fix(#1187): String lengths are uniformly random

Previously string lengths would follow a geometric distribution between
the min and max, where each successive term would have a 3/10 chance of
being the terminal character added to the string.

Therefore in the previous case, regardless of the maximum (where the maximum is at
least 5 greater than the minimum), typically 50% of the values would be
within 2 characters of the minimum length.

This change instead generates up to the maximum length (our current
default is 1,000), and slices randomly between the min and max length.
This change appears to cut performance by about half for the example
given in shorter-than, at the advantage of giving a uniform
distribution.

RandomStringFactory now uses a single string, instead of doing constant
string concatenation. This stops strings being pushed onto the heap.

common/src/main/java/com/scottlogic/deg/common/util/Defaults.java

@@ -32,6 +32,7 @@ public class Defaults {
 
     public static final BigDecimal NUMERIC_MAX = new BigDecimal("1e20");
     public static final BigDecimal NUMERIC_MIN = new BigDecimal("-1e20");
+
     public static final int MAX_STRING_LENGTH = 1000;
     public static final OffsetDateTime ISO_MAX_DATE = OffsetDateTime.of(9999, 12, 31, 23, 59, 59, 999_000_000, ZoneOffset.UTC);
     public static final OffsetDateTime ISO_MIN_DATE = OffsetDateTime.of(1, 1, 1, 0, 0, 0, 0, ZoneOffset.UTC);

generator/src/main/java/com/scottlogic/deg/generator/generation/string/factorys/RandomStringFactory.java

@@ -16,62 +16,46 @@
 
 package com.scottlogic.deg.generator.generation.string.factorys;
 
+import com.scottlogic.deg.common.ValidationException;
 import com.scottlogic.deg.generator.generation.string.StringUtils;
 import com.scottlogic.deg.generator.utils.RandomNumberGenerator;
 import dk.brics.automaton.State;
 import dk.brics.automaton.Transition;
 
+import java.util.LinkedList;
 import java.util.List;
 
 public class RandomStringFactory {
 
-    public String createRandomString(
-            String currentString,
-            State state,
-            int minLength,
-            int maxLength,
-            RandomNumberGenerator random) {
+    public String createRandomString(State state, RandomNumberGenerator random) {
+        List<Integer> validIndices = new LinkedList<>();
 
-        if (finishCreating(currentString, state, minLength, maxLength, random)) {
-            return currentString;
-        }
-
-        List<Transition> transitions = state.getSortedTransitions(false);
-
-        Transition randomTransition = transitions.get(random.nextInt(transitions.size()));
-
-        char randomChar = getRandomChar(random, randomTransition);
-
-        return createRandomString(
-            currentString + randomChar,
-            randomTransition.getDest(),
-            minLength,
-            maxLength,
-            random);
-    }
-
-    private boolean finishCreating(
-            String currentString,
-            State state,
-            int minLength,
-            int maxLength,
-            RandomNumberGenerator random) {
+        StringBuilder builder = new StringBuilder();
 
         if (state.isAccept()) {
-            if (currentString.length() == maxLength) {
-                return true;
-            }
-            if (currentString.length() >= minLength && randomlyStop(random)) {
-                return true;
+            validIndices.add(0);
+        }
+
+        for (int i = 1; !finishCreating(state); i++) {
+            List<Transition> transitions = state.getSortedTransitions(false);
+            Transition randomTransition = transitions.get(random.nextInt(transitions.size()));
+            builder.append(getRandomChar(random, randomTransition));
+            state = randomTransition.getDest();
+            if (state.isAccept()) {
+                validIndices.add(i);
             }
         }
 
-        return state.getTransitions().isEmpty();
-    }
+        if (validIndices.isEmpty()) {
+            throw new ValidationException("No possible states from the current regex");
+        }
 
+        int randomIndex = random.nextInt(validIndices.size());
+        return builder.toString().substring(0, validIndices.get(randomIndex));
+    }
 
-    private boolean randomlyStop(RandomNumberGenerator random) {
-        return random.nextInt(10) < 3; // 3 in 10 chance of stopping
+    private boolean finishCreating(State state) {
+        return state.getTransitions().isEmpty();
     }
 
     /**

generator/src/main/java/com/scottlogic/deg/generator/generation/string/generators/RegexStringGenerator.java

@@ -16,10 +16,13 @@
 
 package com.scottlogic.deg.generator.generation.string.generators;
 
+import com.scottlogic.deg.common.profile.FieldType;
+import com.scottlogic.deg.generator.fieldspecs.FieldSpecFactory;
 import com.scottlogic.deg.generator.generation.string.AutomatonUtils;
 import com.scottlogic.deg.generator.generation.string.iterators.FiniteStringAutomatonIterator;
 import com.scottlogic.deg.generator.generation.string.factorys.InterestingStringFactory;
 import com.scottlogic.deg.generator.generation.string.factorys.RandomStringFactory;
+import com.scottlogic.deg.generator.restrictions.string.StringRestrictions;
 import com.scottlogic.deg.generator.utils.RandomNumberGenerator;
 import dk.brics.automaton.Automaton;
 
@@ -39,6 +42,8 @@ public class RegexStringGenerator implements StringGenerator {
      */
     private static final Map<String, Automaton> containingRegexAutomatonCache = new HashMap<>();
 
+    private static final RegexStringGenerator DEFAULT = (RegexStringGenerator) ((StringRestrictions) FieldSpecFactory.fromType(FieldType.STRING).getRestrictions()).createGenerator();
+
     private Automaton automaton;
     private final String regexRepresentation;
 
@@ -119,7 +124,7 @@ RegexStringGenerator union(RegexStringGenerator otherGenerator) {
     @Override
     public StringGenerator complement() {
         return new RegexStringGenerator(
-            this.automaton.clone().complement(),
+            this.automaton.clone().complement().intersection(DEFAULT.automaton),
             complementaryRepresentation(this.regexRepresentation));
     }
 
@@ -150,10 +155,7 @@ public Stream<String> generateAllValues() {
     public Stream<String> generateRandomValues(RandomNumberGenerator randomNumberGenerator) {
         return Stream.generate(
             () -> randomStringFactory.createRandomString(
-                "",
                 automaton.getInitialState(),
-                1,
-                Integer.MAX_VALUE,
                 randomNumberGenerator));
     }
 

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/AfterConstantTimeConstraint.java

@@ -22,7 +22,7 @@
 import com.scottlogic.deg.generator.restrictions.linear.Limit;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictions;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory;
-import com.scottlogic.deg.generator.utils.Defaults;
+import com.scottlogic.deg.generator.utils.GeneratorDefaults;
 
 import java.time.LocalTime;
 
@@ -50,7 +50,7 @@ public AtomicConstraint negate() {
     public FieldSpec toFieldSpec() {
         final Limit<LocalTime> min = new Limit<>(referenceValue.getValue(), false);
         final LinearRestrictions<LocalTime> timeRestrictions =
-            LinearRestrictionsFactory.createTimeRestrictions(min, Defaults.TIME_MAX_LIMIT);
+            LinearRestrictionsFactory.createTimeRestrictions(min, GeneratorDefaults.TIME_MAX_LIMIT);
         return FieldSpecFactory.fromRestriction(timeRestrictions);
     }
 }

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/AfterConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MAX_LIMIT;
 
 public class AfterConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/AfterOrAtConstraint.java

@@ -27,7 +27,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MAX_LIMIT;
 
 public class AfterOrAtConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/AfterOrEqualToConstantTimeConstraint.java

@@ -22,7 +22,7 @@
 import com.scottlogic.deg.generator.restrictions.linear.Limit;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictions;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory;
-import com.scottlogic.deg.generator.utils.Defaults;
+import com.scottlogic.deg.generator.utils.GeneratorDefaults;
 
 import java.time.LocalTime;
 
@@ -50,7 +50,7 @@ public AtomicConstraint negate() {
     public FieldSpec toFieldSpec() {
         final Limit<LocalTime> min = new Limit<>(referenceValue.getValue(), true);
         final LinearRestrictions<LocalTime> timeRestrictions =
-            LinearRestrictionsFactory.createTimeRestrictions(min, Defaults.TIME_MAX_LIMIT);
+            LinearRestrictionsFactory.createTimeRestrictions(min, GeneratorDefaults.TIME_MAX_LIMIT);
         return FieldSpecFactory.fromRestriction(timeRestrictions);
     }
 }

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/BeforeConstantTimeConstraint.java

@@ -22,7 +22,7 @@
 import com.scottlogic.deg.generator.restrictions.linear.Limit;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictions;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory;
-import com.scottlogic.deg.generator.utils.Defaults;
+import com.scottlogic.deg.generator.utils.GeneratorDefaults;
 
 import java.time.LocalTime;
 
@@ -51,7 +51,7 @@ public FieldSpec toFieldSpec() {
 
         final Limit<LocalTime> max = new Limit<>(referenceValue.getValue(), false);
         final LinearRestrictions<LocalTime> timeRestriction =
-            LinearRestrictionsFactory.createTimeRestrictions(Defaults.TIME_MIN_LIMIT, max);
+            LinearRestrictionsFactory.createTimeRestrictions(GeneratorDefaults.TIME_MIN_LIMIT, max);
         if (timeRestriction.isContradictory()) {
             return FieldSpecFactory.nullOnly();
         }

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/BeforeConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MIN_LIMIT;
 
 public class BeforeConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/BeforeOrAtConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MIN_LIMIT;
 
 public class BeforeOrAtConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/BeforeOrEqualToConstantTimeConstraint.java

@@ -22,7 +22,7 @@
 import com.scottlogic.deg.generator.restrictions.linear.Limit;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictions;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory;
-import com.scottlogic.deg.generator.utils.Defaults;
+import com.scottlogic.deg.generator.utils.GeneratorDefaults;
 
 import java.time.LocalTime;
 
@@ -49,7 +49,7 @@ public AtomicConstraint negate() {
     public FieldSpec toFieldSpec() {
         final Limit<LocalTime> max = new Limit<>(referenceValue.getValue(), true);
         final LinearRestrictions<LocalTime> timeRestriction =
-            LinearRestrictionsFactory.createTimeRestrictions(Defaults.TIME_MIN_LIMIT, max);
+            LinearRestrictionsFactory.createTimeRestrictions(GeneratorDefaults.TIME_MIN_LIMIT, max);
         return FieldSpecFactory.fromRestriction(timeRestriction);
     }
 }

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/GranularToDateConstraint.java

@@ -25,8 +25,8 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MIN_LIMIT;
 
 public class GranularToDateConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/GranularToTimeConstraint.java

@@ -21,7 +21,7 @@
 import com.scottlogic.deg.generator.fieldspecs.FieldSpec;
 import com.scottlogic.deg.generator.fieldspecs.FieldSpecFactory;
 import com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory;
-import com.scottlogic.deg.generator.utils.Defaults;
+import com.scottlogic.deg.generator.utils.GeneratorDefaults;
 
 public class GranularToTimeConstraint implements AtomicConstraint {
     public final TimeGranularity timeGranularity;
@@ -51,8 +51,8 @@ public AtomicConstraint negate() {
     public FieldSpec toFieldSpec() {
         return FieldSpecFactory.fromRestriction(
             LinearRestrictionsFactory.createTimeRestrictions(
-                Defaults.TIME_MIN_LIMIT,
-                Defaults.TIME_MAX_LIMIT,
+                GeneratorDefaults.TIME_MIN_LIMIT,
+                GeneratorDefaults.TIME_MAX_LIMIT,
                 timeGranularity));
     }
 }

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/GreaterThanConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MAX_LIMIT;
 
 public class GreaterThanConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/GreaterThanOrEqualToConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MAX_LIMIT;
 
 public class GreaterThanOrEqualToConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/LessThanConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MIN_LIMIT;
 
 public class LessThanConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/profile/constraints/atomic/LessThanOrEqualToConstraint.java

@@ -26,7 +26,7 @@
 import java.util.Objects;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MIN_LIMIT;
 
 public class LessThanOrEqualToConstraint implements AtomicConstraint {
     public final Field field;

generator/src/main/java/com/scottlogic/deg/generator/restrictions/linear/LinearRestrictionsFactory.java

@@ -22,7 +22,7 @@
 import java.time.OffsetDateTime;
 
 import static com.scottlogic.deg.common.util.Defaults.*;
-import static com.scottlogic.deg.generator.utils.Defaults.*;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.*;
 
 public class LinearRestrictionsFactory {
 

generator/src/main/java/com/scottlogic/deg/generator/utils/Defaults.java → generator/src/main/java/com/scottlogic/deg/generator/utils/GeneratorDefaults.java

@@ -24,7 +24,7 @@
 
 import static com.scottlogic.deg.common.util.Defaults.*;
 
-public class Defaults {
+public class GeneratorDefaults {
 
     public static final Limit<BigDecimal> NUMERIC_MAX_LIMIT = new Limit<>(NUMERIC_MAX, true);
     public static final Limit<BigDecimal> NUMERIC_MIN_LIMIT= new Limit<>(NUMERIC_MIN, true);

generator/src/test/java/com/scottlogic/deg/generator/fieldspecs/FieldSpecTests.java

@@ -33,7 +33,7 @@
 
 import static com.scottlogic.deg.common.profile.FieldType.*;
 import static com.scottlogic.deg.generator.restrictions.string.StringRestrictionsFactory.forMaxLength;
-import static com.scottlogic.deg.generator.utils.Defaults.*;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.*;
 import static org.hamcrest.core.IsEqual.equalTo;
 import static org.hamcrest.core.IsNot.not;
 import static org.junit.Assert.assertFalse;

generator/src/test/java/com/scottlogic/deg/generator/fieldspecs/RestrictionsMergeOperationTest.java

@@ -33,9 +33,9 @@
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createDateTimeRestrictions;
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MIN_LIMIT;
 import static org.mockito.Mockito.*;
 
 class RestrictionsMergeOperationTest {

generator/src/test/java/com/scottlogic/deg/generator/generation/FieldSpecGetFieldValueSourceTests.java

@@ -37,8 +37,8 @@
 
 import static com.scottlogic.deg.common.profile.FieldType.*;
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MIN_LIMIT;
 
 public class FieldSpecGetFieldValueSourceTests {
 

generator/src/test/java/com/scottlogic/deg/generator/generation/fieldvaluesources/RealNumberFieldValueSourceTests.java

@@ -35,8 +35,8 @@
 import java.util.stream.Stream;
 
 import static com.scottlogic.deg.generator.restrictions.linear.LinearRestrictionsFactory.createNumericRestrictions;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.NUMERIC_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.NUMERIC_MIN_LIMIT;
 import static org.hamcrest.CoreMatchers.equalTo;
 import static org.hamcrest.Matchers.*;
 

generator/src/test/java/com/scottlogic/deg/generator/generation/fieldvaluesources/datetime/DateTimeFieldValueSourceTests.java

@@ -34,8 +34,8 @@
 
 import static com.scottlogic.deg.common.util.Defaults.ISO_MAX_DATE;
 import static com.scottlogic.deg.common.util.Defaults.ISO_MIN_DATE;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MAX_LIMIT;
-import static com.scottlogic.deg.generator.utils.Defaults.DATETIME_MIN_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MAX_LIMIT;
+import static com.scottlogic.deg.generator.utils.GeneratorDefaults.DATETIME_MIN_LIMIT;
 import static org.hamcrest.Matchers.equalTo;
 import static org.hamcrest.core.Is.is;
 import static org.hamcrest.core.IsNot.not;