DEP: Always compute checksum_md5 when generating metadata (#844)

docs/src/dataio_3_migration.rst

@@ -195,6 +195,8 @@ Change to this instead 👇:
 Additionally 
 
  - The ``return_symlink`` argument to ``export()`` is deprecated. It is redundant and can be removed.
+ - The ``compute_md5`` argument to ``generate_metadata()`` is deprecated and can be removed, as 
+   an MD5 checksum is always computed by default.
 
 
 Getting partial metadata from generate_metadata() when config is invalid

src/fmu/dataio/_metadata.py

@@ -31,15 +31,13 @@ def _get_meta_filedata(
     obj: types.Inferrable,
     objdata: ObjectDataProvider,
     fmudata: FmuProvider | None,
-    compute_md5: bool,
 ) -> fields.File:
     """Derive metadata for the file."""
     return FileDataProvider(
         dataio=dataio,
         objdata=objdata,
         runpath=fmudata.get_runpath() if fmudata else None,
         obj=obj,
-        compute_md5=compute_md5,
     ).get_metadata()
 
 
@@ -75,7 +73,6 @@ def generate_export_metadata(
     obj: types.Inferrable,
     dataio: ExportData,
     fmudata: FmuProvider | None = None,
-    compute_md5: bool = True,
 ) -> schema.InternalObjectMetadata:
     """
     Main function to generate the full metadata
@@ -119,7 +116,7 @@ def generate_export_metadata(
         ),
         access=_get_meta_access(dataio),
         data=objdata.get_metadata(),
-        file=_get_meta_filedata(dataio, obj, objdata, fmudata, compute_md5),
+        file=_get_meta_filedata(dataio, obj, objdata, fmudata),
         tracklog=fields.Tracklog.initialize(),
         display=_get_meta_display(dataio, objdata),
         preprocessed=dataio.preprocessed,

src/fmu/dataio/dataio.py

@@ -816,6 +816,7 @@ def _export_without_metadata(self, obj: types.Inferrable) -> str:
         filemeta = FileDataProvider(
             dataio=self,
             objdata=objdata,
+            obj=obj,
             runpath=fmudata.get_runpath() if fmudata else None,
         ).get_metadata()
 
@@ -842,17 +843,12 @@ def generate_metadata(
 
         Args:
             obj: XTGeo instance, a Pandas Dataframe instance or other supported object.
-            compute_md5: If True, compute a MD5 checksum for the exported file.
+            compute_md5: Deprecated, a MD5 checksum will always be computed.
             **kwargs: Using other ExportData() input keys is now deprecated, input the
                 arguments when initializing the ExportData() instance instead.
 
         Returns:
             A dictionary with all metadata.
-
-        Note:
-            If the ``compute_md5`` key is False, the ``file.checksum_md5`` will be
-            empty. If true, the MD5 checksum will be generated based on export to
-            a temporary file, which may be time-consuming if the file is large.
         """
 
         logger.info("Generate metadata...")
@@ -865,6 +861,14 @@ def generate_metadata(
                 FutureWarning,
             )
 
+        if not compute_md5:
+            warnings.warn(
+                "Using the 'compute_md5=False' option to prevent an MD5 checksum "
+                "from being computed is now deprecated. This option has no longer "
+                "an effect and will be removed in the near future.",
+                UserWarning,
+            )
+
         self._update_check_settings(kwargs)
 
         if isinstance(obj, (str, Path)):
@@ -879,7 +883,9 @@ def generate_metadata(
         fmudata = self._get_fmu_provider() if self._fmurun else None
 
         return generate_export_metadata(
-            obj, self, fmudata, compute_md5=compute_md5
+            obj=obj,
+            dataio=self,
+            fmudata=fmudata,
         ).model_dump(mode="json", exclude_none=True, by_alias=True)
 
     def export(
@@ -927,7 +933,7 @@ def export(
             self._update_check_settings(kwargs)
             return self._export_without_metadata(obj)
 
-        metadata = self.generate_metadata(obj, compute_md5=True, **kwargs)
+        metadata = self.generate_metadata(obj, **kwargs)
         outfile = Path(metadata["file"]["absolute_path"])
         metafile = outfile.parent / f".{outfile.name}.yml"
 

src/fmu/dataio/providers/_filedata.py

@@ -11,7 +11,7 @@
 from dataclasses import dataclass
 from enum import Enum
 from pathlib import Path
-from typing import TYPE_CHECKING, Final, Optional
+from typing import TYPE_CHECKING, Final
 
 from fmu.dataio._logging import null_logger
 from fmu.dataio._model import enums, fields
@@ -49,9 +49,8 @@ class FileDataProvider(Provider):
     # input
     dataio: ExportData
     objdata: ObjectDataProvider
+    obj: types.Inferrable
     runpath: Path | None = None
-    obj: Optional[types.Inferrable] = None
-    compute_md5: bool = False
 
     @property
     def name(self) -> str:
@@ -88,7 +87,7 @@ def get_metadata(self) -> fields.File:
         return fields.File(
             absolute_path=absolute_path.resolve(),
             relative_path=relative_path,
-            checksum_md5=self._compute_md5() if self.compute_md5 else None,
+            checksum_md5=self._compute_md5(),
         )
 
     def _get_share_folders(self) -> Path:

tests/test_units/test_filedataprovider_class.py

@@ -112,10 +112,7 @@ def test_get_filestem(
     edataobj1.parent = parentname
     edataobj1.name = ""
 
-    fdata = FileDataProvider(
-        edataobj1,
-        objdata,
-    )
+    fdata = FileDataProvider(edataobj1, objdata, regsurf)
 
     stem = fdata._get_filestem()
     assert stem == expected
@@ -163,7 +160,7 @@ def test_get_filestem_shall_fail(
     edataobj1.parent = parentname
     edataobj1.name = ""
 
-    fdata = FileDataProvider(edataobj1, objdata)
+    fdata = FileDataProvider(edataobj1, objdata, regsurf)
 
     with pytest.raises(ValueError) as msg:
         _ = fdata._get_filestem()
@@ -178,7 +175,7 @@ def test_get_share_folders(regsurf, globalconfig2):
     objdata = objectdata_provider_factory(regsurf, edataobj1)
     objdata.name = "some"
 
-    fdata = FileDataProvider(edataobj1, objdata)
+    fdata = FileDataProvider(edataobj1, objdata, regsurf)
     share_folders = fdata._get_share_folders()
     assert isinstance(share_folders, Path)
     assert share_folders == Path(f"share/results/{ExportFolder.maps.value}")
@@ -200,7 +197,7 @@ def test_get_share_folders_with_subfolder(regsurf, globalconfig2):
     objdata = objectdata_provider_factory(regsurf, edataobj1)
     objdata.name = "some"
 
-    fdata = FileDataProvider(edataobj1, objdata)
+    fdata = FileDataProvider(edataobj1, objdata, regsurf)
     share_folders = fdata._get_share_folders()
     assert share_folders == Path("share/results/maps/sub")
 
@@ -229,7 +226,7 @@ def test_filedata_provider(regsurf, tmp_path, globalconfig2):
     objdata.time0 = datetime.strptime(t1, "%Y%m%d")
     objdata.time1 = datetime.strptime(t2, "%Y%m%d")
 
-    fdata = FileDataProvider(cfg, objdata)
+    fdata = FileDataProvider(cfg, objdata, regsurf)
     filemeta = fdata.get_metadata()
 
     assert isinstance(filemeta, fields.File)
@@ -250,6 +247,6 @@ def test_filedata_has_nonascii_letters(regsurf, tmp_path, globalconfig2):
     objdata = objectdata_provider_factory(regsurf, edataobj1)
     objdata.name = "anynõme"
 
-    fdata = FileDataProvider(edataobj1, objdata)
+    fdata = FileDataProvider(edataobj1, objdata, regsurf)
     with pytest.raises(ValueError, match="Path has non-ascii elements"):
         fdata.get_metadata()