#!/bin/sh # update feeds, merge with old feeds. # NOTE: assumes "sfeed_*" executables are in $PATH. # defaults sfeedpath="$HOME/.sfeed/feeds" # used for processing feeds concurrently: wait until ${maxjobs} amount of # feeds are finished at a time. maxjobs=8 # load config (evaluate shellscript). # loadconfig(configfile) loadconfig() { # allow to specify config via argv[1]. if [ "$1" != "" ]; then # get absolute path of config file. config=$(readlink -f "$1") else # default config location. config="$HOME/.sfeed/sfeedrc" fi # config is loaded here to be able to override $sfeedpath or functions. if [ -r "${config}" ]; then . "${config}" else echo "Configuration file \"${config}\" does not exist or is not readable." >&2 echo "See sfeedrc.example for an example." >&2 exit 1 fi } # convert encoding from one encoding to another. # convertencoding(from, to) convertencoding() { # if from != to if [ "$1" != "" ] && [ "$2" != "" ] && [ "$1" != "$2" ]; then iconv -cs -f "$1" -t "$2" 2> /dev/null else # else no convert, just output cat fi } # merge raw files: unique sort by id, title, link. # merge(name, oldfile, newfile) merge() { sort -t ' ' -u -k6,6 -k2,2 -k3,3 "$2" "$3" 2>/dev/null } # filter fields. # filter(name) filter() { cat } # order by timestamp (descending). # order(name) order() { sort -t ' ' -k1rn,1 } # fetch a feed via HTTP/HTTPS etc. # fetchfeed(name, url, feedfile) fetchfeed() { if curl -L --max-redirs 0 -H "User-Agent:" -f -s -m 15 \ -z "$3" "$2" 2>/dev/null; then printf " OK %s %s\n" "$(date +'%H:%M:%S')" "$1" >&2 else printf "FAIL %s %s\n" "$(date +'%H:%M:%S')" "$1" >&2 fi } # fetch and parse feed. # feed(name, feedurl, [basesiteurl], [encoding]) feed() { # wait until ${maxjobs} are finished: throughput using this logic is # non-optimal, but it is simple and portable. [ ${signo} -ne 0 ] && return [ $((curjobs % maxjobs)) -eq 0 ] && wait [ ${signo} -ne 0 ] && return curjobs=$((curjobs + 1)) (name="$1" filename="$(printf '%s' "$1" | tr '/' '_')" feedurl="$2" basesiteurl="$3" tmpfeedfile="${sfeedtmpdir}/${filename}" tmpencfile="" encoding="$4" sfeedfile="${sfeedpath}/${filename}" if [ "${encoding}" != "" ]; then fetchfeed "${name}" "${feedurl}" "${sfeedfile}" | \ convertencoding "${encoding}" "utf-8" else # detect encoding. tmpencfile="${tmpfeedfile}.enc" fetchfeed "${name}" "${feedurl}" "${sfeedfile}" > "${tmpencfile}" detectenc=$(sfeed_xmlenc < "${tmpencfile}") convertencoding "${detectenc}" "utf-8" < "${tmpencfile}" fi | sfeed "${basesiteurl}" | filter "${name}" > "${tmpfeedfile}" # get new data and merge with old. sfeedfilenew="${sfeedpath}/${filename}.new" # new feed data is non-empty. if [ -s "${tmpfeedfile}" ]; then # if file exists, merge if [ -e "${sfeedfile}" ]; then merge "${name}" "${sfeedfile}" "${tmpfeedfile}" | \ order "${name}" > "${sfeedfilenew}" # overwrite old file with updated file mv "${sfeedfilenew}" "${sfeedfile}" else merge "${name}" "/dev/null" "${tmpfeedfile}" | \ order "${name}" > "${sfeedfile}" fi fi) & } cleanup() { # remove temporary directory with feed files. rm -rf "${sfeedtmpdir}" } sighandler() { signo="$1" # ignore TERM signal for myself. trap -- "" TERM # kill all running childs >:D kill -TERM -$$ } feeds() { echo "Configuration file \"${config}\" is invalid or does not contain a \"feeds\" function." >&2 echo "See sfeedrc.example for an example." >&2 } # job counter. curjobs=0 # signal number received for parent. signo=0 # SIGINT: signal to interrupt parent. trap -- "sighandler 2" "INT" # SIGTERM: signal to terminate parent. trap -- "sighandler 15" "TERM" # load config file. loadconfig "$1" # fetch feeds and store in temporary directory. sfeedtmpdir="$(mktemp -d '/tmp/sfeed_XXXXXX')" # make sure path exists. mkdir -p "${sfeedpath}" # fetch feeds specified in config file. feeds # wait till all feeds are fetched (concurrently). [ ${signo} -eq 0 ] && wait # cleanup temporary files etc. cleanup # on signal SIGINT and SIGTERM exit with signal number + 128. [ ${signo} -ne 0 ] && exit $((signo+128)) exit 0