#!/bin/bash 

# checkexternal-traverse-T all URLs on a Webwsite to build a
#   list of external references, then check each one to ascertain
#   which might be dead or otherwise broken. The -a flag forces the
#   script to list all matches, whether they're accessible or not; by
#   default, only unreachable links are shown.

setglobal listall = '0'; setglobal errors = '0'; setglobal checked = '0'

if test $1 = "-a"  {
  setglobal listall = '1'; shift
}

if test -z $1  {
  echo "Usage: $[basename $0] [-a] URL" > !2 ; exit 1
}

trap "$[which rm] -f traverse*.errors reject*.dat traverse*.dat" 0

setglobal outfile = ""$[echo $1 | cut -d/ -f3].errors.ext""
setglobal URLlist = ""$[echo $1 | cut -d/ -f3 | sed 's/www\.//].rejects""

rm -f $outfile     # Ready for new output

if test ! -e $URLlist  {
  echo "File $URLlist not found. Please run checklinks first." > !2; exit 1
}

if test ! -s $URLlist  {
  echo "There don't appear to be any external links ($URLlist is empty)" > !2
  exit 1
}

#### Now, finally, we're ready to begin...

for URL in [$[cat $URLlist | sort | uniq]]
{
  curl -s $URL > /dev/null !2 > !1; setglobal return = $Status
  if test $return -eq 0  {
    if test $listall -eq 1  {
      echo "$URL is fine."
    }
  } else {
    echo "$URL fails with error code $return"
    setglobal errors = $shExpr(' $errors + 1 ')
  }
  setglobal checked = $shExpr(' $checked + 1 ')
}

echo ""
echo "Done. Checked $checked URLs and found $errors errors."
exit 0