U 襡c '@sddlZddlmZddlmZmZmZmZddlmZ ddl m Z ddl m Z ddlmZddZGd d d Zd d Zd ddhZdddZdddZdddZdS)N) _pandas_api)CodecTable concat_tablesschema)_feather) FeatherError)VersioncCstjtdkrtddS)Nz0.17.0z!feather requires pandas >= 0.17.0)rZ loose_versionr ImportErrorr r 3/tmp/pip-unpacked-wheel-qoi2rb4q/pyarrow/feather.py_check_pandas_versionsr c@s6eZdZdZd ddZd ddZdd Zdd d ZdS)FeatherDataseta Encapsulates details of reading a list of Feather files. Parameters ---------- path_or_paths : List[str] A list of file names validate_schema : bool, default True Check that individual file schemas are all the same / compatible TcCs||_||_dSN)pathsvalidate_schema)selfZ path_or_pathsrr r r __init__/szFeatherDataset.__init__NcCsjt|jd|d}|g|_|j|_|jddD].}t||d}|jrR||||j|q0t|jS)a, Read multiple feather files as a single pyarrow.Table Parameters ---------- columns : List[str] Names of columns to read from the file Returns ------- pyarrow.Table Content of the file as a table (of columns) rcolumnsN) read_tablerZ_tablesrrvalidate_schemasappendr)rrZ_filpathtabler r r r3s  zFeatherDataset.read_tablecCs(|j|js$td||j|jdS)Nz-Schema in {!s} was different. {!s} vs {!s})requals ValueErrorformat)rZpiecerr r r rLs zFeatherDataset.validate_schemascCst|j|dj|dS)a Read multiple Parquet files as a single pandas DataFrame Parameters ---------- columns : List[str] Names of columns to read from the file use_threads : bool, default True Use multiple threads when converting to pandas Returns ------- pandas.DataFrame Content of the file as a pandas DataFrame (of columns) r use_threadsr rZ to_pandas)rrr r r r read_pandasSs zFeatherDataset.read_pandas)T)N)NT)__name__ __module__ __qualname____doc__rrrr"r r r r r#s   rcCsN|jdkrdS|jttfkr4td|ntd|t|jdS)NrzqColumn '{}' exceeds 2GB maximum capacity of a Feather binary column. This restriction may be lifted in the futurezkColumn '{}' of type {} was chunked on conversion to Arrow and cannot be currently written to Feather format)Z num_chunkstypeextbinarystringrrstr)namecolr r r check_chunked_overflowhs r.lz4ZzstdZ uncompressedc Cstjr(ttjr(t|tjjr(|}t|r|dkr@d}n|dkrNd}nt dt j ||d}|dkrt |j jD]\}} ||} t| | qxn|}|dkrt|jtt|jkrt d|dk rt d|dk rt d n<|dkrtd rd }n$|dk r |tkr t d |tztj||||||d WnLtk rt|trzt|Wntjk r~YnXYnXdS)a Write a pandas.DataFrame to Feather format. Parameters ---------- df : pandas.DataFrame or pyarrow.Table Data to write out as Feather format. dest : str Local destination path. compression : string, default None Can be one of {"zstd", "lz4", "uncompressed"}. The default of None uses LZ4 for V2 files if it is available, otherwise uncompressed. compression_level : int, default None Use a compression level particular to the chosen compressor. If None use the default compression level chunksize : int, default None For V2 files, the internal maximum size of Arrow RecordBatch chunks when writing the Arrow IPC file format. None means use the default, which is currently 64K version : int, default 2 Feather file version. Version 2 is the current. Version 1 is the more limited legacy format rFr0Nz%Version value should either be 1 or 2)preserve_indexz'cannot serialize duplicate column namesz2Feather V1 files do not support compression optionz0Feather V1 files do not support chunksize optionZ lz4_framer/z1compression="{}" not supported, must be one of {}) compressioncompression_level chunksizeversion)rZ have_pandasr Z has_sparse isinstancepdZSparseDataFrameZto_denseZ is_data_framerrZ from_pandas enumeraternamesr.lenZ column_namessetrZ is_available_FEATHER_SUPPORTED_CODECSrr write_feather Exceptionr+osremoveerror) Zdfdestr2r3r4r5r1rir,r-r r r r=zs^       r=TFcCstt||||dj|dS)a  Read a pandas.DataFrame from Feather format. To read as pyarrow.Table use feather.read_table. Parameters ---------- source : str file path, or file-like object You can use MemoryMappedFile as source, for explicitly use memory map. columns : sequence, optional Only read a specific set of columns. If not provided, all columns are read. use_threads : bool, default True Whether to parallelize reading using multiple threads. If false the restriction is used in the conversion to Pandas as well as in the reading from Feather format. memory_map : boolean, default False Use memory mapping when opening file on disk, when source is a str. Returns ------- df : pandas.DataFrame )r memory_mapr rr!)sourcerr rDr r r read_feathersrFcCstj|||d}|dkr |Sdd|D}ttdd|rL||}nszread_table..cSs|tkSr)inttr r r zread_table..cSs|tkSr)r+rKr r r rM rNcSsg|] }|jqSr )r#)rGrLr r r rI szs      E  U