#!/bin/bash

# file: getwiki.sh
# Pete Nesbitt, Jan 2016 

# Description: script to display the initial summary from wikipedia.
# uses curl to pull initial results


## set some vars for text high-lighting
normal=`tput sgr0`
bold=`tput bold`

######
clear
echo ""
echo ""

# Get search term from user ($WORD)
printf "   => Enter a Term to lookup on Wikipedia:  "
read WORD


# cleanup: rm extra spaces, push all lower, replace spaces with underscores
KEYWORD=`echo "$WORD" |tr -s [:blank:] |tr -t [:upper:] [:lower:] | sed 's/\ /_/g'`

#cleaned up search word but still has spaces for text high-light
FOUND_WORD="`echo "$WORD" |tr -s [:blank:] |tr -t [:upper:] [:lower:]`"

# wiki is using several page layout/styles/classes for summary area.
# Since a "class" type search sometimes fails, will do messier cleanup...

# first, KEYWORD has first letter Uppercased using a ^ at end of var,
# then find 1st 3 paragraphs (1 or 2 is not enough in some cases due to summary info etc),
# then clean out html & blank lines, break up with line returns, then indent

CONTENT_AREA="`curl --silent https://en.wikipedia.org/wiki/${KEYWORD^} | grep -m 3 "<p>" | sed 's/</\n</g' | sed 's|<.*>||g' | tr -d '\n' | tr -d '\t' | sed s/"${FOUND_WORD}"/${bold}\&${normal}/gI | fold -s | sed 's/^/\ \ \ /'`"

# validate word since any number of auto pages may be generated.
echo "${CONTENT_AREA}" | egrep "Other reasons this message may be displayed:|Wikipedia does not have an article with this exact name" &> /dev/null
RET_BAD=$?

if [ ${RET_BAD} -eq 0 ];then
   CONTENT_AREA="   Sorry, there was No Wiki Page Found for:   ${WORD}"
fi

## display results:
echo ""
echo "The Wiki Summary is:"
echo ""

printf "$CONTENT_AREA"
echo ""
echo ""

# only display valid links:
if [ ${RET_BAD} -ne 0 ];then
   echo " Read entire page at https://en.wikipedia.org/wiki/${KEYWORD^}"
   echo ""
fi


# eof