[align=left]/**
* 读取word2003
*/
public String readExtDocContent(String fileName){
// 创建输入流读取DOC文件
InputStream in = null;
WordExtractor extractor = null;
String text = null;
try {
in = new FileInputStream(new File(fileName));
extractor = new WordExtractor(in);
text = extractor.getText();
} catch (FileNotFoundException e) {
e.printStackTrace();
}catch (IOException e) {
e.printStackTrace();
} finally{
try {
if(in != null){
in.close();
}
} catch (IOException e) {
e.printStackTrace();
}
}
return text;
}
/**
* 读取word2007-2013
*/
public String readExtDocxContent(String fileName){
// 创建输入流读取DOC文件
String text = null;
OPCPackage opcPackage = null;
try {
opcPackage = POIXMLDocument.openPackage(fileName);
XWPFDocument xwpf = new XWPFDocument(opcPackage);
POIXMLTextExtractor ex = new XWPFWordExtractor(xwpf);
text = ex.getText();
} catch (Exception e) {
e.printStackTrace();
log.error("转换word异常:" + e);
} finally{
try {
if(opcPackage != null){
opcPackage.close();
}
} catch (IOException e) {
e.printStackTrace();
}
}
return text;
}
/**
* 读取Excel2003
*/
public String readExtXlsContent(String fileName){
String content = "";
try {
HSSFWorkbook workbook = new HSSFWorkbook(new FileInputStream(fileName));
ExcelExtractor extractor = new ExcelExtractor(workbook);
extractor.setFormulasNotResults(true);
extractor.setIncludeSheetNames(false);
content = extractor.getText();
log.debug("content:" + content);
} catch (Exception e) {
e.printStackTrace();
return "";
}
return content;
}
/**
* 读取Excel2007+
*/
public String readExtXlsxContent(String fileName){
String content = "";
try {
XSSFWorkbook workbook = new XSSFWorkbook(new FileInputStream(fileName));
XSSFExcelExtractor extractor = new XSSFExcelExtractor(workbook);
extractor.setFormulasNotResults(true);
extractor.setIncludeSheetNames(false);
content = extractor.getText();
} catch (Exception e) {
e.printStackTrace();
return "";
}
return content;
}
/**
* 读取ppt2003
*/
public String readExtPptContent(String fileName){
String text = "";
try {
SlideShow ss = new SlideShow(new HSLFSlideShow(new FileInputStream(fileName)));
Slide[] slides = ss.getSlides();
for (int i = 0; i < slides.length; i++) {
TextRun[] t = slides[i].getTextRuns();
for (int j = 0; j < t.length; j++) {
text += t[j].getText();
}
}
} catch (Exception e) {
e.printStackTrace();
}
return text;
}
/**
* 读取ppt2007+
*/
public String readExtPptxContent(String fileName){
String text = "";
XMLSlideShow xmlslideshow = null;
try {
xmlslideshow = new XMLSlideShow(new FileInputStream(fileName));
XSLFPowerPointExtractor ppt = new XSLFPowerPointExtractor(xmlslideshow);
text= ppt.getText();
} catch (FileNotFoundException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
}
return text;
}
/**
* 读取pdf
*/
public String readExtPDFContent(String fileName){
String text = "";
try {
PDFParser parser = new PDFParser(new FileInputStream(fileName));
parser.parse();
PDFTextStripper stripper = new PDFTextStripper();
text = stripper.getText(parser.getPDDocument());
parser.getPDDocument().close();
} catch (Exception e) {
e.printStackTrace();
return "";
}
return text;
}[/align]
* 读取word2003
*/
public String readExtDocContent(String fileName){
// 创建输入流读取DOC文件
InputStream in = null;
WordExtractor extractor = null;
String text = null;
try {
in = new FileInputStream(new File(fileName));
extractor = new WordExtractor(in);
text = extractor.getText();
} catch (FileNotFoundException e) {
e.printStackTrace();
}catch (IOException e) {
e.printStackTrace();
} finally{
try {
if(in != null){
in.close();
}
} catch (IOException e) {
e.printStackTrace();
}
}
return text;
}
/**
* 读取word2007-2013
*/
public String readExtDocxContent(String fileName){
// 创建输入流读取DOC文件
String text = null;
OPCPackage opcPackage = null;
try {
opcPackage = POIXMLDocument.openPackage(fileName);
XWPFDocument xwpf = new XWPFDocument(opcPackage);
POIXMLTextExtractor ex = new XWPFWordExtractor(xwpf);
text = ex.getText();
} catch (Exception e) {
e.printStackTrace();
log.error("转换word异常:" + e);
} finally{
try {
if(opcPackage != null){
opcPackage.close();
}
} catch (IOException e) {
e.printStackTrace();
}
}
return text;
}
/**
* 读取Excel2003
*/
public String readExtXlsContent(String fileName){
String content = "";
try {
HSSFWorkbook workbook = new HSSFWorkbook(new FileInputStream(fileName));
ExcelExtractor extractor = new ExcelExtractor(workbook);
extractor.setFormulasNotResults(true);
extractor.setIncludeSheetNames(false);
content = extractor.getText();
log.debug("content:" + content);
} catch (Exception e) {
e.printStackTrace();
return "";
}
return content;
}
/**
* 读取Excel2007+
*/
public String readExtXlsxContent(String fileName){
String content = "";
try {
XSSFWorkbook workbook = new XSSFWorkbook(new FileInputStream(fileName));
XSSFExcelExtractor extractor = new XSSFExcelExtractor(workbook);
extractor.setFormulasNotResults(true);
extractor.setIncludeSheetNames(false);
content = extractor.getText();
} catch (Exception e) {
e.printStackTrace();
return "";
}
return content;
}
/**
* 读取ppt2003
*/
public String readExtPptContent(String fileName){
String text = "";
try {
SlideShow ss = new SlideShow(new HSLFSlideShow(new FileInputStream(fileName)));
Slide[] slides = ss.getSlides();
for (int i = 0; i < slides.length; i++) {
TextRun[] t = slides[i].getTextRuns();
for (int j = 0; j < t.length; j++) {
text += t[j].getText();
}
}
} catch (Exception e) {
e.printStackTrace();
}
return text;
}
/**
* 读取ppt2007+
*/
public String readExtPptxContent(String fileName){
String text = "";
XMLSlideShow xmlslideshow = null;
try {
xmlslideshow = new XMLSlideShow(new FileInputStream(fileName));
XSLFPowerPointExtractor ppt = new XSLFPowerPointExtractor(xmlslideshow);
text= ppt.getText();
} catch (FileNotFoundException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
}
return text;
}
/**
* 读取pdf
*/
public String readExtPDFContent(String fileName){
String text = "";
try {
PDFParser parser = new PDFParser(new FileInputStream(fileName));
parser.parse();
PDFTextStripper stripper = new PDFTextStripper();
text = stripper.getText(parser.getPDDocument());
parser.getPDDocument().close();
} catch (Exception e) {
e.printStackTrace();
return "";
}
return text;
}[/align]