Improve handling of warning when missing content

Address issue #379

The handling of content is moved to _object_data_provider, and also
check that warning is not issued when merging preprocessed data.

src/fmu/dataio/_objectdata_provider.py

@@ -87,14 +87,14 @@
 from dataclasses import dataclass, field
 from datetime import datetime as dt
 from pathlib import Path
-from typing import Any, Optional
+from typing import Any, Optional, Tuple
 from warnings import warn
 
 import numpy as np
 import pandas as pd  # type: ignore
 import xtgeo  # type: ignore
 
-from ._definitions import _ValidFormats, STANDARD_TABLE_INDEX_COLUMNS
+from ._definitions import ALLOWED_CONTENTS, STANDARD_TABLE_INDEX_COLUMNS, _ValidFormats
 from ._utils import generate_description, parse_timedata
 
 try:
@@ -664,6 +664,36 @@ def _derive_from_existing(self):
 
         self.time0, self.time1 = parse_timedata(self.meta_existing["data"])
 
+    def _process_content(self) -> Tuple[str, Optional[dict]]:
+        """Work with the `content` metadata"""
+
+        # content == "unset" is not wanted, but in case metadata has been produced while
+        # doing a preprocessing step first, and this step is re-using metadata, the
+        # check is not done.
+        if self.dataio._usecontent == "unset" and (
+            self.dataio.reuse_metadata_rule is None
+            or self.dataio.reuse_metadata_rule != "preprocessed"
+        ):
+            warn(
+                "The <content> is not provided which defaults to 'unset'. "
+                "It is strongly recommended that content is given explicitly! "
+                f"\n\nValid contents are: {', '.join(ALLOWED_CONTENTS.keys())} "
+                "\n\nThis list can be extended upon request and need.",
+                UserWarning,
+            )
+
+        content = self.dataio._usecontent
+        content_spesific = None
+
+        # Outgoing content is always a string, but it can be given as a dict if content-
+        # specific information is to be included in the metadata.
+        # In that case, it shall be inserted in the data block as a key with name as the
+        # content, e.g. "seismic" or "field_outline"
+        if self.dataio._content_specific is not None:
+            content_spesific = self.dataio._content_specific
+
+        return content, content_spesific
+
     def derive_metadata(self):
         """Main function here, will populate the metadata block for 'data'."""
         logger.info("Derive all metadata for data object...")
@@ -683,14 +713,11 @@ def derive_metadata(self):
         meta["alias"] = nameres.get("alias", None)
         meta["top"] = nameres.get("top", None)
         meta["base"] = nameres.get("base", None)
-        meta["content"] = self.dataio._usecontent
 
-        # Outgoing content is always a string, but it can be given as a dict if content-
-        # specific information is to be included in the metadata.
-        # In that case, it shall be inserted in the data block as a key with name as the
-        # content, e.g. "seismic" or "field_outline"
-        if self.dataio._content_specific is not None:
-            meta[self.dataio._usecontent] = self.dataio._content_specific
+        content, content_spesific = self._process_content()
+        meta["content"] = content
+        if content_spesific:
+            meta[self.dataio._usecontent] = content_spesific
 
         meta["tagname"] = self.dataio.tagname
         meta["format"] = objres["fmt"]

src/fmu/dataio/dataio.py

@@ -134,14 +134,7 @@ def _check_content(proposed: Union[str, dict]) -> Any:
     content_specific = None
     logger.debug("content is %s of type %s", str(content), type(content))
     if content is None:
-        warn(
-            "The <content> is not provided which defaults to 'unset'. "
-            "It is strongly recommended that content is given explicitly! "
-            f"\n\nValid contents are: {', '.join(ALLOWED_CONTENTS.keys())} "
-            "\n\nThis list can be extended upon request and need.",
-            UserWarning,
-        )
-        usecontent = "unset"
+        usecontent = "unset"  # user warnings on this will in _objectdata_provider
 
     elif isinstance(content, str):
         logger.debug("content is a string")

tests/test_units/test_dataio.py

@@ -121,13 +121,26 @@ def test_deprecated_keys(globalconfig1, regsurf, key, value, wtype, expected_msg
 
 def test_content_not_given(globalconfig1, regsurf):
     """When content is not explicitly given, warning shall be issued."""
-    with pytest.warns(match="The <content> is not provided"):
+    with pytest.warns(UserWarning, match="The <content> is not provided"):
         eobj = ExportData(config=globalconfig1)
         mymeta = eobj.generate_metadata(regsurf)
 
     assert mymeta["data"]["content"] == "unset"
 
 
+def test_content_given_init_or_later(globalconfig1, regsurf):
+    """When content is not explicitly given, warning shall be issued."""
+    eobj = ExportData(config=globalconfig1, content="time")
+    mymeta = eobj.generate_metadata(regsurf)
+
+    assert mymeta["data"]["content"] == "time"
+
+    # override by adding content at generate_metadata
+    mymeta = eobj.generate_metadata(regsurf, content="depth")
+
+    assert mymeta["data"]["content"] == "depth"  # last content shall win
+
+
 def test_content_invalid_string(globalconfig1):
     with pytest.raises(ValidationError, match=r"Invalid content"):
         ExportData(config=globalconfig1, content="not_valid")

tests/test_units/test_prerealization_surfaces.py

@@ -125,15 +125,19 @@ def _run_case_fmu(fmurun_w_casemetadata, rmsglobalconfig, surfacepath):
         os.chdir(fmurun_w_casemetadata)
         logger.info("Active folder is %s", fmurun_w_casemetadata)
 
+        casepath = fmurun_w_casemetadata.parent.parent
+
         edata = dataio.ExportData(
             config=rmsglobalconfig,  # read from global config
             fmu_context="case",
-            content="depth",
+            content=None,  # shall be accepted without warning here in this context
             is_observation=True,
         )
         metadata = edata.generate_metadata(
             surfacepath,
+            casepath=casepath,
         )
+        logger.info("Casepath folder is now %s", casepath)
         logger.debug("\n%s", utils.prettyprint_dict(metadata))
         assert (
             metadata["file"]["relative_path"]
@@ -144,6 +148,19 @@ def _run_case_fmu(fmurun_w_casemetadata, rmsglobalconfig, surfacepath):
         assert "TopVolantis" in metadata["data"]["alias"]
         assert "_preprocessed" not in metadata
 
+        # do the actual export (which will copy data to case/share/observations/...)
+        edata.export(
+            surfacepath,
+            casepath=casepath,
+        )
+        assert (
+            casepath
+            / "share"
+            / "observations"
+            / "maps"
+            / ".topvolantis--20240802_20200909.gri.yml"
+        ).exists()
+
     # run two stage process
     mysurf = _export_data_from_rms(rmssetup, rmsglobalconfig, regsurf)
     _run_case_fmu(fmurun_w_casemetadata, rmsglobalconfig, mysurf)