Merge pull request #196 from astrofrog/deterministic-option

Started implementing support for deterministic figure output

Author: astrofrog

README.rst

@@ -288,69 +288,15 @@ Inkscape is required for SVG comparison.
 
 By default, Matplotlib does not produce deterministic output that will have a
 consistent hash every time it is run, or over different Matplotlib versions. In
-order to enforce that the output is deterministic, you will need to set metadata
-as described in the following subsections.
-
-PNG
-^^^
-
-For PNG files, the output can be made deterministic by setting:
-
-.. code:: python
-
-    @pytest.mark.mpl_image_compare(savefig_kwargs={'metadata': {"Software": None}})
-
-PDF
-^^^
-
-For PDF files, the output can be made deterministic by setting:
-
-.. code:: python
-
-    @pytest.mark.mpl_image_compare(savefig_kwargs={'format': 'pdf',
-                                                   'metadata': {"Creator": None,
-                                                                "Producer": None,
-                                                                "CreationDate": None}})
-
-Note that deterministic PDF output can only be achieved with Matplotlib 2.1 and above
-
-EPS
-^^^
-
-For PDF files, the output can be made deterministic by setting:
-
-.. code:: python
-
-    @pytest.mark.mpl_image_compare(savefig_kwargs={'format': 'pdf',
-                                                   'metadata': {"Creator": "test"})
-
-and in addition you will need to set the SOURCE_DATE_EPOCH environment variable to
-a constant value (this is a unit timestamp):
-
-.. code:: python
-
-    os.environ['SOURCE_DATE_EPOCH'] = '1680254601'
-
-You could do this inside the test.
-
-Note that deterministic PDF output can only be achieved with Matplotlib 2.1 and above
-
-SVG
-^^^
-
-For SVG files, the output can be made deterministic by setting:
-
-.. code:: python
-
-    @pytest.mark.mpl_image_compare(savefig_kwargs={'metadata': '{"Date": None}})
-
-and in addition, you should make sure the following rcParam is set to a constant string:
+order to enforce that the output is deterministic, you can set the ``deterministic``
+keyword argument in ``mpl_image_compare``:
 
 .. code:: python
 
-    plt.rcParams['svg.hashsalt'] = 'test'
+    @pytest.mark.mpl_image_compare(deterministic=True)
 
-Note that SVG files can only be used in pytest-mpl with Matplotlib 3.3 and above.
+This does a number of things such as e.g., setting the creation date in the
+metadata to be constant, and avoids hard-coding the Matplotlib in the files.
 
 Test failure example
 --------------------

pytest_mpl/plugin.py

@@ -431,16 +431,13 @@ def generate_baseline_image(self, item, fig):
         """
         Generate reference figures.
         """
-        compare = get_compare(item)
-        savefig_kwargs = compare.kwargs.get('savefig_kwargs', {})
 
         if not os.path.exists(self.generate_dir):
             os.makedirs(self.generate_dir)
 
         baseline_filename = self.generate_filename(item)
         baseline_path = (self.generate_dir / baseline_filename).absolute()
-        fig.savefig(str(baseline_path), **savefig_kwargs)
-
+        self.save_figure(item, fig, baseline_path)
         close_mpl_figure(fig)
 
         return baseline_path
@@ -450,13 +447,9 @@ def generate_image_hash(self, item, fig):
         For a `matplotlib.figure.Figure`, returns the SHA256 hash as a hexadecimal
         string.
         """
-        compare = get_compare(item)
-        savefig_kwargs = compare.kwargs.get('savefig_kwargs', {})
-
-        ext = self._file_extension(item)
 
         imgdata = io.BytesIO()
-        fig.savefig(imgdata, **savefig_kwargs)
+        self.save_figure(item, fig, imgdata)
         out = _hash_file(imgdata)
         imgdata.close()
 
@@ -475,14 +468,13 @@ def compare_image_to_baseline(self, item, fig, result_dir, summary=None):
 
         compare = get_compare(item)
         tolerance = compare.kwargs.get('tolerance', 2)
-        savefig_kwargs = compare.kwargs.get('savefig_kwargs', {})
 
         ext = self._file_extension(item)
 
         baseline_image_ref = self.obtain_baseline_image(item, result_dir)
 
         test_image = (result_dir / f"result.{ext}").absolute()
-        fig.savefig(str(test_image), **savefig_kwargs)
+        self.save_figure(item, fig, test_image)
 
         if ext in ['png', 'svg']:  # Use original file
             summary['result_image'] = test_image.relative_to(self.results_dir).as_posix()
@@ -554,13 +546,55 @@ def load_hash_library(self, library_path):
         with open(str(library_path)) as fp:
             return json.load(fp)
 
+    def save_figure(self, item, fig, filename):
+        if isinstance(filename, Path):
+            filename = str(filename)
+        compare = get_compare(item)
+        savefig_kwargs = compare.kwargs.get('savefig_kwargs', {})
+        deterministic = compare.kwargs.get('deterministic', False)
+
+        original_source_date_epoch = os.environ.get('SOURCE_DATE_EPOCH', None)
+
+        extra_rcparams = {}
+
+        if deterministic:
+
+            # Make sure we don't modify the original dictionary in case is a common
+            # object used by different tests
+            savefig_kwargs = savefig_kwargs.copy()
+
+            if 'metadata' not in savefig_kwargs:
+                savefig_kwargs['metadata'] = {}
+
+            ext = self._file_extension(item)
+
+            if ext == 'png':
+                extra_metadata = {"Software": None}
+            elif ext == 'pdf':
+                extra_metadata = {"Creator": None, "Producer": None, "CreationDate": None}
+            elif ext == 'eps':
+                extra_metadata = {"Creator": "test"}
+                os.environ['SOURCE_DATE_EPOCH'] = '1680254601'
+            elif ext == 'svg':
+                extra_metadata = {"Date": None}
+                extra_rcparams["svg.hashsalt"] = "test"
+
+            savefig_kwargs['metadata'].update(extra_metadata)
+
+        import matplotlib.pyplot as plt
+
+        with plt.rc_context(rc=extra_rcparams):
+            fig.savefig(filename, **savefig_kwargs)
+
+        if original_source_date_epoch is not None:
+            os.environ['SOURCE_DATE_EPOCH'] = original_source_date_epoch
+
     def compare_image_to_hash_library(self, item, fig, result_dir, summary=None):
         hash_comparison_pass = False
         if summary is None:
             summary = {}
 
         compare = get_compare(item)
-        savefig_kwargs = compare.kwargs.get('savefig_kwargs', {})
 
         ext = self._file_extension(item)
 
@@ -601,7 +635,7 @@ def compare_image_to_hash_library(self, item, fig, result_dir, summary=None):
 
         # Save the figure for later summary (will be removed later if not needed)
         test_image = (result_dir / f"result.{ext}").absolute()
-        fig.savefig(str(test_image), **savefig_kwargs)
+        self.save_figure(item, fig, test_image)
         summary['result_image'] = test_image.relative_to(self.results_dir).as_posix()
 
         # Hybrid mode (hash and image comparison)

tests/test_pytest_mpl.py

@@ -704,15 +704,6 @@ def test_formats(pytester, use_hash_library, passes, file_format):
         else:
             pytest.skip('Comparing EPS and PDF files requires ghostscript to be installed')
 
-    if file_format == 'png':
-        metadata = '{"Software": None}'
-    elif file_format == 'pdf':
-        metadata = '{"Creator": None, "Producer": None, "CreationDate": None}'
-    elif file_format == 'eps':
-        metadata = '{"Creator": "test"}'
-    elif file_format == 'svg':
-        metadata = '{"Date": None}'
-
     pytester.makepyfile(
         f"""
         import os
@@ -721,16 +712,9 @@ def test_formats(pytester, use_hash_library, passes, file_format):
         @pytest.mark.mpl_image_compare(baseline_dir=r"{baseline_dir_abs}",
                                        {f'hash_library=r"{hash_library}",' if use_hash_library else ''}
                                        tolerance={DEFAULT_TOLERANCE},
-                                       savefig_kwargs={{'format': '{file_format}',
-                                                        'metadata': {metadata}}})
+                                       deterministic=True,
+                                       savefig_kwargs={{'format': '{file_format}'}})
         def test_format_{file_format}():
-
-            # For reproducible EPS output
-            os.environ['SOURCE_DATE_EPOCH'] = '1680254601'
-
-            # For reproducible SVG output
-            plt.rcParams['svg.hashsalt'] = 'test'
-
             fig = plt.figure()
             ax = fig.add_subplot(1, 1, 1)
             ax.plot([{1 if passes else 3}, 2, 3])