'e-ddlmZddlmZmZ ddlmZmZmZm Z m Z m Z m Z m Z n&#e$rZedeedddZ[wwxYw ddlmZddlmZn#e$rGdd ZeZYnwxYwdd Zddd d efd ZdZefdZddZdS))Table) Expressionfield) DeclarationExecNodeOptionsTableSourceNodeOptionsFilterNodeOptionsProjectNodeOptionsAggregateNodeOptionsOrderByNodeOptionsHashJoinNodeOptionsz@The pyarrow installation is not built with support for 'acero' ()N)ScanNodeOptionscBeZdZGddZGddZdS)DatasetModuleStubceZdZdS)DatasetModuleStub.DatasetN__name__ __module__ __qualname__-/croot/pyarrow_1707330824290/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold/lib/python3.11/site-packages/pyarrow/acero.pyDatasetr2 DrrceZdZdS)!DatasetModuleStub.InMemoryDatasetNrrrrInMemoryDatasetr5rrrN)rrrrrrrrrr1sd                  rrTc ltdt||}d|jjD}tj|tdt |g}|jd}|1tj|tdt|g}|S)Nscan use_threadsc,g|]}t|Sr)r).0fs r z$_dataset_to_decl..?s:::588:::rprojectfilter) rrschemanames from_sequencer _scan_optionsgetr )datasetr#decl projections filter_exprs r_dataset_to_declr3:s vwKPPP Q QD;:W^%9:::K  $ {9&8&E&EFFG  D'++H55K( ;x):;)G)GHH I   KrFc t|ttjfst dt |t|ttjfst dt |it|t tfs|g}t|D] \} } | | < it|tt fs|g}t|D] \} } | | < |j j } |j j } |dks|dkrg} n>|dks|dkrg} n/|dks|dkrfd| D} n|d krfd | D} i}t| D]\} }||vr| ||<i}t| D]\} }||vr| ||<t|tjrt|| }ntd t|}t|tjrt|| }ntd t|}|rt|||| | |pd |pd }nt||||pd |pd }td|||g}|r|dkrzt| }t| }t!| }g}g}t| | zD]\} }| t!| kr||vr{|||||}|t%jdt%j| t%j||zg| |kr||vr|r| |kr ||vr||z }|r| |kr ||vr||z }|||t%j| tdt+||}tj||g}|| }| tkr|S| tjkrtj|St d)a\ Perform join of two tables or datasets. The result will be an output table with the result of the join operation Parameters ---------- join_type : str One of supported join types. left_operand : Table or Dataset The left operand for the join operation. left_keys : str or list[str] The left key (or keys) on which the join operation should be performed. right_operand : Table or Dataset The right operand for the join operation. right_keys : str or list[str] The right key (or keys) on which the join operation should be performed. left_suffix : str, default None Which suffix to add to left column names. This prevents confusion when the columns in left and right operands have colliding names. right_suffix : str, default None Which suffix to add to the right column names. This prevents confusion when the columns in left and right operands have colliding names. use_threads : bool, default True Whether to use multithreading or not. coalesce_keys : bool, default False If the duplicated keys should be omitted from one of the sides in the join result. output_type: Table or InMemoryDataset The output type for the exec plan result. Returns ------- result_table : Table or InMemoryDataset zExpected Table or Dataset, got z left semiz left antiz right semiz right antiinnerz left outercg|]}|v| Srr)r%colright_keys_orders rr'z!_perform_join..s+   C7G,G,GC,G,G,Grz right outercg|]}|v| Srr)r%r7left_keys_orders rr'z!_perform_join..s*   3o+E+EC+E+E+Err" table_source)output_suffix_for_leftoutput_suffix_for_righthashjoin)optionsinputsz full outercoalescer(Unsupported output type) isinstancerdsr TypeErrortypetuplelist enumerater*r+r3rrr setlenappendr_call_fieldr r,to_tabler) join_type left_operand left_keys right_operand right_keys left_suffix right_suffixr# coalesce_keys output_typeidxkey left_columns right_columnsleft_column_keys_indicescolnameright_column_keys_indices left_source right_source join_optsr0left_columns_setright_columns_setright_operand_indexprojected_col_namesr1r7right_key_index projection result_tabler:r8s @@r _perform_joinrkOsP lUBJ$7 8 8PN$|:L:LNNOOO meRZ%8 9 9QO$}:M:MOOPPPO i% / / K i((##S" j4- 0 0" \ j))$$S # &,L!(.MK9 #;#; l " "i<&?&? g  l!:!:    (    m # #    '   "!,//44 W i  03 $W - "!-0055 W j 14 %g ., ++X&|MMM !.2H2V2VWW -,, ' ;OOO " 2=AA    ' y*lM#.#4"$0$6B   ( y*#.#4"$0$6B   I{L.I   D2=l22|,, ..!,//  !,">??$ $ HCS&&&&32J+J+J#**3/// #<s34#6""$Z")#.. 0A/?1A1A2 +++7P0P0P  '$'*=$=$=000;&C (%(,?%?%?///<'C#**3///""%c**! )+7JKK  ($ );<<==[=99Le * * *!,///1222rc tjtdt|tdt|g}|dS)aFilter rows of a table based on the provided expression. The result will be an output table with only the rows matching the provided expression. Parameters ---------- table : Table or Dataset Table or Dataset that should be filtered. expression : Expression The expression on which rows should be filtered. Returns ------- Table r;)r@r)Tr")rr,rr rP)table expressionr0s r _filter_tablerosc"  $N,B5,I,IJJJH&7 &C&CDDD&  D ==T= * **rc t|tjrt|d}nt dt |}t dt |fi|}t j||g}|d}|tkr|S|tj krtj |Std)NTr"r;order_byrC) rDrErr3rrr r,rPrrrF)table_or_dataset sort_keysrYkwargs data_sourcerqr0rjs r _sort_sourcervs"BJ// &'7TJJJ ! 23CDD  :'9)'N'Nv'N'NOOH  $k8%< = =D==T=22Le * * *!,///1222rc tjtdt|tdt||g}||S)Nr; aggregate)keysr")rr,rr rP)rm aggregatesryr#r0s r _group_byr{.s`  $N$:5$A$ABBK!5jt!L!L!LMM&  D ==[= 1 11r)T) pyarrow.librpyarrow.computerrpyarrow._acerorrrr r r r r ImportErrorexcstrpyarrow.datasetr/rEpyarrow._datasetrrr3rkrorvr{rrrrs2--------                      +V33s88VVV          0000000 BBB.#"%# n3n3n3n3b+++0;@3333,222222s&%AAA AA0/A0