fix: try to avoid recording tests

When creating the default config, ignore directories with names that
match the regex .*test.*. This should avoid instrumenting the majority
of test functions. For those that do still get instrumented, have the
test framework integration disable the wrapt wrapped on it, thereby
disabling recording.

_appmap/configuration.py

@@ -7,12 +7,14 @@
 import inspect
 import json
 import os
+import re
 import sys
 from os.path import realpath
 from pathlib import Path
 from textwrap import dedent
 
 import yaml
+from yaml import SafeLoader
 from yaml.parser import ParserError
 
 from _appmap.labels import LabelSet
@@ -50,20 +52,20 @@ def _get_sys_prefix():
     return realpath(sys.prefix)
 
 
+_EXCLUDE_PATTERN = re.compile(r"\..*|node_modules|.*test.*|site-packages")
+
+
 def find_top_packages(rootdir):
     """
-    Scan a directory tree for packages that should appear in the
-    default config file.
+    Scan a directory tree for packages that should appear in the default config file.
 
-    Examine directories in rootdir, to see if they contains an
-    __init__.py.  If it does, add it to the list of packages and don't
-    scan any of its subdirectories.  If it doesn't, scan its
+    Examine each directory in rootdir, to see if it contains an __init__.py.  If it does, add it to
+    the list of packages and don't scan any of its subdirectories.  If it doesn't, scan its
     subdirectories to find __init__.py.
 
-    Some directories are automatically excluded from the search:
-      * sys.prefix
-      * Hidden directories (i.e. those that start with a '.')
-      * node_modules
+    Directory traversal will stop at directories that match _EXCLUDE_PATTERN. Such a directory (and
+    its subdirectories) will not be added to the returned packages, but will in set of excluded
+    directories).
 
     For example, in a directory like this
 
@@ -104,11 +106,11 @@ def find_top_packages(rootdir):
     # build process copies its source to a subdirectory.
     packages = set()
 
-    def excluded(d):
-        excluded = d == "node_modules" or d[0] == "."
-        if excluded:
+    def exclude(d):
+        is_excluded = _EXCLUDE_PATTERN.search(d) is not None
+        if is_excluded:
             logger.trace("excluding dir %s", d)
-        return excluded
+        return is_excluded
 
     sys_prefix = _get_sys_prefix()
 
@@ -123,13 +125,26 @@ def excluded(d):
             packages.add(Path(d).name)
             dirs.clear()
         else:
-            dirs[:] = [d for d in dirs if not excluded(d)]
+            dirs[:] = [d for d in dirs if not exclude(d)]
 
     return packages
 
 class AppMapInvalidConfigException(Exception):
     pass
 
+# We don't have any control over the PyYAML class hierarchy, so we can't control how many ancestors
+# SafeLoader has....
+class _ConfigLoader(SafeLoader):  # pylint: disable=too-many-ancestors
+    def construct_mapping(self, node, deep=False):
+        mapping = super().construct_mapping(node, deep=deep)
+        # Allow record_test_cases to be set using a string (in addition to allowing a boolean).
+        if "record_test_cases" in mapping:
+            val = mapping["record_test_cases"]
+            if isinstance(val, str):
+                mapping["record_test_cases"] = val.lower() == "true"
+        return mapping
+
+
 class Config(metaclass=SingletonMeta):
     """Singleton Config class"""
 
@@ -156,11 +171,16 @@ def name(self):
     def packages(self):
         return self._config["packages"]
 
+    @property
+    def record_test_cases(self):
+        return self._config.get("record_test_cases", False)
+
     @property
     def default(self):
         ret = {
             "name": self.default_name,
             "language": "python",
+            "record_test_cases": False,
             "packages": self.default_packages,
         }
         env = Env.current
@@ -233,7 +253,7 @@ def _load_config(self, show_warnings=False):
             Env.current.enabled = False
             self.file_valid = False
             try:
-                self._config = yaml.safe_load(path.read_text(encoding="utf-8"))
+                self._config = yaml.load(path.read_text(encoding="utf-8"), Loader=_ConfigLoader)
                 if not self._config:
                     # It parsed, but was (effectively) empty.
                     self._config = self.default
@@ -334,7 +354,6 @@ def _check_path_value(self, value):
         except SyntaxError:
             return False
 
-
 def startswith(prefix, sequence):
     """
     Check if a sequence starts with the prefix.
@@ -377,7 +396,8 @@ class DistMatcher(PathMatcher):
     def __init__(self, dist, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self.dist = dist
-        self.files = [str(pp.locate()) for pp in importlib.metadata.files(dist)]
+        dist_files = importlib.metadata.files(dist)
+        self.files = [str(pp.locate()) for pp in dist_files] if dist_files is not None else []
 
     def matches(self, filterable):
         try:

_appmap/test/data/pytest/appmap-no-test-cases.yml

@@ -0,0 +1,4 @@
+name: Simple
+record_test_cases: false
+packages:
+- path: simple

_appmap/test/data/pytest/appmap.yml

@@ -1,3 +1,5 @@
 name: Simple
+record_test_cases: true
 packages:
 - path: simple
+- path: tests

_appmap/test/data/pytest/expected/pytest-numpy1-no-test-cases.appmap.json

@@ -0,0 +1,242 @@
+{
+  "version": "1.9",
+  "metadata": {
+    "language": {
+      "name": "python"
+    },
+    "client": {
+      "name": "appmap",
+      "url": "https://github.com/applandinc/appmap-python"
+    },
+    "app": "Simple",
+    "recorder": {
+      "name": "pytest",
+      "type": "tests"
+    },
+    "source_location": "tests/test_simple.py:5",
+    "name": "hello world",
+    "feature": "Hello world",
+    "test_status": "succeeded"
+  },
+  "events": [
+    {
+      "defined_class": "simple.Simple",
+      "method_id": "hello_world",
+      "path": "simple.py",
+      "lineno": 8,
+      "static": false,
+      "receiver": {
+        "class": "simple.Simple",
+        "kind": "req",
+        "name": "self",
+        "value": "<simple.Simple object at 0xabcdef>"
+      },
+      "parameters": [],
+      "id": 1,
+      "event": "call",
+      "thread_id": 1
+    },
+    {
+      "defined_class": "simple.Simple",
+      "method_id": "hello",
+      "path": "simple.py",
+      "lineno": 2,
+      "static": false,
+      "receiver": {
+        "class": "simple.Simple",
+        "kind": "req",
+        "name": "self",
+        "value": "<simple.Simple object at 0xabcdef>"
+      },
+      "parameters": [],
+      "id": 2,
+      "event": "call",
+      "thread_id": 1
+    },
+    {
+      "return_value": {
+        "class": "builtins.str",
+        "value": "'Hello'"
+      },
+      "parent_id": 2,
+      "id": 3,
+      "event": "return",
+      "thread_id": 1
+    },
+    {
+      "defined_class": "simple.Simple",
+      "method_id": "world",
+      "path": "simple.py",
+      "lineno": 5,
+      "static": false,
+      "receiver": {
+        "class": "simple.Simple",
+        "kind": "req",
+        "name": "self",
+        "value": "<simple.Simple object at 0xabcdef>"
+      },
+      "parameters": [],
+      "id": 4,
+      "event": "call",
+      "thread_id": 1
+    },
+    {
+      "return_value": {
+        "class": "builtins.str",
+        "value": "'world!'"
+      },
+      "parent_id": 4,
+      "id": 5,
+      "event": "return",
+      "thread_id": 1
+    },
+    {
+      "return_value": {
+        "class": "builtins.str",
+        "value": "'Hello world!'"
+      },
+      "parent_id": 1,
+      "id": 6,
+      "event": "return",
+      "thread_id": 1
+    },
+    {
+      "static": false,
+      "receiver": {
+        "kind": "req",
+        "value": "<simple.Simple object at 0xabcdef>",
+        "name": "self",
+        "class": "simple.Simple"
+      },
+      "parameters": [],
+      "id": 7,
+      "event": "call",
+      "thread_id": 1,
+      "defined_class": "simple.Simple",
+      "method_id": "show_numpy_dict",
+      "path": "simple.py",
+      "lineno": 11
+    },
+    {
+      "static": false,
+      "receiver": {
+        "kind": "req",
+        "value": "<simple.Simple object at 0xabcdef>",
+        "name": "self",
+        "class": "simple.Simple"
+      },
+      "parameters": [
+        {
+          "kind": "req",
+          "value": "{0: 'zero', 1: 'one'}",
+          "name": "d",
+          "class": "builtins.dict",
+          "properties": [
+            {
+              "name": "0",
+              "class": "builtins.str"
+            },
+            {
+              "name": "1",
+              "class": "builtins.str"
+            }
+          ],
+          "size": 2
+        }
+      ],
+      "id": 8,
+      "event": "call",
+      "thread_id": 1,
+      "defined_class": "simple.Simple",
+      "method_id": "get_numpy_dict",
+      "path": "simple.py",
+      "lineno": 18
+    },
+    {
+      "return_value": {
+        "value": "{0: 'zero', 1: 'one'}",
+        "class": "builtins.dict",
+        "properties": [
+          {
+            "name": "0",
+            "class": "builtins.str"
+          },
+          {
+            "name": "1",
+            "class": "builtins.str"
+          }
+        ],
+        "size": 2
+      },
+      "parent_id": 8,
+      "id": 9,
+      "event": "return",
+      "thread_id": 1
+    },
+    {
+      "return_value": {
+        "value": "{0: 'zero', 1: 'one'}",
+        "class": "builtins.dict",
+        "properties": [
+          {
+            "name": "0",
+            "class": "builtins.str"
+          },
+          {
+            "name": "1",
+            "class": "builtins.str"
+          }
+        ],
+        "size": 2
+      },
+      "parent_id": 7,
+      "id": 10,
+      "event": "return",
+      "thread_id": 1
+    }
+  ],
+  "classMap": [
+    {
+      "name": "simple",
+      "type": "package",
+      "children": [
+        {
+          "name": "Simple",
+          "type": "class",
+          "children": [
+            {
+              "name": "get_numpy_dict",
+              "type": "function",
+              "location": "simple.py:18",
+              "static": false
+            },
+            {
+              "name": "hello",
+              "type": "function",
+              "location": "simple.py:2",
+              "static": false
+            },
+            {
+              "name": "hello_world",
+              "type": "function",
+              "location": "simple.py:8",
+              "static": false
+            },
+            {
+              "name": "show_numpy_dict",
+              "type": "function",
+              "location": "simple.py:11",
+              "static": false
+            },
+            {
+              "name": "world",
+              "type": "function",
+              "location": "simple.py:5",
+              "static": false
+            }
+          ]
+        }
+      ]
+    }
+  ]
+}