java excel大数据量导入导出与优化(修复excel单元格中有空值异常退出)

package com.hundsun.ta.utils;

import java.io.File;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.io.PrintStream;
import java.lang.reflect.Field;
import java.lang.reflect.InvocationTargetException;
import java.lang.reflect.Method;
import java.math.BigDecimal;
import java.text.DateFormat;
import java.text.SimpleDateFormat;
import java.util.ArrayList;
import java.util.Date;
import java.util.HashMap;
import java.util.Iterator;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;

import javax.xml.parsers.ParserConfigurationException;
import javax.xml.parsers.SAXParser;
import javax.xml.parsers.SAXParserFactory;

import org.apache.commons.beanutils.BeanUtils;
import org.apache.commons.collections.MapUtils;
import org.apache.poi.hssf.usermodel.HSSFDateUtil;
import org.apache.poi.openxml4j.exceptions.OpenXML4JException;
import org.apache.poi.openxml4j.opc.OPCPackage;
import org.apache.poi.ss.usermodel.BuiltinFormats;
import org.apache.poi.ss.usermodel.Cell;
import org.apache.poi.ss.usermodel.DataFormatter;
import org.apache.poi.ss.usermodel.Row;
import org.apache.poi.ss.usermodel.Sheet;
import org.apache.poi.ss.usermodel.Workbook;
import org.apache.poi.ss.usermodel.WorkbookFactory;
import org.apache.poi.xssf.eventusermodel.ReadOnlySharedStringsTable;
import org.apache.poi.xssf.eventusermodel.XSSFReader;
import org.apache.poi.xssf.model.StylesTable;
import org.apache.poi.xssf.streaming.SXSSFSheet;
import org.apache.poi.xssf.streaming.SXSSFWorkbook;
import org.apache.poi.xssf.usermodel.XSSFCellStyle;
import org.apache.poi.xssf.usermodel.XSSFRichTextString;
import org.springframework.util.Assert;
import org.springframework.util.StringUtils;
import org.xml.sax.Attributes;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;
import org.xml.sax.XMLReader;
import org.xml.sax.helpers.DefaultHandler;

/**
 * 通用excel导入导出工具类,支持大数据量
 * 
 * 

* Title: TaExcelUtils *

* *

* Description: *

* *
@author zjhua * * @date 2018年12月6日 */ public class TaExcelUtils { private static final SimpleDateFormat simpleDateFormat = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); private static final DateFormat formater = new SimpleDateFormat("yyyy-MM-dd"); private static final int WINDOW_SIZE = 1000; /** * 写出到excel * * @param fileName * 目标文件名 * @param list * 数据源 * @param keys2titlesMap * @param skipTitle * 是否跳过标题行 标题 * @throws Exception */ public static void exportExcel(String fileName, String sheetName, List> list, LinkedHashMap keys2titlesMap, boolean skipTitle) throws Exception { OutputStream out = new FileOutputStream(fileName); exportExcel(out, sheetName, list, keys2titlesMap, skipTitle); out.flush(); out.close(); } @SuppressWarnings("resource") public static void exportExcel(OutputStream out, String sheetName, List> list, LinkedHashMap keys2titlesMap, boolean skipTitle) throws Exception { SXSSFWorkbook wb = new SXSSFWorkbook(WINDOW_SIZE); Sheet sh = wb.createSheet(sheetName); Iterator keys = keys2titlesMap.keySet().iterator(); int startRow = 1; int i = 0; List keyList = new ArrayList(); Row rowHeader = null; if (!skipTitle) { rowHeader = sh.createRow(0); startRow = 0; } while (keys.hasNext()) { String key = keys.next(); keyList.add(key); String title = keys2titlesMap.get(key); if (!skipTitle) { Cell cellHeader = rowHeader.createCell(i++); cellHeader.setCellValue(title); } } for (int rownum = 1; rownum <= list.size(); rownum++) { Row row = sh.createRow(rownum - startRow); Map dataMap = list.get(rownum - 1); i = 0; for (String key : keyList) { Cell cell = row.createCell(i++); if (dataMap.get(key) instanceof String) { cell.setCellValue(MapUtils.getString(dataMap, key)); } else if (dataMap.get(key) instanceof Date) { cell.setCellValue(simpleDateFormat.format((Date) dataMap.get(key))); } else if (dataMap.get(key) instanceof Number) { cell.setCellValue(MapUtils.getDouble(dataMap, key)); } else if (dataMap.get(key) instanceof Boolean) { cell.setCellValue(MapUtils.getBooleanValue(dataMap, key)); } } if ((rownum - startRow) % WINDOW_SIZE == 0) { ((SXSSFSheet) sh).flushRows(); } } wb.write(out); wb.dispose(); } public static Object getValue(T t, String key) { Method readMethod = org.springframework.beans.BeanUtils.getPropertyDescriptor(t.getClass(), key) .getReadMethod(); try { return readMethod.invoke(t); } catch (IllegalAccessException | IllegalArgumentException | InvocationTargetException e) { e.printStackTrace(); } return null; } /** * 写bean列表到excel * * @param fileName * @param sheetName * @param list * @param keys2titlesMap * @param skipTitle * @throws Exception */ @SuppressWarnings("resource") public static void exportExcelForBean(String fileName, String sheetName, List list, LinkedHashMap keys2titlesMap, boolean skipTitle) throws Exception { SXSSFWorkbook wb = new SXSSFWorkbook(WINDOW_SIZE); Sheet sh = wb.createSheet(sheetName); Iterator keys = keys2titlesMap.keySet().iterator(); OutputStream out = new FileOutputStream(fileName); int i = 0; List keyList = new ArrayList(); Row rowHeader = null; if (!skipTitle) { rowHeader = sh.createRow(0); } while (keys.hasNext()) { String key = keys.next(); keyList.add(key); String title = keys2titlesMap.get(key); if (!skipTitle) { Cell cellHeader = rowHeader.createCell(i++); cellHeader.setCellValue(title); } } for (int rownum = 1; rownum <= list.size(); rownum++) { Row row = sh.createRow(rownum); T dataMap = list.get(rownum - 1); i = 0; for (String key : keyList) { Cell cell = row.createCell(i++); if (org.springframework.beans.BeanUtils.findPropertyType(key, dataMap.getClass()) == String.class) { cell.setCellValue(BeanUtils.getProperty(dataMap, key)); } else if (org.springframework.beans.BeanUtils.findPropertyType(key, dataMap.getClass()) == Date.class) { cell.setCellValue(simpleDateFormat.format((Date) getValue(dataMap, key))); } else if (org.springframework.beans.BeanUtils.findPropertyType(key, dataMap.getClass()) == Number.class) { cell.setCellValue((double) getValue(dataMap, key)); } else if (org.springframework.beans.BeanUtils.findPropertyType(key, dataMap.getClass()) == Boolean.class) { cell.setCellValue((boolean) getValue(dataMap, key)); } } if (rownum % WINDOW_SIZE == 0) { ((SXSSFSheet) sh).flushRows(); } } wb.write(out); out.close(); } /** * * @param inputStream * @param keys * key是汉字,value是英文 * @param skipTitle * @return * @throws Exception */ public static List> importExcel(InputStream inputStream, Map titleMap, boolean skipTitle) throws Exception { return importExcel(inputStream, 0, titleMap, skipTitle); } /** * * @param inputStream * @param keys * key是汉字,value是英文 * @param skipTitle * @return * @throws Exception */ public static List> importExcel(InputStream inputStream, int sheetNo, Map titleMap, boolean skipTitle) throws Exception { String[] keys = new String[titleMap.size()]; keys = titleMap.keySet().toArray(keys); List> tmpResult = importExcel(inputStream, sheetNo, keys, skipTitle); List> result = new ArrayList<>(); for (Map row : tmpResult) { Map e = new HashMap<>(); for (String hanzi : row.keySet()) { e.put(titleMap.get(hanzi), row.get(hanzi)); } result.add(e); } return result; } public static List> importExcel(InputStream inputStream, String[] keys, boolean skipTitle) throws Exception { return importExcel(inputStream, 0, keys, skipTitle); } public static List> importExcel(InputStream inputStream, int sheetNo, String[] keys, boolean skipTitle) throws Exception { Workbook wb = WorkbookFactory.create(inputStream); Sheet sheet = wb.getSheetAt(sheetNo); // 0-based int startRow = skipTitle ? 1 : 0; List> list = new ArrayList>(); for (int rownum = startRow; rownum <= sheet.getLastRowNum(); rownum++) { Row row = sheet.getRow(rownum); if (row == null) { continue; } try { Map map = new HashMap(); for (int cellnum = 0; cellnum < row.getLastCellNum() && cellnum < keys.length; cellnum++) { Cell cell = row.getCell(cellnum); if (cell != null) { int valType = cell.getCellType(); if (valType == Cell.CELL_TYPE_STRING) { map.put(keys[cellnum], cell.getStringCellValue()); } else if (valType == Cell.CELL_TYPE_BOOLEAN) { map.put(keys[cellnum], cell.getBooleanCellValue()); } else if (valType == Cell.CELL_TYPE_NUMERIC) { if (HSSFDateUtil.isCellDateFormatted(cell)) { // 用于转化为日期格式 Date d = cell.getDateCellValue(); map.put(keys[cellnum], formater.format(d)); } else { map.put(keys[cellnum], cell.getNumericCellValue()); } } } } list.add(map); } catch (Exception e) { e.printStackTrace(); System.out.println(rownum + "," + JsonUtils.toJson(row)); } } return list; } /** * 加载excel到map, 3M以内文件适合使用 * * @param fileName * excel文件名 * @param keysString * 每个列对应的key * @return * @throws Exception */ public static List> importExcel(String fileName, String[] keys, boolean skipTitle) throws Exception { return importExcel(new FileInputStream(fileName), keys, skipTitle); } /** * 加载excel到pojo对象,大数据量模式,比如几十万、上百万 * * @param path * excel文件路径 * @param sheetName * sheet页名称 * @param keys * 每列的key,不允许为空,例如["id","name"] * @param clz * bean类 * @param skipTitle * 是否跳过标题行, true:跳过, false:不跳过 * @return * @throws IOException * @throws OpenXML4JException * @throws ParserConfigurationException * @throws SAXException */ public static List readerExcel(String path, String sheetName, String[] keys, Class clz, boolean skipTitle) throws IOException, OpenXML4JException, ParserConfigurationException, SAXException { return XLSXCovertCSVReader.readerExcel(path, sheetName, keys, clz, skipTitle); } /** * 加载excel为hashmap列表,大数据量模式,比如几十万、上百万 * * @param path * excel文件路径 * @param sheetName * sheet页名称 * @param keys * 每列的key,不允许为空 * @param skipTitle * 是否跳过标题行, true:跳过, false:不跳过 * @return List> * @throws IOException * @throws OpenXML4JException * @throws ParserConfigurationException * @throws SAXException */ public static List> readerExcel(String path, String sheetName, String[] keys, boolean skipTitle) throws IOException, OpenXML4JException, ParserConfigurationException, SAXException { return XLSXCovertCSVReader.readerExcel(path, sheetName, keys, skipTitle); } /** * 加载excel到bean, 从第一个sheet页加载, 3M以内文件适合使用 * * @param fileName * excel文件名 * @param keys * 每个列的key * @param clz * bean类名 * @return bean列表 * @throws Exception */ public static List importExcelForBean(String fileName, String[] keys, Class clz, boolean skipTitle) throws Exception { return importExcelForBean(fileName, 0, keys, clz, skipTitle); } public static List importExcelForBean(InputStream inputStream, int sheetNo, String[] keys, Class clz, boolean skipTitle) throws Exception { Workbook wb = WorkbookFactory.create(inputStream); Class type[] = new Class[keys.length]; for (int i = 0; i < keys.length; i++) { Field[] fields = clz.getDeclaredFields(); for (int j = 0; j < fields.length; j++) { if (fields[j].getName().equals(keys[i])) { type[i] = fields[j].getType(); } } } Sheet sheet = wb.getSheetAt(sheetNo); List list = new ArrayList(); int startRow = skipTitle ? 1 : 0; for (int rownum = startRow; rownum <= sheet.getLastRowNum(); rownum++) { Row row = sheet.getRow(rownum); if (row == null) { continue; } T obj = clz.newInstance(); for (int cellnum = 0; cellnum < row.getLastCellNum(); cellnum++) { try { Cell cell = row.getCell(cellnum); if (cell != null) { // int valType = cell.getCellType(); Class cleeType = type[cellnum]; if (cleeType == String.class && cell.getCellType() != Cell.CELL_TYPE_FORMULA) { if (cell.getCellType() == Cell.CELL_TYPE_STRING) BeanUtils.setProperty(obj, keys[cellnum], cell.getStringCellValue()); else BeanUtils.setProperty(obj, keys[cellnum], TaUnitConvertUtil.trimTailZero(cell.getNumericCellValue())); } else if (cleeType == Boolean.class) { BeanUtils.setProperty(obj, keys[cellnum], cell.getBooleanCellValue()); } else if (cleeType == BigDecimal.class || cleeType == Double.class || cleeType == Float.class || cleeType == Long.class || cleeType == Integer.class || cleeType == Short.class || cell.getCellType() == Cell.CELL_TYPE_FORMULA) { if (cell.getCellType() == Cell.CELL_TYPE_STRING) { BeanUtils.setProperty(obj, keys[cellnum], StringUtils.isEmpty(cell.getStringCellValue()) ? 0 : cell.getStringCellValue()); } else BeanUtils.setProperty(obj, keys[cellnum], cell.getNumericCellValue()); } else if (cleeType == Date.class) { Date d = cell.getDateCellValue(); BeanUtils.setProperty(obj, keys[cellnum], formater.format(d)); } } } catch (Exception e) { e.printStackTrace(); System.out.println(rownum + "," + cellnum); } } list.add(obj); } return list; } /** * @see importExcelForBean(String fileName,String[] keys,Class clz), * 指定要加载的sheet页面 * @param fileName * @param sheetNo * @param keys * @param clz * @return * @throws Exception */ public static List importExcelForBean(String fileName, int sheetNo, String[] keys, Class clz, boolean skipTitle) throws Exception { return importExcelForBean(new FileInputStream(new File(fileName)), sheetNo, keys, clz, skipTitle); } public static void main(String[] args) { // LinkedHashMap keys2titlesMap = new LinkedHashMap<>(); // keys2titlesMap.put("id", "序号"); // keys2titlesMap.put("value", "值"); // keys2titlesMap.put("desc", "描述"); // List> list = new ArrayList<>(); // for (int i = 0; i < 1000000; i++) { // Map record = new HashMap<>(); // record.put("id", i); // record.put("value", "值" + i); // record.put("desc", UuidUtil.getTimeBasedUuid() + String.valueOf(i)); // list.add(record); // } String fileName = "d:\\申购费率区间.xlsx"; // try { // TaExcelUtils.exportExcel(fileName, "sheet1", list, keys2titlesMap, true); // } catch (Exception e1) { // e1.printStackTrace(); // } String[] keys = new String[] { "fundCode", "fundName", "shareType", "minBalance", "maxBalance", "fareRatio", "fixedFare", "appFareRate" }; Map keysMap = new HashMap<>(); keysMap.put("基金代码", "fundCode"); keysMap.put("业务代码", "bizCode"); keysMap.put("费用类型", "feeType"); keysMap.put("销售商", "sales"); keysMap.put("份额类别", "Type"); keysMap.put("TA代码", "TaCode"); keysMap.put("销售商最大折扣", "discount"); try { InputStream is = new FileInputStream(new File("d:\\费用分成.xls")); List> result = TaExcelUtils.importExcel(is, 0, keysMap, true); System.out.println("总行数:" + result.size()); // result = TaExcelUtils.importExcel(fileName, keys,true); System.out.println("总行数:" + result.size()); } catch (Exception e) { e.printStackTrace(); } // try { // List result = TaExcelUtils.importExcelForBean(fileName, 0, keys, // Bean.class, false); // System.out.println("bean: " + JsonUtils.toJson(result.subList(0, 100))); // } catch (Exception e) { // e.printStackTrace(); // } try { List> list2 = XLSXCovertCSVReader.readerExcel(fileName, "SQL Results", keys, true); System.out.println(JsonUtils.toJson(list2.subList(0, 100))); System.out.println(list2.size()); } catch (Exception e) { e.printStackTrace(); } } public static final class Bean { private int id; private String name; private String desc; public int getId() { return id; } public void setId(int id) { this.id = id; } public String getName() { return name; } public void setName(String name) { this.name = name; } public String getDesc() { return desc; } public void setDesc(String desc) { this.desc = desc; } } } /** * 使用CVS模式解决XLSX文件,可以有效解决用户模式内存溢出的问题 * 该模式是POI官方推荐的读取大数据的模式,在用户模式下,数据量较大、Sheet较多、或者是有很多无用的空行的情况 * ,容易出现内存溢出,用户模式读取Excel的典型代码如下: FileInputStream file=new * FileInputStream("c:\\test.xlsx"); Workbook wb=new XSSFWorkbook(file); * * * @author zjhua */ class XLSXCovertCSVReader { /** * The type of the data value is indicated by an attribute on the cell. The * value is usually in a "v" element within the cell. */ enum xssfDataType { BOOL, ERROR, FORMULA, INLINESTR, SSTINDEX, NUMBER, } /** * 使用xssf_sax_API处理Excel,请参考: * http://poi.apache.org/spreadsheet/how-to.html#xssf_sax_api *

* Also see Standard ECMA-376, 1st edition, part 4, pages 1928ff, at * http://www.ecma-international.org/publications/standards/Ecma-376.htm *

* A web-friendly version is http://openiso.org/Ecma/376/Part4 */ class MyXSSFSheetHandler extends DefaultHandler { /** * Table with styles */ private StylesTable stylesTable; /** * Table with unique strings */ private ReadOnlySharedStringsTable sharedStringsTable; /** * Destination for data */ private final PrintStream output; /** * Number of columns to read starting with leftmost */ private final int minColumnCount; // Set when V start element is seen private boolean vIsOpen; // Set when cell start element is seen; // used when cell close element is seen. private xssfDataType nextDataType; // Used to format numeric cell values. private short formatIndex; private String formatString; private final DataFormatter formatter; private int thisColumn = -1; private int rowNum = 0; // The last column printed to the output stream private int lastColumnNumber = -1; // Gathers characters as they are seen. private StringBuffer value; private Map record; private List> rows = new ArrayList<>(); private boolean isCellNull = false; /** * Accepts objects needed while parsing. * * @param styles * Table of styles * @param strings * Table of shared strings * @param cols * Minimum number of columns to show * @param target * Sink for output */ public MyXSSFSheetHandler(StylesTable styles, ReadOnlySharedStringsTable strings, int cols, PrintStream target) { this.stylesTable = styles; this.sharedStringsTable = strings; this.minColumnCount = cols; this.output = target; this.value = new StringBuffer(); this.nextDataType = xssfDataType.NUMBER; this.formatter = new DataFormatter(); record = new HashMap<>(this.minColumnCount); rows.clear();// 每次读取都清空行集合 } public List getDataList(Class clz) { List list = new ArrayList<>(); try { T obj = clz.newInstance(); rows.forEach((map) -> { try { BeanUtils.populate(obj, map); list.add(obj); } catch (IllegalAccessException | InvocationTargetException e) { e.printStackTrace(); } }); } catch (InstantiationException | IllegalAccessException e) { e.printStackTrace(); } return list; } /* * (non-Javadoc) * * @see org.xml.sax.helpers.DefaultHandler#startElement(java.lang.String, * java.lang.String, java.lang.String, org.xml.sax.Attributes) */ public void startElement(String uri, String localName, String name, Attributes attributes) throws SAXException { if ("inlineStr".equals(name) || "v".equals(name) || "t".equals(name)) { vIsOpen = true; // Clear contents cache value.setLength(0); } // c => cell else if ("c".equals(name)) { // Get the cell reference String r = attributes.getValue("r"); int firstDigit = -1; for (int c = 0; c < r.length(); ++c) { if (Character.isDigit(r.charAt(c))) { firstDigit = c; break; } } thisColumn = nameToColumn(r.substring(0, firstDigit)); // Set up defaults. this.nextDataType = xssfDataType.NUMBER; this.formatIndex = -1; this.formatString = null; String cellType = attributes.getValue("t"); String cellStyleStr = attributes.getValue("s"); if ("b".equals(cellType)) nextDataType = xssfDataType.BOOL; else if ("e".equals(cellType)) nextDataType = xssfDataType.ERROR; else if ("inlineStr".equals(cellType)) nextDataType = xssfDataType.INLINESTR; else if ("s".equals(cellType)) nextDataType = xssfDataType.SSTINDEX; else if ("str".equals(cellType)) nextDataType = xssfDataType.FORMULA; else if (cellStyleStr != null) { // It's a number, but almost certainly one // with a special style or format int styleIndex = Integer.parseInt(cellStyleStr); XSSFCellStyle style = stylesTable.getStyleAt(styleIndex); this.formatIndex = style.getDataFormat(); this.formatString = style.getDataFormatString(); if (this.formatString == null) this.formatString = BuiltinFormats.getBuiltinFormat(this.formatIndex); } } } /* * (non-Javadoc) * * @see org.xml.sax.helpers.DefaultHandler#endElement(java.lang.String, * java.lang.String, java.lang.String) */ public void endElement(String uri, String localName, String name) throws SAXException { String thisStr = null; // v => contents of a cell if ("v".equals(name) || "t".equals(name)) { // Process the value contents as required. // Do now, as characters() may be called more than once switch (nextDataType) { case BOOL: char first = value.charAt(0); thisStr = first == '0' ? "FALSE" : "TRUE"; break; case ERROR: thisStr = "\"ERROR:" + value.toString() + '"'; break; case FORMULA: // A formula could result in a string value, // so always add double-quote characters. // thisStr = '"' + value.toString() + '"'; thisStr = value.toString(); break; case INLINESTR: // TODO: have seen an example of this, so it's untested. XSSFRichTextString rtsi = new XSSFRichTextString(value.toString()); // thisStr = '"' + rtsi.toString() + '"'; thisStr = rtsi.toString(); break; case SSTINDEX: String sstIndex = value.toString(); try { int idx = Integer.parseInt(sstIndex); XSSFRichTextString rtss = new XSSFRichTextString(sharedStringsTable.getEntryAt(idx)); // thisStr = '"' + rtss.toString() + '"'; thisStr = rtss.toString(); } catch (NumberFormatException ex) { output.println("Failed to parse SST index '" + sstIndex + "': " + ex.toString()); } break; case NUMBER: String n = value.toString(); // 判断是否是日期格式 if (HSSFDateUtil.isADateFormat(this.formatIndex, n)) { Double d = Double.parseDouble(n); Date date = HSSFDateUtil.getJavaDate(d); thisStr = formateDateToString(date); } else if (this.formatString != null) thisStr = formatter.formatRawCellContents(Double.parseDouble(n), this.formatIndex, this.formatString); else { thisStr = StringUtils.trimTrailingCharacter(StringUtils.trimTrailingCharacter(n, '0'), '.'); } break; default: thisStr = "(TODO: Unexpected type: " + nextDataType + ")"; break; } // Output after we've seen the string contents // Emit commas for any fields that were missing on this row if (lastColumnNumber == -1) { lastColumnNumber = 0; } // 判断单元格的值是否为空 if (thisStr == null || "".equals(thisStr)) { isCellNull = true;// 设置单元格是否为空值 } record.put(keys[thisColumn], thisStr); // Update column if (thisColumn > -1) lastColumnNumber = thisColumn; } else if ("row".equals(name)) { // Print out any missing commas if needed if (minColumns > 0) { // Columns are 0 based if (lastColumnNumber == -1) { lastColumnNumber = 0; } // if (isCellNull == false && record[0] != null // && record[1] != null)// 判断是否空行 if (isCellNull == false)// 判断是否空行 { if (rowNum > 0 || !skipTitle) { Map map = new HashMap<>(); map.putAll(record); rows.add(map); } rowNum++; isCellNull = false; record.clear(); } } lastColumnNumber = -1; } } public List> getRows() { return rows; } public void setRows(List> rows) { this.rows = rows; } /** * Captures characters only if a suitable element is open. Originally was just * "v"; extended for inlineStr also. */ public void characters(char[] ch, int start, int length) throws SAXException { if (vIsOpen) value.append(ch, start, length); } /** * Converts an Excel column name like "C" to a zero-based index. * * @param name * @return Index corresponding to the specified name */ private int nameToColumn(String name) { int column = -1; for (int i = 0; i < name.length(); ++i) { int c = name.charAt(i); column = (column + 1) * 26 + c - 'A'; } return column; } private String formateDateToString(Date date) { SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");// 格式化日期 return sdf.format(date); } } // / private OPCPackage xlsxPackage; private int minColumns; private String[] keys; private PrintStream output; private String sheetName; boolean skipTitle; /** * Creates a new XLSX -> CSV converter * * @param pkg * The XLSX package to process * @param output * The PrintStream to output the CSV to * @param minColumns * The minimum number of columns to output, or -1 for no minimum */ public XLSXCovertCSVReader(OPCPackage pkg, PrintStream output, String sheetName, String[] keys, boolean skipTitle) { this.xlsxPackage = pkg; this.output = output; this.skipTitle = skipTitle; this.keys = keys; this.minColumns = keys.length; this.sheetName = sheetName; } /** * Parses and shows the content of one sheet using the specified styles and * shared-strings tables. * * @param styles * @param strings * @param sheetInputStream */ public List> processSheet(StylesTable styles, ReadOnlySharedStringsTable strings, InputStream sheetInputStream) throws IOException, ParserConfigurationException, SAXException { InputSource sheetSource = new InputSource(sheetInputStream); SAXParserFactory saxFactory = SAXParserFactory.newInstance(); SAXParser saxParser = saxFactory.newSAXParser(); XMLReader sheetParser = saxParser.getXMLReader(); MyXSSFSheetHandler handler = new MyXSSFSheetHandler(styles, strings, this.minColumns, this.output); sheetParser.setContentHandler(handler); sheetParser.parse(sheetSource); return handler.getRows(); } /** * 初始化这个处理程序 将 * * @throws IOException * @throws OpenXML4JException * @throws ParserConfigurationException * @throws SAXException */ public List> process() throws IOException, OpenXML4JException, ParserConfigurationException, SAXException { ReadOnlySharedStringsTable strings = new ReadOnlySharedStringsTable(this.xlsxPackage); XSSFReader xssfReader = new XSSFReader(this.xlsxPackage); List> list = null; StylesTable styles = xssfReader.getStylesTable(); XSSFReader.SheetIterator iter = (XSSFReader.SheetIterator) xssfReader.getSheetsData(); while (iter.hasNext()) { InputStream stream = iter.next(); String sheetNameTemp = iter.getSheetName(); if (this.sheetName.equals(sheetNameTemp)) { list = processSheet(styles, strings, stream); stream.close(); } } return list; } /** * 读取Excel * * @param path * 文件路径 * @param sheetName * sheet名称 * @param keys * 字段列表 * @return * @throws SAXException * @throws ParserConfigurationException * @throws OpenXML4JException * @throws IOException */ public static List> readerExcel(String path, String sheetName, String[] keys, boolean skipTitle) throws IOException, OpenXML4JException, ParserConfigurationException, SAXException { Assert.notEmpty(keys, "字段列表不能为空!"); OPCPackage p = OPCPackage.open(path); XLSXCovertCSVReader xlsx2csv = new XLSXCovertCSVReader(p, System.out, sheetName, keys, skipTitle); List> list = xlsx2csv.process(); p.close(); return list; } public static List readerExcel(String path, String sheetName, String[] keys, Class clz, boolean skipTitle) throws IOException, OpenXML4JException, ParserConfigurationException, SAXException { List> list = readerExcel(path, sheetName, keys, skipTitle); List beans = new ArrayList<>(); list.forEach((map) -> { try { T obj = clz.newInstance(); BeanUtils.populate(obj, map); beans.add(obj); } catch (IllegalAccessException | InvocationTargetException | InstantiationException e) { e.printStackTrace(); } }); return beans; } }

 

转载于:https://www.cnblogs.com/zhjh256/p/10083534.html

你可能感兴趣的:(java,大数据)