Clean room implementation of detectCharsetFromBOM.

Drops dependency. Closes GitHub issue #400.
LibrePDF · Aug 28, 2020 · abd765e · abd765e
1 parent 630c9c5
commit abd765e
Show file tree

Hide file tree

Showing 3 changed files with 35 additions and 11 deletions.
diff --git a/README.md b/README.md
@@ -60,7 +60,6 @@ Make sure that your contributions can be released with a dual LGPL and MPL licen
 ## Dependencies ##
 ### Required Dependencies: ###
  - Java 8 or later is required to use OpenPDF. All versions Java 8 to Java OpenJDK 13 have been tested to work.
- - [Juniversalchardet](https://github.com/albfernandez/juniversalchardet)
 
 ### Optional: ###
 
@@ -94,7 +93,7 @@ Significant [Contributors to OpenPDF](https://github.com/LibrePDF/OpenPDF/graphs
   [@ubermichael](https://github.com/ubermichael) - Michael Joyce  
   [@weiyeh](https://github.com/weiyeh)  
   [@SuperPat45](https://github.com/SuperPat45)  
-  [@lapo-luchini](https://github.com/lapo-luchini)  
+  [@lapo-luchini](https://github.com/lapo-luchini)  - Lapo Luchini
   [@MartinKocour](https://github.com/MartinKocour)  - Martin Kocour  
   [@jokimaki](https://github.com/jokimaki)  
   [@sullis](https://github.com/sullis)  

diff --git a/openpdf/pom.xml b/openpdf/pom.xml
@@ -13,16 +13,10 @@
 
     <properties>
         <java-module-name>com.github.librepdf.openpdf</java-module-name>
-        <juniversalchardet.version>2.3.2</juniversalchardet.version>
         <imageio-tiff.version>3.5</imageio-tiff.version>
     </properties>
 
     <dependencies>
-        <dependency>
-            <groupId>com.github.albfernandez</groupId>
-            <artifactId>juniversalchardet</artifactId>
-            <version>${juniversalchardet.version}</version>
-        </dependency>
         <dependency>
             <groupId>org.bouncycastle</groupId>
             <artifactId>bcprov-jdk15on</artifactId>

diff --git a/openpdf/src/main/java/com/lowagie/text/xml/simpleparser/SimpleXMLParser.java b/openpdf/src/main/java/com/lowagie/text/xml/simpleparser/SimpleXMLParser.java
@@ -44,7 +44,6 @@
 package com.lowagie.text.xml.simpleparser;
 
 import com.lowagie.text.error_messages.MessageLocalization;
-import org.mozilla.universalchardet.UniversalDetector;
 
 import java.io.BufferedReader;
 import java.io.ByteArrayOutputStream;
@@ -550,7 +549,39 @@ public static void parse(SimpleXMLDocHandler doc, SimpleXMLDocHandlerComment com
         SimpleXMLParser parser = new SimpleXMLParser(doc, comment, html);
         parser.go(r);
     }
-
+
+    /** Detect charset from BOM, as per <a href="https://unicode.org/faq/utf_bom.html#bom4">Unicode FAQ</a>. */
+    private static String detectCharsetFromBOM(byte[] bom) {
+        // 00 00 FE FF  UTF-32BE
+        // EF BB BF ..  UTF-8
+        // FE FF .. ..  UTF-16BE
+        // FF FE 00 00  UTF-32LE
+        // FF FE .. ..  UTF-16LE
+        switch (bom[0]) {
+            case (byte) 0x00:
+                if (bom[1] == (byte) 0x00 && bom[2] == (byte) 0xFE && bom[3] == (byte) 0xFF)
+                    return "UTF-32BE";
+                break;
+            case (byte) 0xEF:
+                if (bom[1] == (byte) 0xBB && bom[2] == (byte) 0xBF)
+                    return "UTF-8";
+                break;
+            case (byte) 0xFE:
+                if (bom[1] == (byte) 0xFF)
+                    return "UTF-16BE";
+                break;
+            case (byte) 0xFF:
+                if (bom[1] == (byte) 0xFE) {
+                    if (bom[2] == (byte) 0x00 && bom[3] == (byte) 0x00)
+                        return "UTF-32LE";
+                    else
+                        return "UTF-16LE";
+                }
+                break;
+        }
+        return null;
+    }
+
     /**
      * Parses the XML document firing the events to the handler.
      * @param doc the document handler
@@ -562,7 +593,7 @@ public static void parse(SimpleXMLDocHandler doc, InputStream in) throws IOExcep
         int count = in.read(b4);
         if (count != 4)
             throw new IOException(MessageLocalization.getComposedMessage("insufficient.length"));
-        String encoding = UniversalDetector.detectCharsetFromBOM(b4);
+        String encoding = detectCharsetFromBOM(b4);
         if (encoding == null) encoding = "UTF-8"; //UTF-8 is default.
 
         String decl = null;