avoid svg+xml images when cleaning

CHANGELOG.md

@@ -1,4 +1,10 @@
 
+# 0.2.3 (2022-02-15)
+
+### Fixed
+
+- Do not attempt to clean `"image/svg+xml"` images
+
 # 0.2.2 (2022-02-11)
 
 ### Added

setup.py

@@ -7,7 +7,7 @@
     requirements = [c.strip() for c in f.readlines()]
 
 setup(name="nbcleanmeta",
-      version="0.2.2",
+      version="0.2.3",
       description="Le Wagon notebook metadata cleaner tool",
       url="https://github.com/lewagon/nbcleanmeta/",
       author="Sébastien Saunier",

tests/test_cleaner_outputs.py

@@ -18,9 +18,8 @@ def test_outputs_data_trailing_newline_is_present(self, notebook_path):
         code_cells = self._get_code_cells(clean_content)
         for cell in code_cells:
             for output in cell.get('outputs', []):
-                for data_key in output.get('data', {}):
-                    if data_key[:6] == 'image/':
-                        assert output['data'][data_key][-1:] == '\n'
+                if 'image/png' in output.get('data', {}):
+                    assert output['data']['image/png'][-1:] == '\n'
 
     def test_outputs_data_trailing_newline_is_removed(self, notebook_path):
         # Act
@@ -31,6 +30,5 @@ def test_outputs_data_trailing_newline_is_removed(self, notebook_path):
         code_cells = self._get_code_cells(clean_content)
         for cell in code_cells:
             for output in cell.get('outputs', []):
-                for data_key in output.get('data', {}):
-                    if data_key[:6] == 'image/':
-                        assert output['data'][data_key][-1:] != '\n'
+                if 'image/png' in output.get('data', {}):
+                    assert output['data']['image/png'][-1:] != '\n'

wagon_cleaner/run_clean.py

@@ -71,10 +71,9 @@ def edit_notebook(notebook_content, notebook_path, kwargs, delete_notes=False):
             if "execution_count" in output:
                 output["execution_count"] = None
 
-            # clean outputs data image trailing newline
-            for data_key in output.get("data", {}):
-                if data_key[:6] == "image/":
-                    output["data"][data_key] = output["data"][data_key].rstrip("\n")
+            # clean outputs data image trailing newline (avoid "image/svg+xml")
+            if "image/png" in output.get("data", {}):
+                output["data"]["image/png"] = output["data"]["image/png"].rstrip("\n")
 
         # clean cell metadata
         if "metadata" in cell: