kh-ddlZddlZddlmZddlmZddlmZddlm Z m Z m Z m Z m Z mZmZddlmZddlmZGd d eZGd d eZGd deZy)N) defaultdict) HTMLParser)Path)AnyCallableDictListOptionalTupleUnion)default_loader) VisionDatasetceZdZdZdeeefddffd Zdedee ee efddfdZ deddfd Z d eddfd Z xZS) Flickr8kParserzBParser for extracting captions from the Flickr8k dataset web page.rootreturnNcht|||_i|_d|_d|_d|_y)NF)super__init__r annotationsin_table current_tag current_img)selfr __class__s W/var/www/teggl/fontify/venv/lib/python3.12/site-packages/torchvision/datasets/flickr.pyrzFlickr8kParser.__init__s8  24 *.*.tagattrsc,||_|dk(rd|_yy)NtableTrr)rrr s rhandle_starttagzFlickr8kParser.handle_starttags '> DM rc,d|_|dk(rd|_yy)Nr"Fr#)rrs r handle_endtagzFlickr8kParser.handle_endtag"s '>!DM rdatac|jr|dk(rd|_y|jdk(rp|jdd}tj j |j|dz}tj|d}||_g|j|<y|jdk(rF|jr9|j}|j|j|jyyyy)NzImage Not Founda/z_*.jpgrli) rrrsplitospathjoinrglobrappendstrip)rr'img_ids r handle_datazFlickr8kParser.handle_data(s ==((#' !!S(C,dii(1BC6*1-#) +-  (!!T)d.>.>))  (// =/?) r)__name__ __module__ __qualname____doc__r strrrr r r r$r&r5 __classcell__rs@rrr svL /U39- /$ /!3!tE#x}:L4M/N!SW! """ > > >rrc eZdZdZddefdeeefdedee dee de ege fddf fd Z d e de e e ffd Zde fd ZxZS) Flickr8ka?`Flickr8k Entities `_ Dataset. Args: root (str or ``pathlib.Path``): Root directory where images are downloaded to. ann_file (string): Path to annotation file. transform (callable, optional): A function/transform that takes in a PIL image or torch.Tensor, depends on the given loader, and returns a transformed version. E.g, ``transforms.RandomCrop`` target_transform (callable, optional): A function/transform that takes in the target and transforms it. loader (callable, optional): A function to load an image given its path. By default, it uses PIL as its image loader, but users could also pass in ``torchvision.io.decode_image`` for decoding image data into tensors directly. Nrann_file transformtarget_transformloaderrct||||tjj ||_t |j}t|j 5}|j|jddd|j|_ tt|jj|_||_y#1swYSxYw)Nr@rA)rrr.r/ expanduserr?rropenfeedreadrlistsortedkeysidsrB) rrr?r@rArBparserfhrs rrzFlickr8k.__init__Fs EUV**84   * $--  #B KK " #!--t//44678  # #s " CCindexc|j|}|j|}|j|j|}|j|}|j|j |}||fSz Args: index (int): Index Returns: tuple: Tuple (image, target). target is a list of captions for the image. )rLrBr@rrA)rrOr4imgtargets r __getitem__zFlickr8k.__getitem__Zsq%kk&! >> %..%C!!&)  ,**62FF{rc,t|jSNlenrLrs r__len__zFlickr8k.__len__p488}r)r6r7r8r9rr r:rr rrrintr rTrZr;r<s@rr>r>7s $)-/3'5 CIH%  #8,  #$  (sCx,rr>c eZdZdZddefdededeedeedeegefddf fd Z d e de eeffd Z de fd Z xZS) Flickr30ka8`Flickr30k Entities `_ Dataset. Args: root (str or ``pathlib.Path``): Root directory where images are downloaded to. ann_file (string): Path to annotation file. transform (callable, optional): A function/transform that takes in a PIL image or torch.Tensor, depends on the given loader, and returns a transformed version. E.g, ``transforms.RandomCrop`` target_transform (callable, optional): A function/transform that takes in the target and transforms it. loader (callable, optional): A function to load an image given its path. By default, it uses PIL as its image loader, but users could also pass in ``torchvision.io.decode_image`` for decoding image data into tensors directly. Nrr?r@rArBrct ||||tjj ||_t t|_t|j 5}|D]E}|jjd\}} |j|ddj| G dddtt|jj|_||_y#1swYBxYw)NrD r+)rrr.r/rEr?rrIrrFr3r-r2rJrKrLrB) rrr?r@rArBrNliner4captionrs rrzFlickr30k.__init__s EUV**84 't, $--  >B >"&**,"4"4T":  -44W= > > t//44678  > >s !A C..C7rOc0|j|}tjj|j|}|j |}|j |j |}|j|}|j|j|}||fSrQ) rLr.r/r0rrBr@rrA)rrOr4filenamerRrSs rrTzFlickr30k.__getitem__s%77<< 62kk(# >> %..%C!!&)  ,**62FF{rc,t|jSrVrWrYs rrZzFlickr30k.__len__r[r)r6r7r8r9rr:r rrrr\r rTrZr;r<s@rr^r^ts $)-/3'5 H%  #8,  #$  *sCx.rr^)r1r. collectionsr html.parserrpathlibrtypingrrrr r r r folderrvisionrrr>r^rrrmsI #"DDD"!(>Z(>V:}:z< <r