('XPAuthor');
+        expect(TagNames0thIfd[0x9c9d].description([97, 0, 117, 0, 116, 0, 104, 0, 111, 0, 114, 0])).to.equal('author');
     });
 
     it('should have tag XPKeywords', () => {
-        expect(TagNames0thIfd[0x9c9e]).to.equal('XPKeywords');
+        expect(TagNames0thIfd[0x9c9e].name).to.equal('XPKeywords');
+        expect(TagNames0thIfd[0x9c9e].description([107, 0, 101, 0, 121, 0, 119, 0, 111, 0, 114, 0, 100, 0, 115, 0])).to.equal('keywords');
     });
 
     it('should have tag XPSubject', () => {
-        expect(TagNames0thIfd[0x9c9f]).to.equal('XPSubject');
+        expect(TagNames0thIfd[0x9c9f].name).to.equal('XPSubject');
+        expect(TagNames0thIfd[0x9c9f].description([115, 0, 117, 0, 98, 0, 106, 0, 101, 0, 99, 0, 116, 0])).to.equal('subject');
     });
 
     it('should have tag GDALMetadata', () => {