xarray-dbd 0.2.2__tar.gz → 0.2.3__tar.gz

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xarray-dbd
-Version: 0.2.2
+Version: 0.2.3
 Summary: Efficient xarray backend for reading glider DBD files
 Keywords: glider,oceanography,dbd,slocum,xarray,netcdf
 Author-Email: Pat Welch <pat@mousebrains.com>
@@ -209,9 +209,10 @@ Open multiple DBD files as a single concatenated xarray Dataset.
 
 ## Migration from dbdreader
 
-xarray-dbd provides drop-in `DBD` and `MultiDBD` classes that mirror the
-[dbdreader](https://pypi.org/project/dbdreader/) API. For a fully
-transparent swap, alias the import:
+The dbdreader2 API is derived from Lucas Merckelbach's
+[dbdreader](https://github.com/smerckel/dbdreader) library. xarray-dbd
+provides drop-in `DBD` and `MultiDBD` classes that mirror the dbdreader API.
+For a fully transparent swap, alias the import:
 
 ```python
 # Before (dbdreader)
@@ -249,7 +250,7 @@ dbd = xdbd.DBD("file.dcd", cacheDir="cache")
 | `get_xy()`, `get_CTD_sync()` | Yes | Yes |
 | `decimalLatLon` | Yes | Yes |
 | `set_time_limits()` | Yes | Yes |
-| `include_source` | No | Yes |
+| `include_source` | Yes | Yes |
 
 ### Use-case examples
 
@@ -366,9 +367,6 @@ mdbd = dbdreader.MultiDBD(
   individual records within a file. dbdreader also filters by file open
   time, so this is operationally the same for most use cases.
 
-- **`include_source` not supported.** Passing `include_source=True` to
-  `get()` raises `NotImplementedError`.
-
 - **Error handling.** The same `DbdError` exception class and numeric
   error codes (`DBD_ERROR_CACHE_NOT_FOUND`, etc.) are provided for
   compatibility.

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/README.md

@@ -173,9 +173,10 @@ Open multiple DBD files as a single concatenated xarray Dataset.
 
 ## Migration from dbdreader
 
-xarray-dbd provides drop-in `DBD` and `MultiDBD` classes that mirror the
-[dbdreader](https://pypi.org/project/dbdreader/) API. For a fully
-transparent swap, alias the import:
+The dbdreader2 API is derived from Lucas Merckelbach's
+[dbdreader](https://github.com/smerckel/dbdreader) library. xarray-dbd
+provides drop-in `DBD` and `MultiDBD` classes that mirror the dbdreader API.
+For a fully transparent swap, alias the import:
 
 ```python
 # Before (dbdreader)
@@ -213,7 +214,7 @@ dbd = xdbd.DBD("file.dcd", cacheDir="cache")
 | `get_xy()`, `get_CTD_sync()` | Yes | Yes |
 | `decimalLatLon` | Yes | Yes |
 | `set_time_limits()` | Yes | Yes |
-| `include_source` | No | Yes |
+| `include_source` | Yes | Yes |
 
 ### Use-case examples
 
@@ -330,9 +331,6 @@ mdbd = dbdreader.MultiDBD(
   individual records within a file. dbdreader also filters by file open
   time, so this is operationally the same for most use cases.
 
-- **`include_source` not supported.** Passing `include_source=True` to
-  `get()` raises `NotImplementedError`.
-
 - **Error handling.** The same `DbdError` exception class and numeric
   error codes (`DBD_ERROR_CACHE_NOT_FOUND`, etc.) are provided for
   compatibility.

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "xarray-dbd"
-version = "0.2.2"
+version = "0.2.3"
 description = "Efficient xarray backend for reading glider DBD files"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -132,7 +132,7 @@ testpaths = ["tests"]
 python_files = "test_*.py"
 
 [tool.cibuildwheel]
-build = ["cp310-*", "cp311-*", "cp312-*", "cp313-*"]
+build = ["cp310-*", "cp311-*", "cp312-*", "cp313-*", "cp314-*"]
 skip = ["*-musllinux_i686", "*-win32", "pp*"]
 test-skip = ["*-manylinux_*", "*-musllinux_*"]
 test-command = "python -c \"import xarray_dbd; print(xarray_dbd.__version__)\""

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/tests/test_dbdreader2.py

@@ -433,10 +433,57 @@ class TestMultiDBD:
         assert MultiDBD.isScienceDataFile("test.dcd") is False
         assert MultiDBD.isScienceDataFile("test.sbd") is False
 
-    def test_include_source_not_implemented(self):
+    def test_include_source_single_param(self):
         mdbd = MultiDBD(filenames=_all_files(), cacheDir=CACHE_DIR)
-        with pytest.raises(NotImplementedError):
-            mdbd.get("m_depth", include_source=True)
+        result = mdbd.get("m_depth", include_source=True)
+        (t, v), sources = result
+        assert len(sources) == len(t)
+        assert len(sources) == len(v)
+        assert len(t) > 0
+        for src in sources:
+            assert hasattr(src, "filename")
+            assert isinstance(src, DBD)
+        mdbd.close()
+
+    def test_include_source_multi_param(self):
+        mdbd = MultiDBD(filenames=_all_files(), cacheDir=CACHE_DIR)
+        results = mdbd.get("m_depth", "m_heading", include_source=True)
+        assert isinstance(results, list)
+        assert len(results) == 2
+        for (t, v), sources in results:
+            assert len(sources) == len(t)
+            assert len(sources) == len(v)
+        mdbd.close()
+
+    def test_include_source_sources_match_files(self):
+        files = _all_files()
+        mdbd = MultiDBD(filenames=files, cacheDir=CACHE_DIR)
+        (t, v), sources = mdbd.get("m_depth", include_source=True)
+        source_filenames = {src.filename for src in sources}
+        # Every source file must be one of the input files
+        assert source_filenames <= set(files)
+        mdbd.close()
+
+    def test_include_source_max_values(self):
+        mdbd = MultiDBD(filenames=_all_files(), cacheDir=CACHE_DIR)
+        n = 5
+        (t, v), sources = mdbd.get("m_depth", include_source=True, max_values_to_read=n)
+        assert len(t) == n
+        assert len(v) == n
+        assert len(sources) == n
+        mdbd.close()
+
+    def test_continue_on_reading_error_accepted(self):
+        mdbd = MultiDBD(filenames=_all_files(), cacheDir=CACHE_DIR)
+        t, v = mdbd.get("m_depth", continue_on_reading_error=True)
+        assert len(t) > 0
+        mdbd.close()
+
+    def test_continue_on_reading_error_with_include_source(self):
+        mdbd = MultiDBD(filenames=_all_files(), cacheDir=CACHE_DIR)
+        (t, v), sources = mdbd.get("m_depth", include_source=True, continue_on_reading_error=True)
+        assert len(t) > 0
+        assert len(sources) == len(t)
         mdbd.close()
 
     def test_close_guards(self):

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/xarray_dbd/dbdreader2/__init__.py

@@ -26,6 +26,7 @@ from ._errors import (
     DBD_ERROR_NO_FILES_FOUND,
     DBD_ERROR_NO_TIME_VARIABLE,
     DBD_ERROR_NO_VALID_PARAMETERS,
+    DBD_ERROR_READ_ERROR,
     DbdError,
 )
 from ._list import DBDList, DBDPatternSelect
@@ -60,6 +61,7 @@ __all__ = [
     "DBD_ERROR_NO_FILES_FOUND",
     "DBD_ERROR_NO_TIME_VARIABLE",
     "DBD_ERROR_NO_VALID_PARAMETERS",
+    "DBD_ERROR_READ_ERROR",
     "DbdError",
     "LATLON_PARAMS",
     "MultiDBD",

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/xarray_dbd/dbdreader2/_core.py

@@ -496,21 +496,17 @@ class MultiDBD:
         return_nans=False,
         include_source=False,
         max_values_to_read=-1,
+        continue_on_reading_error=False,
     ):
         """Return ``(t, v)`` per parameter."""
         self._check_closed()
 
-        if include_source:
-            raise NotImplementedError("include_source is not yet supported in dbdreader2")
-
         if max_values_to_read > 0 and len(parameters) != 1:
             raise ValueError(
                 "Limiting the values to be read for multiple parameters "
                 "potentially yields undefined behaviour.\n"
             )
 
-        self._ensure_loaded(parameters)
-
         # Validate parameters
         all_known = set(self.parameterNames.get("eng", [])) | set(
             self.parameterNames.get("sci", [])
@@ -523,6 +519,18 @@ class MultiDBD:
                 mesg = f"Parameters {{{','.join(invalid)}}} are unknown glider sensor names."
             raise DbdError(value=DBD_ERROR_NO_VALID_PARAMETERS, mesg=mesg, data=invalid)
 
+        if include_source:
+            return self._get_with_source(
+                parameters,
+                return_nans=return_nans,
+                decimalLatLon=decimalLatLon,
+                discardBadLatLon=discardBadLatLon,
+                max_values_to_read=max_values_to_read,
+                continue_on_reading_error=continue_on_reading_error,
+            )
+
+        self._ensure_loaded(parameters)
+
         results = []
         for p in parameters:
             t, v = self._extract(p, return_nans, decimalLatLon, discardBadLatLon)
@@ -849,6 +857,67 @@ class MultiDBD:
 
         return t, v
 
+    def _get_with_source(
+        self,
+        parameters,
+        *,
+        return_nans,
+        decimalLatLon,  # noqa: N803
+        discardBadLatLon,  # noqa: N803
+        max_values_to_read,
+        continue_on_reading_error=False,
+    ):
+        """Implement ``get(..., include_source=True)`` by iterating per-file DBD objects."""
+        eng_names_set = set(self._eng_param_names)
+
+        results = []
+        for p in parameters:
+            dbds = self.dbds["sci"] if p not in eng_names_set else self.dbds["eng"]
+            t_chunks: list[numpy.ndarray] = []
+            v_chunks: list[numpy.ndarray] = []
+            src_chunks: list[list[DBD]] = []
+            for dbd in dbds:
+                if not dbd.has_parameter(p):
+                    continue
+                try:
+                    tv = dbd.get(
+                        p,
+                        decimalLatLon=decimalLatLon,
+                        discardBadLatLon=discardBadLatLon,
+                        return_nans=return_nans,
+                        check_for_invalid_parameters=False,
+                    )
+                except (RuntimeError, OSError) as e:
+                    if continue_on_reading_error:
+                        logger.warning("Reading from %s returned an error (%s).", dbd.filename, e)
+                        continue
+                    raise
+                t, v = tv
+                if len(t) == 0:
+                    continue
+                t_chunks.append(t)
+                v_chunks.append(v)
+                src_chunks.append([dbd] * len(t))
+            if t_chunks:
+                t_cat = numpy.hstack(t_chunks)
+                v_cat = numpy.hstack(v_chunks)
+                sources: list[DBD] = []
+                for chunk in src_chunks:
+                    sources.extend(chunk)
+            else:
+                t_cat = numpy.array([], dtype=numpy.float64)
+                v_cat = numpy.array([], dtype=numpy.float64)
+                sources = []
+            if max_values_to_read > 0:
+                t_cat = t_cat[:max_values_to_read]
+                v_cat = v_cat[:max_values_to_read]
+                sources = sources[:max_values_to_read]
+            results.append(((t_cat, v_cat), sources))
+
+        if len(parameters) == 1:
+            return results[0]
+        return results
+
     def _resolve_ifun(self, param, factory):
         """Resolve interpolating function factory for a parameter."""
         if factory is None:

{xarray_dbd-0.2.2 → xarray_dbd-0.2.3}/xarray_dbd/dbdreader2/_errors.py

@@ -17,6 +17,7 @@ DBD_ERROR_INVALID_ENCODING = 10
 DBD_ERROR_INVALID_FILE_CRITERION_SPECIFIED = 11
 DBD_ERROR_NO_DATA_TO_INTERPOLATE = 12
 DBD_ERROR_NO_DATA = 13
+DBD_ERROR_READ_ERROR = 14
 
 
 class DbdError(Exception):
@@ -54,6 +55,8 @@ class DbdError(Exception):
             mesg = "One or more parameters that are to be interpolated, does/do not have any data."
         elif self.value == DBD_ERROR_NO_DATA:
             mesg = "One or more parameters do not have any data."
+        elif self.value == DBD_ERROR_READ_ERROR:
+            mesg = "Error reading DBD file."
         else:
             mesg = f"Undefined error. ({self.value})"
         if self.mesg: