o hY@sUddlmZddlZddlZddlZddlZddlZddlZddlZddl m Z ddl m Z ddl mZddlmZmZmZmZmZmZmZmZddlmZeraddlZddlmZdd lmZd Zed Z ddddZ! ddddZ"iZ#de$d<dddZ%dd"d#Z&dd+d,Z'  -ddd4d5Z(dd9d:Z)dd=d>Z*ddBdCZ+ddFdGZ, - -dddMdNZ-ddPdQZ.ddSdTZ/ddVdWZ0ddYdZZ1dd\d]Z2   ^ _dddfdgZ3ddjdkZ4ddodpZ5ej6ddrdsZ7  _ddd{d|Z8dd~dZ9ej6ddddZ:ddZ;ddZ>> infer_storage_options('/mnt/datasets/test.csv') # doctest: +SKIP {"protocol": "file", "path", "/mnt/datasets/test.csv"} >>> infer_storage_options( ... 'hdfs://username:pwd@node:123/mnt/datasets/test.csv?q=1', ... inherit_storage_options={'extra': 'value'}, ... ) # doctest: +SKIP {"protocol": "hdfs", "username": "username", "password": "pwd", "host": "node", "port": 123, "path": "/mnt/datasets/test.csv", "url_query": "q=1", "extra": "value"} z^[a-zA-Z]:[\\/]z^[a-zA-Z0-9]+://Nfile)protocolpath#z^/([a-zA-Z])[:|]([\\/].*)$:)httphttps@rhost)s3s3agcsgsrportusernamepassword url_query url_fragment)rematchrschemefragmentjoinrgroupsnetlocrsplitr'r(r)queryupdate_storage_options)rr parsed_pathrr windows_pathdriveoptionsr:`/var/www/html/construction_image-detection-poc/venv/lib/python3.10/site-packages/fsspec/utils.pyinfer_storage_options&sF       "       r<r9 inheritedNonecCsV|si}t|t|@}|r$|D]}||||kr#td|q||dS)Nz9Collision between inferred and specified storage option: )setgetKeyErrorupdate)r9r= collisions collisionr:r:r;r5{sr5zdict[str, str] compressionsfilename str | NonecCs.tj|dd}|tvrt|SdS)aInfer compression, if available, from filename. Infer a named compression type, if registered and available, from filename extension. This includes builtin (gz, bz2, zip) compressions, as well as optional compressions. See fsspec.compression.register_compression. r!.N)osrsplitextstriplowerrE)rF extensionr:r:r;infer_compressionsrNmax_intfloatCallable[[int], str]cs.|d7}ttt|d fdd }|S) aReturns a function that receives a single integer and returns it as a string padded by enough zero characters to align with maximum possible integer >>> name_f = build_name_function(57) >>> name_f(7) '07' >>> name_f(31) '31' >>> build_name_function(1000)(42) '0042' >>> build_name_function(999)(42) '042' >>> build_name_function(0)(0) '0' g:0yE>iintrrcst|SN)rzfill)rR pad_lengthr:r; name_functionsz*build_name_function..name_functionN)rRrSrr)rSmathceillog10)rOrXr:rVr;build_name_functionsr\r IO[bytes] delimiterbytes blocksizerSboolc Cs|dkrdSd} ||}|sdS|r||n|}z(||vr:||}||t||t|WdSt||krCWdSWn ttfyOYnw|t| d}q )aSeek current file to file start, file end, or byte after delimiter seq. Seeks file to next chunk delimiter, where chunks are defined on file start, a delimiting sequence, and file end. Use file.tell() to see location afterwards. Note that file start is a valid split, so must be at offset > 0 to seek for delimiter. Parameters ---------- file: a file delimiter: bytes a delimiter like ``b'\n'`` or message sentinel, matching file .read() type blocksize: int Number of bytes to read from the file at once. Returns ------- Returns True if a delimiter was found, False if at file start or end. rFNT)tellreadindexseeklenOSError ValueError)rr^r`lastcurrentfullrRr:r:r;seek_delimiters*   " rlFfoffsetlength int | None bytes | None split_beforec Cs|rJ||t||d}|dur|S|}|||8}|||t||d}|}|r:|r:|t|8}|rD|rD|t|8}|}||}|||dusUJ||} | S)aRead a block of bytes from a file Parameters ---------- f: File Open file offset: int Byte offset to start read length: int Number of bytes to read, read through end of file if None delimiter: bytes (optional) Ensure reading starts and stops at delimiter bytestring split_before: bool (optional) Start/stop read *before* delimiter bytestring. If using the ``delimiter=`` keyword argument we ensure that the read starts and stops at delimiter boundaries that follow the locations ``offset`` and ``offset + length``. If ``offset`` is zero then we start at zero, regardless of delimiter. The bytestring returned WILL include the terminating delimiter string. Examples -------- >>> from io import BytesIO # doctest: +SKIP >>> f = BytesIO(b'Alice, 100\nBob, 200\nCharlie, 300') # doctest: +SKIP >>> read_block(f, 0, 13) # doctest: +SKIP b'Alice, 100\nBo' >>> read_block(f, 0, 13, delimiter=b'\n') # doctest: +SKIP b'Alice, 100\nBob, 200\n' >>> read_block(f, 10, 10, delimiter=b'\n') # doctest: +SKIP b'Bob, 200\nCharlie, 300' iN)rerlrcrbrf) rmrnror^rrfound_start_delimstartfound_end_delimendbr:r:r; read_blocks(+         rxargsrkwargscOsV|r||f7}z tt|}W|Sty*tt|dd}Y|Sw)zDeterministic token (modified from dask.base) >>> tokenize([1, 2, '3']) '9d71491b50023b06fc76928e6eddb952' >>> tokenize('Hello') == tokenize('Hello') True F)usedforsecurity)rrencoderh hexdigest)ryrzhr:r:r;tokenize4s  rfilepath%str | os.PathLike[str] | pathlib.PathcCs4t|tr|St|dr|St|dr|jS|S)a6Attempt to convert a path-like object to a string. Parameters ---------- filepath: object to be converted Returns ------- filepath_str: maybe a string version of the object Notes ----- Objects supporting the fspath protocol are coerced according to its __fspath__ method. For backwards compatibility with older Python version, pathlib.Path objects are specially coerced. Any other object is passed through unchanged, which includes bytes, strings, buffers, or anything else that's not even path-like. __fspath__r) isinstancerhasattrrr)rr:r:r;stringify_pathIs   rclsCallable[..., T] Sequence[Any]cCs||i|}||SrT)_determine_worker)rryrzinstr:r:r; make_instanceisrpaths Iterable[str]cspdd|DtddD}d}t|D]tfddD}|s(nq|7dddS) z;For a list of paths, find the shortest prefix common to allcSsg|]}|dqS/)split.0pr:r:r; ssz!common_prefix..cs|]}t|VqdSrT)rfrr:r:r; tsz common_prefix..rc3s$|] }|dkVqdS)rNr:rrRpartsr:r;rws"rN)minrangeallr0)rlmaxrvr:rr; common_prefixqs r list[str]path2str | list[str]existsflattencsttrGd|rfdd|DSt||r%ddds;tdd|Dr;fdd|DSfd d|DSt|tksQJS) aIn bulk file operations, construct a new file tree from a list of files Parameters ---------- paths: list of str The input file tree path2: str or list of str Root to construct the new list in. If this is already a list of str, we just assert it has the right number of elements. exists: bool (optional) For a str destination, it is already exists (and is a dir), files should end up inside. flatten: bool (optional) Whether to flatten the input directory tree structure so that the output files are in the same directory. Returns ------- list of str rcs$g|]}d|ddfqS)rr!)r0rrrr:r;rs$zother_paths..r rcss|] }|d VqdS)rN) startswithrsr:r:r;rszother_paths..csg|] }d|gqSr)r0rrr:r;rcsg|] }|dqS)r replacercprr:r;rr)rrrstriprr3rrf)rrrrr:rr; other_paths~s   robjcCs t|tSrT)r BaseExceptionrr:r:r; is_exception rTypeGuard[IO[bytes]]cstfdddDS)Nc3s|]}t|VqdSrT)r)rattrrmr:r;rszisfilelike..)rccloserb)rrr:rr; isfilelikesrurlcCs0t|}tjd|dd}t|dkr|dSdS)Nz (\:\:|\://)r )maxsplitrr)rr,rrf)rrr:r:r; get_protocols  rrc Cs<ddlm}z t|t|ddWSttfyYdSw)z*Can the given URL be used with open_local?r)get_filesystem_class local_fileF)fsspecrgetattrrrh ImportError)rrr:r:r; can_be_locals rnamec Csl|tjvrtj|}t|dr|jSzt|WSYz ddl}||}|jWSttfy5YdSw)aFor given package name, try to find the version without importing it Import and package.__version__ is still the backup here, so an import *might* happen. Returns either the version string, or None if the package or the version was not readily found. __version__rN) sysmodulesrrr importlib import_modulerAttributeError)rmodrr:r:r;"get_package_version_without_imports    rDEBUGTloggerlogging.Logger | None logger_namelevelclearlogging.LoggercCsh|dur |dur td|pt|}t}td}|||r(|j||| ||S)Nz+Provide either logger object or logger namezD%(asctime)s - %(name)s - %(levelname)s - %(funcName)s -- %(message)s) rhlogging getLogger StreamHandler Formatter setFormatterhandlersr addHandlersetLevel)rrrrhandle formatterr:r:r; setup_loggings    rfsrcCs ||SrT)unstrip_protocol)rrr:r:r;_unstrip_protocolrr origin_namemethodsCallable[[type[T]], type[T]]cs"d fdd dfd d }|S)zqMirror attributes and methods from the given origin_name attribute of the instance to the decorated classmethodrselfrrcst|}t||SrT)r)rrorigin)rr:r; origin_getters  z"mirror_from..origin_getterrtype[T]cs(D]}t|}t||t|q|SrT)rsetattrproperty)rrwrapped_method)rrr:r;wrappers zmirror_from..wrapperN)rrrrrr)rrrrr:)rrrr:)rrrr; mirror_fromsr Iterator[T]ccs |VdSrTr:rr:r:r; nullcontext s rstartslist[int] | intendsmax_gap max_blocksort&tuple[list[str], list[int], list[int]]c Cst|tstt|ts|gt|}t|ts|gt|}t|t|ks/t|t|kr1tt|dkr<|||fSdd|D}|rWddttt|||D\}}}|r|dd}|dd}|dd}tdt|D]Q} || || dkr|ddurqr|| || dks|| |d|ks|dur|| |d|kr||| ||| ||| qr|| |d<qr|||fS|||fS)a}Merge adjacent byte-offset ranges when the inter-range gap is <= `max_gap`, and when the merged byte range does not exceed `max_block` (if specified). By default, this function will re-order the input paths and byte ranges to ensure sorted order. If the user can guarantee that the inputs are already sorted, passing `sort=False` will skip the re-ordering. r cSsg|]}|pdqS)rr:rr:r:r;r/sz'merge_offset_ranges..csrrT)list)rvr:r:r;r2s  z&merge_offset_ranges..Nr!) rr TypeErrorrfrhzipsortedrappend) rrrrrr new_paths new_startsnew_endsrRr:r:r;merge_offset_rangessF              rfilelikec Cs.|}z |ddW||S||w)z+Find length of any open read-mode file-liker)rbre)rposr:r:r; file_sizeXs rwbmodec cstjtj|tj|dd\}}zt|| }|VWdn1s(wYWn"tyPt t  t |Wd1sJwYwt ||dS)z A context manager that opens a temporary file next to `path` and, on exit, replaces `path` with the temporary file, thereby updating `path` atomically. -)dirprefixN) tempfilemkstemprIrdirnamebasenameopenr contextlibsuppressFileNotFoundErrorunlinkr)rrfdfnfpr:r:r; atomic_writeas$      rc Csg}|j}dt|}}||krZ||}|d}|dkr+|r%|d|ur)||n+|dkr5||n!|dkrO|}||krJ||dkrJ|d}||krX||dkrX|d}||krp||dkrp|d}||krp||dksb||kry|d n|||} d | vr| d d } ng} ||dkr|d n|d} |d | |} | dkrn| ||| | d}| d} q|||} | r| | n| dd 7<tt| dddD]*} | | dd| | dkr| | ddd| | dd| | d<| | =qd dd| D} tdd| } |d}| s!|dn5| dkr+|dn+| ddkr;d| dd} n | ddvrFd | } |d| dn|t|||ks||ksaJ|S)Nrr *r!?[!]z\[r\\\rTcss$|] }|ddddVqdS)rrrz\-Nrrr:r:r;rs z_translate..z([&~|])z\\\1z(?!)rH^)rr) rrfrfindrr0r,subescape) patSTAR QUESTION_MARKresaddrRncjstuffchunkskchunkr:r:r; _translatevs~       ,    Ar-cCsDtjjr tjjtjj}ntjj}dttj|}t|dkr&d|dn|}d|d}|d}||}d|d}d }g} t ||} t| d} t | D]C\} } | d krg| | | krc|n|qS| d krw| | | krs|n|qSd | vrt d | r| t| |d || | kr| |qSd| }d |dS)zBTranslate a pathname with shell wildcards to a regular expression.r rrz[^+z(?:.+z)?z.*rz**z:Invalid pattern: '**' can only be an entire path componentz(?s:z)\Z)rIraltsepsepr0mapr,r rfr enumeraterrhextendr-)r!seps escaped_sepsany_sepnot_sepone_last_segment one_segment any_segmentsany_last_segmentsresultsr last_part_idxidxpartr$r:r:r;glob_translates>         rArT)rrrrrr)r9rr=rrr>)rFrrrG)rOrPrrQ)rr]r^r_r`rSrra)NF) rmr]rnrSrorpr^rqrrrarr_)ryrrzrrr)rrrr)rrryrrzrrr)rrrr)FF) rrrrrrarrarr)rrrra)rmrrr)rrrr)rrrra)rrrrG)NNrT) rrrrGrrrrarr)rrrrrr)rrrrrr)rrrr)rNT)rrrrrrrrSrrprrarr)rr]rrS)r)rrrr)= __future__rr rrYrIr,rr functoolsrhashlibrimportlib.metadatartypingrrrr r r r r urllib.parserpathlibtyping_extensionsr fsspec.specrDEFAULT_BLOCK_SIZErr<r5rE__annotations__rNr\rlrxrrrrrrrrrrrrrcontextmanagerrrrrr-rAr:r:r:r;sx   (    V   5 J   -        F  J