public void TestTriangleInequality()
{
var top = Top1000.Take(50).ToArray();
for (var i = 0; i < top.Length; i++)
{
for (var j = 0; j < top.Length; j++)
{
if (j == i)
{
continue;
}
for (var k = 0; k < top.Length; k++)
{
if (k == i || k == j)
{
continue;
}
var string1 = top[i];
var string2 = top[j];
var string3 = top[k];
var editDistance12 = EditDistance.GetEditDistance(string1, string2);
var editDistance13 = EditDistance.GetEditDistance(string1, string3);
var editDistance23 = EditDistance.GetEditDistance(string2, string3);
Assert.True(editDistance13 <= editDistance12 + editDistance23);
}
}
}
}
public void Top1000Test()
{
for (var i = 0; i < Top1000.Length; i++)
{
var source = Top1000[i];
for (var j = 0; j < Top1000.Length; j++)
{
var target = Top1000[j];
var editDistance1 = EditDistance.GetEditDistance(source, target);
if (i == j)
{
Assert.Equal(0, editDistance1);
}
if (editDistance1 == 0)
{
Assert.Equal(i, j);
}
Assert.True(editDistance1 >= 0);
var editDistance2 = EditDistance.GetEditDistance(source, target, editDistance1);
Assert.Equal(editDistance1, editDistance2);
}
}
}
public void EditDistance11()
{
var editDistance = EditDistance.GetEditDistance("book", "moons", 1);
Assert.Equal(editDistance, EditDistance.BeyondThreshold);
VerifyEditDistance("book", "moons", 3);
}
public void EditDistance3()
{
Assert.Equal(EditDistance.GetEditDistance("", "aaa"), 3);
Assert.Equal(EditDistance.GetEditDistance("aaa", ""), 3);
Assert.Equal(EditDistance.GetEditDistance("aaa", "bbb"), 3);
Assert.Equal(EditDistance.GetEditDistance("aaab", "a"), 3);
Assert.Equal(EditDistance.GetEditDistance("a", "aaab"), 3);
Assert.Equal(EditDistance.GetEditDistance("aababbab", "abababaa"), 3);
}
public void EditDistance1()
{
Assert.Equal(EditDistance.GetEditDistance("", "a"), 1);
Assert.Equal(EditDistance.GetEditDistance("a", ""), 1);
Assert.Equal(EditDistance.GetEditDistance("a", "b"), 1);
Assert.Equal(EditDistance.GetEditDistance("ab", "a"), 1);
Assert.Equal(EditDistance.GetEditDistance("a", "ab"), 1);
Assert.Equal(EditDistance.GetEditDistance("aabb", "abab"), 1);
}
void DoTest(string s1, string s2, string expected, Func <char, char, int> cost)
{
var(dist, edits) = EditDistance.GetEditDistance(
s1.ToList(),
s2.ToList(),
cost
);
string actual = $"{dist} {string.Join(",", edits.Select(e => e))}";
Assert.AreEqual(expected, actual);
}
public void GetEditDistance_Returns0_GivenIdenticalStrings()
{
//Arrange
string first = "Some string!";
string second = "Some string!";
int expected = 0;
//Act
int actual = EditDistance.GetEditDistance(first, second);
//Assert
Assert.AreEqual(expected, actual);
}
public void GetEditDistance_Returns5_GivenLongerStringTop()
{
//Arrange
string first = "Some string!12345";
string second = "Some string!";
int expected = 5;
//Act
int actual = EditDistance.GetEditDistance(first, second);
//Assert
Assert.AreEqual(expected, actual);
}
// private
private static void Test(string inStr)
{
// candidates with ED: 1
HashSet <string> candSet1 = GetCandidatesByEd1(inStr);
// print out
Console.WriteLine("-- inStr: [" + inStr + "]");
Console.WriteLine("-- candSet1.size(): " + candSet1.Count);
bool caseFlag = false; // not case sensitive
int dupNo = 0;
// check if the candList correct
foreach (string cand1 in candSet1)
{
int ed = EditDistance.GetEditDistance(inStr, cand1, caseFlag);
// these are errors, should not have any
if ((ed != 1) && (inStr.Equals(cand1) == false))
{
Console.WriteLine(inStr + "|" + cand1 + "|" + ed);
}
// candidate are same as inStr, produced by replace, not 0
if (inStr.ToLower().Equals(cand1) == true)
{
dupNo++;
}
}
Console.WriteLine("-- dupNo: " + dupNo);
//System.out.println(candSet1);
// candidates with ED: 2
/// <summary>
/// HashSet<String> candSet2 = GetCandidatesByEd2(candList1);
/// int exceedNo = 0;
/// for(String cand2:candSet2)
/// {
/// int ed = EditDistance.GetEditDistance(inStr, cand2, caseFlag);
/// if((ed > 2) && (inStr.equals(cand2) == false))
/// {
/// System.out.println(inStr + "|" + cand2 + "|" + ed);
/// exceedNo++;
/// }
/// }
/// System.out.println("-- candSet2.size(): " + candSet2.size());
/// System.out.println("-- exceedNo: " + exceedNo);
///
/// </summary>
}
// test all split
private static void TestSplitUtil(string inStr)
{
Console.WriteLine("====== test split util (no Dic check) ======");
Console.WriteLine("----- inStr: [" + inStr + "]");
// Test possible split combination with 1 split
HashSet <string> splitSet1 = GetSplitSetBy1Space(inStr);
bool caseFlag = false; // not case sensitive
int dupNo = 0;
// print out
Console.WriteLine("----- Check on candList1 -----");
Console.WriteLine("-- splitSet1.size(): " + splitSet1.Count);
Console.WriteLine(splitSet1.ToStringList());
// check if the candList1 correct by edit distance
foreach (string split1 in splitSet1)
{
int ed = EditDistance.GetEditDistance(inStr, split1, caseFlag);
// these are errors: because 1 split should have ed = 1
if ((ed != 1) && (inStr.Equals(split1) == false))
{
Console.WriteLine("**ERR: " + inStr + "|" + split1 + "|" + ed);
}
// candidate are same as inStr, produced by replace, not 0
// check duplicate no.
else if (inStr.Equals(split1) == true)
{
dupNo++;
}
}
Console.WriteLine("-- dupNo: " + dupNo);
HashSet <string> candSet1a = GetSplitSetBySpaces(inStr, 1);
Console.WriteLine("-- candSet1a.size(): " + candSet1a.Count);
Console.WriteLine(candSet1a.ToStringList());
HashSet <string> candSet2a = GetSplitSetBySpaces(inStr, 2);
Console.WriteLine("-- candList2a.size(): " + candSet2a.Count);
Console.WriteLine(candSet2a.ToStringList());
// other test 5 for spliting candidates by hyphen
string testStr = "-123--45-test-123-45.6-";
Console.WriteLine("- testStr: " + testStr);
Console.WriteLine("- splitByHyphen: [" + GetSplitByPunc(testStr, '-') + "]");
}
public void IsCloseTo_ReturnsFalse_GivenInsimilarStringsThreshold0_5()
{
//Arrange
string first = "Some string!";
string second = "Elephant Mongoose!";
double threshold = 0.50;
bool expected = false;
//Act
bool actual = EditDistance.IsCloseTo(first, second, threshold);
//Assert
if (expected != actual)
{
int distance = EditDistance.GetEditDistance(first, second);
throw new Exception($"Edit distance was {distance}");
}
Assert.AreEqual(expected, actual);
}
private static void VerifyEditDistance(string s, string t, int expectedEditDistance)
{
// We want the full edit distance, without bailing out early because we crossed the
// threshold.
var editDistance1 = EditDistance.GetEditDistance(s, t);
Assert.Equal(expectedEditDistance, editDistance1);
// Edit distances are symmetric.
var editDistance2 = EditDistance.GetEditDistance(t, s);
Assert.Equal(editDistance1, editDistance2);
// If we set hte edit distance as our threshold, we should still find the value.
var editDistance3 = EditDistance.GetEditDistance(s, t, editDistance1);
Assert.Equal(editDistance1, editDistance3);
if (editDistance1 > 0)
{
var editDistance4 = EditDistance.GetEditDistance(s, t, editDistance1 - 1);
Assert.Equal(editDistance4, EditDistance.BeyondThreshold);
}
}
private static void AssertEditDistance(int expected, string one, string two)
{
Assert.AreEqual(expected, EditDistance.GetEditDistance(one, two));
Assert.AreEqual(expected, EditDistance.GetEditDistance(two, one));
}
public void MoreEditDistance()
{
Assert.Equal(EditDistance.GetEditDistance("barking", "corkliness"), 6);
}
public double GetEditDistance(string target) => _underlyingObject.GetEditDistance(target);
public static List <ListEdit> GetListEdits(
IReadOnlyList <IListItem> list1,
IReadOnlyList <IListItem> list2
)
{
var result = new List <ListEdit>();
var edits = EditDistance.GetEditDistance(list1, list2, (c1, c2) =>
{
return
(c1 == null || c2 == null ? 1 : // let deletion/insertion of a node to cost 1
c1.Key == c2.Key ? 0 : // encourage reuse of nodes with same Key by giving it no cost,
1); // reuse nodes with different Keys
}).edits;
int targetIdx = 0;
foreach (var(i, j, _) in edits)
{
if (i.HasValue && j.HasValue)
{
var i1 = list1[i.Value];
var i2 = list2[j.Value];
if (i1 != i2)
{
result.Add(new ListEdit
{
Type = EditType.Reuse,
OldItem = i1,
Item = i2,
Index = targetIdx
});
}
if (i2.IsSelected != i1.IsSelected)
{
result.Add(new ListEdit
{
Type = i2.IsSelected ? EditType.Select : EditType.Deselect,
Item = i2,
Index = targetIdx
});
}
++targetIdx;
}
else if (i.HasValue)
{
result.Add(new ListEdit
{
Type = EditType.Delete,
OldItem = list1[i.Value],
Index = targetIdx
});
}
else
{
var i2 = list2[j.Value];
result.Add(new ListEdit
{
Type = EditType.Insert,
Item = i2,
Index = targetIdx
});
if (i2.IsSelected)
{
result.Add(new ListEdit
{
Type = EditType.Select,
Item = i2,
Index = targetIdx
});
}
++targetIdx;
}
}
return(result);
}
public void EditDistance0()
{
Assert.Equal(EditDistance.GetEditDistance("", ""), 0);
Assert.Equal(EditDistance.GetEditDistance("a", "a"), 0);
}
public static List <TreeEdit> GetTreeEdits(ITreeNode root1, ITreeNode root2, Options options = null)
{
options = options ?? Options.@default;
var result = new List <TreeEdit>();
void AddNewNodeEdits(ITreeNode n, ITreeNode parent, int nodeIndex)
{
result.Add(new TreeEdit
{
Type = EditType.Insert,
Node = parent,
NewChild = n,
ChildIndex = nodeIndex
});
if (n.IsExpanded || options.TemporariltyExpandParentToInitChildren)
{
result.Add(new TreeEdit
{
Type = EditType.Expand,
Node = n
});
}
if (n.IsSelected)
{
result.Add(new TreeEdit
{
Type = EditType.Select,
Node = n
});
}
int childIndex = 0;
foreach (var c in n.Children)
{
AddNewNodeEdits(c, n, childIndex++);
}
if (options.TemporariltyExpandParentToInitChildren && !n.IsExpanded)
{
result.Add(new TreeEdit
{
Type = EditType.Collapse,
Node = n
});
}
}
void GetEdits(ITreeNode n1, ITreeNode n2)
{
if (n1 == n2)
{
return;
}
var edits = EditDistance.GetEditDistance(n1.Children, n2.Children, (c1, c2) =>
{
return
(c1 == null || c2 == null ? 1 : // let deletion/insertion of a node cost 1
c1.Key == c2.Key ? 0 : // encourage reuse of nodes with same Key by giving it no cost,
1); // reuse nodes with different Keys
}).edits;
int targetIdx = 0;
foreach (var(i, j, _) in edits)
{
if (i.HasValue && j.HasValue)
{
var c1 = n1.Children[i.Value];
var c2 = n2.Children[j.Value];
if (c1 != c2)
{
result.Add(new TreeEdit
{
Type = EditType.Reuse,
Node = n2,
OldChild = c1,
NewChild = c2,
ChildIndex = targetIdx
});
}
++targetIdx;
GetEdits(c1, c2);
if (c2.IsExpanded != c1.IsExpanded)
{
result.Add(new TreeEdit
{
Type = c2.IsExpanded ? EditType.Expand : EditType.Collapse,
Node = c2
});
}
if (c2.IsSelected != c1.IsSelected)
{
result.Add(new TreeEdit
{
Type = c2.IsSelected ? EditType.Select : EditType.Deselect,
Node = c2
});
}
}
else if (i.HasValue)
{
result.Add(new TreeEdit
{
Type = EditType.Delete,
Node = n2,
OldChild = n1.Children[i.Value],
ChildIndex = targetIdx
});
}
else
{
AddNewNodeEdits(n2.Children[j.Value], n2, targetIdx);
++targetIdx;
}
}
}
GetEdits(root1, root2);
return(result);
}
