rollback ML model

Samsung · Aug 7, 2024 · c9cfeb9 · c9cfeb9
1 parent 29ae9a5
commit c9cfeb9
Show file tree

Hide file tree

Showing 12 changed files with 900 additions and 1,419 deletions.
diff --git a/.github/workflows/check.yml b/.github/workflows/check.yml
@@ -58,7 +58,7 @@ jobs:
     - name: Check ml_model.onnx integrity
       if: ${{ always() && steps.code_checkout.conclusion == 'success' }}
       run: |
-        md5sum --binary credsweeper/ml_model/ml_model.onnx | grep 70a864232576f9b88a08296a5e628208
+        md5sum --binary credsweeper/ml_model/ml_model.onnx | grep 62d92ab2f91a18e861d846a7b8a0c3a7
 
     # # # Python setup
 

diff --git a/credsweeper/ml_model/ml_model.onnx b/credsweeper/ml_model/ml_model.onnx
diff --git a/credsweeper/ml_model/model_config.json b/credsweeper/ml_model/model_config.json
@@ -361,7 +361,6 @@
                     ".adoc",
                     ".asciidoc",
                     ".axaml",
-                    ".backup",
                     ".bash",
                     ".bat",
                     ".bats",
@@ -374,7 +373,6 @@
                     ".cjs",
                     ".cljc",
                     ".cmd",
-                    ".cmm",
                     ".cnf",
                     ".coffee",
                     ".conf",
@@ -396,7 +394,6 @@
                     ".example",
                     ".exs",
                     ".ext",
-                    ".fsproj",
                     ".g4",
                     ".gml",
                     ".gni",
@@ -431,7 +428,7 @@
                     ".log",
                     ".lua",
                     ".m",
-                    ".manifest",
+                    ".map",
                     ".markerb",
                     ".md",
                     ".mdx",
@@ -440,7 +437,6 @@
                     ".ml",
                     ".mlir",
                     ".moo",
-                    ".msg",
                     ".ndjson",
                     ".nix",
                     ".nolint",
@@ -478,7 +474,6 @@
                     ".scala",
                     ".secrets",
                     ".sh",
-                    ".sln",
                     ".snap",
                     ".sql",
                     ".storyboard",
@@ -489,7 +484,6 @@
                     ".tdf",
                     ".template",
                     ".test",
-                    ".testsettings",
                     ".tf",
                     ".tfstate",
                     ".tfvars",
@@ -500,9 +494,7 @@
                     ".travis",
                     ".ts",
                     ".tsx",
-                    ".ttar",
                     ".txt",
-                    ".vsmdi",
                     ".vue",
                     ".xaml",
                     ".xib",
@@ -530,8 +522,7 @@
                     "Salt",
                     "Secret",
                     "Token",
-                    "URL Credentials",
-                    "UUID"
+                    "URL Credentials"
                 ]
             }
         }

diff --git a/tests/__init__.py b/tests/__init__.py
@@ -7,18 +7,18 @@
 NEGLIGIBLE_ML_THRESHOLD = 0.0001
 
 # credentials count after scan
-SAMPLES_CRED_COUNT: int = 360
-SAMPLES_CRED_LINE_COUNT: int = 377
+SAMPLES_CRED_COUNT: int = 363
+SAMPLES_CRED_LINE_COUNT: int = 380
 
 # credentials count after post-processing
-SAMPLES_POST_CRED_COUNT: int = 332
+SAMPLES_POST_CRED_COUNT: int = 321
 
 # with option --doc
-SAMPLES_IN_DOC = 416
+SAMPLES_IN_DOC = 415
 
 # archived credentials that are not found without --depth
 SAMPLES_IN_DEEP_1 = SAMPLES_POST_CRED_COUNT + 24
-SAMPLES_IN_DEEP_2 = SAMPLES_IN_DEEP_1 + 16
+SAMPLES_IN_DEEP_2 = SAMPLES_IN_DEEP_1 + 17
 SAMPLES_IN_DEEP_3 = SAMPLES_IN_DEEP_2 + 1
 
 # well known string with all latin letters

diff --git a/tests/data/__init__.py b/tests/data/__init__.py
@@ -11,14 +11,14 @@
     "sort_output": True,
     "json_filename": "ml_threshold.json",
     "ml_threshold": NEGLIGIBLE_ML_THRESHOLD
-}, {
-    "__cred_count": SAMPLES_IN_DEEP_3,
-    "sort_output": True,
-    "json_filename": "depth_3.json",
-    "depth": 3
 }, {
     "__cred_count": SAMPLES_IN_DOC,
     "sort_output": True,
     "json_filename": "doc.json",
     "doc": True
+}, {
+    "__cred_count": SAMPLES_IN_DEEP_3,
+    "sort_output": True,
+    "json_filename": "depth_3.json",
+    "depth": 3
 }]