java读取docx_Java读取doc、docx、xls、xlsx、ppt、pptx、pdf文件内容

读取文件信息所需依赖

org.apache.poi

poi

4.1.2

org.apache.poi

poi-scratchpad

4.1.2

org.apache.poi

poi-ooxml

4.1.2

org.apache.pdfbox

pdfbox

2.0.12

org.apache.pdfbox

fontbox

2.0.12

读取doc文件内容

public static String readWord(String name)

{

FileInputStream in;

String text = null;

try

{

in = new FileInputStream(name);

WordExtractor extractor = new WordExtractor(in);

text = extractor.getText();

}

catch (FileNotFoundException e)

{

// TODO Auto-generated catch block

e.printStackTrace();

} catch (Exception e) {

// TODO Auto-generated catch block

e.printStackTrace();

}

return text;

}

读取docx文件内容

public static String readDoc(MultipartFile file) {

if (file.isEmpty())return "";

WordExtractor wordExtractor = null;

try {

InputStream inputStream = file.getInputStream();

wordExtractor = new WordExtractor(inputStream);

} catch (IOException e) {

log.warn(e.toString());

e.printStackTrace();

}

return wordExtractor.getText();

}

读取xls文件内容

public static