Parameterising parser merge options (PathParser.setProcessingOptions()).

This is possibly related to #8.
Rothamsted · Jul 14, 2017 · c0d907b · c0d907b
1 parent 2266a8e
commit c0d907b
Show file tree

Hide file tree

Showing 4 changed files with 78 additions and 4 deletions.
diff --git a/modules/tab-parser-2/pom.xml b/modules/tab-parser-2/pom.xml
@@ -74,6 +74,27 @@
          <scope>test</scope>
       </dependency>
 
+			<dependency>
+		    <groupId>net.sourceforge.ondex.modules</groupId>
+		    <artifactId>generic</artifactId>
+		    <version>${project.version}</version>
+		    <scope>test</scope>
+			</dependency>      
+
+      <dependency>
+        <groupId>net.sourceforge.ondex.core</groupId>
+        <artifactId>lucene</artifactId>
+        <version>${project.version}</version>
+        <scope>test</scope>
+      </dependency>
+
+			<dependency>
+				<!-- Cute utility that allows for quick handling of exceptions (eg, checked/unchecked wrapping) -->
+				<groupId>com.machinezoo.noexception</groupId>
+				<artifactId>noexception</artifactId>
+				<version>1.1.0</version>
+			</dependency>      
+
    </dependencies>
 
    <build>

diff --git a/...les/tab-parser-2/src/main/java/net/sourceforge/ondex/plugins/tab_parser_2/TabParser2.java b/...les/tab-parser-2/src/main/java/net/sourceforge/ondex/plugins/tab_parser_2/TabParser2.java
@@ -2,12 +2,17 @@
 
 import java.io.File;
 import java.util.Collections;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
 
 import org.apache.log4j.Logger;
 
 import com.google.common.base.Optional;
+import com.machinezoo.noexception.Exceptions;
 
+import net.sourceforge.ondex.ONDEXPluginArguments;
 import net.sourceforge.ondex.args.ArgumentDefinition;
+import net.sourceforge.ondex.args.BooleanArgumentDefinition;
 import net.sourceforge.ondex.args.FileArgumentDefinition;
 import net.sourceforge.ondex.core.ONDEXConcept;
 import net.sourceforge.ondex.core.ONDEXRelation;
@@ -58,18 +63,40 @@ public ArgumentDefinition<?>[] getArgumentDefinitions ()
       	true, // required 
       	true, // preExisting 
       	false // isDirectory
-      )
+      ),
+      new BooleanArgumentDefinition ( PathParser.MERGE_ACC, "Accession-based concept merging in the parser", false, true ),
+      new BooleanArgumentDefinition ( PathParser.MERGE_NAME, "Name-based concept merging in the parser", false, false ),
+      new BooleanArgumentDefinition ( PathParser.MERGE_GDS, "Data source-based concept merging in the parser", false, false )
 		};
 	}
 
 	public void start () throws Exception
 	{		
-    String tabInputPath = (String) getArguments().getUniqueValue( FileArgumentDefinition.INPUT_FILE );
-    String tabConfigXmlPath = (String) getArguments().getUniqueValue( "configFile" );
+		ONDEXPluginArguments args = getArguments ();
+    String tabInputPath = (String) args.getUniqueValue( FileArgumentDefinition.INPUT_FILE );
+    String tabConfigXmlPath = (String) args.getUniqueValue( "configFile" );
 
     // The mapping parser returns an ONDEX parser straight
     PathParser tabParser = ConfigParser.parseConfigXml ( tabConfigXmlPath, graph, tabInputPath );
 
+    // Merging flags
+    //
+    String[] mergeFlags = (String[]) Stream.of ( PathParser.MERGE_ACC, PathParser.MERGE_NAME, PathParser.MERGE_GDS )
+    .filter ( flag ->
+    	// Keeps it if it's true, else this kind of merging won't happen, because it won't be in the final 
+    	// processing options.
+    	Boolean.TRUE.equals ( 
+    		Exceptions
+        // Wraps InvalidPluginArgumentException for getUniqueValue() into a RuntimeException   			
+    		.wrap ( ex -> new IllegalArgumentException ( "Plug-in argument error: " + ex.getMessage (), ex ) )
+    		.get ( () -> (Boolean) args.getUniqueValue ( flag ) )
+    	)
+    )
+    .collect ( Collectors.toList () )    
+    .toArray ( new String [ 0 ] );
+
+    tabParser.setProcessingOptions ( mergeFlags );
+
     Subgraph newGraph = tabParser.parse ();
     int nconcepts = Optional.fromNullable ( newGraph.getConcepts () ).or ( Collections.<ONDEXConcept>emptySet () ).size ();
     int nrelations = Optional.fromNullable ( newGraph.getRelations () ).or ( Collections.<ONDEXRelation>emptySet () ).size ();

diff --git a/...les/tab-parser-2/src/test/java/net/sourceforge/ondex/plugins/tab_parser_2/ParserTest.java b/...les/tab-parser-2/src/test/java/net/sourceforge/ondex/plugins/tab_parser_2/ParserTest.java
@@ -9,7 +9,6 @@
 import java.util.Set;
 
 import org.apache.log4j.Logger;
-import org.junit.Ignore;
 import org.junit.Test;
 
 import com.google.common.io.Resources;
@@ -188,6 +187,15 @@ public void testDupedRelation () throws Exception
 		pp.setProcessingOptions ( new String [ 0 ] );
 		pp.parse ();
 
+//		Mapping cmap = new Mapping ();
+//		ONDEXPluginArguments args = new ONDEXPluginArguments ( cmap.getArgumentDefinitions () );
+//		args.addOption ( ArgumentNames.IGNORE_AMBIGUOUS_ARG, false );
+//		args.addOption ( ArgumentNames.RELATION_TYPE_ARG, "enc" );
+//		args.addOption ( ArgumentNames.WITHIN_DATASOURCE_ARG, true );		
+//		cmap.setArguments ( args );;
+//		cmap.setONDEXGraph ( graph );
+//		cmap.start ();
+
 		log.info ( "Concepts: " + graph.getConcepts ().size () );
 		log.info ( "Relations: " + graph.getRelations ().size () );
 

diff --git a/modules/tab-parser-2/src/test/resources/duped_relations/duped_rels_wf.xml b/modules/tab-parser-2/src/test/resources/duped_relations/duped_rels_wf.xml
@@ -8,8 +8,26 @@
     <Parser name="tabParser2">
       <Arg name="InputFile">/Users/brandizi/Documents/Work/RRes/ondex_git/ondex-full/ondex-knet-builder/modules/tab-parser-2/src/test/resources/duped_relations/duped_rels.tsv</Arg>
       <Arg name="configFile">/Users/brandizi/Documents/Work/RRes/ondex_git/ondex-full/ondex-knet-builder/modules/tab-parser-2/src/test/resources/duped_relations/duped_rels_cfg.xml</Arg>
+      <Arg name="MERGE_ON_ACCESSIONS">false</Arg>
+      <Arg name="MERGE_ON_NAMES">false</Arg>
+      <Arg name="MERGE_ON_GDS">false</Arg>
       <Arg name="graphId">default</Arg>
     </Parser>
+
+		<!-- Mapping name="lowmemoryaccessionbased">
+		  <Arg name="IgnoreAmbiguity">false</Arg>
+		  <!- - Arg name="RelationType">collapse_me</Arg - ->
+		  <Arg name="WithinDataSourceMapping">false</Arg>
+			<Arg name="graphId">default</Arg>
+	  </Mapping -->
+
+		<!-- Transformer name="relationcollapser">
+      <Arg name="CloneAttributes">true</Arg>
+      <Arg name="CopyTagReferences">true</Arg>
+      <Arg name="graphId">default</Arg>
+      <Arg name="RelationType">enc</Arg>
+    </Transformer -->
+
   </Workflow>
 </Ondex>