o TZh0;@sUddlZddlZddlZddlmZmZddlmZddlm Z m Z m Z m Z m Z mZmZddlZddlZddlmZddlmZddlmZdd lmZdd lmZdd lmZm Z m!Z!e riddl"Z#d d l$m%Z%da&ee e'e(d<ej)dkrzdndZ*e+de+de+de+de+de+de+de+de+de+de+de+de+de+dgZ,eGd d!d!Z-d"e e'fd#d$Z.d%d&d"e/fd'd(Z0d%d&d"e1fd)d*Z2d+ej3d"e1fd,d-Z4d.ee e'e e1e ej3e d&fd"e e1fd/d0Z5dS)1N) dataclassfield)BytesIO) TYPE_CHECKINGAnyClassVarDictListOptionalUnion)config)DownloadConfig)xopen) array_cast) is_local_path)first_non_null_valueno_op_if_value_is_nullstring_to_dict) FeatureType_IMAGE_COMPRESSION_FORMATSlittle<>z|b1|u1zu2zi2zu4zi4zf4zf8c@seZdZUdZdZeed<dZee ed<dZ e e ed<e e e dZe eed <edd d d Ze ed <d dZdee eeejdfdefddZddeddfddZdedee dfffddZdee je je j fde jfddZ!de jde jfddZ"dS)ImageaImage [`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 `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 archived files with sequential access. - An `np.ndarray`: NumPy array representing an image. - A `PIL.Image.Image`: PIL image object. Args: 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("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'} ``` TdecodeNidPIL.Image.Imagedtypebytespathpa_typeF)defaultinitrepr_typecCs|jSN)r$)selfr+N/var/www/html/lang_env/lib/python3.10/site-packages/datasets/features/image.py__call__YszImage.__call__valuereturncCstjrddl}ntdt|trt|}t|tr |ddSt|t r*d|dSt|tj r4t |St||j j r?t |S|ddurVtj|drVd|ddS|ddusd|dduro|d|ddStd|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 isinstancelistnparraystrr"ndarrayencode_np_arrayrencode_pil_imagegetosr#isfile ValueError)r*r.PILr+r+r,encode_example\s(         zImage.encode_examplec CsB|jstdtjrddl}ntd|duri}|d|d}}|dur|dur2td|dt|r=|j |}n^| d d }| tj rMtj ntj}zt||d } || } Wn tyid} Ynwt| d } t|d | d} t| }Wdn1swY|j |}n|j t|}||S)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 r2z::repo_id)tokenrb)download_config)r RuntimeErrorr r3r4r5rArropensplit startswithZ HF_ENDPOINTZHUB_DATASETS_URLZHUB_DATASETS_HFFS_URLrr>rrrreadload) r*r.Ztoken_per_repo_idrBr#bytes_image source_urlpatternrErFrHfr+r+r,decode_examples@    zImage.decode_examplercCs(ddlm}|jr |S|d|ddS)zfIf in the decodable state, return the feature itself, otherwise flatten the feature into a dictionary.r)Valuebinarystringr!)featuresrUr)r*rUr+r+r,flattens z Image.flattenstoragecCstj|jr%tjdgt|td}tjj||gddg| d}ntj |jrJtjdgt|t d}tjj||gddg| d}ntj |jr|j ddkr_|d}ntjdgt|td}|j ddkr{|d}ntjdgt|t d}tjj||gddg| d}n4tj|jrtjdd|Dtd}tjdgt|t d}tjj||gddg| d}t||jS) 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#maskrcSs*g|]}|durtt|dndqS)Nr")r<r8r9).0Zarrr+r+r, s*z&Image.cast_storage..)patypesZ is_stringr\r9lenrV StructArray from_arraysis_nullZ is_binaryrWZ is_structZget_field_indexrZis_list to_pylistrr$r*rZZ bytes_arrayZ path_arrayr+r+r, cast_storages0      zImage.cast_storagecstddtjfdd|Dtd}tjdd|dDtd}tjj||gddg| d }t ||j S) 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()})`. cSs8t|d }|}Wd|S1swY|S)NrG)rrM)r#rSrOr+r+r, path_to_bytess   z*Image.embed_storage..path_to_bytescs8g|]}|dur|ddur|dn|dndqS)Nr"r#r+)r_xrjr+r,r` s*z'Image.embed_storage..r[cSs$g|]}|durtj|ndqSr))r?r#basename)r_r#r+r+r,r`s$r#r"r]) rrar9rgrVrrWrdrerfrr$rhr+rlr, embed_storages   zImage.embed_storager))#__name__ __module__ __qualname____doc__rbool__annotations__rr r:r rrastructrVrWr$rrr(r-r r"dictr8r;rCrTrrYZ StringArrayrdZ ListArrayrirnr+r+r+r,r.s  "$"'9$ .rr/cCsRtjrddl}ntdtdur'|jtt|jj t|jj @atS)Nrr0) r r3r4r5rrr&r7setZOPENkeysZSAVE)rBr+r+r,list_image_compression_formatss  $ryrPrcCsBt}|jtvr |j}n |jdvrdnd}|j||d|S)zmConvert a PIL Image object to bytes using native compression if possible, otherwise use PNG/TIFF compression.)1LZLARGBZRGBAZPNGZTIFF)format)rr}rymodesavegetvalue)rPbufferr}r+r+r,image_to_bytes%s  rcCs.t|dr|jdkr|jddSdt|dS)Nfilenamer1)hasattrrr)rPr+r+r,r=0s r=r9c Cs>tjrddl}ntd|j}|jdkr|jnt}|j}|j}d}|j ddrM|dvr7t d|d|dt d }||krLt d |d |d nB|tvrT|}n;|d kr||t|}t |tvryt |}t d |d |d n|d}|d ksX|durt d|dt|j||}dt|dS)Nrr0=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: r1)r r3r4r5r byteorder_NATIVE_BYTEORDERkinditemsizeshape TypeErrorr8warningswarn_VALID_IMAGE_ARRAY_DTPYESr:rZ fromarrayZastyper) r9rBr Zdtype_byteorderZ dtype_kindZdtype_itemsizeZ dest_dtypeZ dtype_strrPr+r+r,r<7sB   r<objscstjrddl}ntd|rIt|\}}t|tr dd|DSt|tjr3t t fdd|DSt||j j rGt t fdd|DS|S|S)zmEncode a list of objects into a format suitable for creating an extension array of type `ImageExtensionType`.rNr0cSs"g|] }|dur |ddndqS)Nr1r+r_objr+r+r,r`ns"z2objects_to_list_of_image_dicts..cg|]}|qSr+r+rZobj_to_image_dict_funcr+r,r`qcrr+r+rrr+r,r`tr) r r3r4r5rr6r:r8r;rr<rr=)rrB_rr+rr,objects_to_list_of_image_dictsbs    r)6r?sysr dataclassesrriortypingrrrrr r r numpyr8Zpyarrowrarr Zdownload.download_configrZ#download.streaming_download_managerrtablerZutils.file_utilsrZutils.py_utilsrrrr4rBrXrrr:rtrrr rrryr"rrvr=r;r<rr+r+r+r,s\  $      j  +