- * The main usage of this filter is to combine it with a grouping filter and for each group + * The main usage of this filter is to combine it with a transit grouping filter and for each group * make sure there is at least {@code minNumItineraries} and that the best itinerary with respect * to each criterion is kept. So, if the grouping is based on time and riding common trips, then * this filter will use the reminding criterion (transfers, generalized-cost, - * [transit-group-priority]) to filter the grouped set of itineraries. + * [transit-group-priority]) to filter the grouped set of itineraries. DO NOT INCLUDE CRITERIA + * USED TO GROUP THE ITINERARIES, ONLY THE REMINDING CRITERION USED IN THE RAPTOR SEARCH. *
* IMPLEMENTATION DETAILS *
@@ -31,11 +31,14 @@ *
* Note! For a criteria like num-of-transfers or generalized-cost, there is only one set of "best" * itineraries, and usually there are only one or a few itineraries. In case there is more than one, - * picking just one is fine. But, for transit-group-priority there might be more than on set of - * itineraries. For each set, we need to pick one itinerary for the final result. Each of these - * sets may or may not have more than one itinerary. + * picking just one is fine. But, for transit-group-priority there might be more than one optimal + * set of itineraries. For each set, we need to pick one itinerary for the final result. Each of + * these sets may or may not have more than one itinerary. If you group by agency, then there will + * be at least one itinerary for each agency present in the result (simplified, an itinerary may + * consist of legs with different agencies). The transit-group-priority pareto-function used by + * Raptor is reused, so we do not need to worry about the logic here. *
- * Let's discuss an example: + * Let's discuss an example (this example also exists as a unit-test case): *
* minNumItineraries = 4
* comparators = [ generalized-cost, min-num-transfers, transit-group-priority ]
@@ -63,13 +66,13 @@
* The `minNumItineraries` limit is not met, so we need to pick another itinerary, we use the
* sort-order again and add itinerary #0. The result returned is: [#0, #2, #4, #5]
*/
-public class McMinimumNumberItineraryFilter implements RemoveItineraryFlagger {
+public class McMaxLimitFilter implements RemoveItineraryFlagger {
private final String name;
private final int minNumItineraries;
private final List comparators;
- public McMinimumNumberItineraryFilter(
+ public McMaxLimitFilter(
String name,
int minNumItineraries,
List comparators
diff --git a/src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMinimumNumberItineraryFilterTest.java b/src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMaxLimitFilterTest.java
similarity index 96%
rename from src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMinimumNumberItineraryFilterTest.java
rename to src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMaxLimitFilterTest.java
index e094e53cd03..14d0fe62337 100644
--- a/src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMinimumNumberItineraryFilterTest.java
+++ b/src/test/java/org/opentripplanner/routing/algorithm/filterchain/filters/system/mcmax/McMaxLimitFilterTest.java
@@ -14,7 +14,7 @@
import org.opentripplanner.routing.algorithm.raptoradapter.transit.cost.grouppriority.TransitGroupPriority32n;
import org.opentripplanner.transit.model._data.TransitModelForTest;
-class McMinimumNumberItineraryFilterTest {
+class McMaxLimitFilterTest {
private static final TransitModelForTest TEST_MODEL = TransitModelForTest.of();
@@ -44,7 +44,7 @@ class McMinimumNumberItineraryFilterTest {
private static final int TX_1 = 1;
private static final int TX_2 = 2;
- private final McMinimumNumberItineraryFilter subject = new McMinimumNumberItineraryFilter(
+ private final McMaxLimitFilter subject = new McMaxLimitFilter(
"test",
2,
List.of(
@@ -91,7 +91,7 @@ static List> filterTestCases() {
row(EXP_KEEP, COST_MED, TX_0, GROUP_BC, "Best: num-of-transfers")
),
/**
- * This is the example explained in JavaDoc {@link McMinimumNumberItineraryFilter}
+ * This is the example explained in JavaDoc {@link McMaxLimitFilter}
*/
List.of(
row(EXP_DROP, COST_LOW, TX_1, GROUP_A, ""),