vector: implement git_vector_uniq()

The routine remove duplictes from the vector. Only the last added element
of elements with equal keys remains in the vector.

Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>

3 files changed, 44 insertions, 1 deletions

diff --git a/src/vector.c b/src/vector.c
index 8d09350fc..2fc051f0c 100644
--- a/src/vector.c
+++ b/src/vector.c
@@ -162,6 +162,26 @@ int git_vector_remove(git_vector *v, unsigned int idx)
 	return GIT_SUCCESS;
 }
 
+void git_vector_uniq(git_vector *v)
+{
+	git_vector_cmp cmp;
+	unsigned int i, j;
+
+	if (v->length <= 1)
+		return;
+
+	git_vector_sort(v);
+	cmp = v->_cmp ? v->_cmp : strict_comparison;
+
+	for (i = 0, j = 1 ; j < v->length; ++j)
+		if (!cmp(v->contents + i, v->contents + j))
+			v->contents[i] = v->contents[j];
+		else
+			v->contents[++i] = v->contents[j];
+
+	v->length -= j - i - 1;
+}
+
 void git_vector_clear(git_vector *v)
 {
 	assert(v);
diff --git a/src/vector.h b/src/vector.h
index 256452ee5..76778ba4e 100644
--- a/src/vector.h
+++ b/src/vector.h
@@ -32,5 +32,5 @@ GIT_INLINE(void *) git_vector_get(git_vector *v, unsigned int position)
 
 int git_vector_insert(git_vector *v, void *element);
 int git_vector_remove(git_vector *v, unsigned int idx);
-
+void git_vector_uniq(git_vector *v);
 #endif
diff --git a/tests/t00-core.c b/tests/t00-core.c
index ba5188a43..661f92508 100644
--- a/tests/t00-core.c
+++ b/tests/t00-core.c
@@ -73,6 +73,28 @@ BEGIN_TEST(vector1, "don't read past array bounds on remove()")
   git_vector_free(&x);
 END_TEST
 
+static int test_cmp(const void *a, const void *b)
+{
+	int n1 = *(int *)a;
+	int n2 = *(int *)b;
+
+	return n1 - n2;
+}
+
+BEGIN_TEST(vector2, "remove duplicates")
+	git_vector x;
+	must_pass(git_vector_init(&x, 5, test_cmp));
+	must_pass(git_vector_insert(&x, (void *) 0xdeadbeef));
+	must_pass(git_vector_insert(&x, (void *) 0xcafebabe));
+	must_pass(git_vector_insert(&x, (void *) 0xcafebabe));
+	must_pass(git_vector_insert(&x, (void *) 0xdeadbeef));
+	must_pass(git_vector_insert(&x, (void *) 0xcafebabe));
+	must_be_true(x.length == 5);
+	git_vector_uniq(&x);
+	must_be_true(x.length == 2);
+	git_vector_free(&x);
+END_TEST
+
 
 BEGIN_TEST(path0, "get the dirname of a path")
 	char dir[64], *dir2;
@@ -480,6 +502,7 @@ BEGIN_SUITE(core)
 
 	ADD_TEST(vector0);
 	ADD_TEST(vector1);
+	ADD_TEST(vector2);
 
 	ADD_TEST(path0);
 	ADD_TEST(path1);