import_oldnews.sh

   1 #!/bin/sh
   2 #
   3
   4 LC_ALL=C
   5 export LC_ALL
   6 LANG=C
   7 export LANG
   8 LANGUAGE=C
   9 export LANGUAGE
  10
  11 set -u
  12 set -e
  13 umask 0022
  14
  15 HEADLINES=${1}
  16 BODIES=${2}
  17
  18 LINES="$(cat ${HEADLINES} | grep '<li>')"
  19 NLINES=$(echo "${LINES}" | wc -l)
  20
  21 BODYH5="$(grep --line-number '<h5>' ${BODIES})"
  22
  23 for i in $(seq 1 $NLINES); do
  24         l=$(echo "${LINES}" | head -${i} | tail -1 | sed -e 's!^[\t ]*!!')
  25         d=$(echo "${l}" | sed -e 's!^<li>\([^<]*\)<a href=.*!\1!')
  26         n=$(echo "${l}" | sed -e 's!.*a href="#\([^"]*\)".*!\1!')
  27         utctime=$(date --date="${d}" --utc +"%Y%m%d-%H%M%S")
  28         headlinefile="posted_news/${utctime}.${n}.headline.html"
  29         bodyfile="posted_news/${utctime}.${n}.body.html"
  30         echo "${i}: ${headlinefile}"
  31         {
  32                 echo "<!-- BEGIN: ${headlinefile} -->"
  33                 echo "${l}"
  34                 echo "<!-- END: ${headlinefile} -->"
  35         } > ${headlinefile}
  36
  37         boundary=$(echo "${BODYH5}" | grep -A1 "<h5><a name=\"${n}\">" | cut -d ':' -f1)
  38         tmp=$(echo "${boundary}" | wc -l)
  39         case "${tmp}" in
  40         2)
  41                 begin=$(echo "${boundary}" | head -1)
  42                 next=$(echo "${boundary}" | tail -1)
  43                 end=$(expr ${next} - 1)
  44                 len=$(expr ${end} - ${begin})
  45                 ;;
  46         1)
  47                 len="10000"
  48                 ;;
  49         *)
  50                 echo "invalid boundary:"
  51                 echo "${boundary}"
  52                 exit 1
  53                 ;;
  54         esac
  55         echo "${i}: ${bodyfile}"
  56         {
  57                 echo "<!-- BEGIN: ${bodyfile} -->"
  58                 grep -A${len} "<h5><a name=\"${n}\">" ${BODIES}
  59                 echo "<!-- END: ${bodyfile} -->"
  60         } > ${bodyfile}
  61 done