get_tr_pics.sh

#!/usr/bin/env bash

DEFAULT_TR_URL="https://cascadeclimbers.com/forum/topic/100929-tr-baker-river-to-ross-lake-in-technicolor-mystery-ndespair-pioneer-swiss-spectre-challenger-7252017/"

# url for a TR -> broken urls of photos in OP
function original_photo_urls () {
  curl -s $1 | 
    # All of the broken photo urls appear to contain 'plab/data/5'
    # as a substring
    egrep 'plab/data/5' | 
    # For a line containing double-quoted strings, extract just the
    # quoted strings
    egrep -o '\"[^ ]*\"' | 
    # Filter out the quoted strings that don't include 'http'. In the
    # case of an <img/> element, this gets rid of the quoted string for
    # the 'alt' attribute
    egrep 'http' | 
    # Remove the double quotes
    sed 's/"//g'
}

# url of broken photo -> working url
function fix_photo_url () {
  # We know the URL part of the DOM attribute is between
  # the initial 'img=' and the ampersand of the next argument
  # to the imageproxy
  echo $1 | egrep -o 'img=.*&' |
    # Get rid of the 'img=' and the '&'
    # Get rid of the 'www.'
    # Replace 'http' with 'https'
    sed -e 's/img=//g' \
        -e 's/&//g' \
        -e 's/www.//g' \
        -e 's/http/https/g'
}

export -f fix_photo_url

# Download a TR and get its photo urls. Then fix
# the urls and output them one per line. Then download
# the images from the fixed URLs and save them in the
# working directory under their original name.
original_photo_urls ${1:-$DEFAULT_TR_URL} | 
  xargs -I{} bash -c 'fix_photo_url "{}"' |
  parallel -j30 -v curl -s '{}' '>' {/}