o V h-@sddlZddlZddlmZddlmZddlmZddlm Z m Z m Z m Z m Z mZmZddlmZddlmZGd d d eZGd d d eZGd ddeZdS)N) defaultdict) HTMLParser)Path)AnyCallableDictListOptionalTupleUnion)default_loader) VisionDatasetcs|eZdZdZdeeefddffdd Zdedee ee efddfd d Z deddfd d Z d eddfddZ ZS)Flickr8kParserzBParser for extracting captions from the Flickr8k dataset web page.rootreturnNcs,t||_i|_d|_d|_d|_dS)NF)super__init__r annotationsin_table current_tag current_img)selfr __class__O/var/www/vscode/kcb/lib/python3.10/site-packages/torchvision/datasets/flickr.pyrs  zFlickr8kParser.__init__tagattrscCs||_|dkr d|_dSdS)NtableTrr)rrrrrrhandle_starttag zFlickr8kParser.handle_starttagcCsd|_|dkr d|_dSdS)NrFr )rrrrr handle_endtag"r"zFlickr8kParser.handle_endtagdatacCs|jrJ|dkr d|_dS|jdkr3|dd}tj|j|d}t|d}||_g|j |<dS|jdkrL|jrN|j}|j | | dSdSdSdS)NzImage Not Founda/z_*.jpgrli) rrrsplitospathjoinrglobrappendstrip)rr$img_idrrr handle_data(s   zFlickr8kParser.handle_data)__name__ __module__ __qualname____doc__r strrrrr r r!r#r1 __classcell__rrrrr s & rc seZdZdZddefdeeefdedee dee de ege fddf fd d Z d e de e e ffd d Zde fddZZS)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_transformloaderrcstj|||dtj||_t|j}t|j}| | Wdn1s,wY|j |_ t t |j |_||_dS)Nr:r;)rrr*r+ expanduserr9rropenfeedreadrlistsortedkeysidsr<)rrr9r:r;r<parserfhrrrrFs   zFlickr8k.__init__indexcCsN|j|}||}|jdur||}|j|}|jdur#||}||fSz Args: index (int): Index Returns: tuple: Tuple (image, target). target is a list of captions for the image. N)rEr<r:rr;)rrHr0imgtargetrrr __getitem__Zs       zFlickr8k.__getitem__cC t|jSNlenrErrrr__len__p zFlickr8k.__len__)r2r3r4r5r r r6rr rrrintr rLrRr7rrrrr87s(  r8c s|eZdZdZddefdededeedeedeegefddf fd d Z d e de eeffd d Z de fddZ ZS) 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. Nrr9r:r;r<rc stj|||dtj||_tt|_t |j"}|D]}| d\}} |j|dd | qWdn1s?wYtt |j|_||_dS)Nr= r')rrr*r+r>r9rrBrr?r/r)r.rCrDrEr<) rrr9r:r;r<rGliner0captionrrrrs   zFlickr30k.__init__rHcCs^|j|}tj|j|}||}|jdur||}|j|}|jdur+||}||fSrI) rEr*r+r,rr<r:rr;)rrHr0filenamerJrKrrrrLs       zFlickr30k.__getitem__cCrMrNrOrQrrrrRrSzFlickr30k.__len__)r2r3r4r5r r6r rrrrTr rLrRr7rrrrrUts( rU)r-r* collectionsr html.parserrpathlibrtypingrrrrr r r folderr visionrrr8rUrrrrs   $  +=