update

2020-11-17 11:57:38 +01:00 · 2020-11-17 11:57:38 +01:00 · b7d1313dbc
commit b7d1313dbc
parent 32dfa62505
5 changed files with 268 additions and 1 deletions
--- a/office-alexander-logistics-app/pom.xml
+++ b/office-alexander-logistics-app/pom.xml
@ -236,6 +236,13 @@
 			<scope>compile</scope>
 		</dependency>
 		
+		<!-- Apache PDFBox -->
+		<dependency>
+			<groupId>org.apache.pdfbox</groupId>
+			<artifactId>pdfbox</artifactId>
+			<version>2.0.19</version>
+			<scope>compile</scope>
+		</dependency>
 		
 	</dependencies>
 </project>
--- a/office-alexander-logistics-app/src/main/java/com/alexanderlogistics/XMLPDFAdapter.java
+++ b/office-alexander-logistics-app/src/main/java/com/alexanderlogistics/XMLPDFAdapter.java
@ -0,0 +1,219 @@
+package com.alexanderlogistics;
+
+import java.io.ByteArrayOutputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.util.List;
+import java.util.Map;
+import java.util.logging.Logger;
+import java.util.regex.Pattern;
+
+import javax.inject.Inject;
+
+import org.apache.pdfbox.cos.COSInputStream;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDDocumentNameDictionary;
+import org.apache.pdfbox.pdmodel.PDEmbeddedFilesNameTreeNode;
+import org.apache.pdfbox.pdmodel.common.PDNameTreeNode;
+import org.apache.pdfbox.pdmodel.common.filespecification.PDComplexFileSpecification;
+import org.apache.pdfbox.pdmodel.common.filespecification.PDEmbeddedFile;
+import org.imixs.archive.core.SnapshotService;
+import org.imixs.workflow.FileData;
+import org.imixs.workflow.ItemCollection;
+import org.imixs.workflow.SignalAdapter;
+import org.imixs.workflow.engine.WorkflowService;
+import org.imixs.workflow.exceptions.AdapterException;
+import org.imixs.workflow.exceptions.PluginException;
+
+/**
+ * The 
+ * 
+ * @version 1.0
+ * @author rsoika
+ */
+public class XMLPDFAdapter implements SignalAdapter {
+
+    public static final String FILE_PATTERN_PDF = ".[pP][dD][fF]";
+    public static final String FILE_PATTERN_XML = ".[xX][mM][lL]";
+
+    private static Logger logger = Logger.getLogger(XMLPDFAdapter.class.getName());
+
+  
+    @Inject
+    WorkflowService workflowService;
+
+    @Inject
+    SnapshotService snapshotService;
+
+    /**
+     * This method posts a text from an attachment to the Imixs-ML Analyse service
+     * endpoint
+     */
+    @SuppressWarnings("unchecked")
+    @Override
+    public ItemCollection execute(ItemCollection document, ItemCollection event) throws AdapterException {
+
+        logger.info("......starting xml extraction..."  );
+        try {
+             byte[] xmlData = getXMLFile(document, FILE_PATTERN_PDF);
+             
+             
+        } catch (PluginException e) {
+            // TODO Auto-generated catch block
+            e.printStackTrace();
+        }
+
+        return document;
+    }
+    
+    
+
+    /**
+     * This method searches attached PDF files of a workitem and extracts an
+     * embedded XML file.
+     * <p>
+     * The method only returns the first embedded xml file and does not support
+     * multiple xml files embedded in one pdf file.
+     * 
+     * @param document
+     * @param filePattern
+     * @return
+     * @throws PluginException
+     */
+    public static byte[] getXMLFile(ItemCollection document, String file_pattern) throws PluginException {
+
+        // verify all attached PDF files
+        List<String> filenames = document.getFileNames();
+        for (String filename : filenames) {
+
+            if (Pattern.compile(file_pattern).matcher(filename).find()) {
+
+                logger.info("...extract embedded XML from '" + filename + "'");
+
+                // we only support the first embedded XML file here!
+                FileData pdfFileData = document.getFileData(filename);
+                return getFirstEmbeddedXML(pdfFileData.getContent());
+
+            }
+        }
+
+        // no data found
+        return null;
+    }
+    
+    /**
+     * Helper method to extract the first XML file from a list of files
+     * 
+     * @param names
+     * @return
+     * @throws IOException
+     */
+    private static byte[] extractFirstXMLFile(Map<String, PDComplexFileSpecification> names) throws IOException {
+        for (Map.Entry<String, PDComplexFileSpecification> entry : names.entrySet()) {
+            PDComplexFileSpecification fileSpec = entry.getValue();
+            String filename = fileSpec.getFile();
+
+            // test if file is a xml file
+            if (Pattern.compile(FILE_PATTERN_XML).matcher(filename).find()) {
+                // we found an xml file!
+                PDEmbeddedFile embeddedFile = getEmbeddedFile(fileSpec);
+                COSInputStream inStream = embeddedFile.createInputStream();
+
+                return streamToByteArray(inStream);
+
+            }
+        }
+
+        // no XML file found
+        return null;
+    }
+
+    /**
+     * This method extract the first XML file form a pdf file content
+     * 
+     * @param content
+     * @return a byte array with the xml data or null if no xml file was found in
+     *         the pdf file.
+     */
+    private static byte[] getFirstEmbeddedXML(byte[] content) {
+        PDDocument doc = null;
+        byte[] result = null;
+        try {
+
+            doc = PDDocument.load(content);
+
+            PDDocumentNameDictionary namesDictionary = new PDDocumentNameDictionary(doc.getDocumentCatalog());
+            PDEmbeddedFilesNameTreeNode efTree = namesDictionary.getEmbeddedFiles();
+            if (efTree != null) {
+                Map<String, PDComplexFileSpecification> names = efTree.getNames();
+                if (names != null) {
+                    result = extractFirstXMLFile(names);
+                } else {
+                    List<PDNameTreeNode<PDComplexFileSpecification>> kids = efTree.getKids();
+                    for (PDNameTreeNode<PDComplexFileSpecification> node : kids) {
+                        names = node.getNames();
+                        result = extractFirstXMLFile(names);
+                    }
+                }
+            }
+        } catch (IOException e) {
+            logger.warning("unable to load embedded xml : " + e.getMessage());
+
+        } finally {
+            if (doc != null) {
+                try {
+                    doc.close();
+                } catch (IOException e) {
+
+                    e.printStackTrace();
+                }
+            }
+        }
+
+        return result;
+    }
+
+    /**
+     * Helper method to extract platform specific embedded file format.
+     * 
+     * @param fileSpec
+     * @return
+     */
+    private static PDEmbeddedFile getEmbeddedFile(PDComplexFileSpecification fileSpec) {
+        PDEmbeddedFile embeddedFile = null;
+        if (fileSpec != null) {
+            embeddedFile = fileSpec.getEmbeddedFileUnicode();
+            if (embeddedFile == null) {
+                embeddedFile = fileSpec.getEmbeddedFileDos();
+            }
+            if (embeddedFile == null) {
+                embeddedFile = fileSpec.getEmbeddedFileMac();
+            }
+            if (embeddedFile == null) {
+                embeddedFile = fileSpec.getEmbeddedFileUnix();
+            }
+            if (embeddedFile == null) {
+                embeddedFile = fileSpec.getEmbeddedFile();
+            }
+        }
+        return embeddedFile;
+    }
+    
+    /**
+     * This method converts a inputStream into a byte array.
+     * 
+     * @param ins
+     * @return
+     * @throws IOException
+     */
+    public static byte[] streamToByteArray(InputStream ins) throws IOException {
+        ByteArrayOutputStream baos = new ByteArrayOutputStream();
+        byte[] byteBuffer = new byte[1024];
+        int len;
+        while ((len = ins.read(byteBuffer)) > -1) {
+            baos.write(byteBuffer, 0, len);
+        }
+        baos.flush();
+        return baos.toByteArray();
+    }
+}
--- a/office-alexander-logistics-app/src/main/webapp/WEB-INF/beans.xml
+++ b/office-alexander-logistics-app/src/main/webapp/WEB-INF/beans.xml
@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<beans xmlns="http://xmlns.jcp.org/xml/ns/javaee"
+       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+       xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/beans_2_0.xsd"
+       bean-discovery-mode="all" version="2.0">
+</beans>
--- a/pom.xml
+++ b/pom.xml
@ -29,7 +29,7 @@
 		<org.imixs.marty.version>4.1.8</org.imixs.marty.version>
 		<org.imixs.office.version>4.4.0</org.imixs.office.version>
 		<org.imixs.adapters.version>2.2.3</org.imixs.adapters.version>
-		<org.imixs.archive.version>2.2.4</org.imixs.archive.version>
+		<org.imixs.archive.version>2.2.7</org.imixs.archive.version>
 		<org.imixs.melman.version>1.0.19</org.imixs.melman.version>
 		<org.imixs.ml.version>1.0.1</org.imixs.ml.version>
 		<microprofile.version>2.2</microprofile.version>
--- a/workflow/posteingang-de-1.0.0.bpmn
+++ b/workflow/posteingang-de-1.0.0.bpmn
@ -39,6 +39,7 @@
  </bpmn2:extensionElements>
  <bpmn2:signal id="Signal_1" name="org.imixs.archive.documents.OCRDocumentAdapter"/>
  <bpmn2:signal id="Signal_2" name="org.imixs.ml.workflow.MLAdapter"/>
+  <bpmn2:signal id="Signal_3" name="com.alexanderlogistics.XMLPDFAdapter"/>
  <bpmn2:collaboration id="Collaboration_1" name="Default Collaboration">
    <bpmn2:participant id="Participant_1" name="Posteingang" processRef="Process_1"/>
    <bpmn2:textAnnotation id="TextAnnotation_1">
@ -92,6 +93,7 @@ Possible ImageTypes are:
        <bpmn2:flowNodeRef>StartEvent_1</bpmn2:flowNodeRef>
        <bpmn2:flowNodeRef>IntermediateCatchEvent_8</bpmn2:flowNodeRef>
        <bpmn2:flowNodeRef>EndEvent_1</bpmn2:flowNodeRef>
+        <bpmn2:flowNodeRef>IntermediateCatchEvent_1</bpmn2:flowNodeRef>
      </bpmn2:lane>
    </bpmn2:laneSet>
    <bpmn2:task id="Task_1" imixs:processid="100" name="Erfassung">
@ -122,6 +124,7 @@ Possible ImageTypes are:
      </bpmn2:extensionElements>
      <bpmn2:documentation id="Documentation_3"><![CDATA[<textblock><itemvalue>$workflowgroup</itemvalue> - <itemvalue>$workflowstatus</itemvalue></textblock>]]></bpmn2:documentation>
      <bpmn2:incoming>SequenceFlow_13</bpmn2:incoming>
+      <bpmn2:incoming>SequenceFlow_26</bpmn2:incoming>
      <bpmn2:outgoing>SequenceFlow_3</bpmn2:outgoing>
    </bpmn2:task>
    <bpmn2:exclusiveGateway id="ExclusiveGateway_1" gatewayDirection="Diverging">
@ -1124,6 +1127,26 @@ Betrag: <itemvalue>_amount</itemvalue> (Brutto € <itemvalue>_amount_brutto</it
    <bpmn2:sequenceFlow id="SequenceFlow_2" sourceRef="BoundaryEvent_1" targetRef="IntermediateCatchEvent_3"/>
    <bpmn2:sequenceFlow id="SequenceFlow_8" sourceRef="BoundaryEvent_2" targetRef="IntermediateCatchEvent_8"/>
    <bpmn2:sequenceFlow id="SequenceFlow_25" sourceRef="BoundaryEvent_3" targetRef="IntermediateCatchEvent_10"/>
+    <bpmn2:intermediateCatchEvent id="IntermediateCatchEvent_1" imixs:activityid="50" name="XML Test">
+      <bpmn2:extensionElements>
+        <imixs:item name="txtactivityresult" type="CDATA">
+          <imixs:value><![CDATA[ <item name="PDFXMLExtractor">
+    <filename>*.xml</filename>
+    <report>myReport</report>
+</item>]]></imixs:value>
+        </imixs:item>
+      </bpmn2:extensionElements>
+      <bpmn2:outgoing>SequenceFlow_26</bpmn2:outgoing>
+      <bpmn2:dataOutput id="DataOutput_7" name="Signal_7_Output"/>
+      <bpmn2:dataOutputAssociation id="DataOutputAssociation_7">
+        <bpmn2:sourceRef>DataOutput_7</bpmn2:sourceRef>
+      </bpmn2:dataOutputAssociation>
+      <bpmn2:outputSet id="OutputSet_7" name="Output Set 7">
+        <bpmn2:dataOutputRefs>DataOutput_7</bpmn2:dataOutputRefs>
+      </bpmn2:outputSet>
+      <bpmn2:signalEventDefinition id="SignalEventDefinition_7" signalRef="Signal_3"/>
+    </bpmn2:intermediateCatchEvent>
+    <bpmn2:sequenceFlow id="SequenceFlow_26" sourceRef="IntermediateCatchEvent_1" targetRef="Task_1"/>
  </bpmn2:process>
  <bpmndi:BPMNDiagram id="BPMNDiagram_1" name="Default Collaboration Diagram">
    <bpmndi:BPMNPlane id="BPMNPlane_1" bpmnElement="Collaboration_1">
@ -1285,6 +1308,12 @@ Betrag: <itemvalue>_amount</itemvalue> (Brutto € <itemvalue>_amount_brutto</it
          <dc:Bounds height="315.0" width="529.0" x="446.0" y="600.0"/>
        </bpmndi:BPMNLabel>
      </bpmndi:BPMNShape>
+      <bpmndi:BPMNShape id="BPMNShape_IntermediateCatchEvent_1" bpmnElement="IntermediateCatchEvent_1">
+        <dc:Bounds height="36.0" width="36.0" x="242.0" y="182.0"/>
+        <bpmndi:BPMNLabel id="BPMNLabel_54">
+          <dc:Bounds height="14.0" width="50.0" x="235.0" y="218.0"/>
+        </bpmndi:BPMNLabel>
+      </bpmndi:BPMNShape>
      <bpmndi:BPMNEdge id="BPMNEdge_SequenceFlow_1" bpmnElement="SequenceFlow_1" sourceElement="BPMNShape_ExclusiveGateway_1" targetElement="BPMNShape_IntermediateCatchEvent_2">
        <di:waypoint xsi:type="dc:Point" x="426.0" y="296.0"/>
        <di:waypoint xsi:type="dc:Point" x="426.0" y="188.0"/>
@ -1438,6 +1467,12 @@ Betrag: <itemvalue>_amount</itemvalue> (Brutto € <itemvalue>_amount_brutto</it
        <di:waypoint xsi:type="dc:Point" x="739.0" y="438.0"/>
        <bpmndi:BPMNLabel id="BPMNLabel_52"/>
      </bpmndi:BPMNEdge>
+      <bpmndi:BPMNEdge id="BPMNEdge_SequenceFlow_26" bpmnElement="SequenceFlow_26" sourceElement="BPMNShape_IntermediateCatchEvent_1" targetElement="BPMNShape_Task_1">
+        <di:waypoint xsi:type="dc:Point" x="278.0" y="200.0"/>
+        <di:waypoint xsi:type="dc:Point" x="296.0" y="200.0"/>
+        <di:waypoint xsi:type="dc:Point" x="296.0" y="296.0"/>
+        <bpmndi:BPMNLabel id="BPMNLabel_55"/>
+      </bpmndi:BPMNEdge>
    </bpmndi:BPMNPlane>
    <bpmndi:BPMNLabelStyle id="BPMNLabelStyle_1">
      <dc:Font name="arial" size="9.0"/>