shell-scripts/subs2transcript - master - telnov.ob/cerberus

shell-scripts/subs2transcript
84 строки2 KB
Daniel J Wilcox
old shell scripts
21 янв 2026, 15:43
21 янв 2026, 15:43dfaba3d
#!/bin/sh

# convert subtitles to a transcript

# script usage
usage()
{
echo "\
# convert subtitles to a transcript

$(basename "$0") -i infile.(srt|vtt)"
exit 2
}

# error messages
NOTFILE_ERR='not a file'
INVALID_OPT_ERR='Invalid option:'
REQ_ARG_ERR='requires an argument'
WRONG_ARGS_ERR='wrong number of arguments passed to script'
FILE_EXT_ERR='has the wrong file extension'
NOT_TEXT_FILE_ERR='is not a text file'

# if script is run arguments pass and check the options with getopts,
# else display script usage and exit
[ $# -gt 0 ] || usage "${WRONG_ARGS_ERR}"

# getopts check and validate options
while getopts ':i:o:h' opt
do
  case ${opt} in
     i) infile="${OPTARG}"
	[ -f "${infile}" ] || usage "${infile} ${NOTFILE_ERR}";;
     o) outfile="${OPTARG}";;
     h) usage;;
     \?) usage "${INVALID_OPT_ERR} ${OPTARG}" 1>&2;;
     :) usage "${INVALID_OPT_ERR} ${OPTARG} ${REQ_ARG_ERR}" 1>&2;;
  esac
done
shift $((OPTIND-1))

# infile, infile name and extension
infile_nopath="${infile##*/}"
infile_name="${infile_nopath%.*}"
infile_ext="${infile##*.}"

# text file regular expressions for case statement
srt='[Ss][Rr][Tt]'
vtt='[Vv][Tt][Tt]'

# file command check input file mime type
infile_check="$(file --mime-type -b "${infile}")"

# text mimetype
subs_mime='text/plain'

# check the text file mime type is a text
case "${infile_check}" in
    ${subs_mime});;
    *) usage "${infile_check} ${NOT_TEXT_FILE_ERR}";;
esac

# outfile file recording destination
outfile_default="${infile_name}-transcript-$(date +"%Y-%m-%d-%H-%M-%S").txt"

# srt subs
srt_subs () {
sed 's/^\([0-9:.,[:space:]>-]\)*//;s/ \{1,\}/ /g;s/\.[[:space:]]/./g;/^$/d' "${infile}" \
| awk -v RS="." '!/^\n$/ { print $0 "." }' ORS='\n\n' \
| fmt -sp -w 120 > "${outfile:=${outfile_default}}"
}

# vtt subs
vtt_subs () {
sed '1,3d;s/^\([0-9:.[:space:]>-]\)*//;s/ \{1,\}/ /g;s/\.[[:space:]]/./g;/^$/d' "${infile}" \
| awk -v RS="." '!/^\n$/ { print $0 "." }' ORS='\n\n' \
| fmt -sp -w 120 > "${outfile:=${outfile_default}}"
}

# check the subtitle file extension before passing to the correct function
case "${infile_ext}" in
    ${srt}) srt_subs "${infile}";;
    ${vtt}) vtt_subs "${infile}";;
    *) usage "${infile} ${FILE_EXT_ERR}";;
esac