修复部分ofd文件格式特殊，用java标准库zip包无法解压的问题。引入apache commons compress来解压

purewind · Mar 1, 2022 · c0d7a2e · c0d7a2e
1 parent 9d32841
commit c0d7a2e
Show file tree

Hide file tree

Showing 9 changed files with 102 additions and 41 deletions.
diff --git a/ofdrw-converter/src/test/java/OFD2HTMLTest.java b/ofdrw-converter/src/test/java/OFD2HTMLTest.java
@@ -20,8 +20,11 @@ public class OFD2HTMLTest {
     public void convertHtml() {
         try {
             // 1. 提供文档
-            Path ofdIn = Paths.get("src/test/resources/n.ofd");
-            Path htmlOut = Paths.get("target/n.html");
+//            Path ofdIn = Paths.get("src/test/resources/n.ofd");
+//            Path htmlOut = Paths.get("target/n.html");
+            //发票示例.ofd
+            Path ofdIn = Paths.get("src/test/resources/发票示例.ofd");
+            Path htmlOut = Paths.get("target/发票示例.ofd/发票示例.html");
             // 2. [可选]配置字体，别名，扫描目录等
             // FontLoader.getInstance().addAliasMapping(null, "小标宋体", "方正小标宋简体", "方正小标宋简体")
             // FontLoader.getInstance().scanFontDir(new File("src/test/resources/fonts"));

diff --git a/ofdrw-converter/src/test/java/OFD2IMGTest.java b/ofdrw-converter/src/test/java/OFD2IMGTest.java
@@ -19,22 +19,31 @@ public class OFD2IMGTest {
     public void test() throws IOException {
 
         //为不规范的字体名创建映射
-//        FontLoader.getInstance()
-//                .addAliasMapping("仿宋简体", "方正仿宋简体")
-//                .addAliasMapping("仿宋", "方正仿宋简体")
-//                .addAliasMapping("小标宋体", "方正小标宋简体")
-//                .addAliasMapping("KaiTi_GB2312", "楷体");
-        FontLoader.getInstance();
+        FontLoader.getInstance()
+                .addAliasMapping(null, "小标宋体", "方正小标宋简体", "方正小标宋简体")
+                .addAliasMapping(null, "KaiTi_GB2312", "楷体", "楷体")
+                .addAliasMapping(null, "楷体", "KaiTi", "KaiTi")
+
+                .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
+                .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
+                .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun")
+                .addSimilarFontReplaceRegexMapping(null, ".*SimSun.*", null, "SimSun")
+                .addSimilarFontReplaceRegexMapping(null, ".*Song.*", null, "宋体")
+                .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun");
+
+        FontLoader.getInstance().scanFontDir(new File("src/main/resources/fonts"));
+        FontLoader.setSimilarFontReplace(true);
         long start = System.currentTimeMillis();
-        toPng("src/test/resources/helloworld.ofd", "target/helloworld.ofd");
-        toPng("src/test/resources/999.ofd", "target/999.ofd");
-        toPng("src/test/resources/zsbk.ofd", "target/zsbk.ofd");
-        toPng("src/test/resources/ano.ofd", "target/ano.ofd");
-        toPng("src/test/resources/文字横向-数科.ofd", "target/文字横向-数科.ofd");
-        toPng("src/test/resources/z.ofd", "target/z.ofd");
-        toPng("src/test/resources/不规范资源路径.ofd", "target/不规范资源路径.ofd");
-        toPng("src/test/resources/V4RideRight.ofd", "target/V4RideRight.ofd");
-        toPng("src/test/resources/发票示例.ofd", "target/发票示例.ofd");
+//        toPng("src/test/resources/helloworld.ofd", "target/helloworld.ofd");
+//        toPng("src/test/resources/999.ofd", "target/999.ofd");
+//        toPng("src/test/resources/zsbk.ofd", "target/zsbk.ofd");
+//        toPng("src/test/resources/ano.ofd", "target/ano.ofd");
+//        toPng("src/test/resources/文字横向-数科.ofd", "target/文字横向-数科.ofd");
+//        toPng("src/test/resources/z.ofd", "target/z.ofd");
+//        toPng("src/test/resources/不规范资源路径.ofd", "target/不规范资源路径.ofd");
+        toPng("src/test/resources/安证通手动盖章.ofd", "target/安证通手动盖章.ofd");
+//        toPng("src/test/resources/金格信创自动盖章.ofd", "target/金格信创自动盖章.ofd");
+//        toPng("src/test/resources/发票示例.ofd", "target/发票示例.ofd");
         System.out.printf(">> 总计花费: %dms\n", System.currentTimeMillis() - start);
     }
 

diff --git a/ofdrw-converter/src/test/java/OFD2PDFTest.java b/ofdrw-converter/src/test/java/OFD2PDFTest.java
@@ -3,6 +3,7 @@
 import org.ofdrw.converter.FontLoader;
 import org.ofdrw.converter.GeneralConvertException;
 
+import java.io.File;
 import java.nio.file.Paths;
 
 public class OFD2PDFTest {
@@ -15,23 +16,30 @@ public void convertPdf() {
 //        Path src = Paths.get("src/test/resources/zsbk.ofd");
 //        Path dst = Paths.get("target/zsbk.pdf");
 
+//为不规范的字体名创建映射
         FontLoader.getInstance()
-            .addAliasMapping(null, "小标宋体", "方正小标宋简体", "方正小标宋简体")
-            .addAliasMapping(null, "KaiTi_GB2312", "楷体", "楷体")
+                .addAliasMapping(null, "小标宋体", "方正小标宋简体", "方正小标宋简体")
+                .addAliasMapping(null, "KaiTi_GB2312", "楷体", "楷体")
+                .addAliasMapping(null, "楷体", "KaiTi", "KaiTi")
 
-            .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
-            .addSimilarFontReplaceRegexMapping(null, ".*SimSun.*", null, "SimSun")
-            .addSimilarFontReplaceRegexMapping(null, ".*Song.*", null, "宋体")
-            .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun");
+                .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
+                .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
+                .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun")
+                .addSimilarFontReplaceRegexMapping(null, ".*SimSun.*", null, "SimSun")
+                .addSimilarFontReplaceRegexMapping(null, ".*Song.*", null, "宋体")
+                .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun");
 
+        FontLoader.getInstance().scanFontDir(new File("src/main/resources/fonts"));
         FontLoader.setSimilarFontReplace(true);
-
+        long start = System.currentTimeMillis();
         try {
 //            ConvertHelper.toPdf(src, dst);
-//            ConvertHelper.toPdf(Paths.get("src/test/resources/signout.ofd"), Paths.get("target/signout.pdf"));
-            ConvertHelper.toPdf(Paths.get("src/test/resources/zsbk.ofd"), Paths.get("target/zsbk.pdf"));
-//            ConvertHelper.toPdf(Paths.get("src/test/resources/999.ofd"), Paths.get("target/999.pdf"));
+//            ConvertHelper.toPdf(Paths.get("src/test/resources/发票示例.ofd"), Paths.get("target/发票示例.pdf"));
+            ConvertHelper.toPdf(Paths.get("src/test/resources/金格信创自动盖章.ofd"), Paths.get("target/金格信创自动盖章.pdf"));
 
+//            ConvertHelper.toPdf(Paths.get("src/test/resources/zsbk.ofd"), Paths.get("target/zsbk.pdf"));
+//            ConvertHelper.toPdf(Paths.get("src/test/resources/999.ofd"), Paths.get("target/999.pdf"));
+            System.out.printf(">> 总计花费: %dms\n", System.currentTimeMillis() - start);
         } catch (GeneralConvertException e) {
             e.printStackTrace();
         }

diff --git a/ofdrw-converter/src/test/java/OFD2SVGTest.java b/ofdrw-converter/src/test/java/OFD2SVGTest.java
@@ -3,6 +3,7 @@
 import org.ofdrw.converter.SVGMaker;
 import org.ofdrw.reader.OFDReader;
 
+import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.OutputStream;
@@ -26,6 +27,7 @@ public void test() throws IOException {
         FontLoader.getInstance()
             .addAliasMapping(null, "小标宋体", "方正小标宋简体", "方正小标宋简体")
             .addAliasMapping(null, "KaiTi_GB2312", "楷体", "楷体")
+            .addAliasMapping(null, "楷体", "KaiTi", "KaiTi")
 
             .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
             .addSimilarFontReplaceRegexMapping(null, ".*Kai.*", null, "楷体")
@@ -34,18 +36,18 @@ public void test() throws IOException {
             .addSimilarFontReplaceRegexMapping(null, ".*Song.*", null, "宋体")
             .addSimilarFontReplaceRegexMapping(null, ".*MinionPro.*", null, "SimSun");
 
+        FontLoader.getInstance().scanFontDir(new File("src/main/resources/fonts"));
         FontLoader.setSimilarFontReplace(true);
-
-
         long start = System.currentTimeMillis();
 //        toSVG("src/test/resources/999.ofd", "target/999.ofd");
 //        toSVG("src/test/resources/zsbk.ofd", "target/zsbk.ofd");
 //        toSVG("src/test/resources/ano.ofd", "target/ano.ofd");
 //        toSVG("src/test/resources/文字横向-数科.ofd", "target/文字横向-数科.ofd");
 //        toSVG("src/test/resources/z.ofd", "target/z.ofd");
-//        toSVG("src/test/resources/发票示例.ofd", "target/发票示例.ofd");
-        toSVG("src/test/resources/signout.ofd", "target/signout.ofd");
-        toSVG("src/test/resources/n.ofd", "target/n.ofd");
+//        toSVG("src/test/resources/金格信创自动盖章.ofd", "target/安证通手动盖章.ofd");
+        toSVG("src/test/resources/发票示例.ofd", "target/发票示例.ofd");
+//        toSVG("src/test/resources/signout.ofd", "target/signout.ofd");
+//        toSVG("src/test/resources/n.ofd", "target/n.ofd");
 
 //        toPng("src/test/resources/不规范资源路径.ofd", "target/不规范资源路径.ofd");
         System.out.printf(">> 总计花费: %dms\n", System.currentTimeMillis() - start);

diff --git a/ofdrw-converter/src/test/resources/安证通手动盖章.ofd b/ofdrw-converter/src/test/resources/安证通手动盖章.ofd
diff --git a/ofdrw-converter/src/test/resources/金格信创自动盖章.ofd b/ofdrw-converter/src/test/resources/金格信创自动盖章.ofd
diff --git a/ofdrw-reader/pom.xml b/ofdrw-reader/pom.xml
@@ -19,6 +19,11 @@
             <version>${project.version}</version>
         </dependency>
 
+        <dependency>
+            <groupId>org.apache.commons</groupId>
+            <artifactId>commons-compress</artifactId>
+            <version>1.20</version>
+        </dependency>
         <dependency>
             <groupId>org.junit.jupiter</groupId>
             <artifactId>junit-jupiter-api</artifactId>

diff --git a/ofdrw-reader/src/main/java/org/ofdrw/reader/OFDReader.java b/ofdrw-reader/src/main/java/org/ofdrw/reader/OFDReader.java
@@ -104,7 +104,7 @@ public OFDReader(Path ofdFile) throws IOException {
         }
         workDir = Files.createTempDirectory("ofd-tmp-");
         // 解压文档，到临时的工作目录
-        ZipUtil.unZipFiles(ofdFile.toFile(), workDir.toAbsolutePath().toString() + File.separator);
+        ZipUtil.unZipFileByApacheCommonCompress(ofdFile.toFile(), workDir.toAbsolutePath().toString() + File.separator);
         ofdDir = new OFDDir(workDir);
         // 创建资源定位器
         rl = new ResourceLocator(ofdDir);

diff --git a/ofdrw-reader/src/main/java/org/ofdrw/reader/ZipUtil.java b/ofdrw-reader/src/main/java/org/ofdrw/reader/ZipUtil.java
@@ -1,18 +1,22 @@
 package org.ofdrw.reader;
 
+import org.apache.commons.compress.archivers.ArchiveEntry;
+import org.apache.commons.compress.archivers.zip.ZipArchiveInputStream;
+import org.apache.commons.compress.utils.IOUtils;
+
 import java.io.*;
 import java.nio.charset.Charset;
 import java.nio.file.Files;
-import java.nio.file.Path;
 import java.nio.file.Paths;
+import java.util.Enumeration;
 import java.util.zip.ZipEntry;
+import java.util.zip.ZipFile;
 import java.util.zip.ZipInputStream;
 
 public class ZipUtil {
-
     /**
      * 解压许可最大字节数，为了防止 ZIP炸弹攻击
-     *
+     * <p>
      * 默认值： 100M
      */
     private static long MaxSize = 100 * 1024 * 1024;
@@ -103,11 +107,7 @@ public static void unZipFiles(InputStream src, String descDir) throws IOExceptio
      * @throws IOException 文件操作IO异常
      */
     public static void unZipFiles(File zipFile, String descDir) throws IOException {
-
-        try (InputStream in = new FileInputStream(zipFile)) {
-            unZipFiles(in, descDir);
-        }
-
+        unZipFileByApacheCommonCompress(zipFile, descDir);
     }
 
     /**
@@ -121,4 +121,38 @@ private static void pathValid(String targetDir, String filePath) throws IOExcept
         if (!filePath.startsWith(targetDir))
             throw new IOException(String.format("不合法的路径：%s", filePath));
     }
+
+    /**
+     * 使用apache common compress库 解压zipFile，能支持更多zip包解压的特性
+     * @param srcFile 带解压的源文件
+     * @param descDir 解压到目录
+     * @throws IOException
+     */
+    public static void unZipFileByApacheCommonCompress(File srcFile, String descDir) throws IOException {
+        File pathFile = new File(descDir).getCanonicalFile();
+        if (!pathFile.exists() && !pathFile.mkdirs()) {
+            throw new IOException("failed to create directory " + pathFile);
+        }
+        try (ZipFile zipFile = new ZipFile(srcFile)) {
+            ZipEntry entry = null;
+            Enumeration<? extends ZipEntry> entries = zipFile.entries();
+            while (entries.hasMoreElements()) {
+                entry = entries.nextElement();
+                File f = new File(pathFile, entry.getName());
+                if (entry.isDirectory()) {
+                    if (!f.isDirectory() && !f.mkdirs()) {
+                        throw new IOException("failed to create directory " + f);
+                    }
+                } else {
+                    File parent = f.getParentFile();
+                    if (!parent.isDirectory() && !parent.mkdirs()) {
+                        throw new IOException("failed to create directory " + parent);
+                    }
+                    try (OutputStream o = Files.newOutputStream(f.toPath())) {
+                        IOUtils.copy(zipFile.getInputStream(entry), o);
+                    }
+                }
+            }
+        }
+    }
 }