Add a ignore tag step based on paths #688

Signed-off-by: Thomas Druez <tdruez@nexb.com>

scanpipe/pipelines/develop_to_deploy.py

@@ -88,6 +88,7 @@ def tag_empty_and_ignored_files(self):
         tag.tag_empty_codebase_resources(self.project)
         tag.tag_ignored_filenames(self.project, filenames=d2d.IGNORE_FILENAMES)
         tag.tag_ignored_extensions(self.project, extensions=d2d.IGNORE_EXTENSIONS)
+        tag.tag_ignored_paths(self.project, paths=d2d.IGNORE_PATHS)
 
     def checksum_match(self):
         """Match using SHA1 checksum."""

scanpipe/pipes/d2d.py

@@ -32,8 +32,8 @@
 TO = "to/"
 
 IGNORE_FILENAMES = ("packageinfo",)
-
 IGNORE_EXTENSIONS = ()
+IGNORE_PATHS = ("gradleTest/",)
 
 
 def get_inputs(project):

scanpipe/pipes/tag.py

@@ -20,6 +20,8 @@
 # ScanCode.io is a free software code scanning tool from nexB Inc. and others.
 # Visit https://github.com/nexB/scancode.io for support and download.
 
+from django.db.models import Q
+
 
 def tag_empty_codebase_resources(project):
     """Tags empty files as ignored."""
@@ -41,3 +43,13 @@ def tag_ignored_extensions(project, extensions):
     """Tag codebase resource as `ignored` status from list of `extensions`."""
     qs = project.codebaseresources.no_status().filter(extension__in=extensions)
     return qs.update(status="ignored-extension")
+
+
+def tag_ignored_paths(project, paths):
+    """Tag codebase resource as `ignored` status from list of `paths`."""
+    lookups = Q()
+    for path in paths:
+        lookups |= Q(path__contains=path)
+
+    qs = project.codebaseresources.no_status().filter(lookups)
+    return qs.update(status="ignored-path")

scanpipe/tests/pipes/test_tag.py

@@ -61,3 +61,10 @@ def test_scanpipe_pipes_tag_tag_ignored_extensions(self):
         self.resource2.refresh_from_db()
         self.assertEqual("", self.resource1.status)
         self.assertEqual("ignored-extension", self.resource2.status)
+
+    def test_scanpipe_pipes_tag_tag_ignored_paths(self):
+        tag.tag_ignored_paths(self.project1, paths=["dir/"])
+        self.resource1.refresh_from_db()
+        self.resource2.refresh_from_db()
+        self.assertEqual("ignored-path", self.resource1.status)
+        self.assertEqual("ignored-path", self.resource2.status)