partially revert 0f13423, 1f3e66a and b917561

This reverts a majority of the initial implementation of the partial L* algorithm, including the partial observation table and some of the SLI-specific interfaces.

While the concept of state local inputs showed huge improvements regarding query performance, the implementation via specific partial versions of learning algorithms was hard to maintain, as essentially every currently existing learning algorithm would have to be re-implemented.

Instead, this commit introduces the SLI concept as a SUL filter and introduces a special StateLocalInputSULOracle which early-answers queries that would traverse unavailable inputs with a previously specified symbol. This way, queries that would traverse undefined input symbols still won't be executed on the SUL but the SUL appears as a 'total' Mealy system to the learner, allowing one to use every currently existing Mealy learner as-is.

The only 'drawback' of this approach is that the internal datastructures of the learners are bigger (consume more memory) compared to a 'partial' version of the learner, since the system appears to be total when it is not. But since this is often not the limiting factor in practical AAL, this is a worthwhile trade-off.

Author: mtf90

CHANGELOG.md

@@ -15,6 +15,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 * JSR305 annotations have been replaced with checker-framework annotations.
   * LearnLib (incl. AutomataLib) now follows checker-framework's convention that (non-annotated) types are usually considered non-null unless explicitly annotated with `@Nullable`.
   * LearnLib (incl. AutomataLib) no longer has a (runtime-) dependency on JSR305 (and other `javax.*`) annotations or includes them in the distribution artifact. This now makes LearnLib (incl. AutomataLib) compliant with [Oracle's binary code license](https://www.oracle.com/downloads/licenses/binary-code-license.html) and allows LearnLib (incl. AutomataLib) artifacts as-is to be bundled in binary distributions with Oracle's JDKs/JREs.
+* A lot of code for inferring partial Mealy machines (esp. `PartialLStarMealy` and `PartialObservationTable`) has been removed/refactored. The concept of state local inputs is now implemented as a SUL filter and introduces a special `StateLocalInputSULOracle` which early-answers queries that would traverse unavailable inputs with a previously specified symbol. This way, queries that would traverse undefined input symbols still won't be executed on the SUL but the SUL appears as a 'total' Mealy system to the learner, allowing one to use every currently existing Mealy learner as-is. See the in-tree examples for more information.
 
 ### Removed
 

algorithms/active/lstar/pom.xml

@@ -70,11 +70,6 @@ limitations under the License.
             <artifactId>automata-util</artifactId>
         </dependency>
 
-        <dependency>
-            <groupId>org.slf4j</groupId>
-            <artifactId>slf4j-api</artifactId>
-        </dependency>
-
         <!-- build -->
         <dependency>
             <groupId>com.github.misberner.buildergen</groupId>

algorithms/active/lstar/src/main/java/de/learnlib/algorithms/lstar/AbstractAutomatonLStar.java

@@ -76,7 +76,7 @@ public A getHypothesisModel() {
     protected abstract A exposeInternalHypothesis();
 
     @Override
-    public void startLearning() {
+    public final void startLearning() {
         super.startLearning();
         updateInternalHypothesis();
     }
@@ -125,14 +125,11 @@ protected void updateInternalHypothesis() {
                 I input = alphabet.getSymbol(i);
 
                 Row<I> succ = sp.getSuccessor(i);
+                int succId = succ.getRowContentId();
 
-                if (succ != null) {
-                    int succId = succ.getRowContentId();
+                S succState = stateInfos.get(succId).getState();
 
-                    S succState = stateInfos.get(succId).getState();
-
-                    setTransition(state, input, succState, sp, i);
-                }
+                setTransition(state, input, succState, sp, i);
             }
         }
     }

algorithms/active/lstar/src/main/java/de/learnlib/algorithms/lstar/AbstractLStar.java

@@ -25,7 +25,6 @@
 import de.learnlib.api.algorithm.feature.GlobalSuffixLearner;
 import de.learnlib.api.oracle.MembershipOracle;
 import de.learnlib.api.query.DefaultQuery;
-import de.learnlib.datastructure.observationtable.AbstractObservationTable;
 import de.learnlib.datastructure.observationtable.GenericObservationTable;
 import de.learnlib.datastructure.observationtable.Inconsistency;
 import de.learnlib.datastructure.observationtable.OTLearner;
@@ -59,7 +58,7 @@ public abstract class AbstractLStar<A, I, D>
 
     protected final Alphabet<I> alphabet;
     protected final MembershipOracle<I, D> oracle;
-    protected AbstractObservationTable<I, D> table;
+    protected GenericObservationTable<I, D> table;
 
     /**
      * Constructor.

algorithms/active/lstar/src/main/java/de/learnlib/algorithms/lstar/AbstractLStarState.java

@@ -17,7 +17,7 @@
 
 import java.io.Serializable;
 
-import de.learnlib.datastructure.observationtable.AbstractObservationTable;
+import de.learnlib.datastructure.observationtable.GenericObservationTable;
 
 /**
  * Class that contains all data that represent the internal state of the {@link AbstractLStar} learner.
@@ -31,13 +31,13 @@
  */
 public abstract class AbstractLStarState<I, D> implements Serializable {
 
-    private final AbstractObservationTable<I, D> observationTable;
+    private final GenericObservationTable<I, D> observationTable;
 
-    AbstractLStarState(final AbstractObservationTable<I, D> observationTable) {
+    AbstractLStarState(final GenericObservationTable<I, D> observationTable) {
         this.observationTable = observationTable;
     }
 
-    AbstractObservationTable<I, D> getObservationTable() {
+    GenericObservationTable<I, D> getObservationTable() {
         return observationTable;
     }
 }

algorithms/active/lstar/src/main/java/de/learnlib/algorithms/lstar/AutomatonLStarState.java

@@ -18,7 +18,7 @@
 import java.util.List;
 
 import de.learnlib.algorithms.lstar.AbstractAutomatonLStar.StateInfo;
-import de.learnlib.datastructure.observationtable.AbstractObservationTable;
+import de.learnlib.datastructure.observationtable.GenericObservationTable;
 
 /**
  * Class that contains all data that represent the internal state of the {@link AbstractAutomatonLStar} learner and its
@@ -40,7 +40,7 @@ public class AutomatonLStarState<I, D, AI, S> extends AbstractLStarState<I, D> {
     private final AI hypothesis;
     private final List<StateInfo<S, I>> stateInfos;
 
-    AutomatonLStarState(final AbstractObservationTable<I, D> observationTable,
+    AutomatonLStarState(final GenericObservationTable<I, D> observationTable,
                         final AI hypothesis,
                         final List<StateInfo<S, I>> stateInfos) {
         super(observationTable);