GitCommit object now contains diff stats information (#334)

GitCommit now contains a new changed_files_stats attribute which
is a dictionary mapping changed file paths to a corresponding
GitChangedFileStats object which contains the number of additions and
deletions in that file.

10 files changed, 193 insertions, 17 deletions

diff --git a/qa/base.py b/qa/base.py
index 7a7ccd9..ce66fbf 100644
--- a/qa/base.py
+++ b/qa/base.py
@@ -75,11 +75,23 @@ class BaseTestCase(TestCase):
         return tmp_git_repo
 
     @staticmethod
-    def create_file(parent_dir):
+    def create_file(parent_dir, content=None):
         """Creates a file inside a passed directory. Returns filename."""
         test_filename = "test-fïle-" + str(uuid4())
-        # pylint: disable=consider-using-with
-        open(os.path.join(parent_dir, test_filename), "a", encoding=DEFAULT_ENCODING).close()
+        full_path = os.path.join(parent_dir, test_filename)
+
+        if content:
+            if isinstance(content, bytes):
+                open_kwargs = {"mode": "wb"}
+            else:
+                open_kwargs = {"mode": "w", "encoding": DEFAULT_ENCODING}
+
+            with open(full_path, **open_kwargs) as f:  # pylint: disable=unspecified-encoding
+                f.write(content)
+        else:
+            # pylint: disable=consider-using-with
+            open(full_path, "a", encoding=DEFAULT_ENCODING).close()
+
         return test_filename
 
     @staticmethod
@@ -95,7 +107,9 @@ class BaseTestCase(TestCase):
         tmp_config = self.create_tmpfile(contents)
         return self.create_environment({"GIT_CONFIG": tmp_config})
 
-    def create_simple_commit(self, message, out=None, ok_code=None, env=None, git_repo=None, tty_in=False):
+    def create_simple_commit(
+        self, message, *, file_contents=None, out=None, ok_code=None, env=None, git_repo=None, tty_in=False
+    ):
         """Creates a simple commit with an empty test file.
         :param message: Commit message for the commit."""
 
@@ -108,7 +122,7 @@ class BaseTestCase(TestCase):
         environment = self.create_environment(env)
 
         # Create file and add to git
-        test_filename = self.create_file(git_repo)
+        test_filename = self.create_file(git_repo, file_contents)
         git("add", test_filename, _cwd=git_repo)
         # https://amoffat.github.io/sh/#interactive-callbacks
         if not ok_code:
@@ -132,8 +146,15 @@ class BaseTestCase(TestCase):
         # Not using a context manager to avoid unnecessary indentation in test code
         tmpfile, tmpfilepath = tempfile.mkstemp()
         self.tmpfiles.append(tmpfilepath)
-        with open(tmpfile, "w", encoding=DEFAULT_ENCODING) as f:
+
+        if isinstance(content, bytes):
+            open_kwargs = {"mode": "wb"}
+        else:
+            open_kwargs = {"mode": "w", "encoding": DEFAULT_ENCODING}
+
+        with open(tmpfile, **open_kwargs) as f:  # pylint: disable=unspecified-encoding
             f.write(content)
+
         return tmpfilepath
 
     @staticmethod
diff --git a/qa/expected/test_commits/test_lint_staged_msg_filename_1 b/qa/expected/test_commits/test_lint_staged_msg_filename_1
index 882a82c..dd565b3 100644
--- a/qa/expected/test_commits/test_lint_staged_msg_filename_1
+++ b/qa/expected/test_commits/test_lint_staged_msg_filename_1
@@ -68,10 +68,10 @@ DEBUG: gitlint.cli Using --msg-filename.
 DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
 DEBUG: gitlint.cli Linting 1 commit(s)
 DEBUG: gitlint.lint Linting commit [SHA UNKNOWN]
+DEBUG: gitlint.git ('diff', '--staged', '--numstat', '-r')
 DEBUG: gitlint.git ('config', '--get', 'user.name')
 DEBUG: gitlint.git ('config', '--get', 'user.email')
 DEBUG: gitlint.git ('rev-parse', '--abbrev-ref', 'HEAD')
-DEBUG: gitlint.git ('diff', '--staged', '--name-only', '-r')
 DEBUG: gitlint.lint Commit Object
 --- Commit Message ----
 WIP: from fïle test.
@@ -85,6 +85,8 @@ is-squash-commit: False
 is-revert-commit: False
 Branches: ['main']
 Changed Files: {changed_files}
+Changed Files Stats:
+  {changed_files_stats}
 -----------------------
 1: T3 Title has trailing punctuation (.): "WIP: from fïle test."
 1: T5 Title contains the word 'WIP' (case-insensitive): "WIP: from fïle test."
diff --git a/qa/expected/test_commits/test_lint_staged_stdin_1 b/qa/expected/test_commits/test_lint_staged_stdin_1
index 9d6153a..f999e39 100644
--- a/qa/expected/test_commits/test_lint_staged_stdin_1
+++ b/qa/expected/test_commits/test_lint_staged_stdin_1
@@ -70,10 +70,10 @@ DEBUG: gitlint.cli Stdin detected and not ignored. Using as input.
 DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
 DEBUG: gitlint.cli Linting 1 commit(s)
 DEBUG: gitlint.lint Linting commit [SHA UNKNOWN]
+DEBUG: gitlint.git ('diff', '--staged', '--numstat', '-r')
 DEBUG: gitlint.git ('config', '--get', 'user.name')
 DEBUG: gitlint.git ('config', '--get', 'user.email')
 DEBUG: gitlint.git ('rev-parse', '--abbrev-ref', 'HEAD')
-DEBUG: gitlint.git ('diff', '--staged', '--name-only', '-r')
 DEBUG: gitlint.lint Commit Object
 --- Commit Message ----
 WIP: Pïpe test.
@@ -87,6 +87,8 @@ is-squash-commit: False
 is-revert-commit: False
 Branches: ['main']
 Changed Files: {changed_files}
+Changed Files Stats:
+  {changed_files_stats}
 -----------------------
 1: T3 Title has trailing punctuation (.): "WIP: Pïpe test."
 1: T5 Title contains the word 'WIP' (case-insensitive): "WIP: Pïpe test."
diff --git a/qa/expected/test_config/test_config_from_env_1 b/qa/expected/test_config/test_config_from_env_1
index bbbdcc2..c828067 100644
--- a/qa/expected/test_config/test_config_from_env_1
+++ b/qa/expected/test_config/test_config_from_env_1
@@ -72,8 +72,8 @@ DEBUG: gitlint.cli Linting 1 commit(s)
 DEBUG: gitlint.git ('log', '{commit_sha}', '-1', '--pretty=%aN%x00%aE%x00%ai%x00%P%n%B')
 DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
 DEBUG: gitlint.lint Linting commit {commit_sha}
+DEBUG: gitlint.git ('diff-tree', '--no-commit-id', '--numstat', '-r', '--root', '{commit_sha}')
 DEBUG: gitlint.git ('branch', '--contains', '{commit_sha}')
-DEBUG: gitlint.git ('diff-tree', '--no-commit-id', '--name-only', '-r', '--root', '{commit_sha}')
 DEBUG: gitlint.lint Commit Object
 --- Commit Message ----
 WIP: Thïs is a title thåt is a bit longer.
@@ -90,6 +90,8 @@ is-squash-commit: False
 is-revert-commit: False
 Branches: ['main']
 Changed Files: {changed_files}
+Changed Files Stats:
+  {changed_files_stats}
 -----------------------
 1: CC1 Body does not contain a 'Signed-off-by' line
 1: CT1 Title does not start with one of fix, feat, chore, docs, style, refactor, perf, test, revert, ci, build
diff --git a/qa/expected/test_config/test_config_from_env_2 b/qa/expected/test_config/test_config_from_env_2
index 6049f32..a923f22 100644
--- a/qa/expected/test_config/test_config_from_env_2
+++ b/qa/expected/test_config/test_config_from_env_2
@@ -68,10 +68,10 @@ DEBUG: gitlint.cli Using --msg-filename.
 DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
 DEBUG: gitlint.cli Linting 1 commit(s)
 DEBUG: gitlint.lint Linting commit [SHA UNKNOWN]
+DEBUG: gitlint.git ('diff', '--staged', '--numstat', '-r')
 DEBUG: gitlint.git ('config', '--get', 'user.name')
 DEBUG: gitlint.git ('config', '--get', 'user.email')
 DEBUG: gitlint.git ('rev-parse', '--abbrev-ref', 'HEAD')
-DEBUG: gitlint.git ('diff', '--staged', '--name-only', '-r')
 DEBUG: gitlint.lint Commit Object
 --- Commit Message ----
 WIP: msg-fïlename test.
@@ -85,5 +85,6 @@ is-squash-commit: False
 is-revert-commit: False
 Branches: ['main']
 Changed Files: []
+Changed Files Stats: {{}}
 -----------------------
 DEBUG: gitlint.cli Exit Code = 3
diff --git a/qa/expected/test_config/test_config_from_file_debug_1 b/qa/expected/test_config/test_config_from_file_debug_1
index f136348..80d26e8 100644
--- a/qa/expected/test_config/test_config_from_file_debug_1
+++ b/qa/expected/test_config/test_config_from_file_debug_1
@@ -69,8 +69,8 @@ DEBUG: gitlint.cli Linting 1 commit(s)
 DEBUG: gitlint.git ('log', '{commit_sha}', '-1', '--pretty=%aN%x00%aE%x00%ai%x00%P%n%B')
 DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
 DEBUG: gitlint.lint Linting commit {commit_sha}
+DEBUG: gitlint.git ('diff-tree', '--no-commit-id', '--numstat', '-r', '--root', '{commit_sha}')
 DEBUG: gitlint.git ('branch', '--contains', '{commit_sha}')
-DEBUG: gitlint.git ('diff-tree', '--no-commit-id', '--name-only', '-r', '--root', '{commit_sha}')
 DEBUG: gitlint.lint Commit Object
 --- Commit Message ----
 WIP: Thïs is a title thåt is a bit longer.
@@ -87,6 +87,8 @@ is-squash-commit: False
 is-revert-commit: False
 Branches: ['main']
 Changed Files: {changed_files}
+Changed Files Stats:
+  {changed_files_stats}
 -----------------------
 1: T1 Title exceeds max length (42>20)
 1: T5 Title contains the word 'WIP' (case-insensitive)
diff --git a/qa/expected/test_gitlint/test_commit_binary_file_1 b/qa/expected/test_gitlint/test_commit_binary_file_1
new file mode 100644
index 0000000..8e87051
--- /dev/null
+++ b/qa/expected/test_gitlint/test_commit_binary_file_1
@@ -0,0 +1,92 @@
+DEBUG: gitlint.cli To report issues, please visit https://github.com/jorisroovers/gitlint/issues
+DEBUG: gitlint.cli Platform: {platform}
+DEBUG: gitlint.cli Python version: {python_version}
+DEBUG: gitlint.git ('--version',)
+DEBUG: gitlint.cli Git version: {git_version}
+DEBUG: gitlint.cli Gitlint version: {gitlint_version}
+DEBUG: gitlint.cli GITLINT_USE_SH_LIB: {GITLINT_USE_SH_LIB}
+DEBUG: gitlint.cli DEFAULT_ENCODING: {DEFAULT_ENCODING}
+DEBUG: gitlint.cli Configuration
+config-path: None
+[GENERAL]
+extra-path: None
+contrib: []
+ignore: 
+ignore-merge-commits: True
+ignore-fixup-commits: True
+ignore-fixup-amend-commits: True
+ignore-squash-commits: True
+ignore-revert-commits: True
+ignore-stdin: False
+staged: False
+fail-without-commits: False
+verbosity: 3
+debug: True
+target: {target}
+[RULES]
+  I1: ignore-by-title
+     ignore=all
+     regex=None
+  I2: ignore-by-body
+     ignore=all
+     regex=None
+  I3: ignore-body-lines
+     regex=None
+  I4: ignore-by-author-name
+     ignore=all
+     regex=None
+  T1: title-max-length
+     line-length=72
+  T2: title-trailing-whitespace
+  T6: title-leading-whitespace
+  T3: title-trailing-punctuation
+  T4: title-hard-tab
+  T5: title-must-not-contain-word
+     words=WIP
+  T7: title-match-regex
+     regex=None
+  T8: title-min-length
+     min-length=5
+  B1: body-max-line-length
+     line-length=80
+  B5: body-min-length
+     min-length=20
+  B6: body-is-missing
+     ignore-merge-commits=True
+  B2: body-trailing-whitespace
+  B3: body-hard-tab
+  B4: body-first-line-empty
+  B7: body-changed-file-mention
+     files=
+  B8: body-match-regex
+     regex=None
+  M1: author-valid-email
+     regex=[^@ ]+@[^@ ]+\.[^@ ]+
+
+DEBUG: gitlint.cli No --msg-filename flag, no or empty data passed to stdin. Using the local repo.
+DEBUG: gitlint.git ('log', '-1', '--pretty=%H')
+DEBUG: gitlint.cli Linting 1 commit(s)
+DEBUG: gitlint.git ('log', '{commit_sha}', '-1', '--pretty=%aN%x00%aE%x00%ai%x00%P%n%B')
+DEBUG: gitlint.git ('config', '--get', 'core.commentchar')
+DEBUG: gitlint.lint Linting commit {commit_sha}
+DEBUG: gitlint.git ('diff-tree', '--no-commit-id', '--numstat', '-r', '--root', '{commit_sha}')
+DEBUG: gitlint.git ('branch', '--contains', '{commit_sha}')
+DEBUG: gitlint.lint Commit Object
+--- Commit Message ----
+Sïmple commit
+
+--- Meta info ---------
+Author: gitlint-test-user <gitlint@test.com>
+Date:   {commit_date}
+is-merge-commit:  False
+is-fixup-commit:  False
+is-fixup-amend-commit: False
+is-squash-commit: False
+is-revert-commit: False
+Branches: ['main']
+Changed Files: {changed_files}
+Changed Files Stats:
+  {changed_files_stats}
+-----------------------
+3: B6 Body message is missing
+DEBUG: gitlint.cli Exit Code = 1
diff --git a/qa/test_commits.py b/qa/test_commits.py
index 8581504..d40c211 100644
--- a/qa/test_commits.py
+++ b/qa/test_commits.py
@@ -155,7 +155,15 @@ class CommitsTests(BaseTestCase):
 
         # Determine variable parts of expected output
         expected_kwargs = self.get_debug_vars_last_commit()
-        expected_kwargs.update({"changed_files": sorted([filename1, filename2])})
+        filenames = sorted([filename1, filename2])
+        expected_kwargs.update(
+            {
+                "changed_files": filenames,
+                "changed_files_stats": (
+                    f"{filenames[0]}: 0 additions, 0 deletions\n  {filenames[1]}: 0 additions, 0 deletions"
+                ),
+            }
+        )
 
         # It's not really possible to determine the "Date: ..." line that is part of the debug output as this date
         # is not taken from git but instead generated by gitlint itself. As a workaround, we extract the date from the
@@ -199,7 +207,15 @@ class CommitsTests(BaseTestCase):
 
         # Determine variable parts of expected output
         expected_kwargs = self.get_debug_vars_last_commit()
-        expected_kwargs.update({"changed_files": sorted([filename1, filename2])})
+        filenames = sorted([filename1, filename2])
+        expected_kwargs.update(
+            {
+                "changed_files": filenames,
+                "changed_files_stats": (
+                    f"{filenames[0]}: 0 additions, 0 deletions\n  {filenames[1]}: 0 additions, 0 deletions"
+                ),
+            }
+        )
 
         # It's not really possible to determine the "Date: ..." line that is part of the debug output as this date
         # is not taken from git but instead generated by gitlint itself. As a workaround, we extract the date from the
diff --git a/qa/test_config.py b/qa/test_config.py
index c9ea4e7..1225f6a 100644
--- a/qa/test_config.py
+++ b/qa/test_config.py
@@ -73,7 +73,13 @@ class ConfigTests(BaseTestCase):
             output = gitlint("--config", config_path, "--debug", _cwd=target_repo, _tty_in=True, _ok_code=[5])
 
             expected_kwargs = self.get_debug_vars_last_commit(git_repo=target_repo)
-            expected_kwargs.update({"config_path": config_path, "changed_files": [filename]})
+            expected_kwargs.update(
+                {
+                    "config_path": config_path,
+                    "changed_files": [filename],
+                    "changed_files_stats": f"{filename}: 0 additions, 0 deletions",
+                }
+            )
             self.assertEqualStdout(
                 output, self.get_expected("test_config/test_config_from_file_debug_1", expected_kwargs)
             )
@@ -103,7 +109,9 @@ class ConfigTests(BaseTestCase):
         )
         output = gitlint(_env=env, _cwd=self.tmp_git_repo, _tty_in=True, _ok_code=[5])
         expected_kwargs = self.get_debug_vars_last_commit(git_repo=target_repo)
-        expected_kwargs.update({"changed_files": [filename]})
+        expected_kwargs.update(
+            {"changed_files": [filename], "changed_files_stats": f"{filename}: 0 additions, 0 deletions"}
+        )
 
         self.assertEqualStdout(output, self.get_expected("test_config/test_config_from_env_1", expected_kwargs))
 
diff --git a/qa/test_gitlint.py b/qa/test_gitlint.py
index bad58a1..6af19c8 100644
--- a/qa/test_gitlint.py
+++ b/qa/test_gitlint.py
@@ -191,7 +191,12 @@ class IntegrationTests(BaseTestCase):
         # necessary but seems good for consistency.
         env = self.create_tmp_git_config("[user]\n  email = test-emåil@foo.com\n")
         output = gitlint(
-            "--staged", "--msg-filename", tmp_commit_msg_file, _ok_code=[self.GIT_CONTEXT_ERROR_CODE], _env=env
+            "--staged",
+            "--msg-filename",
+            tmp_commit_msg_file,
+            _ok_code=[self.GIT_CONTEXT_ERROR_CODE],
+            _env=env,
+            _cwd=self.tmp_git_repo,
         )
         expected = "Missing git configuration: please set user.name\n"
         self.assertEqualStdout(output, expected)
@@ -201,7 +206,12 @@ class IntegrationTests(BaseTestCase):
         tmp_commit_msg_file = self.create_tmpfile("WIP: msg-fïlename NO email test.")
         env = self.create_tmp_git_config("[user]\n  name = test åuthor\n")
         output = gitlint(
-            "--staged", "--msg-filename", tmp_commit_msg_file, _ok_code=[self.GIT_CONTEXT_ERROR_CODE], _env=env
+            "--staged",
+            "--msg-filename",
+            tmp_commit_msg_file,
+            _ok_code=[self.GIT_CONTEXT_ERROR_CODE],
+            _env=env,
+            _cwd=self.tmp_git_repo,
         )
         expected = "Missing git configuration: please set user.email\n"
         self.assertEqualStdout(output, expected)
@@ -227,3 +237,23 @@ class IntegrationTests(BaseTestCase):
             echo("WIP: Pïpe test."), "--staged", _cwd=empty_git_repo, _tty_in=False, _err_to_out=True, _ok_code=[3]
         )
         self.assertEqualStdout(output, expected)
+
+    def test_commit_binary_file(self):
+        """When committing a binary file, git shows somewhat different output in diff commands,
+        this test ensures gitlint deals with that correctly"""
+        binary_filename = self.create_simple_commit("Sïmple commit", file_contents=bytes([0x48, 0x00, 0x49, 0x00]))
+        output = gitlint(
+            "--debug",
+            _ok_code=1,
+            _cwd=self.tmp_git_repo,
+        )
+
+        expected_kwargs = self.get_debug_vars_last_commit()
+        expected_kwargs.update(
+            {
+                "changed_files": [binary_filename],
+                "changed_files_stats": (f"{binary_filename}: None additions, None deletions"),
+            }
+        )
+        expected = self.get_expected("test_gitlint/test_commit_binary_file_1", expected_kwargs)
+        self.assertEqualStdout(output, expected)