nJptcPaq

#!/bin/bash
# Download MTGO Vintage results from magic.wizards.com

# Get script's directory
ROOT=$(dirname $(readlink -f ${0}))

# Store the date of the most recent results in the script's root
DATE_FILE="${ROOT}/vintage-last-checked"

# If ./last-checked exists, get the date of the last results. Otherwise,
# don't restrict the search window. By default, this appears to return the
# most recent 6 results.
FROM_DATE=""
[ -f ${DATE_FILE} ] && FROM_DATE=$(egrep "^[0-9/]{10}$" ${DATE_FILE})

# URL definitions. Use magic.wizards.com for both searches and event links.
URL_BASE="https://magic.wizards.com/en/"
SEARCH_PATH="section-articles-see-more-ajax?l=en&f=9041&search-result-theme=&fromDate=${FROM_DATE}&word=Vintage"
RESULT_PATH="articles/archive/mtgo-standings"

# Query the recent tournament results.  Clean up the output enough to extract
# links.  Sort them chronologically, ascending.
#
# Note: We must reverse the results manually intead of using the website's
# search function because reversing the sort in the curl request returns results
# from ~2014 if no fromDate is defined.
SEARCH=$(curl -s "${URL_BASE}${SEARCH_PATH}" \
             | sed -r "s/\\\n/\n/g" \
             | egrep -o "href.*$" \
             | sed -r "s/^.*mtgo-standings(.+)\\.*$/\1/g" \
             | awk -F\\ '{print $2}' \
             | tr '\n' ' ' \
             | tac -rs' ')

# Loop over the results.  Extract the date from the end for later use, then
# format the result as a link and print it.
for S in ${SEARCH}; do
    D=$(echo "${S}" | egrep -o "[0-9-]{10}$")

    # Clean up the result: strip the leading '/', convert to proper case,
    # and replace dashes with spaces.
    T=$(echo "${S:0:-10}" | tr -d '/' | tr '-' ' ' | sed 's/[^ ]\+/\L\u&/g')
    echo "${T}${D}"
done

# Reformat the date of the last entry and save it to ${DATE_FILE}
[ ! -z ${D} ] && date -d "${D}+1 day" "+%m/%d/%Y" > ${DATE_FILE} 2>/dev/null