XlsSaxAnalyser.java 8.0 KB
Newer Older
1
package com.alibaba.excel.analysis.v03;
J
update  
jipengfei.jpf 已提交
2

Z
zhuangjiaju 已提交
3 4 5 6
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
Z
zhuangjiaju 已提交
7 8
import java.util.Map;
import java.util.TreeMap;
Z
zhuangjiaju 已提交
9 10 11 12 13 14 15 16

import org.apache.poi.hssf.eventusermodel.EventWorkbookBuilder;
import org.apache.poi.hssf.eventusermodel.FormatTrackingHSSFListener;
import org.apache.poi.hssf.eventusermodel.HSSFEventFactory;
import org.apache.poi.hssf.eventusermodel.HSSFListener;
import org.apache.poi.hssf.eventusermodel.HSSFRequest;
import org.apache.poi.hssf.eventusermodel.MissingRecordAwareHSSFListener;
import org.apache.poi.hssf.eventusermodel.dummyrecord.LastCellOfRowDummyRecord;
Z
zhuangjiaju 已提交
17
import org.apache.poi.hssf.record.ExtendedFormatRecord;
Z
zhuangjiaju 已提交
18
import org.apache.poi.hssf.record.Record;
Z
zhuangjiaju 已提交
19 20
import org.apache.poi.hssf.record.StyleRecord;
import org.apache.poi.hssf.usermodel.HSSFCellStyle;
Z
zhuangjiaju 已提交
21 22 23 24 25
import org.apache.poi.hssf.usermodel.HSSFWorkbook;
import org.apache.poi.poifs.filesystem.POIFSFileSystem;

import com.alibaba.excel.analysis.ExcelExecutor;
import com.alibaba.excel.analysis.v03.handlers.BlankOrErrorRecordHandler;
26
import com.alibaba.excel.analysis.v03.handlers.BofRecordHandler;
Z
zhuangjiaju 已提交
27 28 29 30 31
import com.alibaba.excel.analysis.v03.handlers.FormulaRecordHandler;
import com.alibaba.excel.analysis.v03.handlers.LabelRecordHandler;
import com.alibaba.excel.analysis.v03.handlers.MissingCellDummyRecordHandler;
import com.alibaba.excel.analysis.v03.handlers.NoteRecordHandler;
import com.alibaba.excel.analysis.v03.handlers.NumberRecordHandler;
32 33
import com.alibaba.excel.analysis.v03.handlers.RkRecordHandler;
import com.alibaba.excel.analysis.v03.handlers.SstRecordHandler;
34
import com.alibaba.excel.context.AnalysisContext;
Z
zhuangjiaju 已提交
35
import com.alibaba.excel.enums.CellDataTypeEnum;
36
import com.alibaba.excel.exception.ExcelAnalysisException;
Z
zhuangjiaju 已提交
37
import com.alibaba.excel.metadata.CellData;
38 39
import com.alibaba.excel.read.listener.event.EachRowAnalysisFinishEvent;
import com.alibaba.excel.read.metadata.ReadSheet;
Z
zhuangjiaju 已提交
40
import com.alibaba.excel.read.metadata.holder.ReadRowHolder;
41
import com.alibaba.excel.read.metadata.holder.ReadWorkbookHolder;
clevertension's avatar
clevertension 已提交
42
import com.alibaba.excel.util.CollectionUtils;
43

J
update  
jipengfei.jpf 已提交
44
/**
Z
zhuangjiaju 已提交
45 46 47 48 49 50 51 52 53 54 55
 * /** * A text extractor for Excel files. *
 * <p>
 * * Returns the textual content of the file, suitable for * indexing by something like Lucene, but not really *
 * intended for display to the user. *
 * </p>
 * *
 * <p>
 * * To turn an excel file into a CSV or similar, then see * the XLS2CSVmra example *
 * </p>
 * * * @see <a href=
 * "http://svn.apache.org/repos/asf/poi/trunk/src/examples/src/org/apache/poi/hssf/eventusermodel/examples/XLS2CSVmra.java">XLS2CSVmra</a>
J
update  
jipengfei.jpf 已提交
56 57 58
 *
 * @author jipengfei
 */
Z
zhuangjiaju 已提交
59
public class XlsSaxAnalyser implements HSSFListener, ExcelExecutor {
clevertension's avatar
clevertension 已提交
60 61 62 63
    private boolean outputFormulaValues = true;
    private POIFSFileSystem fs;
    private int lastRowNumber;
    private int lastColumnNumber;
Z
zhuangjiaju 已提交
64
    private boolean notAllEmpty = false;
clevertension's avatar
clevertension 已提交
65 66 67 68 69
    /**
     * For parsing Formulas
     */
    private EventWorkbookBuilder.SheetRecordCollectingListener workbookBuildingListener;
    private FormatTrackingHSSFListener formatListener;
Z
zhuangjiaju 已提交
70
    private Map<Integer, CellData> records;
71
    private List<ReadSheet> sheets = new ArrayList<ReadSheet>();
clevertension's avatar
clevertension 已提交
72 73
    private HSSFWorkbook stubWorkbook;
    private List<XlsRecordHandler> recordHandlers = new ArrayList<XlsRecordHandler>();
74
    private AnalysisContext analysisContext;
Z
zhuangjiaju 已提交
75

76
    public XlsSaxAnalyser(AnalysisContext context) throws IOException {
J
update  
jipengfei.jpf 已提交
77
        this.analysisContext = context;
Z
zhuangjiaju 已提交
78
        this.records = new TreeMap<Integer, CellData>();
79 80 81 82 83 84
        ReadWorkbookHolder readWorkbookHolder = analysisContext.readWorkbookHolder();
        if (readWorkbookHolder.getFile() != null) {
            this.fs = new POIFSFileSystem(readWorkbookHolder.getFile());
        } else {
            this.fs = new POIFSFileSystem(readWorkbookHolder.getInputStream());
        }
Z
zhuangjiaju 已提交
85

J
update  
jipengfei.jpf 已提交
86 87
    }

88
    @Override
89
    public List<ReadSheet> sheetList() {
J
update  
jipengfei.jpf 已提交
90 91 92 93 94 95 96
        return sheets;
    }

    @Override
    public void execute() {
        MissingRecordAwareHSSFListener listener = new MissingRecordAwareHSSFListener(this);
        formatListener = new FormatTrackingHSSFListener(listener);
clevertension's avatar
clevertension 已提交
97 98 99 100 101 102
        workbookBuildingListener = new EventWorkbookBuilder.SheetRecordCollectingListener(formatListener);
        if (workbookBuildingListener != null && stubWorkbook == null) {
            stubWorkbook = workbookBuildingListener.getStubHSSFWorkbook();
        }

        init();
J
update  
jipengfei.jpf 已提交
103 104 105 106 107 108 109 110 111 112 113 114 115 116

        HSSFEventFactory factory = new HSSFEventFactory();
        HSSFRequest request = new HSSFRequest();
        if (outputFormulaValues) {
            request.addListenerForAllRecords(formatListener);
        } else {
            request.addListenerForAllRecords(workbookBuildingListener);
        }

        try {
            factory.processWorkbookEvents(request, fs);
        } catch (IOException e) {
            throw new ExcelAnalysisException(e);
        }
Z
zhuangjiaju 已提交
117 118 119 120
        // Sometimes tables lack the end record of the last column
        if (!records.isEmpty()) {
            endRow();
        }
J
update  
jipengfei.jpf 已提交
121 122 123 124 125
    }

    private void init() {
        lastRowNumber = 0;
        lastColumnNumber = 0;
Z
zhuangjiaju 已提交
126
        records = new TreeMap<Integer, CellData>();
127
        sheets = new ArrayList<ReadSheet>();
clevertension's avatar
clevertension 已提交
128
        buildXlsRecordHandlers();
J
update  
jipengfei.jpf 已提交
129 130
    }

131
    @Override
J
update  
jipengfei.jpf 已提交
132 133 134
    public void processRecord(Record record) {
        int thisRow = -1;
        int thisColumn = -1;
Z
zhuangjiaju 已提交
135
        CellData cellData = null;
Z
zhuangjiaju 已提交
136
        for (XlsRecordHandler handler : this.recordHandlers) {
clevertension's avatar
clevertension 已提交
137 138 139 140
            if (handler.support(record)) {
                handler.processRecord(record);
                thisRow = handler.getRow();
                thisColumn = handler.getColumn();
Z
zhuangjiaju 已提交
141 142
                cellData = handler.getCellData();
                if (cellData != null) {
Z
zhuangjiaju 已提交
143
                    records.put(thisColumn, cellData);
Z
zhuangjiaju 已提交
144
                }
J
update  
jipengfei.jpf 已提交
145
                break;
clevertension's avatar
clevertension 已提交
146
            }
J
update  
jipengfei.jpf 已提交
147 148
        }
        // If we got something to print out, do so
Z
zhuangjiaju 已提交
149 150 151 152
        if (cellData != null) {
            if (analysisContext.currentReadHolder().globalConfiguration().getAutoTrim()
                && CellDataTypeEnum.STRING == cellData.getType()) {
                cellData.setStringValue(cellData.getStringValue().trim());
J
update  
jipengfei.jpf 已提交
153
            }
Z
zhuangjiaju 已提交
154
            if (CellDataTypeEnum.EMPTY != cellData.getType()) {
J
update  
jipengfei.jpf 已提交
155 156 157 158
                notAllEmpty = true;
            }
        }

clevertension's avatar
clevertension 已提交
159 160 161 162 163
        // Handle new row
        if (thisRow != -1 && thisRow != lastRowNumber) {
            lastColumnNumber = -1;
        }

J
update  
jipengfei.jpf 已提交
164 165 166 167 168 169 170 171
        // Update column and row count
        if (thisRow > -1) {
            lastRowNumber = thisRow;
        }
        if (thisColumn > -1) {
            lastColumnNumber = thisColumn;
        }

clevertension's avatar
clevertension 已提交
172 173 174 175
        processLastCellOfRow(record);
    }

    private void processLastCellOfRow(Record record) {
J
update  
jipengfei.jpf 已提交
176 177
        // Handle end of row
        if (record instanceof LastCellOfRowDummyRecord) {
Z
zhuangjiaju 已提交
178 179 180 181 182 183 184 185 186 187 188 189
            endRow();
        }
    }

    private void endRow() {
        if (lastColumnNumber == -1) {
            lastColumnNumber = 0;
        }
        if (notAllEmpty) {
            analysisContext.readRowHolder(
                new ReadRowHolder(lastRowNumber, analysisContext.readSheetHolder().getGlobalConfiguration()));
            analysisContext.readSheetHolder().notifyEndOneRow(new EachRowAnalysisFinishEvent(records), analysisContext);
J
update  
jipengfei.jpf 已提交
190
        }
Z
zhuangjiaju 已提交
191 192
        records.clear();
        lastColumnNumber = -1;
J
update  
jipengfei.jpf 已提交
193
    }
clevertension's avatar
clevertension 已提交
194 195 196 197

    private void buildXlsRecordHandlers() {
        if (CollectionUtils.isEmpty(recordHandlers)) {
            recordHandlers.add(new BlankOrErrorRecordHandler());
198
            recordHandlers.add(new BofRecordHandler(workbookBuildingListener, analysisContext, sheets));
clevertension's avatar
clevertension 已提交
199 200 201 202
            recordHandlers.add(new FormulaRecordHandler(stubWorkbook, formatListener));
            recordHandlers.add(new LabelRecordHandler());
            recordHandlers.add(new NoteRecordHandler());
            recordHandlers.add(new NumberRecordHandler(formatListener));
203 204
            recordHandlers.add(new RkRecordHandler());
            recordHandlers.add(new SstRecordHandler());
clevertension's avatar
clevertension 已提交
205 206 207 208
            recordHandlers.add(new MissingCellDummyRecordHandler());
            Collections.sort(recordHandlers);
        }

Z
zhuangjiaju 已提交
209
        for (XlsRecordHandler x : recordHandlers) {
clevertension's avatar
clevertension 已提交
210 211 212
            x.init();
        }
    }
J
update  
jipengfei.jpf 已提交
213
}