chore: resolve doctest issues

pyproject.toml

@@ -234,6 +234,7 @@ minversion = "7.0"
 addopts = "--verbose --doctest-modules -ra --cov macaron"  # Consider adding --pdb
 doctest_optionflags = "IGNORE_EXCEPTION_DETAIL"
 testpaths = [
+    "src/macaron",  # this is included to run doctest on examples
     "tests",
 ]
 env = [

src/macaron/config/defaults.py

@@ -65,16 +65,18 @@ def get_list(
         --------
         Given the following ``defaults.ini``
 
-        .. code-block::
+        .. code-block:: ini
 
             [git]
             allowed_hosts =
                 github.com
                 boo.com gitlab.com
                 host com
 
-        >>> config_parser.get_list("git", "allowed_hosts")
-        ['github.com', 'boo.com gitlab.com', 'host com']
+        .. code-block:: python3
+
+            allowed_hosts = config_parser.get_list("git", "allowed_hosts")
+            # allowed_hosts == ['github.com', 'boo.com gitlab.com', 'host com']
         """
         try:
             value = self.get(section, item)

src/macaron/output_reporter/jinja2_extensions.py

@@ -118,14 +118,13 @@ def j2_filter_get_flatten_dict(data: Any, has_key: bool = False) -> dict | Any:
     Examples
     --------
     >>> j2_filter_get_flatten_dict(
-        {
-            "A": [1,2,3],
-            "B": {
-                "C": ["blah", "bar", "foo"]
-            }
-        }
-    )
-    {'A': {'0': 1, '1': 2, '2': 3}, 'B': {'C': {'0': 'blah', '1': 'bar', '2': 'foo'}}}
+    ... {
+    ...     "A": [1, 2, 3],
+    ...     "B": {
+    ...         "C": ["blah", "bar", "foo"],
+    ...     },
+    ... })
+    {'A': {0: 1, 1: 2, 2: 3}, 'B': {'C': {0: 'blah', 1: 'bar', 2: 'foo'}}}
     """
     if isinstance(data, (str, int, bool, float)):
         if has_key:

src/macaron/output_reporter/results.py

@@ -98,8 +98,23 @@ def get_summary(self) -> dict:
 
         Examples
         --------
-        >>> record.get_summary()
-        {'id': 'apache/maven', 'description': 'Analysis completed', 'report': 'apache.html', 'status': SCMStatus.AVAILABLE}
+        >>> from pprint import pprint
+        >>> from macaron.config.target_config import Configuration
+        >>> from macaron.output_reporter.scm import SCMStatus
+        >>> from macaron.output_reporter.results import Record
+        >>> record = Record(
+        ...     record_id="apache/maven",
+        ...     description="Analysis completed",
+        ...     pre_config=Configuration({}),
+        ...     status=SCMStatus.MISSING_SCM,
+        ...     context=None,
+        ...     dependencies=[],
+        ... )
+        >>> pprint(record.get_summary())
+        {'description': 'Analysis completed',
+         'id': 'apache/maven',
+         'repo_url_status': <SCMStatus.MISSING_SCM: 'MISSING REPO URL'>,
+         'report': ''}
         """
         return {
             "id": self.record_id,

src/macaron/policy_engine/souffle_code_generator.py

@@ -95,8 +95,11 @@ def table_to_declaration(table: Table) -> str:
 
     Examples
     --------
-    >>> tbl = Table("example", Column("id", Integer), Column("hello", String)
-    >>> assert table_to_declaration(tbl) == '.decl "example" (id: number, hello: symbol)'
+    >>> from sqlalchemy import Column, MetaData, Table
+    >>> from sqlalchemy.sql.sqltypes import Boolean, Integer, String, Text
+    >>> metadata = MetaData()
+    >>> tbl = Table("_example", metadata, Column("id", Integer, nullable=False), Column("hello", String))
+    >>> assert table_to_declaration(tbl) == '.decl example (id: number, hello: symbol)'
 
     Parameters
     ----------

src/macaron/slsa_analyzer/git_service/api_client.py

@@ -193,10 +193,12 @@ def get_repo_workflow_data(self, full_name: str, workflow_name: str) -> dict:
         The following call to this method will perform a query to
         ``https://api.github.com/repos/owner/repo/actions/workflows/build.yml``
 
-        >>> gh_client.get_repo_workflow_data(
-            full_name="owner/repo",
-            workflow_name="build.yml"
-        )
+        .. code-block: python3
+
+            gh_client.get_repo_workflow_data(
+                full_name="owner/repo",
+                workflow_name="build.yml",
+            )
         """
         logger.debug("Query for %s workflow in repo %s", workflow_name, full_name)
         url = f"{GhAPIClient._REPO_END_POINT}/{full_name}/actions/workflows/{workflow_name}"
@@ -239,12 +241,14 @@ def get_workflow_runs(
         ``https://api.github/com/repos/owner/repo/actions/runs?1&branch=master&created=>=
         2022-03-11T16:44:40Z&per_page=100``
 
-        >>> gh_client.get_workflow_runs(
-            full_name="owner/repo",
-            branch_name="master",
-            created_after="2022-03-11T16:44:40Z",
-            page=1
-        )
+        .. code-block: python3
+
+            gh_client.get_workflow_runs(
+                full_name="owner/repo",
+                branch_name="master",
+                created_after="2022-03-11T16:44:40Z",
+                page=1,
+            )
         """
         logger.debug("Query for runs data in repo %s", full_name)
         query_params: dict = {
@@ -290,10 +294,12 @@ def get_workflow_run_jobs(self, full_name: str, run_id: str) -> dict:
         ``https://api.github/com/repos/{full_name}/
         actions/runs/<run_id>/jobs``
 
-        >>> gh_client.get_workflow_run_jobs(
-            full_name="owner/repo",
-            run_id=<run_id>
-        )
+        .. code-block: python3
+
+            gh_client.get_workflow_run_jobs(
+                full_name="owner/repo",
+                run_id=<run_id>,
+            )
         """
         logger.debug("Query GitHub to get run jobs for %s with run ID %s", full_name, run_id)
 
@@ -325,10 +331,12 @@ def get_workflow_run_for_date_time_range(self, full_name: str, datetime_range: s
         The following call to this method will perform a query to
         ``https://api.github/com/repos/owner/repo/actions/runs?created=2022-11-05T20:38:40..2022-11-05T20:38:58``
 
-        >>> e.g., gh_client.get_workflow_run_for_date_time_range(
-            full_name="owner/repo",
-            created=2022-11-05T20:38:40..2022-11-05T20:38:58
-        )
+        .. code-block: python3
+
+            gh_client.get_workflow_run_for_date_time_range(
+                full_name="owner/repo",
+                created=2022-11-05T20:38:40..2022-11-05T20:38:58,
+            )
         """
         logger.debug("Query GitHub to get run details for %s at %s", full_name, datetime_range)
         query_params = {"created": datetime_range}
@@ -362,10 +370,12 @@ def get_commit_data_from_hash(self, full_name: str, commit_hash: str) -> dict:
         The following call to this method will perform a query to:
         ``https://api.github.com/repos/owner/repo/commits/6dcb09b5b57875f334f61aebed695e2e4193db5e``
 
-        >>> gh_client.get_commit_data_from_hash(
-            full_name="owner/repo",
-            commit_hash="6dcb09b5b57875f334f61aebed695e2e4193db5e"
-        )
+        .. code-block:: python3
+
+            gh_client.get_commit_data_from_hash(
+                full_name="owner/repo",
+                commit_hash="6dcb09b5b57875f334f61aebed695e2e4193db5e",
+            )
         """
         logger.debug("Query for commit %s 's data in repo %s", commit_hash, full_name)
         url = f"{GhAPIClient._REPO_END_POINT}/{full_name}/commits/{commit_hash}"
@@ -396,10 +406,12 @@ def search(self, target: str, query: str) -> dict:
         The following call to this method will perform a query to:
         ``https://api.github.com/search/code?q=addClass+in:file+language:js+repo:jquery/jquery``
 
-        >>> gh_client.search(
-            target="repositories",
-            query="q=addClass+in:file+language:js+repo:jquery/jquery"
-        )
+        .. code-block:: python3
+
+            gh_client.search(
+                target="repositories",
+                query="q=addClass+in:file+language:js+repo:jquery/jquery",
+            )
         """
         logger.debug("Search %s with query: %s", target, query)
         url = f"{GhAPIClient._SEARCH_END_POINT}/{target}?{query}"
@@ -457,7 +469,9 @@ def get_repo_data(self, full_name: str) -> dict:
         --------
         To get the repo data from repository ``apache/maven``:
 
-        >>> gh_client.get_repo_data("apache/maven")
+        .. code-block:: python3
+
+            gh_client.get_repo_data("apache/maven")
         """
         logger.debug("Get data of repository %s", full_name)
         url = f"{GhAPIClient._REPO_END_POINT}/{full_name}"
@@ -488,7 +502,8 @@ def get_file_link(self, full_name: str, commit_sha: str, file_path: str) -> str:
 
         Examples
         --------
-        >>> get_files_link("owner/repo", "5aaaaa43caabbdbc26c254df8f3aaa7bb3f4ec01", ".travis_ci.yml")
+        >>> api_client = GhAPIClient(profile={"headers": "", "query": []})
+        >>> api_client.get_file_link("owner/repo", "5aaaaa43caabbdbc26c254df8f3aaa7bb3f4ec01", ".travis_ci.yml")
         'https://github.com/owner/repo/blob/5aaaaa43caabbdbc26c254df8f3aaa7bb3f4ec01/.travis_ci.yml'
         """
         return f"https://github.com/{full_name}/blob/{commit_sha}/{file_path}"
@@ -508,6 +523,7 @@ def get_relative_path_of_workflow(self, workflow_name: str) -> str:
 
         Examples
         --------
+        >>> api_client = GhAPIClient(profile={"headers": "", "query": []})
         >>> api_client.get_relative_path_of_workflow("build.yaml")
         '.github/workflows/build.yaml'
         """

src/macaron/slsa_analyzer/git_url.py

@@ -348,7 +348,10 @@ def get_repo_complete_name_from_url(url: str) -> str:
 
     Examples
     --------
-    >>> get_repo_complete_name_from_url("github.com/apache/maven")
+    >>> from macaron.config.defaults import load_defaults
+    >>> load_defaults("")
+    True
+    >>> get_repo_complete_name_from_url("https://github.com/apache/maven")
     'github.com/apache/maven'
     """
     remote_url = get_remote_vcs_url(url)

src/macaron/slsa_analyzer/registry.py

@@ -291,10 +291,10 @@ def _validate_check_id_format(check_id: Any) -> bool:
 
         Examples
         --------
-        >>> Register._validate_check_id_format("mcn_the_check_name_123")
+        >>> Registry._validate_check_id_format("mcn_the_check_name_123")
         True
 
-        >>> Register._validate_check_id_format("Some_Thing', '', '%(*$)")
+        >>> Registry._validate_check_id_format("Some_Thing', '', '%(*$)")
         False
         """
         if (not isinstance(check_id, str)) or (not Registry._id_format.match(check_id)):

src/macaron/util.py

@@ -232,7 +232,9 @@ def copy_file_bulk(file_list: list, src_path: str, target_path: str) -> bool:
     --------
     ``file.txt`` will be copied from ``src/foo/bar/file.txt`` to ``target/foo/bar/file.txt``
 
-    >>> copy_file_bulk(['foo/bar/file.txt'], 'src', 'target')
+    .. code-block:: python3
+
+        copy_file_bulk(["foo/bar/file.txt"], "src", "target")
     """
     for file in file_list:
         file_src_path = os.path.join(src_path, file)