pysdmx 1.10.1__py3-none-any.whl → 1.12.0__py3-none-any.whl

pysdmx/io/json/sdmxjson2/writer/_helper.py

@@ -0,0 +1,118 @@
+"""Writer interface for SDMX-JSON 2.0.0 Reference Metadata messages."""
+
+from pathlib import Path
+from typing import Literal, Optional, Sequence, Union
+
+import msgspec
+
+from pysdmx.io.json.sdmxjson2.messages import (
+    JsonMetadataMessage,
+    JsonStructureMessage,
+)
+from pysdmx.model import MetadataReport, encoders
+from pysdmx.model.__base import MaintainableArtefact
+from pysdmx.model.message import Header, MetadataMessage, StructureMessage
+
+
+def write_metadata_msg(
+    reports: Sequence[MetadataReport],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+    msg_version: Literal["2.0.0", "2.1"] = "2.0.0",
+) -> Optional[str]:
+    """Write metadata reports in requested SDMX-JSON version.
+
+    Args:
+        reports: The reference metadata reports to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+        msg_version: The desired version of SDMX-JSON. Defaults to 2.0.0.
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    if not header:
+        header = Header()
+    sm = MetadataMessage(header, reports)
+    jsm = JsonMetadataMessage.from_model(sm, msg_version)
+
+    encoder = msgspec.json.Encoder(enc_hook=encoders)
+    serialized_data = encoder.encode(jsm)
+
+    # Apply pretty-printing if requested
+    if prettyprint:
+        serialized_data = msgspec.json.format(serialized_data, indent=4)
+
+    # If output_path is provided, write to file
+    if output_path:
+        # Convert to Path object if string
+        if isinstance(output_path, str):
+            output_path = Path(output_path)
+
+        # Create parent directories if they don't exist
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+
+        # Write to file
+        with open(output_path, "wb") as f:
+            f.write(serialized_data)
+        return None
+    else:
+        # Return as string
+        return serialized_data.decode("utf-8")
+
+
+def write_structure_msg(
+    structures: Sequence[MaintainableArtefact],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+    msg_version: Literal["2.0.0", "2.1"] = "2.0.0",
+) -> Optional[str]:
+    """Write maintainable SDMX artefacts in SDMX-JSON 2.0.0.
+
+    Args:
+        structures: The maintainable SDMX artefacts to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+        msg_version: The desired version of SDMX-JSON. Defaults to 2.0.0.
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    if not header:
+        header = Header()
+    sm = StructureMessage(header, structures)
+    jsm = JsonStructureMessage.from_model(sm, msg_version)
+
+    encoder = msgspec.json.Encoder(enc_hook=encoders)
+    serialized_data = encoder.encode(jsm)
+
+    # Apply pretty-printing if requested
+    if prettyprint:
+        serialized_data = msgspec.json.format(serialized_data, indent=4)
+
+    # If output_path is provided, write to file
+    if output_path:
+        # Convert to Path object if string
+        if isinstance(output_path, str):
+            output_path = Path(output_path)
+
+        # Create parent directories if they don't exist
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+
+        # Write to file
+        with open(output_path, "wb") as f:
+            f.write(serialized_data)
+        return None
+    else:
+        # Return as string
+        return serialized_data.decode("utf-8")

pysdmx/io/json/sdmxjson2/writer/v2_0/__init__.py

@@ -0,0 +1 @@
+"""Collection of writers for SDMX-JSON 2.0.0 messages."""

pysdmx/io/json/sdmxjson2/writer/v2_0/metadata.py

@@ -0,0 +1,33 @@
+"""Writer interface for SDMX-JSON 2.0.0 Reference Metadata messages."""
+
+from pathlib import Path
+from typing import Optional, Sequence, Union
+
+from pysdmx.io.json.sdmxjson2.writer._helper import write_metadata_msg
+from pysdmx.model import MetadataReport
+from pysdmx.model.message import Header
+
+
+def write(
+    reports: Sequence[MetadataReport],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+) -> Optional[str]:
+    """Write metadata reports in SDMX-JSON 2.0.0.
+
+    Args:
+        reports: The reference metadata reports to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    return write_metadata_msg(
+        reports, output_path, prettyprint, header, "2.0.0"
+    )

pysdmx/io/json/sdmxjson2/writer/v2_0/structure.py

@@ -0,0 +1,33 @@
+"""Writer interface for SDMX-JSON 2.0.0 Structure messages."""
+
+from pathlib import Path
+from typing import Optional, Sequence, Union
+
+from pysdmx.io.json.sdmxjson2.writer._helper import write_structure_msg
+from pysdmx.model.__base import MaintainableArtefact
+from pysdmx.model.message import Header
+
+
+def write(
+    structures: Sequence[MaintainableArtefact],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+) -> Optional[str]:
+    """Write maintainable SDMX artefacts in SDMX-JSON 2.0.0.
+
+    Args:
+        structures: The maintainable SDMX artefacts to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    return write_structure_msg(
+        structures, output_path, prettyprint, header, "2.0.0"
+    )

pysdmx/io/json/sdmxjson2/writer/v2_1/__init__.py

@@ -0,0 +1 @@
+"""Collection of writers for SDMX-JSON 2.1.0 messages."""

pysdmx/io/json/sdmxjson2/writer/v2_1/metadata.py

@@ -0,0 +1,31 @@
+"""Writer interface for SDMX-JSON 2.1.0 Reference Metadata messages."""
+
+from pathlib import Path
+from typing import Optional, Sequence, Union
+
+from pysdmx.io.json.sdmxjson2.writer._helper import write_metadata_msg
+from pysdmx.model import MetadataReport
+from pysdmx.model.message import Header
+
+
+def write(
+    reports: Sequence[MetadataReport],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+) -> Optional[str]:
+    """Write metadata reports in SDMX-JSON 2.1.0.
+
+    Args:
+        reports: The reference metadata reports to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    return write_metadata_msg(reports, output_path, prettyprint, header, "2.1")

pysdmx/io/json/sdmxjson2/writer/v2_1/structure.py

@@ -0,0 +1,33 @@
+"""Writer interface for SDMX-JSON 2.1.0 Structure messages."""
+
+from pathlib import Path
+from typing import Optional, Sequence, Union
+
+from pysdmx.io.json.sdmxjson2.writer._helper import write_structure_msg
+from pysdmx.model.__base import MaintainableArtefact
+from pysdmx.model.message import Header
+
+
+def write(
+    structures: Sequence[MaintainableArtefact],
+    output_path: Optional[Union[str, Path]] = None,
+    prettyprint: bool = True,
+    header: Optional[Header] = None,
+) -> Optional[str]:
+    """Write maintainable SDMX artefacts in SDMX-JSON 2.1.0.
+
+    Args:
+        structures: The maintainable SDMX artefacts to be serialized.
+        output_path: The path to save the JSON file. If None or empty, the
+            serialized content is returned as a string instead.
+        prettyprint: Whether to format the JSON output with indentation (True)
+            or output compact JSON without extra whitespace (False).
+        header: The header to be used in the SDMX-JSON message
+            (will be generated if no header is supplied).
+
+    Returns:
+        The JSON string if output_path is None or empty, None otherwise.
+    """
+    return write_structure_msg(
+        structures, output_path, prettyprint, header, "2.1"
+    )

pysdmx/io/reader.py

@@ -80,7 +80,10 @@ def read_sdmx(  # noqa: C901
         header = read_header(input_str, validate=validate)
         # SDMX-ML 3.1 Structure
         result_structures = read_structure(input_str, validate=validate)
-    elif read_format == Format.STRUCTURE_SDMX_JSON_2_0_0:
+    elif read_format in (
+        Format.STRUCTURE_SDMX_JSON_2_0_0,
+        Format.STRUCTURE_SDMX_JSON_2_1_0,
+    ):
         from pysdmx.io.json.sdmxjson2.reader.structure import (
             read as read_struct,
         )
@@ -90,7 +93,10 @@ def read_sdmx(  # noqa: C901
         result_structures = (
             struct_msg.structures if struct_msg.structures else []
         )
-    elif read_format == Format.REFMETA_SDMX_JSON_2_0_0:
+    elif read_format in (
+        Format.REFMETA_SDMX_JSON_2_0_0,
+        Format.REFMETA_SDMX_JSON_2_1_0,
+    ):
         from pysdmx.io.json.sdmxjson2.reader.metadata import (
             read as read_refmeta,
         )
@@ -174,7 +180,10 @@ def read_sdmx(  # noqa: C901
         return Message(header=header, data=result_data)
     elif read_format == Format.REGISTRY_SDMX_ML_2_1:
         return Message(header=header, submission=result_submission)
-    elif read_format == Format.REFMETA_SDMX_JSON_2_0_0:
+    elif read_format in (
+        Format.REFMETA_SDMX_JSON_2_0_0,
+        Format.REFMETA_SDMX_JSON_2_1_0,
+    ):
         return Message(header=header, reports=reports)
     # TODO: Ensure we have changed the signature of the structure readers
     return Message(header=header, structures=result_structures)

pysdmx/io/writer.py

@@ -24,10 +24,16 @@ WRITERS = {
     Format.DATA_SDMX_ML_3_1: "pysdmx.io.xml.sdmx31.writer.structure_specific",
     Format.STRUCTURE_SDMX_ML_3_1: "pysdmx.io.xml.sdmx31.writer.structure",
     Format.STRUCTURE_SDMX_JSON_2_0_0: (
-        "pysdmx.io.json.sdmxjson2.writer.structure"
+        "pysdmx.io.json.sdmxjson2.writer.v2_0.structure"
+    ),
+    Format.STRUCTURE_SDMX_JSON_2_1_0: (
+        "pysdmx.io.json.sdmxjson2.writer.v2_1.structure"
     ),
     Format.REFMETA_SDMX_JSON_2_0_0: (
-        "pysdmx.io.json.sdmxjson2.writer.metadata"
+        "pysdmx.io.json.sdmxjson2.writer.v2_0.metadata"
+    ),
+    Format.REFMETA_SDMX_JSON_2_1_0: (
+        "pysdmx.io.json.sdmxjson2.writer.v2_1.metadata"
     ),
 }
 
@@ -36,9 +42,13 @@ STRUCTURE_WRITERS = (
     Format.STRUCTURE_SDMX_ML_3_0,
     Format.STRUCTURE_SDMX_ML_3_1,
     Format.STRUCTURE_SDMX_JSON_2_0_0,
+    Format.STRUCTURE_SDMX_JSON_2_1_0,
 )
 
-REFMETA_WRITERS = (Format.REFMETA_SDMX_JSON_2_0_0,)
+REFMETA_WRITERS = (
+    Format.REFMETA_SDMX_JSON_2_0_0,
+    Format.REFMETA_SDMX_JSON_2_1_0,
+)
 
 
 def write_sdmx(

pysdmx/io/xml/__ss_aux_reader.py

@@ -1,7 +1,7 @@
 """SDMX XML StructureSpecificData reader aux module."""
 
 import itertools
-from typing import Any, Dict
+from typing import Any, Dict, List
 
 import numpy as np
 import pandas as pd
@@ -40,23 +40,22 @@ def _reading_str_series(dataset: Dict[str, Any]) -> pd.DataFrame:
     return df
 
 
-def _reading_group_data(dataset: Dict[str, Any]) -> pd.DataFrame:
+def _reading_group_data(dataset: Dict[str, Any]) -> List[pd.DataFrame]:
     # Structure Specific Group Data
-    test_list = []
-    df = None
+    group_dfs = []
     dataset[GROUP] = add_list(dataset[GROUP])
     for data in dataset[GROUP]:
-        test_list.append(dict(data.items()))
-        test_list, df = __process_df(test_list, df)
-    test_list, df = __process_df(test_list, df, is_end=True)
+        group_dict = dict(data.items())
+        group_df = pd.DataFrame([group_dict])
 
-    cols_to_delete = [x for x in df.columns if ":type" in x]
-    for x in cols_to_delete:
-        del df[x]
+        # Remove :type columns
+        cols_to_delete = [x for x in group_df.columns if ":type" in x]
+        for x in cols_to_delete:
+            del group_df[x]
 
-    df = df.drop_duplicates(keep="first").reset_index(drop=True)
+        group_dfs.append(group_df)
 
-    return df
+    return group_dfs
 
 
 def _get_at_att_str(dataset: Dict[str, Any]) -> Dict[str, Any]:
@@ -76,11 +75,34 @@ def _parse_structure_specific_data(
         # Structure Specific Series
         df = _reading_str_series(dataset)
         if GROUP in dataset:
-            df_group = _reading_group_data(dataset)
-            common_columns = list(
-                set(df.columns).intersection(set(df_group.columns))
-            )
-            df = pd.merge(df, df_group, on=common_columns, how="left")
+            group_dfs = _reading_group_data(dataset)
+            original_columns = df.columns.tolist()
+            for group_df in group_dfs:
+                # Find non-NaN columns in this group
+                non_nan_cols = [
+                    col
+                    for col in group_df.columns
+                    if not group_df[col].isna().all()
+                ]
+
+                # Merge keys are intersection of original and non-NaN cols
+                merge_cols = list(
+                    set(original_columns).intersection(set(non_nan_cols))
+                )
+
+                group_df = group_df.drop_duplicates(merge_cols, keep="first")
+                df = pd.merge(
+                    df,
+                    group_df,
+                    on=merge_cols,
+                    how="left",
+                    suffixes=("", "_drop"),
+                )
+                for col in list(df.columns):
+                    if col.endswith("_drop"):
+                        original = col[:-5]
+                        df[original] = df[original].fillna(df[col])
+                        df.drop(col, axis=1, inplace=True)
     elif OBS in dataset:
         dataset[OBS] = add_list(dataset[OBS])
         # Structure Specific All dimensions