o z3iC@sUddlZddlZddlZddlmZmZddlmZddlm Z ddl m Z m Z m Z mZmZddlZddlZddlmZddlmZdd lmZdd lmZmZdd lmZmZm Z e rgddl!Z"d d l#m$Z$da%ee&e'e(d<ej)dkrxdndZ*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)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|u1zu2zi2zu4zi4zf4zf8c@s*eZdZUdZdZeeed<dZe ed<e dddZ eeed<d Z e eed <eeed Ze eed <e dddd Zeed<ddZdeeeeeejd fdefddZddedd fddZdedeedfffddZdeejej ej!fdej fddZ"ddej dej fddZ#dS) 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)rinitr_typecCs|jSN)r&)selfr+e/lsinfo/ai/hellotax_ai/llm_service/venv_embed/lib/python3.10/site-packages/datasets/features/image.py__call___szImage.__call__valuereturncCstjrddl}ntdt|trt|}t|tr |ddSt|t r.t| ddSt|t t fr:d|dSt|tj rDt|St||jjrOt|S|ddurftj|drfd|ddS|ddust|ddur|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$zUAn image sample should have one of 'path' or 'bytes' but they are missing or None in .)r PIL_AVAILABLE PIL.Image ImportError isinstancelistnparraystrrabsoluter$ bytearrayndarrayencode_np_arrayrencode_pil_imagegetosr%isfile ValueError)r*r.PILr+r+r,encode_examplebs,         zImage.encode_examplec Cs~|jstdtjrddl}ddl}ntd|duri}|d|d}}|dur|dur6td|dt|rA|j |}nV| d d }| tj rQtjntj}t||} | durd|| d nd} t| d } t|d | d} t| }Wdn1swY|j |}n|j t|}|||j jjjdur|j|}|jr|j|jkr||j}|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$zCAn image should have one of 'path' or 'bytes' but both are None in r2::repo_idtokenrbdownload_config)r RuntimeErrorr r3r4 PIL.ImageOpsr5rCrropensplit startswith HF_ENDPOINTHUB_DATASETS_URLHUB_DATASETS_HFFS_URLrr@r rrreadloadgetexifExifTagsBase OrientationImageOpsexif_transposerconvert) r*r.token_per_repo_idrDr%bytes_image source_urlpatternsource_url_fieldsrJrMfr+r+r,decode_examplesD      zImage.decode_examplercCs(ddlm}|jr |S|d|ddS)zfIf in the decodable state, return the feature itself, otherwise flatten the feature into a dictionary.r)Valuebinarystringr#)featuresrgr)r*rgr+r+r,flattens z Image.flattenstoragec CsLtj|jr&z |t}Wntjy%}ztd||d}~wwtj|jrKtj dgt |t d}tj j ||gddg|d}ntj|jrwt|t }tj dgt |td}tj j ||gddg|d}ntj|jrtj dgt |td}tj j ||gddg|d}ntj|jr|jddkr|d}ntj dgt |t d}|jddkr|d}ntj dgt |td}tj j ||gddg|d}n5tj|jr tj dd |Dt d}tj dgt |td}tj j ||gddg|d}t||jS) aCast an Arrow array to the Image arrow storage type. The Arrow types that can be converted to the Image pyarrow storage type are: - `pa.string()` - it must contain the "path" data - `pa.large_string()` - it must contain the "path" data (will be cast to string if possible) - `pa.binary()` - it must contain the image bytes - `pa.struct({"bytes": pa.binary()})` - `pa.struct({"path": pa.string()})` - `pa.struct({"bytes": pa.binary(), "path": pa.string()})` - order doesn't matter - `pa.list(*)` - it must contain the image array data Args: storage (`Union[pa.StringArray, pa.StructArray, pa.ListArray]`): PyArrow array to cast. Returns: `pa.StructArray`: Array in the Image arrow storage type, that is `pa.struct({"bytes": pa.binary(), "path": pa.string()})`. zvFailed to cast large_string to string for Image feature. This can happen if string values exceed 2GB. Original error: Ntyper$r%maskrcSs*g|]}|durtt|dndqS)Nr$)r>r8r9).0arrr+r+r, s*z&Image.cast_storage..)patypesis_large_stringrncastri ArrowInvalidrC is_stringr9lenrh StructArray from_arraysis_nullis_large_binaryr is_binary is_structget_field_indexris_list to_pylistr&)r*rle bytes_array path_arrayr+r+r, cast_storagesV       zImage.cast_storagecsduritfddtjfdd|Dtd}tjdd|dDtd}tjj||gd dg| 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()})`. Ncs|dd}|tjrtjntj}t||}|dur#|dnd}t|d}t |d|d }| WdS1s@wYdS)NrFrGrHrIrKrL) rQrRr rSrTrUrr@r rrV)r%rbrcrdrJrMre)r_r+r, path_to_bytes!s  $z*Image.embed_storage..path_to_bytescs8g|]}|dur|ddur|dn|dndqS)Nr$r%r+)rqx)rr+r,rs.s*z'Image.embed_storage..rmcSs$g|]}|durtj|ndqSr))rAr%basename)rqr%r+r+r,rs5s$r%r$ro) rrtr9rrhrrir{r|r}rr&)r*rlr_rrr+)rr_r, embed_storages    zImage.embed_storager))$__name__ __module__ __qualname____doc__rr r:__annotations__rboolrr r"rrtstructrhrir&rr(r-r r$r<dictr8r=rErfrk StringArrayr{ ListArrayrrr+r+r+r,r.s ' $$)=$ >rr/cCsRtjrddl}ntdtdur'|jtt|jj t|jj @atS)Nrr0) r r3r4r5rrr'r7setOPENkeysSAVE)rDr+r+r,list_image_compression_formats<s  $rrar!cCsBt}|jtvr |j}n |jdvrdnd}|j||d|S)zmConvert a PIL Image object to bytes using native compression if possible, otherwise use PNG/TIFF compression.)1LLARGBRGBAPNGTIFF)format)rrrrsavegetvalue)rabufferrr+r+r,image_to_bytesIs  rcCs.t|dr|jdkr|jddSdt|dS)Nfilenamer1)hasattrrr)rar+r+r,r?Ts r?r9c Cs>tjrddl}ntd|j}|jdkr|jnt}|j}|j}d}|j ddrM|dvr7t d|d|dt d }||krLt d |d |d nB|tvrT|}n;|d kr||t|}t |tvryt |}t d |d |d n|d}|d ksX|durt d|dt|j||}dt|dS)Nrr0=r )uizUnsupported array dtype z for image encoding. Only z' is supported for multi-channel arrays.rzDowncasting array dtype z to z to be compatible with 'Pillow'rzCannot downcast dtype z- to a valid image dtype. Valid image dtypes: r1)r r3r4r5r" byteorder_NATIVE_BYTEORDERkinditemsizeshape TypeErrorr8warningswarn_VALID_IMAGE_ARRAY_DTPYESr:r fromarrayastyper) r9rDr"dtype_byteorder dtype_kinddtype_itemsize dest_dtype dtype_strrar+r+r,r>[sB   r>objscstjrddl}ntd|rIt|\}}t|tr dd|DSt|tjr3t t fdd|DSt||j j rGt t fdd|DS|S|S)zmEncode a list of objects into a format suitable for creating an extension array of type `ImageExtensionType`.rNr0cSs"g|] }|dur |ddndqS)Nr1r+rqobjr+r+r,rss"z2objects_to_list_of_image_dicts..cg|]}|qSr+r+robj_to_image_dict_funcr+r,rscrr+r+rrr+r,rsr) r r3r4r5rr6r:r8r=rr>rr?)rrD_rr+rr,objects_to_list_of_image_dictss    r)6rAsysr dataclassesrriorpathlibrtypingrrrr r numpyr8pyarrowrtrr download.download_configr tablerutils.file_utilsrrutils.py_utilsrrrr4rDrjrrr7r:rrrr"rrrr$rrr?r=r>rr+r+r+r,s^         +