a h@sddlZddlZddlmZddlmZddlmZddlm Z m Z m Z m Z ddl mZddlmZGd d d eZGd d d eZGd ddeZdS)N) defaultdict) HTMLParser)Path)AnyCallableOptionalUnion)default_loader) VisionDatasetcsreZdZdZeeefddfdd Zeee ee efddddZ edd d d Z edd d dZ ZS)Flickr8kParserzBParser for extracting captions from the Flickr8k dataset web page.N)rootreturncs,t||_i|_d|_d|_d|_dS)NF)super__init__r annotationsin_table current_tag current_img)selfr  __class__I/var/www/auris/lib/python3.9/site-packages/torchvision/datasets/flickr.pyrs  zFlickr8kParser.__init__)tagattrsrcCs||_|dkrd|_dS)NtableTrr)rrrrrrhandle_starttagszFlickr8kParser.handle_starttag)rrcCsd|_|dkrd|_dS)NrFr)rrrrr handle_endtag"szFlickr8kParser.handle_endtag)datarcCs|jr|dkrd|_nv|jdkrb|dd}tj|j|d}t|d}||_g|j |<n*|jdkr|jr|j}|j | | dS)NzImage Not Founda/z_*.jpgrli) rrrsplitospathjoinr globrappendstrip)rr img_idrrr handle_data(s  zFlickr8kParser.handle_data)__name__ __module__ __qualname____doc__rstrrrlisttuplerrrr- __classcell__rrrrr s  "r csxeZdZdZddefeeefeee ee e ege fddfdd Z e e e e fdddZe d d d ZZS) 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. Nr ann_file transformtarget_transformloaderrcstj|||dtj||_t|j}t|j}| | Wdn1sX0Y|j |_ t t |j |_||_dS)Nr9r:)rrr&r' expanduserr8r r openfeedreadrr3sortedkeysidsr;)rr r8r9r:r;parserfhrrrrFs  ,zFlickr8k.__init__indexrcCsN|j|}||}|jdur(||}|j|}|jdurF||}||fSz Args: index (int): Index Returns: tuple: Tuple (image, target). target is a list of captions for the image. N)rCr;r9rr:)rrGr,imgtargetrrr __getitem__Zs       zFlickr8k.__getitem__rcCs t|jSNlenrCrrrr__len__pszFlickr8k.__len__)r.r/r0r1r rr2rrrrrintr4rKrQr5rrrrr67s  r6cspeZdZdZddefeeeeeeeegefddfdd Z e e eefdddZ e d d d Z 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. Nr7c stj|||dtj||_tt|_t |jD}|D].}| d\}} |j|dd | q:Wdn1s~0Ytt |j|_||_dS)Nr< r#)rrr&r'r=r8rr3rr>r+r%r*rArBrCr;) rr r8r9r:r;rEliner,captionrrrrs  8zFlickr30k.__init__rFcCs^|j|}tj|j|}||}|jdur8||}|j|}|jdurV||}||fSrH) rCr&r'r(r r;r9rr:)rrGr,filenamerIrJrrrrKs       zFlickr30k.__getitem__rLcCs t|jSrMrNrPrrrrQszFlickr30k.__len__)r.r/r0r1r r2rrrrrRr4rKrQr5rrrrrSts rS)r)r& collectionsr html.parserrpathlibrtypingrrrrfolderr Zvisionr r r6rSrrrrs     +=