update tests

.gitignore

@@ -159,3 +159,4 @@ output
 **/*parquet
 **/checkpoints
 *code-workspace
+.ruff_cache

cobertura.xml

@@ -1,5 +1,5 @@
 <?xml version="1.0" ?>
-<coverage version="7.4.3" timestamp="1708808380781" lines-valid="602" lines-covered="450" line-rate="0.7475" branches-covered="0" branches-valid="0" branch-rate="0" complexity="0">
+<coverage version="7.4.3" timestamp="1708858280150" lines-valid="612" lines-covered="460" line-rate="0.7516" branches-covered="0" branches-valid="0" branch-rate="0" complexity="0">
 	<!-- Generated by coverage.py: https://coverage.readthedocs.io/en/7.4.3 -->
 	<!-- Based on https://raw.githubusercontent.com/cobertura/web/master/htdocs/xml/coverage-04.dtd -->
 	<sources>
@@ -12,45 +12,55 @@
 					<methods/>
 					<lines>
 						<line number="1" hits="1"/>
-						<line number="3" hits="1"/>
+						<line number="4" hits="1"/>
 						<line number="5" hits="1"/>
-						<line number="26" hits="1"/>
+						<line number="30" hits="1"/>
+						<line number="35" hits="1"/>
+						<line number="36" hits="1"/>
 					</lines>
 				</class>
 				<class name="test_exact_hash.py" filename="tests/test_exact_hash.py" complexity="0" line-rate="1" branch-rate="0">
 					<methods/>
 					<lines>
 						<line number="1" hits="1"/>
-						<line number="3" hits="1"/>
+						<line number="4" hits="1"/>
 						<line number="5" hits="1"/>
-						<line number="26" hits="1"/>
+						<line number="30" hits="1"/>
+						<line number="35" hits="1"/>
+						<line number="36" hits="1"/>
 					</lines>
 				</class>
 				<class name="test_minhash.py" filename="tests/test_minhash.py" complexity="0" line-rate="1" branch-rate="0">
 					<methods/>
 					<lines>
 						<line number="1" hits="1"/>
-						<line number="3" hits="1"/>
+						<line number="4" hits="1"/>
 						<line number="5" hits="1"/>
-						<line number="26" hits="1"/>
+						<line number="30" hits="1"/>
+						<line number="35" hits="1"/>
+						<line number="36" hits="1"/>
 					</lines>
 				</class>
 				<class name="test_simhash.py" filename="tests/test_simhash.py" complexity="0" line-rate="1" branch-rate="0">
 					<methods/>
 					<lines>
 						<line number="1" hits="1"/>
-						<line number="3" hits="1"/>
+						<line number="4" hits="1"/>
 						<line number="5" hits="1"/>
-						<line number="26" hits="1"/>
+						<line number="30" hits="1"/>
+						<line number="35" hits="1"/>
+						<line number="36" hits="1"/>
 					</lines>
 				</class>
 				<class name="test_suffix_array.py" filename="tests/test_suffix_array.py" complexity="0" line-rate="1" branch-rate="0">
 					<methods/>
 					<lines>
 						<line number="1" hits="1"/>
-						<line number="3" hits="1"/>
+						<line number="4" hits="1"/>
 						<line number="5" hits="1"/>
-						<line number="26" hits="1"/>
+						<line number="30" hits="1"/>
+						<line number="36" hits="1"/>
+						<line number="37" hits="1"/>
 					</lines>
 				</class>
 			</classes>
@@ -184,12 +194,12 @@
 						<line number="41" hits="1"/>
 						<line number="101" hits="1"/>
 						<line number="104" hits="1"/>
-						<line number="108" hits="1"/>
-						<line number="112" hits="1"/>
+						<line number="109" hits="1"/>
 						<line number="115" hits="1"/>
-						<line number="116" hits="1"/>
-						<line number="120" hits="1"/>
-						<line number="121" hits="1"/>
+						<line number="118" hits="1"/>
+						<line number="119" hits="1"/>
+						<line number="123" hits="1"/>
+						<line number="126" hits="1"/>
 					</lines>
 				</class>
 				<class name="simhash.py" filename="text_dedup/simhash.py" complexity="0" line-rate="0.6347" branch-rate="0">

docs/source/conf.py

@@ -11,7 +11,7 @@
 sys.path.insert(0, os.path.abspath("../.."))
 
 project = "text-dedup"
-copyright = "2023, Chenghao Mou"
+copyright = "Apache 2.0"
 author = "Chenghao Mou"
 
 # -- General configuration ---------------------------------------------------

tests/test_bloom_filter.py

@@ -30,3 +30,7 @@ def test_bloom_filter():
     assert (
         "88803" in result.stdout and "47049" in result.stdout
     ), f"Expected before and after are not present in the output: {result.stdout}"
+
+    # remove the output and input
+    subprocess.run(["rm", "-rf", ".cache"])
+    subprocess.run(["rm", "-rf", ".temp-output"])

tests/test_exact_hash.py

@@ -30,3 +30,7 @@ def test_exact_hash():
     assert (
         "88803" in result.stdout and "47049" in result.stdout
     ), f"Expected before and after are not present in the output: {result.stdout}"
+
+    # remove the output and input
+    subprocess.run(["rm", "-rf", ".cache"])
+    subprocess.run(["rm", "-rf", ".temp-output"])

tests/test_minhash.py

@@ -30,3 +30,7 @@ def test_minhash():
     assert (
         "88803" in result.stdout and "44092" in result.stdout
     ), f"Expected before and after are not present in the output: {result.stdout}"
+
+    # remove the output and input
+    subprocess.run(["rm", "-rf", ".cache"])
+    subprocess.run(["rm", "-rf", ".temp-output"])

tests/test_simhash.py

@@ -30,3 +30,7 @@ def test_simhash():
     assert (
         "88803" in result.stdout and "45589" in result.stdout
     ), f"Expected before and after are not present in the output: {result.stdout}"
+
+    # remove the output and input
+    subprocess.run(["rm", "-rf", ".cache"])
+    subprocess.run(["rm", "-rf", ".temp-output"])

tests/test_suffix_array.py

@@ -31,3 +31,7 @@ def test_suffix_array():
         "180332342 bytes (88803)" in result.stdout
         and "51305898 bytes (29254)" in result.stdout
     ), f"Expected before and after are not present in the output: {result.stdout}"
+
+    # remove the output and input
+    subprocess.run(["rm", "-rf", ".cache"])
+    subprocess.run(["rm", "-rf", ".temp-output"])