Mirakurun/mirakcチャンネル設定出力

scan_ch_mirak.sh

#!/bin/sh
# Mirakurun/mirakc channel yaml output  by uru (https://twitter.com/uru_2)
#
# require:
#   stz2012/epgdump
#   xmllint (libxml2 tools)
#   ISDB-T/S record software
#     recpt1, dvbv5-zap & Chinachu/dvbconf-for-isdb

set -eu
export LANG=C.UTF-8

# example: recpt1
readonly RECORD_COMMAND_GR='recpt1 --device /dev/px4video2 <channel> <record_second> <output>'
readonly RECORD_COMMAND_BSCS='recpt1 --device /dev/px4video0 --lnb 15 <channel> <record_second> <output>'
# example: dvbv5-zap
#readonly RECORD_COMMAND_GR='dvbv5-zap -a 1 -c dvbv5_channels_isdbt.conf -s -r -P <channel> -t <record_second> -o <output>'
#readonly RECORD_COMMAND_BSCS='dvbv5-zap -a 0 -c dvbv5_channels_isdbs.conf -s -r -P <channel> -t <record_second> -o <output>'
readonly CONF_MIRAKURUN='channels.yml'
readonly CONF_MIRAKC='channels_mirakc.yml'
readonly RECORD_SECOND=10

# dump epg
dump_epg() {
  ch="${1}"
  type="${2}"
  output_ts="${workdir}/${ch}.ts"
  output_xml="${workdir}/${ch}.xml"

  cmd_base=''
  if [ "${type}" = 'GR' ]; then
    cmd_base="${RECORD_COMMAND_GR}"
  elif [ "${type}" = 'BS' ] || [ "${type}" = 'CS' ]; then
    cmd_base="${RECORD_COMMAND_BSCS}"
  fi
  if [ -z "${cmd_base}" ]; then
    return 0
  fi

  echo "[Channel: ${ch}]" >&2

  # generate record command
  rec_cmd="$(echo "${cmd_base}" | sed -e "s/<channel>/${ch}/" -e "s/<record_second>/${RECORD_SECOND}/" -e "s;<output>;${output_ts};")"

  # execute record command
  set +e
  ${rec_cmd}
  set -e

  # dump epg
  opt_type=""
  if [ -f "${output_ts}" ]; then
    if [ "${type}" = 'GR' ]; then
      opt_type="${ch}"
    elif [ "${type}" = 'BS' ]; then
      opt_type='/BS'
    elif [ "${type}" = 'CS' ]; then
      opt_type='/CS'
    fi
    epgdump "${opt_type}" "${output_ts}" "${output_xml}"

    rm -f "${output_ts}"
  fi

  return 0
}

# parse epg
parse_epg() {
  epg_xml="${1}"
  output="${2}"
  div="${3}"

  for i in $(seq 1 "$(xmllint --xpath 'count(/tv/channel)' "${epg_xml}")"); do
    tp="$(xmllint --xpath "string(/tv/channel[${i}]/@tp)" "${epg_xml}")"
    service="$(xmllint --xpath "string(/tv/channel[${i}]/service_id/text())" "${epg_xml}")"
    name="$(xmllint --xpath "string(/tv/channel[${i}]/display-name/text())" "${epg_xml}")"

    if [ -z "${name}" ]; then
      continue
    fi

    # sort order
    sort_ch=''
    if [ "${div}" = '1' ]; then
      sort_ch="$(printf "%02d" "$(basename "${epg_xml}" .xml)")"
    elif [ "${div}" = '2' ]; then
      sort_ch="$(printf "%02d" "$(basename "${epg_xml}" .xml | cut -c 2-)")"
    elif [ "${div}" = '3' ]; then
      sort_ch="$(echo "${tp}" | sed -e 's/BS//' -e 's/_/ /' | awk '{printf("%02d%d", $1, $2)}')"
    elif [ "${div}" = '4' ]; then
      sort_ch="$(printf "%02d" "$(echo "${tp}" | cut -c 3-)")"
    fi
    sort_service="$(printf "%05d" "${service}")"

    echo "${div}|${sort_ch}|${sort_service}|${tp}|${service}|${name}" >> "${output}"

    if [ "${div}" = '1' ] || [ "${div}" = '2' ]; then
      # GR first service only
      break
    fi
  done

  return 0
}



# find require tools
if [ ! "$(which epgdump)" ]; then
  echo 'require epgdump' >&2
  return 1
fi
if [ ! "$(which xmllint)" ]; then
  echo 'require xmllint' >&2
  return 1
fi

# create working
workdir="$(mktemp -d)"

# dump epg GR channels
for c in $(seq 13 62); do
  ch="${c}"
  dump_epg "${ch}" GR
done
for c in $(seq 13 63); do
  ch="C${c}"
  dump_epg "${ch}" GR
done

# dump epg BS channels
for c in $(seq -w 1 2 23); do
  ch="BS${c}_0"
  dump_epg "${ch}" BS
done

# dump epg CS channels
for c in $(seq 2 2 24); do
  ch="CS${c}"
  dump_epg "${ch}" CS
done

ch_temp="${workdir}/scan_ch.tmp"
touch "${ch_temp}"

# parse GR epg
for f in $(find "${workdir}/" -type f -regex '.*\/[0-9]*\.xml$'); do
  parse_epg "${f}" "${ch_temp}" 1
done
for f in $(find "${workdir}/" -type f -regex '.*\/C[0-9]*\.xml$'); do
  parse_epg "${f}" "${ch_temp}" 2
done

# parse BS epg
for f in $(find "${workdir}/" -type f -regex '.*\/BS[0-9]*_[0-2]\.xml$'); do
  parse_epg "${f}" "${ch_temp}" 3
done

# parse CS epg
for f in $(find "${workdir}/" -type f -regex '.*\/CS[0-9]*\.xml$'); do
  parse_epg "${f}" "${ch_temp}" 4
done

# generate mirakurun config
rm -f mirakurun_ch.yml
sort < "${ch_temp}" | uniq | awk -F '|' '{
  type = "";
  if ($1 == "1" || $1 == "2") {
    type = "GR";
  } else if ($1 == "3") {
    type = "BS";
  } else if ($1 == "4") {
    type = "CS";
  }
  printf("- name: '\''%s'\''\n", $6);
  printf("  type: '\''%s'\''\n", type);
  printf("  channel: '\''%s'\''\n", $4);
  if ($1 == "3" || $1 == "4") {
    printf("  serviceId: %s\n", $5);
  }
  printf("\n");
}' >> "${CONF_MIRAKURUN}"

# generate mirakc config
rm -f mirakc_ch.yml
echo "channels:" >> mirakc_ch.yml
sort < "${ch_temp}" | uniq | awk -F '|' '{
  type = "";
  if ($1 == "1" || $1 == "2") {
    type = "GR";
  } else if ($1 == "3") {
    type = "BS";
  } else if ($1 == "4") {
    type = "CS";
  }
  printf("  - name: '\''%s'\''\n", $6);
  printf("    type: '\''%s'\''\n", type);
  printf("    channel: '\''%s'\''\n", $4);
  if ($1 == "3" || $1 == "4") {
    printf("    services: [%s]\n", $5);
  }
  printf("\n");
}' >> "${CONF_MIRAKC}"

# drop working
rm -rf "${workdir}"

echo 'channel scan done.' >&2
echo "output ${CONF_MIRAKURUN}, ${CONF_MIRAKC}" >&2
return 0

https://www.jifu-labo.net/2019/02/mirakurn_channels_config/ との相違点

• Perl5関連が不要
• recdvb, dvbv5-zapにも対応可能 (RECORD_COMMAND_GR, RECORD_COMMAND_BSCSの編集が必要)
  • recdvbは一部環境でチャンネルスキャン時に指定秒数で終了しない不具合が発生したため非推奨
• Mirakurunだけでなくmirakc向けの出力も行う
• channels.ymlの設定マージ機能はない (出力されたファイルを手でマージしてくれ)