deltacat 2.0.0b11__py3-none-any.whl → 2.0.0b12__py3-none-any.whl

deltacat/utils/polars.py

@@ -1,58 +1,156 @@
 import logging
-from typing import Optional, List, Dict, Callable, Union
+import bz2
+import gzip
+from functools import partial
+from typing import Optional, List, Dict, Callable, Union, Iterable, Any
 
 import polars as pl
+import pyarrow as pa
+import pyarrow.fs as pafs
 
 from fsspec import AbstractFileSystem
 from ray.data.datasource import FilenameProvider
 
 from deltacat import logs
+from deltacat.utils.filesystem import resolve_path_and_filesystem
+from deltacat.utils.common import ContentTypeKwargsProvider, ReadKwargsProvider
+from deltacat.utils.performance import timed_invocation
 
-from deltacat.types.media import ContentType
+from deltacat.types.media import (
+    ContentType,
+    ContentEncoding,
+    DELIMITED_TEXT_CONTENT_TYPES,
+    TABULAR_CONTENT_TYPES,
+)
+from deltacat.types.partial_download import PartialFileDownloadParams
 
 logger = logs.configure_deltacat_logger(logging.getLogger(__name__))
 
+# Encoding to file initialization function mapping
+ENCODING_TO_FILE_INIT: Dict[str, Callable] = {
+    ContentEncoding.GZIP.value: partial(gzip.open, mode="rb"),
+    ContentEncoding.BZIP2.value: partial(bz2.open, mode="rb"),
+    ContentEncoding.IDENTITY.value: lambda file_path: file_path,
+}
+
 
 def write_json(
     table: pl.DataFrame,
     path: str,
     *,
-    filesystem: Optional[AbstractFileSystem] = None,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
     fs_open_kwargs: Dict[str, any] = {},
     **write_kwargs,
 ) -> None:
-    if not filesystem:
-        table.write_ndjson(path, **write_kwargs)
+    # Check if the path already indicates compression to avoid double compression
+    should_compress = path.endswith(".gz")
+
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path, filesystem)
+        with filesystem.open_output_stream(path, **fs_open_kwargs) as f:
+            if should_compress:
+                # Path ends with .gz, PyArrow filesystem automatically compresses
+                table.write_ndjson(f, **write_kwargs)
+            else:
+                # No compression indicated, write uncompressed
+                table.write_ndjson(f, **write_kwargs)
     else:
         with filesystem.open(path, "wb", **fs_open_kwargs) as f:
-            table.write_ndjson(f, **write_kwargs)
+            if should_compress:
+                # For fsspec filesystems, we need to apply compression explicitly
+                with pa.CompressedOutputStream(f, ContentEncoding.GZIP.value) as out:
+                    table.write_ndjson(out, **write_kwargs)
+            else:
+                # No compression indicated, write uncompressed
+                table.write_ndjson(f, **write_kwargs)
+
+
+def content_type_to_writer_kwargs(content_type: str) -> Dict[str, any]:
+    """
+    Returns writer kwargs for the given content type when writing with polars.
+    """
+    if content_type == ContentType.UNESCAPED_TSV.value:
+        return {
+            "separator": "\t",
+            "include_header": False,
+            "null_value": "",
+            "quote_style": "never",  # Equivalent to QUOTE_NONE in pandas
+        }
+    if content_type == ContentType.TSV.value:
+        return {
+            "separator": "\t",
+            "include_header": False,
+            "quote_style": "necessary",
+        }
+    if content_type == ContentType.CSV.value:
+        return {
+            "separator": ",",
+            "include_header": False,
+            "quote_style": "necessary",
+        }
+    if content_type == ContentType.PSV.value:
+        return {
+            "separator": "|",
+            "include_header": False,
+            "quote_style": "necessary",
+        }
+    if content_type in {
+        ContentType.PARQUET.value,
+        ContentType.FEATHER.value,
+        ContentType.JSON.value,
+        ContentType.AVRO.value,
+        ContentType.ORC.value,
+    }:
+        return {}
+    raise ValueError(f"Unsupported content type: {content_type}")
 
 
 def write_csv(
     table: pl.DataFrame,
     path: str,
     *,
-    filesystem: Optional[AbstractFileSystem] = None,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
     fs_open_kwargs: Dict[str, any] = {},
-    **write_kwargs,
+    **kwargs,
 ) -> None:
-    if not filesystem:
-        table.write_csv(path, **write_kwargs)
+    """
+    Write a polars DataFrame to a CSV file (or other delimited text format).
+    """
+    # Check if the path already indicates compression to avoid double compression
+    should_compress = path.endswith(".gz")
+
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path, filesystem)
+        with filesystem.open_output_stream(path, **fs_open_kwargs) as f:
+            if should_compress:
+                # Path ends with .gz, PyArrow filesystem automatically compresses
+                table.write_csv(f, **kwargs)
+            else:
+                # No compression indicated, write uncompressed
+                table.write_csv(f, **kwargs)
     else:
         with filesystem.open(path, "wb", **fs_open_kwargs) as f:
-            table.write_csv(f, **write_kwargs)
+            if should_compress:
+                # For fsspec filesystems, we need to apply compression explicitly
+                with pa.CompressedOutputStream(f, ContentEncoding.GZIP.value) as out:
+                    table.write_csv(out, **kwargs)
+            else:
+                # No compression indicated, write uncompressed
+                table.write_csv(f, **kwargs)
 
 
 def write_avro(
     table: pl.DataFrame,
     path: str,
     *,
-    filesystem: Optional[AbstractFileSystem] = None,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
     fs_open_kwargs: Dict[str, any] = {},
     **write_kwargs,
 ) -> None:
-    if not filesystem:
-        table.write_avro(path, **write_kwargs)
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path, filesystem)
+        with filesystem.open_output_stream(path, **fs_open_kwargs) as f:
+            table.write_avro(f, **write_kwargs)
     else:
         with filesystem.open(path, "wb", **fs_open_kwargs) as f:
             table.write_avro(f, **write_kwargs)
@@ -62,25 +160,75 @@ def write_parquet(
     table: pl.DataFrame,
     path: str,
     *,
-    filesystem: Optional[AbstractFileSystem] = None,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
     fs_open_kwargs: Dict[str, any] = {},
     **write_kwargs,
 ) -> None:
-    if not filesystem:
-        table.write_parquet(path, **write_kwargs)
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path, filesystem)
+        with filesystem.open_output_stream(path, **fs_open_kwargs) as f:
+            table.write_parquet(f, **write_kwargs)
     else:
         with filesystem.open(path, "wb", **fs_open_kwargs) as f:
             table.write_parquet(f, **write_kwargs)
 
 
+def write_feather(
+    table: pl.DataFrame,
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    **kwargs,
+) -> None:
+    """
+    Write a polars DataFrame to a Feather file.
+    """
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path, filesystem)
+        with filesystem.open_output_stream(path, **fs_open_kwargs) as f:
+            table.write_ipc(f, **kwargs)
+    else:
+        with filesystem.open(path, "wb", **fs_open_kwargs) as f:
+            table.write_ipc(f, **kwargs)
+
+
+def write_orc(
+    table: pl.DataFrame,
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    **write_kwargs,
+) -> None:
+    """
+    Write a polars DataFrame to an ORC file by delegating to PyArrow implementation.
+    """
+    from deltacat.utils.pyarrow import write_orc as pyarrow_write_orc
+
+    # Convert polars DataFrame to PyArrow Table
+    pa_table = table.to_arrow()
+
+    # Delegate to PyArrow write_orc implementation
+    pyarrow_write_orc(
+        pa_table,
+        path,
+        filesystem=filesystem,
+        fs_open_kwargs=fs_open_kwargs,
+        **write_kwargs,
+    )
+
+
 CONTENT_TYPE_TO_PL_WRITE_FUNC: Dict[str, Callable] = {
-    # TODO (pdames): add support for other delimited text content types as
-    #  pyarrow adds support for custom delimiters, escaping, and None value
-    #  representations to pyarrow.csv.WriteOptions.
-    ContentType.AVRO.value: write_avro,
+    ContentType.UNESCAPED_TSV.value: write_csv,
+    ContentType.TSV.value: write_csv,
     ContentType.CSV.value: write_csv,
+    ContentType.PSV.value: write_csv,
     ContentType.PARQUET.value: write_parquet,
+    ContentType.FEATHER.value: write_feather,
     ContentType.JSON.value: write_json,
+    ContentType.AVRO.value: write_avro,
+    ContentType.ORC.value: write_orc,
 }
 
 
@@ -108,21 +256,504 @@ def dataframe_size(table: pl.DataFrame) -> int:
 def dataframe_to_file(
     table: pl.DataFrame,
     base_path: str,
-    file_system: Optional[AbstractFileSystem],
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]],
     block_path_provider: Union[Callable, FilenameProvider],
     content_type: str = ContentType.PARQUET.value,
+    schema: Optional[pa.Schema] = None,
     **kwargs,
 ) -> None:
     """
-    Writes the given Pyarrow Table to a file.
+    Writes the given Polars DataFrame to a file.
     """
     writer = CONTENT_TYPE_TO_PL_WRITE_FUNC.get(content_type)
+    writer_kwargs = content_type_to_writer_kwargs(content_type)
+    writer_kwargs.update(kwargs)
     if not writer:
         raise NotImplementedError(
-            f"Pyarrow writer for content type '{content_type}' not "
+            f"Polars writer for content type '{content_type}' not "
             f"implemented. Known content types: "
-            f"{CONTENT_TYPE_TO_PL_WRITE_FUNC.keys}"
+            f"{CONTENT_TYPE_TO_PL_WRITE_FUNC.keys()}"
         )
     path = block_path_provider(base_path)
-    logger.debug(f"Writing table: {table} with kwargs: {kwargs} to path: {path}")
-    writer(table, path, filesystem=file_system, **kwargs)
+    logger.debug(f"Writing table: {table} with kwargs: {writer_kwargs} to path: {path}")
+    writer(table, path, filesystem=filesystem, **writer_kwargs)
+
+
+def write_table(
+    table: pl.DataFrame,
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_type: str = ContentType.PARQUET.value,
+    **kwargs,
+) -> None:
+    """
+    Write a polars DataFrame to a file in the specified format.
+    """
+    writer = CONTENT_TYPE_TO_PL_WRITE_FUNC.get(content_type)
+    writer_kwargs = content_type_to_writer_kwargs(content_type)
+    writer_kwargs.update(kwargs)
+    if not writer:
+        raise NotImplementedError(
+            f"Polars writer for content type '{content_type}' not "
+            f"implemented. Known content types: "
+            f"{CONTENT_TYPE_TO_PL_WRITE_FUNC.keys()}"
+        )
+    writer(
+        table,
+        path,
+        filesystem=filesystem,
+        fs_open_kwargs=fs_open_kwargs,
+        **writer_kwargs,
+    )
+
+
+CONTENT_TYPE_TO_PL_READ_FUNC: Dict[str, Callable] = {
+    ContentType.UNESCAPED_TSV.value: pl.read_csv,
+    ContentType.TSV.value: pl.read_csv,
+    ContentType.CSV.value: pl.read_csv,
+    ContentType.PSV.value: pl.read_csv,
+    ContentType.PARQUET.value: pl.read_parquet,
+    ContentType.FEATHER.value: pl.read_ipc,
+    ContentType.JSON.value: pl.read_ndjson,
+    ContentType.AVRO.value: pl.read_avro,
+}
+
+
+class ReadKwargsProviderPolarsStringTypes(ContentTypeKwargsProvider):
+    """ReadKwargsProvider impl that reads columns of delimited text files
+    as UTF-8 strings (i.e. disables type inference). Useful for ensuring
+    lossless reads of UTF-8 delimited text datasets and improving read
+    performance in cases where type casting is not required."""
+
+    def __init__(self, include_columns: Optional[Iterable[str]] = None):
+        self.include_columns = include_columns
+
+    def _get_kwargs(self, content_type: str, kwargs: Dict[str, Any]) -> Dict[str, Any]:
+        if content_type in DELIMITED_TEXT_CONTENT_TYPES:
+            include_columns = (
+                self.include_columns if self.include_columns else kwargs.get("columns")
+            )
+            if not include_columns:
+                # read all columns as strings - disable schema inference
+                kwargs["infer_schema"] = False
+            else:
+                # read only the included columns as strings
+                kwargs["schema_overrides"] = {
+                    column_name: pl.Utf8 for column_name in include_columns
+                }
+        return kwargs
+
+
+def content_type_to_reader_kwargs(content_type: str) -> Dict[str, Any]:
+    if content_type == ContentType.UNESCAPED_TSV.value:
+        return {
+            "separator": "\t",
+            "has_header": False,
+            "null_values": [""],
+            "quote_char": None,
+        }
+    if content_type == ContentType.TSV.value:
+        return {"separator": "\t", "has_header": False}
+    if content_type == ContentType.CSV.value:
+        return {"separator": ",", "has_header": False}
+    if content_type == ContentType.PSV.value:
+        return {"separator": "|", "has_header": False}
+    if content_type in {
+        ContentType.PARQUET.value,
+        ContentType.FEATHER.value,
+        ContentType.ORC.value,
+        ContentType.JSON.value,
+        ContentType.AVRO.value,
+    }:
+        return {}
+    raise ValueError(f"Unsupported content type: {content_type}")
+
+
+def _add_column_kwargs(
+    content_type: str,
+    column_names: Optional[List[str]],
+    include_columns: Optional[List[str]],
+    kwargs: Dict[str, Any],
+):
+    if content_type in DELIMITED_TEXT_CONTENT_TYPES:
+        if column_names:
+            kwargs["new_columns"] = column_names
+        if include_columns:
+            kwargs["columns"] = include_columns
+    else:
+        if content_type in TABULAR_CONTENT_TYPES:
+            if include_columns:
+                kwargs["columns"] = include_columns
+        else:
+            if include_columns:
+                logger.warning(
+                    f"Ignoring request to include columns {include_columns} "
+                    f"for non-tabular content type {content_type}"
+                )
+
+
+def concat_dataframes(dataframes: List[pl.DataFrame]) -> Optional[pl.DataFrame]:
+    if dataframes is None or not len(dataframes):
+        return None
+    if len(dataframes) == 1:
+        return next(iter(dataframes))
+    return pl.concat(dataframes)
+
+
+def append_column_to_table(
+    table: pl.DataFrame,
+    column_name: str,
+    column_value: Any,
+) -> pl.DataFrame:
+    return table.with_columns(pl.lit(column_value).alias(column_name))
+
+
+def select_columns(
+    table: pl.DataFrame,
+    column_names: List[str],
+) -> pl.DataFrame:
+    return table.select(column_names)
+
+
+def file_to_dataframe(
+    path: str,
+    content_type: str,
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    column_names: Optional[List[str]] = None,
+    include_columns: Optional[List[str]] = None,
+    pl_read_func_kwargs_provider: Optional[ReadKwargsProvider] = None,
+    partial_file_download_params: Optional[PartialFileDownloadParams] = None,
+    fs_open_kwargs: Dict[str, Any] = {},
+    **kwargs,
+) -> pl.DataFrame:
+    """
+    Read a file into a Polars DataFrame using any filesystem.
+
+    Args:
+        path: The file path to read
+        content_type: The content type of the file (e.g., ContentType.CSV.value)
+        content_encoding: The content encoding (default: IDENTITY)
+        filesystem: The filesystem to use (if None, will be inferred from path)
+        column_names: Optional column names to assign
+        include_columns: Optional columns to include in the result
+        pl_read_func_kwargs_provider: Optional kwargs provider for customization
+        fs_open_kwargs: Optional kwargs for filesystem open operations
+        **kwargs: Additional kwargs passed to the reader function
+
+    Returns:
+        pl.DataFrame: The loaded DataFrame
+    """
+    logger.debug(
+        f"Reading {path} to Polars. Content type: {content_type}. "
+        f"Encoding: {content_encoding}"
+    )
+
+    pl_read_func = CONTENT_TYPE_TO_READ_FN.get(content_type)
+    if not pl_read_func:
+        raise NotImplementedError(
+            f"Polars reader for content type '{content_type}' not "
+            f"implemented. Known content types: "
+            f"{list(CONTENT_TYPE_TO_READ_FN.keys())}"
+        )
+
+    reader_kwargs = content_type_to_reader_kwargs(content_type)
+    _add_column_kwargs(content_type, column_names, include_columns, reader_kwargs)
+
+    # Merge with provided kwargs
+    reader_kwargs.update(kwargs)
+
+    if pl_read_func_kwargs_provider:
+        reader_kwargs = pl_read_func_kwargs_provider(content_type, reader_kwargs)
+
+    logger.debug(f"Reading {path} via {pl_read_func} with kwargs: {reader_kwargs}")
+
+    dataframe, latency = timed_invocation(
+        pl_read_func,
+        path,
+        filesystem=filesystem,
+        fs_open_kwargs=fs_open_kwargs,
+        content_encoding=content_encoding,
+        **reader_kwargs,
+    )
+    logger.debug(f"Time to read {path} into Polars DataFrame: {latency}s")
+    return dataframe
+
+
+def read_csv(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        if content_encoding == ContentEncoding.IDENTITY.value:
+            with filesystem.open_input_stream(path, **fs_open_kwargs) as f:
+                return pl.read_csv(f, **read_kwargs)
+        else:
+            # For compressed files with PyArrow, we need to be careful because PyArrow
+            # may auto-decompress some formats. Try to read directly first.
+            try:
+                with filesystem.open_input_stream(path, **fs_open_kwargs) as f:
+                    # Try reading as if it's already decompressed by PyArrow
+                    return pl.read_csv(f, **read_kwargs)
+            except Exception:
+                # If that fails, try manual decompression
+                with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+                    input_file_init = ENCODING_TO_FILE_INIT.get(
+                        content_encoding, lambda x: x
+                    )
+                    with input_file_init(f) as input_file:
+                        content = input_file.read()
+                        if isinstance(content, str):
+                            content = content.encode("utf-8")
+                        return pl.read_csv(content, **read_kwargs)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_csv(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    if isinstance(content, str):
+                        content = content.encode("utf-8")
+                    return pl.read_csv(content, **read_kwargs)
+
+
+def read_parquet(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_parquet(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_parquet(content, **read_kwargs)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_parquet(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_parquet(content, **read_kwargs)
+
+
+def read_ipc(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_ipc(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_ipc(content, **read_kwargs)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_ipc(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_ipc(content, **read_kwargs)
+
+
+def read_ndjson(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        if content_encoding == ContentEncoding.IDENTITY.value:
+            with filesystem.open_input_stream(path, **fs_open_kwargs) as f:
+                return pl.read_ndjson(f, **read_kwargs)
+        else:
+            # For compressed files with PyArrow, we need to be careful because PyArrow
+            # may auto-decompress some formats. Try to read directly first.
+            try:
+                with filesystem.open_input_stream(path, **fs_open_kwargs) as f:
+                    # Try reading as if it's already decompressed by PyArrow
+                    return pl.read_ndjson(f, **read_kwargs)
+            except Exception:
+                # If that fails, try manual decompression
+                with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+                    input_file_init = ENCODING_TO_FILE_INIT.get(
+                        content_encoding, lambda x: x
+                    )
+                    with input_file_init(f) as input_file:
+                        content = input_file.read()
+                        if isinstance(content, str):
+                            content = content.encode("utf-8")
+                        return pl.read_ndjson(content, **read_kwargs)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_ndjson(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    if isinstance(content, str):
+                        content = content.encode("utf-8")
+                    return pl.read_ndjson(content, **read_kwargs)
+
+
+def read_avro(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_avro(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_avro(content, **read_kwargs)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                return pl.read_avro(f, **read_kwargs)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content as bytes and pass to polars
+                    content = input_file.read()
+                    return pl.read_avro(content, **read_kwargs)
+
+
+def read_orc(
+    path: str,
+    *,
+    filesystem: Optional[Union[AbstractFileSystem, pafs.FileSystem]] = None,
+    fs_open_kwargs: Dict[str, any] = {},
+    content_encoding: str = ContentEncoding.IDENTITY.value,
+    **read_kwargs,
+) -> pl.DataFrame:
+    """
+    Read an ORC file using pandas and convert to polars since polars doesn't have native ORC support.
+    """
+    import pandas as pd
+
+    if not filesystem or isinstance(filesystem, pafs.FileSystem):
+        path, filesystem = resolve_path_and_filesystem(path)
+        with filesystem.open_input_file(path, **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                pd_df = pd.read_orc(f, **read_kwargs)
+                return pl.from_pandas(pd_df)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content and pass to pandas
+                    content = input_file.read()
+                    import io
+
+                    pd_df = pd.read_orc(io.BytesIO(content), **read_kwargs)
+                    return pl.from_pandas(pd_df)
+    else:
+        # fsspec AbstractFileSystem
+        with filesystem.open(path, "rb", **fs_open_kwargs) as f:
+            # Handle compression
+            if content_encoding == ContentEncoding.IDENTITY.value:
+                pd_df = pd.read_orc(f, **read_kwargs)
+                return pl.from_pandas(pd_df)
+            else:
+                input_file_init = ENCODING_TO_FILE_INIT.get(
+                    content_encoding, lambda x: x
+                )
+                with input_file_init(f) as input_file:
+                    # Read decompressed content and pass to pandas
+                    content = input_file.read()
+                    import io
+
+                    pd_df = pd.read_orc(io.BytesIO(content), **read_kwargs)
+                    return pl.from_pandas(pd_df)
+
+
+# New mapping for encoding-aware reader functions used by file_to_dataframe
+CONTENT_TYPE_TO_READ_FN: Dict[str, Callable] = {
+    ContentType.UNESCAPED_TSV.value: read_csv,
+    ContentType.TSV.value: read_csv,
+    ContentType.CSV.value: read_csv,
+    ContentType.PSV.value: read_csv,
+    ContentType.PARQUET.value: read_parquet,
+    ContentType.FEATHER.value: read_ipc,
+    ContentType.JSON.value: read_ndjson,
+    ContentType.AVRO.value: read_avro,
+    ContentType.ORC.value: read_orc,
+}