CSV importer

From OpenKM Documentation
Revision as of 18:01, 31 December 2013 by Jllort (talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Import metadata from csv file. The script can be execute from administration scripting ( or used as crontab script too ) to import metadata values to OpenKM.

Description:

  • There's a metadata file with two columns, one for document path and other for metadatada value.
  • Script search document by document name. In theory only there's one document with same name in the system, otherside script shows error. Here could be used document path from first csv column directly, but to get more complete example has been used search engine to find document.
  • For each document found, script add metadata property group with CSV value.

Nota clasica.png The script can be easily changed to use more than two columns.


CSV File CSV file has two colum, first column contains document path and second column has metadata value. You can download from here File:Metadata.csv.zip.

User guide 535.png


Property Group definition

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE property-groups PUBLIC "-//OpenKM//DTD Property Groups 2.1//EN"
"http://www.openkm.com/dtd/property-groups-2.1.dtd">
<property-groups>
    <property-group label="Metadata" name="okg:metadata">
        <input label="Document ID" name="okp:metadata.value" type="text"/>
    </property-group>
</property-groups>


import java.io.FileReader;
import java.io.Reader;
import java.util.Iterator;
import java.util.List;
import java.util.Collection;

import com.googlecode.jcsv.CSVStrategy;
import com.googlecode.jcsv.reader.CSVReader;
import com.googlecode.jcsv.reader.internal.CSVReaderBuilder;
import com.googlecode.jcsv.reader.internal.DefaultCSVEntryParser;

import com.openkm.dao.bean.QueryParams;
import com.openkm.bean.QueryResult;
import com.openkm.api.OKMSearch;
import com.openkm.util.FileLogger;
import com.openkm.api.OKMPropertyGroup;
import com.openkm.util.PathUtils;

String grpName = "okg:metadata";
String FILE_LOG_NAME = "CSVLOG";
String META_PATH = "/home/openkm/csv/";
String META_FILE_NAME = "metadata.csv";
int uniqueFileName = 0;
int valueColumn = 1;

// Format defintion
char delimiter = ',';
char quoteCharacter = '"';
char commentIndicator = '#';
boolean skipHeader = true;
boolean ignoreEmptyLines = true;
CSVStrategy strategy = new CSVStrategy(delimiter, quoteCharacter, commentIndicator, skipHeader, ignoreEmptyLines);
// File reader
Reader reader = new FileReader(META_PATH + META_FILE_NAME);
// CSV reader		
CSVReader csvParser = new CSVReaderBuilder(reader).strategy(strategy).entryParser(new DefaultCSVEntryParser()).build();
List data = csvParser.readAll();
int count = 1;
int countFound = 0;
int countNotDocument = 0;
int moreThanOneDocumentFound = 0;
int notFound = 0;
int noName = 0;

for (Iterator it = data.listIterator(); it.hasNext();) {
        String[] row = (String[]) it.next();
        String docPath = row[uniqueFileName];
    	print(count + ">>>> " + docPath);

        if (docPath != null && !docPath.equals("")) {
            QueryParams queryParams = new QueryParams();
            queryParams.setDomain(QueryParams.DOCUMENT);
            queryParams.setName(PathUtils.getName(docPath));
            Collection results = OKMSearch.getInstance().find(null, queryParams);

            if (results.size() == 1) {
                QueryResult queryResult = (QueryResult) results.iterator().next();
                if (queryResult.getDocument() != null) {
                    print("found");
                    countFound++;
                    // Add Group
		    OKMPropertyGroup.getInstance().addGroup(null, docPath, grpName);
		    // Add metadata
		    Map map = new HashMap();
		    map.put("okp:metadata.value", row[valueColumn]);
		    OKMPropertyGroup.getInstance().setPropertiesSimple(null, docPath, grpName, map);
                } else {
                    print("error is not document");
                    countNotDocument++;
                }
            } else if (results.size() > 1) {
                print("error more than one document found can not decide");
                moreThanOneDocumentFound++;
            } else {
                print("not found");
                notFound++;
            }
        } else {
            print("error document has no name");
            noName++;
        }

        print("</br>");
        
        //FileLogger.info(FILE_LOG_NAME, "Document name ''{0}'' to ''{1}''", row[0], row[posDocRevNo]);
        count++;
}

print("Total:" + count + "</br>");
print("Found:" + countFound + "</br>");
print("Error not document:" + countNotDocument + "</br>");
print("Error more then one document found:" + moreThanOneDocumentFound + "</br>");
print("Error not found:" + notFound + "</br>");
print("Error name empty:" + notFound + "</br>");


Exemple

Register property group:

User guide 536.png


Files at /okm:root:

User guide 537.png


Execute script:

User guide 538.png


Script results:

User guide 539.png


Imported metadata:

User guide 540.png