DUfFddlZddlZddlZddlZddlZddlZddl m Z ddl m Z ddl mZmZmZmZmZmZddlmZmZddlmZmZmZejd Zeeddkrdd lm Z d\d Z!ej"d Z#dZ$dZ%e$e%ej&dddeedddZ'e$e%ej&dddeddej&dddedej&d d!d"ed#ej&d$d%d&dd'd(ej&d)d*d+eddd,Z(e$e%ej&dddeddej&d$d%d&dd'd(ej&d)d*d+eddej&d-d.e)d/dd0Z*e$e%ej&d1d2d3ed4ej&d5d6d7ed4ej&ddded4ej&d d!d"ed4ej&d$d%d&dd'd(ej&d8e)d dd9:ej&d;dej&d?ej+d@dAgd@dBdCej&dDdEe)dFej&dddGedHej&dIdJeKej&dLdMedNdej&dOdPej,dd'ddQKej&dRdSdd'>ej&dTdUVej&dWdXdd'dYZd[Z-dS)]N)compose)Pool)HDF5AggregatorPairixAggregatorTabixAggregatoraggregate_records create_coolersanitize_records)cli get_logger) parse_binsparse_field_paramparse_kv_list_param.) get_handle#cDg}|std|}d\}}t|dr|jj}|jj}n.t|dr|j}|j}ntd|d}||rz|}t|tr| }| | |d}||z||fS)aReturns a header from the stream and an the reaminder of the stream with the actual data. Parameters ---------- instream : a file object An input stream. comment_char : str The character prepended to header lines (use '@' when parsing sams, '#' when parsing pairsams). Returns ------- header : list The header lines, stripped of terminal spaces and newline characters. remainder_stream : stream/file-like object Stream with the remaining lines. zPlease, provide a comment char!)NNbufferpeekz7Cannot find the peek() function of the provided stream!r ) ValueErrorencodehasattrrrreadline startswith isinstancebytesdecodeappendstrip) instream comment_charheader comment_byteread_fpeek_f readline_f current_peeklines M/var/www/html/software/conda/lib/python3.11/site-packages/cooler/cli/cload.py get_headerr,s1$F <:;;;&&((LNFFx""T%_- 6 " "T& RSSS6!99L  ! !, / / !z|| dE " " !;;==D djjll###vayy   ! !, / / ! 8 cdS)z Create a cooler from genomic pairs and bins. Choose a subcommand based on the format of the input contact list. Nr/r-r+cloadr0Ns  Dr-cttjdtdtjdtjdddtjdd tjd d tjdd |S)NbinsBINS)typemetavar pairs_pathT)exists allow_dash PAIRS_PATH cool_path COOL_PATH)r5z --metadataz+Path to JSON file containing user metadata.helpz --assemblyz)Name of genome assembly (e.g. hg19, mm10))r0commandclickargumentstrPathoptionfuncs r+register_subcommandrFZs ==??8vC888 ENZt===$    A{K@@@EL$+X (!L           r-cD|jd|_|S)Na BINS : One of the following : 1. Path to a chromsizes file, 2. Bin size in bp : Path to BED file defining the genomic bin segmentation. PAIRS_PATH : Path to contacts (i.e. read pairs) file. COOL_PATH : Output COOL file path or URI.)__doc__formatrDs r+ add_arg_helprJqs'<&& 1  DL Kr-z --chunksizez-czFControl the number of pixels handled by each worker process at a time.gחAT)r=r4default show_defaultc Nt|\}}|;t|5}tj|}dddn #1swxYwYt j|d5}t ||||} t||| ||dddddS#1swxYwYdS)z Bin a hiclib HDF5 contact list (frag) file. {} hiclib on BitBucket: . NrTmetadataassemblyordered)ropenjsonloadh5pyFilerr ) r2r6r:rPrQ chunksize chromsizesfh5pairsiterators r+hiclibr]s8("$''J (^^ $qy||H $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ :s # #  w!':tYGG                           s#AA  A %(BB!Bz--nprocz-pz.Number of processes to split the work between.z--chrom2z-c2zchrom2 field number (one-based))r=r4rKz--pos2z-p2zpos2 field number (one-based)z --zero-basedz-0zPositions are zero-basedF)r=is_flagrKrLz --max-splitz-szDivide the pairs from each chromosome into at most this many chunks. Smaller chromosomes will be split less frequently or not at all. Increase ths value if large chromosomes dominate the workload on multiple processors.c \tt} t|\} }|;t|5} t j| }dddn #1swxYwY t } |dkr/t|} | d|d| j } i}d|vr|ddz |d<d|vr|ddz |d<t|| |f| | |d |}t|||||d |dkr| dSdS#|dkr| wwxYw) z Bin a tabix-indexed contact list file. {} See also: 'cooler csort' to sort and index a contact list file Tabix manpage: . Nr Using  coreschrom2C2pos2P2)map is_one_basedn_chunksTrO) r__name__rrSrTrUririnfoimaprr close)r2r6r:rPrQnproc zero_based max_splitkwargsloggerrYrZmap_funcpooloptsr\s r+tabixrxsp ! !F!$''J (^^ $qy||H $ $ $ $ $ $ $ $ $ $ $ $ $ $ $! 199;;D KK.... / / /yH v  )A-DJ V  !+DJ"    (.               199 JJLLLLL 95199 JJLLLL sAA A%B DD+z --block-charzICharacter separating contig names in the block names of the pairix index.|c tt} t|\} }|;t|5} t j| }dddn #1swxYwY t } |dkr/t|} | d|d| j } t|| || | ||}t|||||d|dkr| dSdS#|dkr| wwxYw)z Bin a pairix-indexed contact list file. {} See also: 'cooler csort' to sort and index a contact list file Pairix on GitHub: . Nr rcrd)rirjrk block_charTrO) rrlrrSrTrUrirrmrnrr ro)r2r6r:rPrQrprqrrr{rtrYrZrurvr\s r+pairixr| sqt ! !F!$''J (^^ $qy||H $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ 199;;D KK.... / / /yH#   (.!             199 JJLLLLL 95199 JJLLLL sAA A%A(C++Dz--chrom1z-c1zchrom1 field number (one-based))r=r4requiredz--pos1z-p1zpos1 field number (one-based)z--comment-charz1Comment character that indicates lines to ignore.)r4rKrLr=z--no-symmetric-upperz-NztCreate a complete square matrix without implicit symmetry. This allows for distinct upper- and lower-triangle values)r=rarKz--input-copy-statusuniqueduplexaCopy status of input data when using symmetric-upper storage. | `unique`: Incoming data comes from a unique half of a symmetric map, regardless of how the coordinates of a pair are ordered. `duplex`: Incoming data contains upper- and lower-triangle duplicates. All input records that map to the lower triangle will be discarded! | If you wish to treat lower- and upper-triangle input data as distinct, use the ``--no-symmetric-upper`` option. )r4rKr=rLz--fieldaSpecify quantitative input fields to aggregate into value columns using the syntax ``--field =``. Optionally, append ``:`` followed by ``dtype=`` to specify the data type (e.g. float), and/or ``agg=`` to specify an aggregation function different from sum (e.g. mean). Field numbers are 1-based. Passing 'count' as the target name will override the default behavior of storing pair counts. Repeat the ``--field`` option for each additional field.)r=r4multiplezSize in number of lines/records of data chunks to read and process from the input stream at a time. These chunks will be saved as temporary partial coolers and then merged.iz --mergebufzmTotal number of pixel records to buffer per epoch of merging data. Defaults to the same value as `chunksize`.)r=r4z --max-mergez3Maximum number of chunks to merge in a single pass.z --temp-dirzaCreate temporary files in a specified directory. Pass ``-`` to use the platform default temp dir.)r7 file_okaydir_okayr8z--no-delete-tempz,Do not delete temporary files when finished.z--storage-optionszOptions to modify the data filter pipeline. Provide as a comma-separated list of key-value pairs of the form 'k1=v1,k2=v2,...'. See http://docs.h5py.org/en/stable/high/dataset.html#filter-pipeline for more details.r<z--appendz-az_Pass this flag to append the output cooler to an existing file instead of overwriting the file.)rarKr=c 2(t|\}}| | } | }d}|r|dkrd}n|dkrd}|;t|5}tj|}dddn #1swxYwYgd}tt jtt jd}i(dD]2}||dkrtjd| ||d z (|<3g}i}i}t| r| D]}t|\}}}}|$| | |d vr|||<&tjd | ||vr| |||vr| ||(|<| |||<|||<||||<d ||<d|vr| d|Ht|} | D]5}!t| |!trt| |!| |!<6nd} i}|dkr t j}"nlt%t&dd kr7t%t&d dkrt)|ddd}"nt)|ddj}"t-|"\}#}"t/j|"fd(fd|D||d| d|}$t3|dd| |dd}%t5|dd}&t7|&|%}'t9||t;|'|$fid|d|d|d|d | d!|d"| d#| d$dd%dd&dd'dd(|d)| d*dd+|rd,nd-dS).z Bin any text file or stream of pairs. Pairs data need not be sorted. Accepts compressed files. To pipe input from stdin, set PAIRS_PATH to '-'. {} Nr~reflectrdrop)chrom1pos1rergrzField numbers start at 1) param_hintr >countbin1_idbin2_idzA field number is required.sumr-rrNinfer)mode compression c g|] }| Sr/r/).0nameinput_field_numberss r+ zpairs..rsIIIt$T*IIIr-T)sepusecolsnamesdtyper\rXpairs)schema decode_chromsrj tril_actionsortvalidateF)aggrrcolumnsdtypesrPrQmergebuf max_mergetemp_dir delete_temp boundscheck triucheckdupcheck ensure_sortedsymmetric_upperh5optsrRraw)rrSrTrUrAnpint64r? BadParameterlenrr rrlisttuplesysstdinint_pandas_versionrhandler,pdread_csvr r rr ri))r2r6r:rPrQrqr#input_copy_statusno_symmetric_upperfieldrXrrno_delete_temprstorage_optionsr rsrYrrrZinput_field_namesinput_field_dtypesroutput_field_namesoutput_field_dtypes aggregationsargcolnumrrrkeyf_in_readersanitize aggregatepipeliners) @r+rrisp"$''J,,OK!  ( (#KK ( * * K (^^ $qy||H $ $ $ $ $ $ $ $ $ $ $ $ $ $ $  455 $<1  $%?DQQQ Q$*4L1$4D!!L 5zz"+! +! +C'8'='= $D&%~K) ???05'-,5#,,,!((...---"))$///(.  % +0"4(,1#D)%( T""%* T""(((!!'***"$_55 1 1C&+t,, 1#F3K00s  1FSy _Q  A % %#oa.@*A*AA*E*E*3GDDDQG*3GDDDKGAt [   IIII7HIII      F  #^ H"l$UKKKIy(++H  Hf#"  #"   )'&&E%e (!"v#$%& #SS'sA""A&)A&)r).rr?rVnumpyrpandasr simplejsonrTcytoolzr multiprocessrcreaterrrr r r r r_utilrrr __version__splitrrpandas.io.commonrr,groupr0rFrJrCr]rxrAr|ChoicerBrr/r-r+rsb EEEEEEEEEE.&&s++3qQ++++++////d     .       Q CJJ    8    9  =CQR e9Q #        ==  F=@    9   #          77  J7t  =CRV e9d=CRV e9d #     <  @    x* + +  :     ?    & '  1   >   % 454D Q Q Q   7        ' vv   '    vvvvr-