j_dZddlmZddlZddlZddlZddlmZmZm Z ddl m Z m Z ddl mZddlmZddlmZmZdd lmZdd lmZdd lmZmZdd lmZdd lmZmZm Z m!Z!m"Z"erddl#m$Z$m%Z%m&Z&m'Z'm(Z(m)Z)ddZ* d ddZ+GddZ,Gdde,Z-Gdde,Z. d ddZ/eddddej`dddf d dZ1y)!zparquet compat) annotationsN) TYPE_CHECKINGAnyLiteral)catch_warningsfilterwarnings)lib)import_optional_dependency)AbstractMethodErrorPandas4Warning) set_module)check_dtype_backend) DataFrame get_option)arrow_table_to_pandas) IOHandles get_handle is_fsspec_urlis_urlstringify_path) DtypeBackendFilePathParquetCompressionOptions ReadBufferStorageOptions WriteBufferBaseImplc$|dk(r td}|dk(r,ttg}d}|D] } |cStd||dk(r tS|dk(r tSt d #t$r}|dt |zz }Yd}~dd}~wwxYw) zreturn our implementationautozio.parquet.enginez - NzUnable to find a usable engine; tried using: 'pyarrow', 'fastparquet'. A suitable version of pyarrow or fastparquet is required for parquet support. Trying to import the above resulted in these errors:pyarrow fastparquetz.engine must be one of 'pyarrow', 'fastparquet')r PyArrowImplFastParquetImpl ImportErrorstr ValueError)engineengine_classes error_msgs engine_classerrs R/root/aria/tools/markitdown-venv/lib/python3.12/site-packages/pandas/io/parquet.py get_enginer.4s /0 %7 * 1L 1#~% 1  C l    } =   E FF% 1gC00  1sA++ B4B  Bc,t|}|tdd}tdd}|#t||jr |rOt d|!t||j j rn!tdt|jt|rk|i|5td}td} |jj|\}}|Mtd}|jj|fi|xsi\}}n|rt!|r|d k7r td d} |sN|sLt|t"r%>t%D"N :/9F!6!6!6"#2#8b" B &"8DDLSTTG  ~s + n-  D%   7B &&7r/  s7E;;FFc0eZdZeddZddZdddZy)rc:t|ts tdy)Nz+to_parquet only supports IO with DataFrames)r7rr')dfs r-validate_dataframezBaseImpl.validate_dataframes"i(JK K)c t|Nr )selfrQrD compressionkwargss r-writezBaseImpl.write !$''rSNc t|rUrV)rWrDcolumnsrYs r-readz BaseImpl.readr[rS)rQrreturnNonerU)r_r)r= __module__ __qualname__ staticmethodrRrZr^rSr-rrs LL((rSczeZdZddZ d ddZddej dddf ddZy) r#c<tddddl}ddl}||_y)Nr!z(pyarrow is required for parquet support.extrar)r pyarrow.parquet(pandas.core.arrays.arrow.extension_typesapi)rWr!pandass r-__init__zPyArrowImpl.__init__s!" G   8rSNc |j|d|jddi} ||| d<|jjj|fi| } |j rNdt j|j i} | jj} i| | } | j| } t|||d|du\}}}t|tjrmt|drat|j t"t$frAt|j t$r|j j'}n |j } |-|jj(j*| |f|||d|n+|jj(j,| |f||d|||j/yy#||j/wwxYw) Nschemapreserve_index PANDAS_ATTRSwb)r6rHrIname)rXpartition_cols filesystem)rXru)rRpoprkTable from_pandasattrsjsondumpsrometadatareplace_schema_metadatarNr7ioBufferedWriterhasattrrsr&bytesdecodeparquetwrite_to_dataset write_tableclose)rWrQrDrXindexr6rtrurYfrom_pandas_kwargstable df_metadataexisting_metadatamerged_metadatarJrMs r-rZzPyArrowImpl.writes #.6 8T8R-S  38 / 0***2D1CD 88)4::bhh+?@K % 5 5 B!2BkBO11/BE.A  +!- / + ~r'8'8 9/>..e =.--u5!/!4!4!;!;!=!/!4!4 )1  11"!,#1)   -  ,,"!,)   " #w" #s AF::Gc "d|d<t|||d\} } } |jjj| f|||d|} t 5t ddt t| ||} ddd| jjrKd | jjvr3| jjd } tj|  _  | | jSS#1swYxYw#| | jwwxYw) NTuse_pandas_metadatar3)r6rH)r]rufiltersr0make_block is deprecated) dtype_backendto_pandas_kwargss PANDAS_ATTRS)rNrkr read_tablerrr rror|rzloadsryr)rWrDr]rrr6rurrYrJrMpa_tableresultrs r-r^zPyArrowImpl.reads)-$%.A  + / +  2txx''22%   H ! ." /"/%5 ''"hoo&>&>>"*//":":?"KK#'::k#:FL" #%  $" #s$5C9 C-/A*C9-C62C99Dr_r`snappyNNNN)rQrrDzFilePath | WriteBuffer[bytes]rXrr bool | Noner6StorageOptions | Nonertlist[str] | Noner_r`)rDtypeBackend | lib.NoDefaultr6rrzdict[str, Any] | Noner_r)r=rarbrmrZr no_defaultr^rdrSr-r#r#s 2:!15+/@ @ ,@ / @  @ / @ )@  @ J69nn1526. 4 . / . 0.  . rSr#cTeZdZddZ d ddZ d d dZy) r$c,tdd}||_y)Nr"z,fastparquet is required for parquet support.rg)r rk)rWr"s r-rmzFastParquetImpl.__init__"s1 !O rSNc  |j|d|vr | tdd|vr|jd}|d|d<| tdt |}t |rt d fd|d<n r td td 5|jj||f|||d |dddy#1swYyxYw) N partition_onzYCannot use both partition_on and partition_cols. Use partition_cols for partitioning datahive file_scheme9filesystem is not implemented for the fastparquet engine.r2cPj|dfixsijS)Nrr)open)rD_r2r6s r-z'FastParquetImpl.write..Ms.+&++d3.4"3dfrS open_withz?storage_options passed with file object or non-fsspec file pathT)record)rX write_indexr) rRr'rvr9rrr rrkrZ) rWrQrDrXrrtr6rurYr2s ` @r-rZzFastParquetImpl.write*s # V #(BK  V ##ZZ7N  %$*F= !  !%K  d#  /9F#F; Q 4 (  DHHNN (!+        s #B>>Cc ,i}|jdtj} d|d<| tjur td| t d| t dt |}d} t |r1td} | j|dfi|xsij|d <nJt|tr:tjj|st|dd| } | j } |j"j$|fi|} t'5t)d d t*| j,d||d |cddd| | j/SS#1swYnxYw | | j/yy#| | j/wwxYw)NrF pandas_nullszHThe 'dtype_backend' argument is not supported for the fastparquet enginerz?to_pandas_kwargs is not implemented for the fastparquet engine.r2r3rGr4r0r)r]rrd)rvr rr'r9rrr rrGr7r&rCrDrErrFrk ParquetFilerrr to_pandasr) rWrDr]rr6rurrYparquet_kwargsrrMr2 parquet_files r-r^zFastParquetImpl.read_s*, ?CNNC ).~&  .%   !%K   '%Q d#  /9F#.6;;tT#Uo>SQS#U#X#XN4 c "277==+>!dE?G>>D /488//GGL! ." .|--#W8>   " #   " #w" #s$3'E>&E E>E&"E>>Frr)rQrrXz*Literal['snappy', 'gzip', 'brotli'] | Noner6rr_r`)NNNNN)r6rr dict | Noner_r)r=rarbrmrZr^rdrSr-r$r$!s{CK153 3@ 3/3 3p15(,7 / 7 &7  7 rSr$rc t|tr|g}t|} |tjn|} | j || f|||||d||,t| tjsJ| j Sy)a Write a DataFrame to the parquet format. Parameters ---------- df : DataFrame path : str, path object, file-like object, or None, default None String, path object (implementing ``os.PathLike[str]``), or file-like object implementing a binary ``write()`` function. If None, the result is returned as bytes. If a string, it will be used as Root Directory path when writing a partitioned dataset. The engine fastparquet does not accept file-like objects. engine : {'auto', 'pyarrow', 'fastparquet'}, default 'auto' Parquet library to use. If 'auto', then the option ``io.parquet.engine`` is used. The default ``io.parquet.engine`` behavior is to try 'pyarrow', falling back to 'fastparquet' if 'pyarrow' is unavailable. When using the ``'pyarrow'`` engine and no storage options are provided and a filesystem is implemented by both ``pyarrow.fs`` and ``fsspec`` (e.g. "s3://"), then the ``pyarrow.fs`` filesystem is attempted first. Use the filesystem keyword with an instantiated fsspec filesystem if you wish to use its implementation. compression : {'snappy', 'gzip', 'brotli', 'lz4', 'zstd', None}, default 'snappy'. Name of the compression to use. Use ``None`` for no compression. index : bool, default None If ``True``, include the dataframe's index(es) in the file output. If ``False``, they will not be written to the file. If ``None``, similar to ``True`` the dataframe's index(es) will be saved. However, instead of being saved as values, the RangeIndex will be stored as a range in the metadata so it doesn't require much space and is faster. Other indexes will be included as columns in the file output. partition_cols : str or list, optional, default None Column names by which to partition the dataset. Columns are partitioned in the order they are given. Must be None if path is not a string. storage_options : dict, optional Extra options that make sense for a particular storage connection, e.g. host, port, username, password, etc. For HTTP(S) URLs the key-value pairs are forwarded to ``urllib.request.Request`` as header options. For other URLs (e.g. starting with "s3://", and "gcs://") the key-value pairs are forwarded to ``fsspec.open``. Please see ``fsspec`` and ``urllib`` for more details, and for more examples on storage options refer `here `_. filesystem : fsspec or pyarrow filesystem, default None Filesystem object to use when reading the parquet file. Only implemented for ``engine="pyarrow"``. .. versionadded:: 2.1.0 **kwargs Additional keyword arguments passed to the engine: * For ``engine="pyarrow"``: passed to :func:`pyarrow.parquet.write_table` or :func:`pyarrow.parquet.write_to_dataset` (when using partition_cols) * For ``engine="fastparquet"``: passed to :func:`fastparquet.write` Returns ------- bytes if no path argument is provided else None N)rXrrtr6ru)r7r&r.r~BytesIOrZgetvalue) rQrDr(rXrr6rtrurYimpl path_or_bufs r- to_parquetrsV.#&() f DAESWKDJJ   %'    |+rzz222##%%rSrlc bt|} t|| j|f||||||d|S)aE Load a parquet object from the file path, returning a DataFrame. The function automatically handles reading the data from a parquet file and creates a DataFrame with the appropriate structure. Parameters ---------- path : str, path object or file-like object String, path object (implementing ``os.PathLike[str]``), or file-like object implementing a binary ``read()`` function. The string could be a URL. Valid URL schemes include http, ftp, s3, gs, and file. For file URLs, a host is expected. A local file could be: ``file://localhost/path/to/table.parquet``. A file URL can also be a path to a directory that contains multiple partitioned parquet files. Both pyarrow and fastparquet support paths to directories as well as file URLs. A directory path could be: ``file://localhost/path/to/tables`` or ``s3://bucket/partition_dir``. engine : {'auto', 'pyarrow', 'fastparquet'}, default 'auto' Parquet library to use. If 'auto', then the option ``io.parquet.engine`` is used. The default ``io.parquet.engine`` behavior is to try 'pyarrow', falling back to 'fastparquet' if 'pyarrow' is unavailable. When using the ``'pyarrow'`` engine and no storage options are provided and a filesystem is implemented by both ``pyarrow.fs`` and ``fsspec`` (e.g. "s3://"), then the ``pyarrow.fs`` filesystem is attempted first. Use the filesystem keyword with an instantiated fsspec filesystem if you wish to use its implementation. columns : list, default=None If not None, only these columns will be read from the file. storage_options : dict, optional Extra options that make sense for a particular storage connection, e.g. host, port, username, password, etc. For HTTP(S) URLs the key-value pairs are forwarded to ``urllib.request.Request`` as header options. For other URLs (e.g. starting with "s3://", and "gcs://") the key-value pairs are forwarded to ``fsspec.open``. Please see ``fsspec`` and ``urllib`` for more details, and for more examples on storage options refer `here `_. dtype_backend : {'numpy_nullable', 'pyarrow'} Back-end data type applied to the resultant :class:`DataFrame` (still experimental). If not specified, the default behavior is to not use nullable data types. If specified, the behavior is as follows: * ``"numpy_nullable"``: returns nullable-dtype-backed :class:`DataFrame` * ``"pyarrow"``: returns pyarrow-backed nullable :class:`ArrowDtype` :class:`DataFrame` .. versionadded:: 2.0 filesystem : fsspec or pyarrow filesystem, default None Filesystem object to use when reading the parquet file. Only implemented for ``engine="pyarrow"``. .. versionadded:: 2.1.0 filters : List[Tuple] or List[List[Tuple]], default None To filter out data. Filter syntax: [[(column, op, val), ...],...] where op is [==, =, >, >=, <, <=, !=, in, not in] The innermost tuples are transposed into a set of filters applied through an `AND` operation. The outer list combines these sets of filters through an `OR` operation. A single list of tuples can also be used, meaning that no `OR` operation between set of filters is to be conducted. Using this argument will NOT result in row-wise filtering of the final partitions unless ``engine="pyarrow"`` is also specified. For other engines, filtering is only performed at the partition level, that is, to prevent the loading of some row-groups and/or files. .. versionadded:: 2.1.0 to_pandas_kwargs : dict | None, default None Keyword arguments to pass through to :func:`pyarrow.Table.to_pandas` when ``engine="pyarrow"``. .. versionadded:: 3.0.0 **kwargs Additional keyword arguments passed to the engine: * For ``engine="pyarrow"``: passed to :func:`pyarrow.parquet.read_table` * For ``engine="fastparquet"``: passed to :meth:`fastparquet.ParquetFile.to_pandas` Returns ------- DataFrame DataFrame based on parquet file. See Also -------- DataFrame.to_parquet : Create a parquet object that serializes a DataFrame. Examples -------- >>> original_df = pd.DataFrame({"foo": range(5), "bar": range(5, 10)}) >>> original_df foo bar 0 0 5 1 1 6 2 2 7 3 3 8 4 4 9 >>> df_parquet_bytes = original_df.to_parquet() >>> from io import BytesIO >>> restored_df = pd.read_parquet(BytesIO(df_parquet_bytes)) >>> restored_df foo bar 0 0 5 1 1 6 2 2 7 3 3 8 4 4 9 >>> restored_df.equals(original_df) True >>> restored_bar = pd.read_parquet(BytesIO(df_parquet_bytes), columns=["bar"]) >>> restored_bar bar 0 5 1 6 2 7 3 8 4 9 >>> restored_bar.equals(original_df[["bar"]]) True The function uses `kwargs` that are passed directly to the engine. In the following example, we use the `filters` argument of the pyarrow engine to filter the rows of the DataFrame. Since `pyarrow` is the default engine, we can omit the `engine` argument. Note that the `filters` argument is implemented by the `pyarrow` engine, which can benefit from multithreading and also potentially be more economical in terms of memory. >>> sel = [("foo", ">", 2)] >>> restored_part = pd.read_parquet(BytesIO(df_parquet_bytes), filters=sel) >>> restored_part foo bar 0 3 8 1 4 9 )r]rr6rrur)r.rr^) rDr(r]r6rrurrrYrs r- read_parquetrsN@ f D & 499  '#)    rS)r(r&r_r)Nr3F) rDz1FilePath | ReadBuffer[bytes] | WriteBuffer[bytes]rGrr6rrHr&rIboolr_zVtuple[FilePath | ReadBuffer[bytes] | WriteBuffer[bytes], IOHandles[bytes] | None, Any])NrrNNNN)rQrrDz$FilePath | WriteBuffer[bytes] | Noner(r&rXrrrr6rrtrrurr_z bytes | None)rDzFilePath | ReadBuffer[bytes]r(r&r]rr6rrrrurrz&list[tuple] | list[list[tuple]] | Nonerrr_r)2__doc__ __future__rr~rzrCtypingrrrwarningsrr pandas._libsr pandas.compat._optionalr pandas.errorsr r pandas.util._decoratorsr pandas.util._validatorsrrlrrpandas.io._utilrpandas.io.commonrrrrrpandas._typingrrrrrrr.rNrr#r$rrrrdrSr-rs"   >/7 2GJ.2 <' ;<' <'+<'  <'  <'  <'~ ( (| (| ~u hu t26-5-1'+`` .` `+ `  ` + `%```F H $-125..6:$(k &k kk+ k 0 k  k4k"kkkrS