Make FastXmlSerializer more suitable to persist arbitrary strings

- Encode '\u000' - '\u001F', so KXmlParser can read them properly.
Otherwise KXmlParser will ignore CRs/LFs in attributes, and CRs
in text.

- Originally FastXmlSerializer would throw if a string contains
dangling surrogate pairs.  Now we REPLACE them with.

Bug 27792649

Change-Id: I10c547dad2475b68f60e9e8208d9a3eae8e20063

2 files changed, 117 insertions, 9 deletions

diff --git a/core/java/com/android/internal/util/FastXmlSerializer.java b/core/java/com/android/internal/util/FastXmlSerializer.java
index 7a040801a9a5..3c1d2d6ce581 100644
--- a/core/java/com/android/internal/util/FastXmlSerializer.java
+++ b/core/java/com/android/internal/util/FastXmlSerializer.java
@@ -28,6 +28,7 @@ import java.nio.CharBuffer;
 import java.nio.charset.Charset;
 import java.nio.charset.CharsetEncoder;
 import java.nio.charset.CoderResult;
+import java.nio.charset.CodingErrorAction;
 import java.nio.charset.IllegalCharsetNameException;
 import java.nio.charset.UnsupportedCharsetException;
 
@@ -38,14 +39,14 @@ import java.nio.charset.UnsupportedCharsetException;
  */
 public class FastXmlSerializer implements XmlSerializer {
     private static final String ESCAPE_TABLE[] = new String[] {
-        null,     null,     null,     null,     null,     null,     null,     null,  // 0-7
-        null,     null,     null,     null,     null,     null,     null,     null,  // 8-15
-        null,     null,     null,     null,     null,     null,     null,     null,  // 16-23
-        null,     null,     null,     null,     null,     null,     null,     null,  // 24-31
-        null,     null,     """, null,     null,     null,     "&",  null,  // 32-39
-        null,     null,     null,     null,     null,     null,     null,     null,  // 40-47
-        null,     null,     null,     null,     null,     null,     null,     null,  // 48-55
-        null,     null,     null,     null,     "<",   null,     ">",   null,  // 56-63
+        "�",   "",   "",   "",  "",    "",   "",  "",  // 0-7
+        "",   "	",   "
",  "", "",   "
",  "", "", // 8-15
+        "",  "",  "",  "", "",   "",  "", "", // 16-23
+        "",  "",  "",  "", "",   "",  "", "", // 24-31
+        null,     null,     """, null,     null,     null,     "&",  null,   // 32-39
+        null,     null,     null,     null,     null,     null,     null,     null,   // 40-47
+        null,     null,     null,     null,     null,     null,     null,     null,   // 48-55
+        null,     null,     null,     null,     "<",   null,     ">",   null,   // 56-63
     };
 
     private static final int BUFFER_LEN = 8192;
@@ -310,7 +311,9 @@ public class FastXmlSerializer implements XmlSerializer {
             throw new IllegalArgumentException();
         if (true) {
             try {
-                mCharset = Charset.forName(encoding).newEncoder();
+                mCharset = Charset.forName(encoding).newEncoder()
+                        .onMalformedInput(CodingErrorAction.REPLACE)
+                        .onUnmappableCharacter(CodingErrorAction.REPLACE);
             } catch (IllegalCharsetNameException e) {
                 throw (UnsupportedEncodingException) (new UnsupportedEncodingException(
                         encoding).initCause(e));
diff --git a/core/tests/utiltests/src/com/android/internal/util/FastXmlSerializerTest.java b/core/tests/utiltests/src/com/android/internal/util/FastXmlSerializerTest.java
index 5f36c2d0c29b..3cef33621a01 100644
--- a/core/tests/utiltests/src/com/android/internal/util/FastXmlSerializerTest.java
+++ b/core/tests/utiltests/src/com/android/internal/util/FastXmlSerializerTest.java
@@ -16,16 +16,32 @@
 
 package com.android.internal.util;
 
+import android.test.suitebuilder.annotation.LargeTest;
+import android.test.suitebuilder.annotation.SmallTest;
+import android.util.Log;
+import android.util.Xml;
+
 import junit.framework.TestCase;
 
+import org.xmlpull.v1.XmlPullParser;
 import org.xmlpull.v1.XmlSerializer;
 
+import java.io.ByteArrayInputStream;
 import java.io.ByteArrayOutputStream;
+import java.nio.charset.StandardCharsets;
 
 /**
  * Tests for {@link FastXmlSerializer}
  */
+@SmallTest
 public class FastXmlSerializerTest extends TestCase {
+    private static final String TAG = "FastXmlSerializerTest";
+
+    private static final boolean ENABLE_DUMP = false; // DO NOT SUBMIT WITH TRUE.
+
+    private static final String ROOT_TAG = "root";
+    private static final String ATTR = "attr";
+
     public void testEmptyText() throws Exception {
         final ByteArrayOutputStream stream = new ByteArrayOutputStream();
 
@@ -44,4 +60,93 @@ public class FastXmlSerializerTest extends TestCase {
         assertEquals("<?xml version='1.0' encoding='utf-8' standalone='yes' ?>\n"
                 + "<string name=\"meow\"></string>\n", stream.toString());
     }
+
+    private boolean checkPreserved(String description, String str) {
+        boolean ok = true;
+        byte[] data;
+        try (final ByteArrayOutputStream baos = new ByteArrayOutputStream()) {
+            final XmlSerializer out = new FastXmlSerializer();
+            out.setOutput(baos, StandardCharsets.UTF_16.name());
+            out.startDocument(null, true);
+
+            out.startTag(null, ROOT_TAG);
+            out.attribute(null, ATTR, str);
+            out.text(str);
+            out.endTag(null, ROOT_TAG);
+
+            out.endDocument();
+            baos.flush();
+            data = baos.toByteArray();
+        } catch (Exception e) {
+            Log.e(TAG, "Unable to serialize: " + description, e);
+            return false;
+        }
+
+        if (ENABLE_DUMP) {
+            Log.d(TAG, "Dump:");
+            Log.d(TAG, new String(data));
+        }
+
+        try (final ByteArrayInputStream baos = new ByteArrayInputStream(data)) {
+            XmlPullParser parser = Xml.newPullParser();
+            parser.setInput(baos, StandardCharsets.UTF_16.name());
+
+            int type;
+            String tag = null;
+            while ((type = parser.next()) != XmlPullParser.END_DOCUMENT) {
+                if (type == XmlPullParser.START_TAG) {
+                    tag = parser.getName();
+                    if (ROOT_TAG.equals(tag)) {
+                        String read = parser.getAttributeValue(null, ATTR);
+                        if (!str.equals(read)) {
+                            Log.e(TAG, "Attribute not preserved: " + description
+                                    + " input=\"" + str + "\", but read=\"" + read + "\"");
+                            ok = false;
+                        }
+                    }
+                }
+                if (type == XmlPullParser.TEXT && ROOT_TAG.equals(tag)) {
+                    String read = parser.getText();
+                    if (!str.equals(parser.getText())) {
+                        Log.e(TAG, "Text not preserved: " + description
+                                + " input=\"" + str + "\", but read=\"" + read + "\"");
+                        ok = false;
+                    }
+                }
+            }
+        } catch (Exception e) {
+            Log.e(TAG, "Unable to parse: " + description, e);
+            return false;
+        }
+        return ok;
+    }
+
+    private boolean check(String description, String str) throws Exception {
+        boolean ok = false;
+        ok |= checkPreserved(description, str);
+        ok |= checkPreserved(description + " wrapped with spaces" ,"  " + str + "  ");
+        return ok;
+    }
+
+    @LargeTest
+    public void testAllCharacters() throws Exception {
+        boolean ok = true;
+        for (int i = 0; i < 0xffff; i++) {
+            if (0xd800 <= i && i <= 0xdfff) {
+                // Surrogate pair characters.
+                continue;
+            }
+            ok &= check("char: " + i, String.valueOf((char) i));
+        }
+        // Dangling surrogate pairs. We can't preserve them.
+        assertFalse(check("+ud800", "\ud800"));
+        assertFalse(check("+udc00", "\udc00"));
+
+        for (int i = 0xd800; i < 0xdc00; i ++) {
+            for (int j = 0xdc00; j < 0xe000; j++) {
+                ok &= check("char: " + i, String.valueOf((char) i) + String.valueOf((char) j));
+            }
+        }
+        assertTrue("Some tests failed.  See logcat for details.", ok);
+    }
 }