html文件转换成pdf和word

1、html文件转成pdf

采用jar包有itext-asian.jar、itextpdf-5.5.5.jar、itext-pdfa-5.5.5.jar、itext-xtra-5.5.5.jar，为了保持html页面的全部格式，需要进行相关设置，代码如下：

private boolean convert2Pdf(InputStream htmlInputStream, String pdfFile, String padding) throws FileNotFoundException, DocumentException  {
String[] paddings = padding.split(",");
float a1 = (float) (Float.valueOf(paddings[0]) * 0.77);  //为了设置pdf的上下左右页边距
float a2 = (float) (Float.valueOf(paddings[1]) * 0.77);
float a3 = (float) (Float.valueOf(paddings[2]) * 0.77);
float a4 = (float) (Float.valueOf(paddings[3]) * 0.77);
com.itextpdf.text.Document document = new com.itextpdf.text.Document(
PageSize.A4, a1, a2, a3, a4);
PdfWriter pdfwriter = PdfWriter.getInstance(document, new FileOutputStream(pdfFile));

document.open();
CSSResolver cssResolver = new StyleAttrCSSResolver();

// HTML
XMLWorkerFontProvider fontProvider = new XMLWorkerFontProvider(XMLWorkerFontProvider.DONTLOOKFORFONTS);
fontProvider.register("resources/garial.ttf", "Linrial");  //前面的ttf字体需要系统支持，后一个参数是html文件中的字体格式
fontProvider.register("resources/fzst_gb18030_20101201.ttf", "FZSongTi_GB18030");
fontProvider.register("resources/fzfangsong.ttf", "FZFangSong");
fontProvider.register("resources/fzheiti.ttf", "FZHeiTi");
fontProvider.register("resources/fzkaiti.ttf", "FZKaiTi");
fontProvider.register("resources/gtimes.ttf", "LinTimes");
CssAppliers cssAppliers = new CssAppliersImpl(fontProvider);
HtmlPipelineContext htmlContext = new HtmlPipelineContext(cssAppliers);
htmlContext.setTagFactory(Tags.getHtmlTagProcessorFactory());

final String IMG_PATH = pdfFile.substring(0, pdfFile.indexOf("/temp")) + "temp/";  //指定html文件的图片路径
htmlContext.setImageProvider(new AbstractImageProvider() {
public String getImageRootPath() {
return IMG_PATH;
}
});
// Pipelines
PdfWriterPipeline pdf = new PdfWriterPipeline(document, pdfwriter);
HtmlPipeline html = new HtmlPipeline(htmlContext, pdf);
CssResolverPipeline css = new CssResolverPipeline(cssResolver, html);

// XML Worker
XMLWorker worker = new XMLWorker(css, true);
XMLParser p = new XMLParser(worker);
try {
p.parse(htmlInputStream, Charset.forName("UTF-8"));
} catch (IOException e) {
e.printStackTrace();
return false;
} finally {
document.close();
}
return true;
}

使用方式

convert2Pdf(new ByteArrayInputStream(html.getBytes()), "/root/123.pdf", padding)，其中html为html文件，需要转换为inputstream

2、html转为word

采用openoffice或者中标office，需要最新版本，jar包为jodconverter-2.2.2.jar、jodconverter-cli-2.2.2.jar需要把openoffice和web服务安装在同一个机器上

private boolean startOfficeService() {
final String paramStr = "\'socket,host=localhost,port=8100;urp;\'";
String pro = "/opt/neoshineoffice/program/soffice.bin -headless -accept="
+ paramStr;
System.out.println("---中标Office后台服务启动中...>>>");
try {
Runtime.getRuntime().exec(new String[] { "/bin/sh", "-c", pro }); // 创建一个office服务进程；
} catch (Exception e) {
System.out.println("中标Office后台服务启动失败<<<");
e.printStackTrace();
return false;
}
System.out.println("中标Office后台服务启动成功<<<");
return true;
}

private com.artofsolving.jodconverter.openoffice.connection.OpenOfficeConnection connectOfficeService() {
System.out.println("开始准备连接OFFICE服务...>>>");
com.artofsolving.jodconverter.openoffice.connection.OpenOfficeConnection connection = new com.artofsolving.jodconverter.openoffice.connection.SocketOpenOfficeConnection(
"localhost", 8100);
Boolean flag = false;
int count = 0; // 初始连接OFFICE失败后再次连接的次数；
while (!flag) {
try {
count++;
Thread.sleep((count + 1) * 1000);
connection.connect();
flag = true;
System.out.println(connection);

} catch (Exception e) {
System.out.println("OFFICE服务连接失败<<<");
System.out.println("第" + count + "次重启OFFICE服务连接");
if (count > 5) { // 连接OFFICE失败5次后，停止连接；
e.printStackTrace();
System.out.println("-----无法启动OFFICE服务连接，请检查OFFICE安装情况！----错误信息："+ e.getMessage());
return null;
}
}
}
System.out.println("OFFICE服务连接成功<<<");
return connection;
}

if (startOfficeService()) {
OpenOfficeConnection conn = connectOfficeService();
if (conn != null) {
DocumentConverter convert = new OpenOfficeDocumentConverter(conn);
convert.convert(new File(rootPath + "/temp/resulted.html"), new File(rootPath + "/temp/" + reportName + ".doc"));
conn.disconnect();
}
}