Attempt to support self-managed GitLab instances

README.rst

@@ -95,18 +95,22 @@ you should be able to call the executable, like so:
 ``howfairis`` supports URLs from the following code repository platforms:
 
 1. ``https://github.com``
-2. ``https://gitlab.com`` (not including self-hosted instances)
+2. ``https://gitlab.com``
+3. self-hosted GitLab instances
+
+Self-hosted GitLab instances will rely on the required features being supported,
+expect errors if this is not the case.
 
 Docker
 ---------------
 
-You can run howfairis Docker image using the command below.
+There is a howfairis Docker image that can be obtrained using the command below.
 
 .. code:: console
 
     docker pull fairsoftware/howfairis
 
-You can run howfairis Docker image using the command below.
+You can then run the howfairis Docker image using the command below.
 
 .. code:: console
 
@@ -242,7 +246,10 @@ Which then shows something like:
 
       * https://github.com
 
-      * https://gitlab.com (not including any self-hosted instances)
+      * https://gitlab.com
+
+      * a self-hosted GitLab instance (provided the required features are
+        supported, expect errors)
 
     Options:
       -b, --branch TEXT               Which git branch to use. Also accepts other
@@ -258,6 +265,11 @@ Which then shows something like:
       -i, --ignore-repo-config        Ignore any configuration files on the
                                       remote.
 
+      -s, --self-hosted               Use a self-hosted GitLab instance, guessing
+                                      the FQDN to use. This may not work,
+                                      depending on the server version. Expect
+                                      errors if unsupported features are requried.
+
       -p, --path TEXT                 Relative path (on the remote). Use this if
                                       you want howfairis to look for a README and
                                       a configuration file in a subdirectory.

howfairis/cli/cli.py

@@ -24,6 +24,9 @@
               help="Show default configuration and exit.")
 @click.option("-i", "--ignore-repo-config", default=False, is_flag=True,
               help="Ignore any configuration files on the remote.")
+@click.option("-s", "--self-hosted", default=False, is_flag=True,
+              help="Use a self-hosted GitLab instance, guessing the FQDN to use. This may not work, depending on the "
+                   "server version. Expect errors if unsupported features are requried.")
 @click.option("-p", "--path", default=None, type=click.STRING,
               help="Relative path (on the remote). Use this if you want howfairis to look for a "
                    "README and a configuration file in a subdirectory.")
@@ -40,15 +43,17 @@
 @click.option("-v", "--version", default=False, is_flag=True,
               help="Show version and exit.")
 @click.argument("url", required=False)
-def cli(url=None, branch=None, user_config_filename=None, repo_config_filename=None, path=None,
+def cli(url=None, branch=None, user_config_filename=None, repo_config_filename=None, self_hosted=False, path=None,
         show_trace=False, json_output=False, version=False, ignore_repo_config=False, show_default_config=False, quiet=False):
 
     """Determine compliance with recommendations from fair-software.eu for the repository at URL. The following
     code repository platforms are supported:
 
     * https://github.com
 
-    * https://gitlab.com (not including any self-hosted instances)
+    * https://gitlab.com
+
+    * a self-hosted GitLab instance (provided the required features are supported, expect errors)
     """
 
     if show_trace is False:
@@ -66,13 +71,15 @@ def cli(url=None, branch=None, user_config_filename=None, repo_config_filename=N
         quiet = True
 
     init_terminal_colors()
-    repo = Repo(url, branch, path)
+    repo = Repo(url, branch, self_hosted, path)
 
-    print_feedback_about_repo_args(url, branch, path, is_quiet=quiet)
-    print_feedback_about_config_args(ignore_repo_config, repo_config_filename, user_config_filename, is_quiet=quiet)
+    print_feedback_about_repo_args(url, branch, self_hosted, path, is_quiet=quiet)
+    print_feedback_about_config_args(ignore_repo_config, repo_config_filename,
+                                     user_config_filename, is_quiet=quiet)
 
-    checker = Checker(repo, user_config_filename=user_config_filename, repo_config_filename=repo_config_filename,
-                    ignore_repo_config=ignore_repo_config, is_quiet=quiet)
+    checker = Checker(repo, user_config_filename=user_config_filename,
+                      repo_config_filename=repo_config_filename,
+                      ignore_repo_config=ignore_repo_config, is_quiet=quiet)
 
     previous_compliance = checker.readme.get_compliance()
     current_compliance = checker.check_five_recommendations()

howfairis/cli/print_feedback_about_repo_args.py

@@ -1,4 +1,4 @@
-def print_feedback_about_repo_args(url, branch, path, is_quiet=False):
+def print_feedback_about_repo_args(url, branch, self_hosted, path, is_quiet=False):
     """  """
 
     assert url is not None, "Expected URL to not be emtpy."
@@ -9,6 +9,9 @@ def print_feedback_about_repo_args(url, branch, path, is_quiet=False):
         if url is not None:
             print("url: " + url)
 
+        if self_hosted:
+            print("using self-hosted instance")
+
         if branch is not None:
             print("branch: " + branch)
 

howfairis/mixins/license_mixin.py

@@ -41,7 +41,7 @@ def has_license(self):
             result = True
 
         if self.repo.platform == Platform.GITLAB:
-            url = f"https://gitlab.com/{self.repo.owner}/{self.repo.repo}"
+            url = f"https://{self.repo.instance}/{self.repo.owner}/{self.repo.repo}"
 
             try:
                 response = get_from_platform(

howfairis/repo.py

@@ -16,11 +16,13 @@ class Repo:
         url: URL of a code repository such as https://github.com/fair-software/howfairis
         branch: Branch to checkout. Defaults to default branch of the repository.
             Can also be a commit SHA-1 hash or tag.
+        self_hosted: Flag if self-hosted GitLab instance such as my.gitlab.tld is used
         path: Path inside repository. Defaults to root.
 
     Attributes:
         url (str): URL of a code repository,
         branch (str, None): Branch to checkout. If None then :attr:`Repo.default_branch` will be used.
+        instance (str): FQDN of a self-hosted GitLab instance. 'gitlab.com' by default.
         path (str): Path inside repository.
         platform (.code_repository_platforms.Platform): Detected code repository platform of repo.
         owner (str): Owner of the repo.
@@ -34,15 +36,18 @@ class Repo:
 
     # pylint: disable=too-many-instance-attributes
     def __init__(
-        self, url: str, branch: Optional[str] = None, path: Optional[str] = None
+        self, url: str, branch: Optional[str] = None,
+        self_hosted: Optional[bool] = False,
+        path: Optional[str] = None
     ):
 
         # run assertions on user input
-        Repo._check_assertions(url)
+        Repo._check_assertions(url, self_hosted)
 
         # assign arguments to instance members
         self.url = url
         self.branch = branch
+        self.instance = "gitlab.com" if not self_hosted else urlparse(url).netloc
         self.path = "" if path is None else "/" + path.strip("/")
 
         # assign remaining members as needed
@@ -55,21 +60,28 @@ def __init__(
         self.reuse_url = self._derive_reuse_url()
 
     @staticmethod
-    def _check_assertions(url):
+    def _check_assertions(url, self_hosted):
         assert url.startswith("https://"), "url should start with https://"
-        assert True in [
-            url.startswith("https://github.com"),
-            url.startswith("https://gitlab.com"),
-        ], "Repository should be on github.com or on gitlab.com."
-        assert re.search(
-            "^https://git(hub|lab).com/[^/]+/[^/]+", url
+        if self_hosted:
+            assert False in [
+                url.startswith("https://github.com"),
+                url.startswith("https://gitlab.com"),
+            ], f"Repository should be on self-hosted GitLab."
+        else:
+            assert True in [
+                url.startswith("https://github.com"),
+                url.startswith(f"https://gitlab.com"),
+            ], f"Repository should be on github.com or on gitlab.com."
+        assert (
+            re.search("^https://git(hub|lab).com/[^/]+/[^/]+", url) or
+            re.search("^https://"+urlparse(url).netloc+"/[^/]+/[^/]+", url)
         ), "url is not a repository"
 
     def _derive_api(self):
         if self.platform == Platform.GITHUB:
             api = f"https://api.github.com/repos/{self.owner}/{self.repo}"
         elif self.platform == Platform.GITLAB:
-            api = f"https://gitlab.com/api/v4/projects/{self.owner}%2F{self.repo}"
+            api = f"https://{self.instance}/api/v4/projects/{self.owner}%2F{self.repo}"
         return api
 
     def _derive_reuse_url(self):
@@ -89,7 +101,7 @@ def _derive_owner_and_repo(self):
         elif self.platform == Platform.GITLAB:
             try:
                 owner, repo = (
-                    self.url.replace("https://gitlab.com", "").strip("/").split("/")[:2]
+                    self.url.replace(f"https://{self.instance}", "").strip("/").split("/")[:2]
                 )
             except ValueError as ex:
                 raise ValueError("Bad value for input argument URL.") from ex
@@ -103,7 +115,7 @@ def _derive_platform(self):
         if self.url.startswith("https://github.com"):
             return Platform.GITHUB
 
-        if self.url.startswith("https://gitlab.com"):
+        if self.url.startswith(f"https://{self.instance}"):
             return Platform.GITLAB
 
         return None
@@ -122,7 +134,7 @@ def _derive_raw_url_format_string(self):
             )
 
         elif self.platform == Platform.GITLAB:
-            raw_url_format_string = f"https://gitlab.com/{self.owner}/{self.repo}/-/raw/{branch}{self.path}/{{0}}"
+            raw_url_format_string = f"https://{self.instance}/{self.owner}/{self.repo}/-/raw/{branch}{self.path}/{{0}}"
 
         return raw_url_format_string