bi*@UddlZddlZddlZddlmZmZddlmZddlm Z ddl m Z m Z m Z mZmZddlZddlZddlmZddlmZdd lmZdd lmZmZdd lmZmZm Z e r ddl!Z"d d l#m$Z$da%ee&e'e(d<ejRdk(rdndZ*ejVdejVdejVdejVdejVdejVdejVdejVdejVdejVdejVdejVdejVdejVdgZ,eGd d!Z-d"e&e'fd#Z.d$d%d"e/fd&Z0d$d%d"e1fd'Z2d(ejfd"e1fd)Z4d*ee&e'e&e1e&ejfe&d%fd"e&e1fd+Z5y),N) dataclassfield)BytesIO)Path) TYPE_CHECKINGAnyClassVarOptionalUnion)config)DownloadConfig) array_cast) is_local_pathxopen)first_non_null_valueno_op_if_value_is_nullstring_to_dict) FeatureType_IMAGE_COMPRESSION_FORMATSlittle<>z|b1|u1zu2zi2zu4zi4zf4zf8ceZdZUdZdZeeed<dZe ed<e ddZ eeed<d Z e eed <ejej ej"d Ze eed <e ddd Zeed<dZdeeeeeej6d fdefdZddedd fdZdedeedfffdZdeej>ej@ejBfdej@fdZ"ddej@dej@fdZ#y)Imagea=Image [`Feature`] to read image data from an image file. Input: The Image feature accepts as input: - A `str`: Absolute path to the image file (i.e. random access is allowed). - A `pathlib.Path`: path to the image file (i.e. random access is allowed). - A `dict` with the keys: - `path`: String with relative path of the image file to the archive file. - `bytes`: Bytes of the image file. This is useful for parquet or webdataset files which embed image files. - An `np.ndarray`: NumPy array representing an image. - A `PIL.Image.Image`: PIL image object. Output: The Image features output data as `PIL.Image.Image` objects. Args: mode (`str`, *optional*): The mode to convert the image to. If `None`, the native mode of the image is used. decode (`bool`, defaults to `True`): Whether to decode the image data. If `False`, returns the underlying dictionary in the format `{"path": image_path, "bytes": image_bytes}`. Examples: ```py >>> from datasets import load_dataset, Image >>> ds = load_dataset("AI-Lab-Makerere/beans", split="train") >>> ds.features["image"] Image(decode=True, id=None) >>> ds[0]["image"] >>> ds = ds.cast_column('image', Image(decode=False)) {'bytes': None, 'path': '/root/.cache/huggingface/datasets/downloads/extracted/b0a21163f78769a2cf11f58dfc767fb458fc7cea5c05dccc0144a2c0f0bc1292/train/healthy/healthy_train.85.jpg'} ``` NmodeTdecodeF)defaultrepridPIL.Image.Imagedtypebytespathpa_type)r initr!_typec|jSN)r()selfs R/home/cdr/jupyterlab/.venv/lib/python3.12/site-packages/datasets/features/image.py__call__zImage.__call___s ||valuereturnctjrddl}n tdt |t rt j|}t |tr|ddSt |trt|jddSt |ttfrd|dSt |t jr t|St ||jjr t!|S|j#d6t$j&j)|drd|j#ddS|j#d|j#d#|j#d|j#ddSt+d|d ) aEncode example into a format for Arrow. Args: value (`str`, `np.ndarray`, `PIL.Image.Image` or `dict`): Data passed as input to Image feature. Returns: `dict` with "path" and "bytes" fields rN4To support encoding images, please install 'Pillow'.r'r&r'r%r&zUAn image sample should have one of 'path' or 'bytes' but they are missing or None in .)r PIL_AVAILABLE PIL.Image ImportError isinstancelistnparraystrrabsoluter& bytearrayndarrayencode_np_arrayrencode_pil_imagegetosr'isfile ValueError)r-r1PILs r.encode_examplezImage.encode_examplebsE    TU U eT "HHUOE eS !!D1 1 t $ 01DA A y1 2 51 1 rzz *"5) ) syy /#E* * YYv  *rww~~eFm/L!599V+<= = YYw  +uyy/@/L"YYw/69JK Kghmgnnop r0cp|js tdtjr ddl}ddl}n t d|i}|d|d}}||td|dt|r|jj|}n|jd d }|jtjrtjntj}t!||} | |j#| d nd} t%| } t'|d | 5} t)| j+}ddd|jj|}n$|jjt)|}|j-|j/j#|jj0j2j4|j6j9|}|j:r4|j:|j:k7r|j=|j:}|S#1swYxYw)aqDecode example image file into image data. Args: value (`str` or `dict`): A string with the absolute image file path, a dictionary with keys: - `path`: String with absolute or relative image file path. - `bytes`: The bytes of the image file. token_per_repo_id (`dict`, *optional*): To access and decode image files from private repositories on the Hub, you can pass a dictionary repo_id (`str`) -> token (`bool` or `str`). Returns: `PIL.Image.Image` zMDecoding is disabled for this feature. Please use Image(decode=True) instead.rNz4To support decoding images, please install 'Pillow'.r'r&zCAn image should have one of 'path' or 'bytes' but both are None in r6::repo_idtokenrbdownload_config)r RuntimeErrorr r7r8 PIL.ImageOpsr9rGrropensplit startswith HF_ENDPOINTHUB_DATASETS_URLHUB_DATASETS_HFFS_URLrrDrrrreadloadgetexifExifTagsBase OrientationImageOpsexif_transposerconvert) r-r1token_per_repo_idrHr'bytes_image source_urlpatternsource_url_fieldsrOrRfs r.decode_examplezImage.decode_examples${{no o     TU U  $ " V}eGnf >| #fglfmmn!opp &IINN40E!%D!1"!5J&001C1CD//#99 )7z7(K%O`Ol)--.? .JKrv'55&AOtT?K3q!(!23IINN62EIINN76?3E ==?  syy1166BB C OLL//6E 99ejj0MM$)),E 33s H,,H5rcLddlm}|jr|S|d|ddS)zfIf in the decodable state, return the feature itself, otherwise flatten the feature into a dictionary.r)Valuebinarystringr%)featuresrmr)r-rms r.flattenz Image.flattens2#{{  xh r0storagec tjj|jrltjdgt |ztj }tjj||gddg|j}ntjj|jrltjdgt |ztj}tjj||gddg|j}n&tjj|jr|jjddk\r|jd}n6tjdgt |ztj }|jjddk\r|jd}n6tjdgt |ztj}tjj||gddg|j}ntjj|jrtj|j!Dcgc]'}|!t#t%j|dnd)c}tj }tjdgt |ztj}tjj||gddg|j}t'||j(Scc}w)a'Cast an Arrow array to the Image arrow storage type. The Arrow types that can be converted to the Image pyarrow storage type are: - `pa.string()` - it must contain the "path" data - `pa.binary()` - it must contain the image bytes - `pa.struct({"bytes": pa.binary()})` - `pa.struct({"path": pa.string()})` - `pa.struct({"bytes": pa.binary(), "path": pa.string()})` - order doesn't matter - `pa.list(*)` - it must contain the image array data Args: storage (`Union[pa.StringArray, pa.StructArray, pa.ListArray]`): PyArrow array to cast. Returns: `pa.StructArray`: Array in the Image arrow storage type, that is `pa.struct({"bytes": pa.binary(), "path": pa.string()})`. Ntyper&r'maskr)patypes is_stringrur=lenrn StructArray from_arraysis_null is_binaryro is_structget_field_indexris_list to_pylistrBr<rr()r-rr bytes_array path_arrayarrs r. cast_storagezImage.cast_storagest& 88  gll +((D6CL#8ryy{KKnn00+w1G'SYIZahapapar0sG XX   -4&3w<"7biikJJnn00':1FRXHY`g`o`o`q0rG XX   -||++G49%mmG4  hhvG '<299;O ||++F3q8$]]62 XXtfs7|&;"))+N nn00+z1JWV\L]dkdsdsdu0vG XX  gll +((ahararatuZ]CO#/8QUUuYY[K4&3w<"7biikJJnn00j)GV+<;CVCVCX1G'4<<00vs ,M'c ritfd}tj|jDcgc]}||d ||dn|dndc}tj}tj|j djDcgc]%}|t jj|nd'c}tj}tjj||gddg|j}t||jScc}wcc}w)a8Embed image files into the Arrow array. Args: storage (`pa.StructArray`): PyArrow array to embed. Returns: `pa.StructArray`: Array in the Image arrow storage type, that is `pa.struct({"bytes": pa.binary(), "path": pa.string()})`. Ncr|jdd}|jtjrtjntj }t ||}|j|dnd}t|}t|d|5}|jcdddS#1swYyxYw)NrKrLrMrNrPrQ) rVrWr rXrYrZrrDrrr[)r'rgrhrirOrRrjrds r. path_to_bytesz*Image.embed_storage..path_to_bytessD)"-J+5+@+@ASAS+T''Z`ZvZv !/z7 C K\Kh%))*;I*FGnrE,59OtT?C qvvx   s B--B6r&r'rtrv)rrxr=rrnrrEr'basenameror|r}r~rr()r-rrrdrxrr'rs ` r. embed_storagezImage.embed_storages%  $ "     hh!**, UVTaQwZ-?qy)QwZgkk   XXNUmm\bNcNmNmNo pdt'7RWW  d #T A p ..,,k:-FRXHY`k`s`s`u,v'4<<00  qs D/ *D4r,)$__name__ __module__ __qualname____doc__rr r>__annotations__rboolrr"r$r rxstructrnror(rr*r/r r&r@dictr<rArIrkrq StringArrayr| ListArrayrrr0r.rr.s=%ND(3-FDd7B 7,E8C=,&RYYibiik'RSGXc]SwU?E3?'E#uirzzSd*d$e'jn'R;D;EV;z  }d3 3E.FFG  ,1E".."..",,*V$W,1\^\j\j,1\&1R^^&1PRP^P^&1r0rr2cVtjrddl}n tdt~|j j tt|j jjt|j jjzatS)Nrr4) r r7r8r9rrr)r;setOPENkeysSAVE)rHs r.list_image_compression_formatsr,sr PQQ") %)#ciinn.A.A.C*Ds399>>K^K^K`Ga*a%b" %%r0rfr#ct}|jtvr |j}n|jdvrdnd}|j |||j S)zmConvert a PIL Image object to bytes using native compression if possible, otherwise use PNG/TIFF compression.)1LLARGBRGBAPNGTIFF)format)rrrrsavegetvalue)rfbufferrs r.image_to_bytesr9sR YF ||577**(GGV JJvfJ% ?? r0crt|dr|jdk7r|jddSdt|dS)Nfilenamer5)hasattrrr)rfs r.rCrCDs6uj!enn&:66~e'<==r0r=ctjrddl}n td|j}|j dk7r |j nt }|j}|j}d}|jddrM|dvrtd|d|dtjd }||k7rtjd |d |d n|tvr|}n|d k\ri||zt|z}tj|tvr2tj|}tjd |d |d n |dz}|d k\ri|td|dt|j j#|j%|}dt'|dS)Nrr4=r )uizUnsupported array dtype z for image encoding. Only z' is supported for multi-channel arrays.rzDowncasting array dtype z to z to be compatible with 'Pillow'rzCannot downcast dtype z- to a valid image dtype. Valid image dtypes: r5)r r7r8r9r$ byteorder_NATIVE_BYTEORDERkinditemsizeshape TypeErrorr<warningswarn_VALID_IMAGE_ARRAY_DTPYESr>r fromarrayastyper) r=rHr$dtype_byteorder dtype_kinddtype_itemsize dest_dtype dtype_strrfs r.rBrBKs PQQ KKE).C)?eooEVOJ^^NJ {{12 Z '*5'1KJ7JJIxx "&??XXi0  8tJ%#8 99r0objsctjrddl}n td|rt |\}}t |t r|Dcgc] }||ddnd c}St |tjr%tt}|Dcgc] }|| c}St ||jjr%tt}|Dcgc] }|| c}S|S|Scc}wcc}wcc}w)zmEncode a list of objects into a format suitable for creating an extension array of type `ImageExtensionType`.rNr4r5) r r7r8r9rr:r>r<rArrBrrC)rrH_objobj_to_image_dict_funcs r.objects_to_list_of_image_dictsrvsPQQ %d+3 c3 Y]^RUCOS40M^ ^ c2:: &%;O%L ";?@C*3/@ @ SYY__ -%;rrrr$rrrr&rrrCrArBrrr0r.rs (@@53YY%37HT#Y/6==H4C# BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO BHHUO$ z1z1 z1z &S &+>->$>(:2::(:$(:V S 4:tBJJ'7>O9PP Q $Zr0