Merge pull request #621 from syjer/more-robust-data-uri-base64

more robust data uri with base64 encoded data. Handle whitespaces.
danfickle · Dec 18, 2020 · fcf4a1d · fcf4a1d
2 parents 7373154 + cef71c3
commit fcf4a1d
Show file tree

Hide file tree

Showing 6 changed files with 79 additions and 10 deletions.
diff --git a/openhtmltopdf-core/src/main/java/com/openhtmltopdf/swing/NaiveUserAgent.java b/openhtmltopdf-core/src/main/java/com/openhtmltopdf/swing/NaiveUserAgent.java
@@ -26,7 +26,6 @@
 import java.net.URISyntaxException;
 import java.net.URL;
 import java.nio.charset.StandardCharsets;
-import java.util.Base64;
 import java.util.HashMap;
 import java.util.LinkedHashMap;
 import java.util.Map;
@@ -136,7 +135,7 @@ public FSStream getUrl(String url) {
 				String data = url.substring(idxSeparator+1);
 				byte[] res;
 				if (url.indexOf("base64,") == idxSeparator - 6 /* 6 = "base64,".length */) {
-					res = Base64.getMimeDecoder().decode(data);
+					res = ImageUtil.fromBase64Encoded(data);
 				} else {
 					res = data.getBytes(StandardCharsets.UTF_8);
 				}
@@ -438,6 +437,10 @@ public static class DefaultUriResolver implements FSUriResolver {
 		public String resolveURI(String baseUri, String uri) {
 			if (uri == null || uri.isEmpty())
 				return null;
+
+			if (uri.startsWith("data:")) {
+				return uri; //bypass URI "formatting" check for data uri, as we may have whitespace in the base64 encoded data
+			}
 
 			try {
 				URI possiblyRelative = new URI(uri);

diff --git a/openhtmltopdf-core/src/main/java/com/openhtmltopdf/util/ImageUtil.java b/openhtmltopdf-core/src/main/java/com/openhtmltopdf/util/ImageUtil.java
@@ -25,6 +25,7 @@
 import java.util.HashMap;
 import java.util.Map;
 import java.util.logging.Level;
+import java.util.regex.Pattern;
 
 /**
  * Static utility methods for working with images. Meant to suggest "best practices" for the most straightforward
@@ -202,7 +203,13 @@ public static BufferedImage createTransparentImage(int width, int height) {
         g2d.dispose();
         return bi;
     }
-
+
+    private static final Pattern WHITE_SPACE = Pattern.compile("\\s+");
+
+    public static byte[] fromBase64Encoded(String b64encoded) {
+        return Base64.getMimeDecoder().decode(WHITE_SPACE.matcher(b64encoded).replaceAll(""));
+    }
+
     /**
      * Get the binary content of an embedded base 64 image.
      *
@@ -213,7 +220,7 @@ public static byte[] getEmbeddedBase64Image(String imageDataUri) {
         int b64Index = imageDataUri.indexOf("base64,");
         if (b64Index != -1) {
             String b64encoded = imageDataUri.substring(b64Index + "base64,".length());
-            return Base64.getMimeDecoder().decode(b64encoded);
+            return fromBase64Encoded(b64encoded);
         } else {
             XRLog.log(Level.SEVERE, LogMessageId.LogMessageId0Param.LOAD_EMBEDDED_DATA_URI_MUST_BE_ENCODED_IN_BASE64);
         }

diff --git a/...htmltopdf-examples/src/main/resources/visualtest/expected/more-robust-data-uri-base64.pdf b/...htmltopdf-examples/src/main/resources/visualtest/expected/more-robust-data-uri-base64.pdf
diff --git a/openhtmltopdf-examples/src/main/resources/visualtest/html/more-robust-data-uri-base64.html b/openhtmltopdf-examples/src/main/resources/visualtest/html/more-robust-data-uri-base64.html
@@ -0,0 +1,56 @@
+<html>
+<head>
+    <style>
+    @page {margin:0;size: 70px 70px}
+    html, body,img {margin:0;padding:0;}
+    </style>
+</head>
+
+<body>
+<img style="height:64px; width:64px;border:1px solid red;"
+     src="data:image/png;base64,     iVBORw0KGgoAAAANSUhEUgAAAEAAAABACAIAAAAlC+aJAAAACXBIWXMAAA7EAA
+
+
+     AOxAGVKw4bAAAE2klEQVRoge2Za0xbZRjH/x0wb
+
+
+
+
+     mMggzIQmFTAYYjD6WSr2eUD29R42eIy2eYlojHT
+
+     aEycuuiSRWN08QNZBJIZx2VGg1GJjhh1cZMi47K5sDBXx7g
+     OCi1QbqVSRunaxw/zAuSc95z3LeZI0t+n5rz/5/k///a05/Q9O
+     iLCYmaJ1gP4Sy   CA1gQCaE0ggNYs+gDBIkWuUbScxFgvHFZ4PQrioBD
+     EGZCQieQc6DNE7JjouK /EDiuqDsBYgJV3IDoJQSEKeu8MRq5hqA29zbC3IzgUGZuxLh8h4c
+     JDz4G46DxLZXtouIuvajZeDzV/ Tcd3U0M5zUyJ9/kbngAddfTps+R2+
+
+     e/6b4xznxH5/OmkOoC9g8r2Lsz0/+D1UG0JVb1O9g66clqsh7oAE4N0fDe
+     NWcQ8FLj8A320jeo+FqtW9zNqKkbea7glZWG+dvPIfhBxaUi5W6xaRYDOeuiA
+     23LFDJTxXMdQm3C1UgDyoe4Ytr4hbAAA
+     432sVZsZU+MIWirWWylAYwWy8hAeLStoq8HJt1mXM+cQ
+     irYDQGO5tIAI9z+PlBzFWSVhBhjrRXcT1
+     j/D0lyqRss3aK/FQKtsE/LhyL
+     1orxUbkQ07gAVpudAxNSNdA
+
+     PDVq+g8Ky24ef6ER2PwKipfFJ
+     uSAfNeaMyC2FSW4A87nHYAIB9yHpPWxKcjNhVjFgSFoN0kOqcs
+     zHfXOYgoPUsQpcfKLIQtxz1PICpBWpO8BssT/3qdsVloSBbMT2DwKjbuV2iw4320mWAwsjSb
+           9iMxCz3n8fA7vPMpwgww40JopEKDmGTkPqWgMWyAYYP8sg7kU+ggj/wp5LuBkDDhvhwk3on+FuFq+Q
+
+
+           CjvYhdJdyXg9BIeKaFq+UDuEYRuUK4Lx8r0oTvJvj/E4/3ofWMmJks0Ym4PiFWy
+           h/A68GFStxwi/ktOPIBImLg6Jc4HmfAXY/AVPzfzcSFfAB9JqYcGOmWWMrZiYkBNJ0
+           QMexuEqmSh3kKbXsTP34AyW2LXYWwmdFVz21Yd4y7hAkzQJwBhvVoqpBe3X4QphI4rBxu
+           k8P+XLMkUfoSGwvQ8yuunZNYWhaPR9/Dd4c53DrqsGYHh14FSgF0S7DzQ5iKpFf1GbjdiG8PwudV5db6E
+           zI2SRxXWS6Fip/R8GgkrEbHL9KrxueQlI3qQ8p9rpzCMr3E7a3Diq4GJGYpd5BE1d7FtJNK99DEgKzgzFE
+           yFbM6jPZIbytNDFBpPtk7VY0hheqNLZuZqg6wBDVFVLaXLlXT9OSc424XNZRRxZM03D2/xHKRSvP92qgk4tn
+           crT6ErK3I3CIrmBzBhUrYzHNuzoJDkb4Ra3dhacQccVcDGsvxeCEiYrhPm9lwhHW7qOJpGmj15w0jIiIfNZTRF
+           y8tyEYl5+60c5BK88k1Lm7odtGXr5CphHxe8Saz4AxARJZm+vwFQXvnEJXvo7YakVoZ+AMQUWMFnS7krrKaqTS
+           feptFHOXhf0Jzk1NHEBaFLS+rElsvw1SE4FA88BZibhWxk0c0ANRlIB/qP4Htdzx0GFHxgkZM/AgA4Pt3EZ+O+
+           /bNPz45jP7f0HMeNjNW58FY4M+IbP
+
+           wL4JnGz0fRd3HOEzufF5GxWLUOSdlIXev/iGz8C/A/YNE/
+           6A4E0JpAAK0JBNCaQACtCQTQmkAArQkE0Jo/AaHeXcfZZth5AAAAAElFTkSuQmCC    "
+     alt="Smiley"/>
+</body>
+</html>
diff --git a/...-examples/src/test/java/com/openhtmltopdf/visualregressiontests/VisualRegressionTest.java b/...-examples/src/test/java/com/openhtmltopdf/visualregressiontests/VisualRegressionTest.java
@@ -1332,6 +1332,14 @@ public void testIssue615InfiniteLoopBreakWord() throws IOException {
         assertTrue(vt.runTest("issue-615-infinite-loop-break-word"));
     }
 
+    /**
+     * Test that we allow whitespaces inside the base64 encoded data uri.
+     */
+    @Test
+    public void testMoreRobustDataUriBase64() throws IOException {
+        assertTrue(vt.runTest("more-robust-data-uri-base64"));
+    }
+
     // TODO:
     // + Elements that appear just on generated overflow pages.
     // + content property (page counters, etc)

diff --git a/openhtmltopdf-pdfbox/src/main/java/com/openhtmltopdf/pdfboxout/PdfBoxUserAgent.java b/openhtmltopdf-pdfbox/src/main/java/com/openhtmltopdf/pdfboxout/PdfBoxUserAgent.java
@@ -23,14 +23,12 @@
 import java.io.ByteArrayOutputStream;
 import java.io.IOException;
 import java.io.InputStream;
-import java.net.URI;
 import java.util.Locale;
 import java.util.logging.Level;
 
 import com.openhtmltopdf.layout.SharedContext;
 import com.openhtmltopdf.resource.ImageResource;
 import com.openhtmltopdf.swing.NaiveUserAgent;
-import com.openhtmltopdf.util.ImageUtil;
 import com.openhtmltopdf.util.LogMessageId;
 import com.openhtmltopdf.util.XRLog;
 
@@ -77,10 +75,7 @@ public ImageResource getImageResource(String uriStr) {
 
         if (is != null) {
             try {
-                URI uri = new URI(uriStr);
-                if (uri.getPath() != null
-                    && uri.getPath().toLowerCase(Locale.US)
-                                .endsWith(".pdf")) {
+                if (uriStr.toLowerCase(Locale.US).endsWith(".pdf")) {
                     // TODO: Implement PDF AS IMAGE
                     // PdfReader reader = _outputDevice.getReader(uri);
                     // PDFAsImage image = new PDFAsImage(uri);