Throw ERROR when file areas point at same file

src/main/java/gov/nasa/pds/tools/util/FileReferencedMapList.java

@@ -49,11 +49,9 @@ public class FileReferencedMapList {
   // Class to hold a map of every labels that referenced a physical file.
   // If a file is referenced by more than one label, an error should be thrown.
   class FileReferencedMap {
-    private URL url = null;
     private List<String> labelNames = new ArrayList<>();
 
-    public FileReferencedMap(URL url) {
-      this.url = url;
+    public FileReferencedMap() {
     }
 
     public void setLabels(String labelName) {
@@ -86,7 +84,7 @@ public FileReferencedMap setLabels(URL url, String labelName) {
     } else {
       // If fileReferencedMaps does not contain an existing, url, create a new
       // FileReferencedMap and add labelName to the new empty list.
-      fileReferencedMap = new FileReferencedMap(url);
+      fileReferencedMap = new FileReferencedMap();
       fileReferencedMap.setLabels(labelName);
       fileReferencedMaps.put(url, fileReferencedMap);
     }

src/main/java/gov/nasa/pds/tools/validate/CrossLabelFileAreaReferenceChecker.java

@@ -0,0 +1,61 @@
+package gov.nasa.pds.tools.validate;
+
+import java.io.IOException;
+import java.nio.file.Path;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.List;
+import javax.xml.parsers.DocumentBuilderFactory;
+import javax.xml.parsers.ParserConfigurationException;
+import javax.xml.transform.dom.DOMSource;
+import org.w3c.dom.Document;
+import org.xml.sax.SAXException;
+import gov.nasa.pds.tools.util.LabelUtil;
+
+public class CrossLabelFileAreaReferenceChecker {
+  final private static HashMap<String,List<String>> knownRefs = new HashMap<String,List<String>>();
+  private static String resolve (String name, ValidationTarget target) {
+    if (!name.startsWith ("/")) {
+      name = Path.of(target.getUrl().getPath()).getParent().resolve(name).toString();
+    }
+    return name;
+  }
+  /**
+   * @param name - the file being referenced by the file area
+   * @param target - the label being validated
+   * @return true if name and target are unique and only known references and false otherwise
+   * All of these exceptions should not happen because they are getting the LID from the target.
+   * Would not have made it this far if that could not have happened already, So, just pass them
+   * back to the called and let them handle it with their own generic exception handler/message.
+   * @throws IOException
+   * @throws ParserConfigurationException
+   * @throws SAXException
+   */
+  public static boolean add (String name, ValidationTarget target) throws IOException, ParserConfigurationException, SAXException {
+    boolean success = false;
+    DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
+    Document xml = dbf.newDocumentBuilder().parse(target.getUrl().openStream());
+    DOMSource domSource = new DOMSource(xml);    
+    for (String lid : LabelUtil.getLogicalIdentifiers (domSource, target.getUrl())) {
+      if (lid.contains("::")) {
+        lid = lid.substring (0, lid.indexOf("::"));
+      }
+      if (!knownRefs.keySet().contains (resolve(name, target))) {
+        knownRefs.put(resolve(name, target), (List<String>)Arrays.asList(lid, target.getUrl().getPath()));
+        success = true;
+      } else {
+        success = false;
+      }
+    }
+    return success;
+  }
+  public static String getOtherId (String name, ValidationTarget target) {
+    return knownRefs.get(resolve(name, target)).get(0);
+  }
+  public static String getOtherFilename (String name, ValidationTarget target) {
+    return knownRefs.get(resolve(name, target)).get(1);
+  }
+  public static void reset() {
+    knownRefs.clear();
+  }
+}

src/main/java/gov/nasa/pds/tools/validate/ProblemType.java

@@ -24,6 +24,8 @@ public enum ProblemType {
 
   MISSING_REFERENCED_FILE("error.label.missing_file"),
 
+  DUPLICATED_FILE_AREA_REFERENCE("error.label.file_areas_duplicated_reference"),
+
   INVALID_LABEL("error.validation.invalid_label"),
 
   EMPTY_FOLDER("error.validation.empty_folder", ProblemCategory.EXECUTION),

src/main/java/gov/nasa/pds/tools/validate/rule/pds4/FileReferenceValidationRule.java

@@ -45,6 +45,7 @@
 import gov.nasa.pds.tools.util.PDFUtil;
 import gov.nasa.pds.tools.util.Utility;
 import gov.nasa.pds.tools.util.XMLExtractor;
+import gov.nasa.pds.tools.validate.CrossLabelFileAreaReferenceChecker;
 import gov.nasa.pds.tools.validate.ProblemDefinition;
 import gov.nasa.pds.tools.validate.ProblemType;
 import gov.nasa.pds.tools.validate.ValidationProblem;
@@ -280,6 +281,15 @@ private boolean validate(NodeInfo xml) {
                 name = child.getStringValue();
                 this.fileMapping.put(name, "");
                 LOG.debug("FileReferenceValidationRule:validate:name {}", name);
+                if (!CrossLabelFileAreaReferenceChecker.add (name, target)) {
+                  this.getListener().addProblem(
+                      new ValidationProblem(
+                          new ProblemDefinition(ExceptionType.ERROR, ProblemType.DUPLICATED_FILE_AREA_REFERENCE,
+                              "This file area references " + name + " that is already used by label "
+                              + CrossLabelFileAreaReferenceChecker.getOtherId(name, target)
+                              + " in file " + CrossLabelFileAreaReferenceChecker.getOtherFilename(name, target)),
+                          target, fileObject.getLineNumber(), -1));
+                };
               } else if ("md5_checksum".equals(child.getLocalPart())) {
                 checksum = child.getStringValue();
                 LOG.debug("FileReferenceValidationRule:validate:checksum {}", checksum);

src/test/java/cucumber/StepDefs.java

@@ -12,6 +12,7 @@
 import org.slf4j.LoggerFactory;
 import com.google.gson.Gson;
 import com.google.gson.JsonObject;
+import gov.nasa.pds.tools.validate.CrossLabelFileAreaReferenceChecker;
 import gov.nasa.pds.tools.validate.ProblemType;
 import gov.nasa.pds.validate.ValidateLauncher;
 import gov.nasa.pds.validate.constants.TestConstants;
@@ -78,6 +79,8 @@ void tearDown() throws Exception {
     this.launcher.flushValidators();
     // It seems the launcher does not completely flush any references to schematron
     // which causes problem for subsequent tests.
+    // get rid of the cross references
+    CrossLabelFileAreaReferenceChecker.reset();
   }
 
   private void createManifestFileDo(String testPath) {

src/test/resources/features/developer.feature

@@ -9,6 +9,10 @@ Scenario Outline: Execute validate command for tests below.
   Examples:
  | testName                                             | testDir     | messageCount | messageText         | problemEnum     | resourceDir          | reportDir     | commandArgs                                                                | refOutputValue | 
 
+# Validate#755
+|"NASA-PDS/validate#755 Detect and report file areas referencing same file" | "github755" | 2 | "2 errors expected" | "DUPLICATED_FILE_AREA_REFERENCE" | "src/test/resources"   | "target/test" | "-r {reportDir}/report_github755.json -s json -skip-content-validation -t {resourceDir}/github755/m221011.0013.xml {resourceDir}/github755/m221011.0014.xml {resourceDir}/github755/m221011.0015.xml {resourceDir}/github755/m221011.0030.xml" | "report_github755.json" |
+|"NASA-PDS/validate#755 No common file referenced" | "github755" | 0 | "0 errors expected" | "totalErrors" | "src/test/resources"   | "target/test" | "-r {reportDir}/report_github755.1.json -s json -skip-content-validation -t {resourceDir}/github755/m221011.0013.xml {resourceDir}/github755/m221011.0015.xml" | "report_github755.1.json" |
+
 # Validate#754
 |"NASA-PDS/validate#754 Success differentiating file area offsets" | "github754" | 0 | "1 warnings expected" | "DATA_OBJECTS_OUT_OF_ORDER" | "src/test/resources"   | "target/test" | "-r {reportDir}/report_github754.json -s json --skip-context-validation  -t {resourceDir}/github754/Cassini_ISS_CB2_Jupiter_global_map_2.xml" | "report_github754.json" |
 
@@ -263,7 +267,7 @@ Scenario Outline: Execute validate command for tests below.
 # https://github.com/NASA-PDS/validate/issues/188 As a user, I want to validate a bundle that uses multiple versions of the Information Model / Discipline LDDs
  |"NASA-PDS/validate#188 VALID" | "github188" | 0 | "0 errors message expected" | "totalErrors" | "src/test/resources" | "target/test" | "--skip-content-validation -r {reportDir}/report_github188_label_valid_both.json  -s json -t {resourceDir}/github188/bundle_cassini-huygens-coradar.xml {resourceDir}/github188/BILQH07S314_D065_T008S02_V02_without_Missing_Area_tag.xml" | "report_github188_label_valid_both.json" |
 
- |"NASA-PDS/validate#217 Success with Errors: Tables Types" | "github217" | 15 | "15 error message expected" | "totalErrors" | "src/test/resources" | "target/test" | "-r {reportDir}/report_github217.json  -s json -t {resourceDir}/github217" | "report_github217.json" |
+ |"NASA-PDS/validate#217 Success with Errors: Tables Types" | "github217" | 65 | "65 error message expected" | "totalErrors" | "src/test/resources" | "target/test" | "-r {reportDir}/report_github217.json  -s json -t {resourceDir}/github217" | "report_github217.json" |
 
 # https://github.com/NASA-PDS/validate/issues/210 As a user, I want validate to raise a WARNING when differing versions of IM are used within a bundle
  |"NASA-PDS/validate#210 WITH_WARNING" | "github210" | 1 | "1 warning message expected" | "totalWarnings" | "src/test/resources" | "target/test" | "--skip-content-validation -r {reportDir}/report_github210_label_valid_both_with_warning.json  -s json -t {resourceDir}/github210/bundle_cassini-huygens-coradar.xml {resourceDir}/github210/BILQH07S314_D065_T008S02_V02_without_Missing_Area_tag.xml" | "report_github210_label_valid_both_with_warning.json" |

src/test/resources/github755/m221011.0013.xml

@@ -0,0 +1,203 @@
+<?xml version="1.0" encoding="utf-8"?>
+<?xml-model
+        href="https://pds.nasa.gov/pds4/pds/v1/PDS4_PDS_1E00.sch"
+        schematypens="http://purl.oclc.org/dsdl/schematron"?>
+<?xml-model
+        href="https://pds.nasa.gov/pds4/disp/v1/PDS4_DISP_1B00.sch"
+        schematypens="http://purl.oclc.org/dsdl/schematron"?>
+<?xml-model
+        href="https://pds.nasa.gov/pds4/img/v1/PDS4_IMG_1E00_1800.sch"
+        schematypens="http://purl.oclc.org/dsdl/schematron"?>
+<?xml-model
+        href="https://pds.nasa.gov/pds4/geom/v1/PDS4_GEOM_1E00_1810.sch"
+        schematypens="http://purl.oclc.org/dsdl/schematron"?>
+<Product_Observational
+        xmlns="http://pds.nasa.gov/pds4/pds/v1"
+        xmlns:disp="http://pds.nasa.gov/pds4/disp/v1"
+        xmlns:img="http://pds.nasa.gov/pds4/img/v1"
+        xmlns:geom="http://pds.nasa.gov/pds4/geom/v1"
+        xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+        xsi:schemaLocation="http://pds.nasa.gov/pds4/pds/v1
+                        https://pds.nasa.gov/pds4/pds/v1/PDS4_PDS_1E00.xsd
+                        http://pds.nasa.gov/pds4/disp/v1
+                        https://pds.nasa.gov/pds4/disp/v1/PDS4_DISP_1B00.xsd
+                        http://pds.nasa.gov/pds4/img/v1
+                        https://pds.nasa.gov/pds4/img/v1/PDS4_IMG_1E00_1800.xsd
+                        http://pds.nasa.gov/pds4/geom/v1
+                        https://pds.nasa.gov/pds4/geom/v1/PDS4_GEOM_1E00_1810.xsd">
+    <Identification_Area>
+        <logical_identifier>urn:nasa:pds:dart_teleobs:data_mroraw:m221011.0013</logical_identifier>
+        <version_id>1.0</version_id>
+        <title>Magdalena Ridge Observatory 2.4-meter Telescope Raw Image: m221011.0013</title>
+        <information_model_version>1.14.0.0</information_model_version>
+        <product_class>Product_Observational</product_class>
+        <Modification_History>
+            <Modification_Detail>
+                <modification_date>2023-08-01</modification_date>
+                <version_id>1.0</version_id>
+                <description>Initial PDS4 Delivery</description>
+            </Modification_Detail>
+        </Modification_History>
+    </Identification_Area>
+    <Observation_Area>
+        <Time_Coordinates>
+            <start_date_time>2022-10-11T08:49:36.32Z</start_date_time>
+            <stop_date_time>2022-10-11T08:49:51.32Z</stop_date_time>
+        </Time_Coordinates>
+        <Investigation_Area>
+            <name>Double Asteroid Redirection Test</name>
+            <type>Mission</type>
+            <Internal_Reference>
+                <lid_reference>urn:nasa:pds:context:investigation:mission.double_asteroid_redirection_test</lid_reference>
+                <reference_type>data_to_investigation</reference_type>
+            </Internal_Reference>
+        </Investigation_Area>
+        <Observing_System>
+            <Observing_System_Component>
+                <name>Magdalena Ridge Observatory (MRO)</name>
+                <type>Host</type>
+                <Internal_Reference>
+                    <lid_reference>urn:nasa:pds:context:facility:observatory.magdalena_ridge</lid_reference>
+                    <reference_type>is_facility</reference_type>
+                </Internal_Reference>
+            </Observing_System_Component>
+            <Observing_System_Component>
+                <name>MRO 2.4-meter Telescope</name>
+                <type>Telescope</type>
+                <description>
+                    The MRO 2.4-meter Telescope is a Ritchey-Chretien telescope with a 2.4-meter mirror located in the 
+                    Cibola National Forest in New Mexico. It is owned and operated by New Mexico Institute of Mining and 
+                    Technology.
+                </description>
+                <Internal_Reference>
+                    <lid_reference>urn:nasa:pds:context:telescope:magdalena_ridge.mro2m4</lid_reference>
+                    <reference_type>is_telescope</reference_type>
+                </Internal_Reference>
+            </Observing_System_Component>
+            <Observing_System_Component>
+                <name>MRO2K</name>
+                <type>Instrument</type>
+                <description>
+                    MRO2K, the visible-wavelength direct-imaging CCD for the MRO 2.4-meter Telescope, is an Andor iKon-L 936. 
+                    It is a 2048 x 2048 array delivering a 27.6 x 27.6 mm active image area. The CCD images are a 4.5 arc 
+                    minute square field of view with 13.5-micron pixels.
+                </description>
+                <Internal_Reference>
+                    <lid_reference>urn:nasa:pds:context:instrument:magdalena_ridge.mro2m4.andor_2kx2k_ccd</lid_reference>
+                    <reference_type>is_instrument</reference_type>
+                </Internal_Reference>
+            </Observing_System_Component>
+        </Observing_System>
+        <Target_Identification>
+            <name>Sky</name>
+            <type>Calibration Field</type>
+        </Target_Identification>
+        <Discipline_Area>
+            <disp:Display_Settings>
+                <Local_Internal_Reference>
+                    <local_identifier_reference>m221011.0013</local_identifier_reference>
+                    <local_reference_type>display_settings_to_array</local_reference_type>
+                </Local_Internal_Reference>
+                <disp:Display_Direction>
+                    <disp:horizontal_display_axis>Sample</disp:horizontal_display_axis>
+                    <disp:horizontal_display_direction>Left to Right</disp:horizontal_display_direction>
+                    <disp:vertical_display_axis>Line</disp:vertical_display_axis>
+                    <disp:vertical_display_direction>Bottom to Top</disp:vertical_display_direction>
+                </disp:Display_Direction>
+            </disp:Display_Settings>
+            <img:Exposure>
+                <img:exposure_duration unit="s">15.0</img:exposure_duration>
+            </img:Exposure>
+            <img:Optical_Filter>
+                <img:filter_name>R</img:filter_name>
+                <img:bandwidth unit="nm">158</img:bandwidth>
+                <img:center_filter_wavelength unit="nm">640</img:center_filter_wavelength>
+            </img:Optical_Filter>
+            <geom:Geometry>
+                <geom:Image_Display_Geometry>
+                    <Local_Internal_Reference>
+                        <local_identifier_reference>m221011.0013</local_identifier_reference>
+                        <local_reference_type>display_to_data_object</local_reference_type>
+                    </Local_Internal_Reference>
+                    <geom:Display_Direction>
+                        <geom:horizontal_display_axis>Sample</geom:horizontal_display_axis>
+                        <geom:horizontal_display_direction>Left to Right</geom:horizontal_display_direction>
+                        <geom:vertical_display_axis>Line</geom:vertical_display_axis>
+                        <geom:vertical_display_direction>Bottom to Top</geom:vertical_display_direction>
+                    </geom:Display_Direction>
+                    <geom:Object_Orientation_RA_Dec>
+                        <geom:right_ascension_angle unit="deg">89.60996</geom:right_ascension_angle>
+                        <geom:declination_angle unit="deg">-16.43276</geom:declination_angle>
+                        <geom:celestial_north_clock_angle unit="deg">0</geom:celestial_north_clock_angle>
+                        <geom:Reference_Frame_Identification>
+                            <geom:name>J2000</geom:name>
+                            <geom:comment>equinox of RA and DEC</geom:comment>
+                        </geom:Reference_Frame_Identification>
+                    </geom:Object_Orientation_RA_Dec>
+                </geom:Image_Display_Geometry>
+            </geom:Geometry>
+        </Discipline_Area>
+    </Observation_Area>
+    <Reference_List>
+        <Internal_Reference>
+            <lid_reference>urn:nasa:pds:dart_teleobs:document_mro:mro_2p4m_dart_pds_sis</lid_reference>
+            <reference_type>data_to_document</reference_type>
+            <comment>
+                MRO DART Uncalibrated/Calibrated Data Product Software Interface Specification document which describes the data products in this collection.
+            </comment>
+        </Internal_Reference>
+        <Internal_Reference>
+            <lid_reference>urn:nasa:pds:dart_teleobs:document_mro:mro_description</lid_reference>
+            <reference_type>data_to_document</reference_type>
+            <comment>Overview of the MRO 2.4-meter Telescope.</comment>
+        </Internal_Reference>
+        <Internal_Reference>
+            <lid_reference>urn:nasa:pds:dart_teleobs:document_mro:mro2k_description</lid_reference>
+            <reference_type>data_to_document</reference_type>
+            <comment>Short description of the MRO2K instrument.</comment>
+        </Internal_Reference>
+    </Reference_List>
+    <File_Area_Observational>
+        <File>
+            <file_name>m221011.0013.fits</file_name>
+            <comment>Raw MRO image file</comment>
+        </File>
+        <Header>
+            <offset unit="byte">0</offset>
+            <object_length unit="byte">5760</object_length>
+            <parsing_standard_id>FITS 3.0</parsing_standard_id>
+            <description>Fits header.</description>
+        </Header>
+        <Array_2D_Image>
+            <local_identifier>m221011.0013</local_identifier>
+            <offset unit="byte">5760</offset>
+            <axes>2</axes>
+            <axis_index_order>Last Index Fastest</axis_index_order>
+            <Element_Array>
+                <data_type>SignedMSB2</data_type>
+                <unit>DN</unit>
+                <value_offset>32768.0</value_offset>
+            </Element_Array>
+            <Axis_Array>
+                <axis_name>Line</axis_name>
+                <elements>512</elements>
+                <sequence_number>1</sequence_number>
+            </Axis_Array>
+            <Axis_Array>
+                <axis_name>Sample</axis_name>
+                <elements>512</elements>
+                <sequence_number>2</sequence_number>
+            </Axis_Array>
+        </Array_2D_Image>
+    </File_Area_Observational>
+    <File_Area_Observational_Supplemental>
+        <File>
+            <file_name>comps_221011.png</file_name>
+            <comment>Identifies the reference stars used in the photometric reduction.</comment>
+        </File>
+        <Encoded_Image>
+            <offset unit="byte">0</offset>
+            <encoding_standard_id>PNG</encoding_standard_id>
+        </Encoded_Image>
+    </File_Area_Observational_Supplemental>
+</Product_Observational>