fix: creating a Parse.File with base64 string fails for some file types (#1467)

Author: RazvanCristian

integration/test/ParseFileTest.js

@@ -66,6 +66,7 @@ describe('Parse.File', () => {
 
   it('can get file data from base64', async () => {
     const file = new Parse.File('parse-server-logo', { base64: 'ParseA==' });
+    await file.save();
     let data = await file.getData();
     assert.equal(data, 'ParseA==');
     file._data = null;
@@ -79,6 +80,7 @@ describe('Parse.File', () => {
     const file = new Parse.File('parse-server-logo', {
       base64: 'data:image/jpeg;base64,ParseA==',
     });
+    await file.save();
     let data = await file.getData();
     assert.equal(data, 'ParseA==');
     file._data = null;

src/ParseFile.js

@@ -42,7 +42,15 @@ export type FileSource =
       type: string,
     };
 
-const dataUriRegexp = /^data:([a-zA-Z]+\/[-a-zA-Z0-9+.]+)(;charset=[a-zA-Z0-9\-\/]*)?;base64,/;
+const base64Regex = new RegExp(
+  '([0-9a-zA-Z+/]{4})*(([0-9a-zA-Z+/]{2}==)|([0-9a-zA-Z+/]{3}=))',
+  'i'
+);
+
+const dataUriRegex = new RegExp(
+  `^data:([a-zA-Z]+\\/[-a-zA-Z0-9+.]+(;[a-z-]+=[a-zA-Z0-9+.-]+)?)?(;base64)?,(${base64Regex.source})*$`,
+  'i'
+);
 
 function b64Digit(number: number): string {
   if (number < 26) {
@@ -137,26 +145,30 @@ class ParseFile {
           type: specifiedType,
         };
       } else if (data && typeof data.base64 === 'string') {
-        const base64 = data.base64;
-        const commaIndex = base64.indexOf(',');
-
-        if (commaIndex !== -1) {
-          const matches = dataUriRegexp.exec(base64.slice(0, commaIndex + 1));
-          // if data URI with type and charset, there will be 4 matches.
-          this._data = base64.slice(commaIndex + 1);
-          this._source = {
-            format: 'base64',
-            base64: this._data,
-            type: matches[1],
-          };
-        } else {
-          this._data = base64;
-          this._source = {
-            format: 'base64',
-            base64: base64,
-            type: specifiedType,
-          };
+        // Check if data URI or base64 string is valid
+        const validationRegex = new RegExp(base64Regex.source + '|' + dataUriRegex.source, 'i');
+        if (!validationRegex.test(data.base64)) {
+          throw new Error(
+            'Cannot create a Parse.File without valid data URIs or base64 encoded data.'
+          );
+        }
+
+        const base64 = data.base64.split(',').slice(-1)[0];
+        let type =
+          specifiedType || data.base64.split(';').slice(0, 1)[0].split(':').slice(1, 2)[0] || '';
+
+        // https://tools.ietf.org/html/rfc2397
+        // If <mediatype> is omitted, it defaults to text/plain;charset=US-ASCII.
+        // As a shorthand, "text/plain" can be omitted but the charset parameter supplied.
+        if (dataUriRegex.test(data.base64)) {
+          type = type || 'text/plain';
         }
+
+        this._source = {
+          format: 'base64',
+          base64,
+          type,
+        };
       } else {
         throw new TypeError('Cannot create a Parse.File with that data.');
       }

src/__tests__/ParseFile-test.js

@@ -66,14 +66,30 @@ describe('ParseFile', () => {
     expect(file._source.type).toBe('');
   });
 
-  it('can extract data type from base64', () => {
+  it('can set the default type to be text/plain when using base64', () => {
     const file = new ParseFile('parse.txt', {
+      base64: 'data:;base64,ParseA==',
+    });
+    expect(file._source.base64).toBe('ParseA==');
+    expect(file._source.type).toBe('text/plain');
+  });
+
+  it('can extract data type from base64', () => {
+    const file = new ParseFile('parse.png', {
       base64: 'data:image/png;base64,ParseA==',
     });
     expect(file._source.base64).toBe('ParseA==');
     expect(file._source.type).toBe('image/png');
   });
 
+  it('can extract data type from base64 with a filename parameter', () => {
+    const file = new ParseFile('parse.pdf', {
+      base64: 'data:application/pdf;filename=parse.pdf;base64,ParseA==',
+    });
+    expect(file._source.base64).toBe('ParseA==');
+    expect(file._source.type).toBe('application/pdf');
+  });
+
   it('can create files with file uri', () => {
     const file = new ParseFile('parse-image', {
       uri: 'http://example.com/image.png',
@@ -136,6 +152,12 @@ describe('ParseFile', () => {
     expect(function () {
       new ParseFile('parse.txt', 'string');
     }).toThrow('Cannot create a Parse.File with that data.');
+
+    expect(function () {
+      new ParseFile('parse.txt', {
+        base64: 'abc',
+      });
+    }).toThrow('Cannot create a Parse.File without valid data URIs or base64 encoded data.');
   });
 
   it('throws with invalid base64', () => {