44 lines
1.2 KiB
Plaintext
44 lines
1.2 KiB
Plaintext
|
#!/bin/sh
|
||
|
|
||
|
if [ -z "$(command -v recode)" ]; then
|
||
|
echo "recode needed."
|
||
|
exit
|
||
|
fi
|
||
|
|
||
|
|
||
|
## Convert all 'bad' encoding to UTF-8/LF. WARNING: It will fail for encodings
|
||
|
## other the one explicitly supported below.
|
||
|
while read -r i; do
|
||
|
CODING=$(file "$i")
|
||
|
|
||
|
if [ -n "$(echo $CODING | grep 'ISO-8859')" ]; then
|
||
|
echo "ISO-8859: [$i]"
|
||
|
recode latin1..utf-8 "$i"
|
||
|
|
||
|
elif [ -n "$(echo $CODING | grep 'Non-ISO extended-ASCII')" ]; then
|
||
|
echo "cp1252: [$i]"
|
||
|
recode cp1252..utf-8 "$i"
|
||
|
|
||
|
elif [ -n "$(echo $CODING | grep 'UTF-16 Unicode text')" ]; then
|
||
|
echo "UTF-16: [$i]"
|
||
|
recode utf-16..utf-8 "$i"
|
||
|
|
||
|
elif [ -n "$(echo $CODING | grep 'UTF-8 Unicode (with BOM)')" ]; then
|
||
|
echo "UTF-8 BOM: [$i]"
|
||
|
sed -i '1s/^.//' "$i"
|
||
|
## sed -i is not the fastest depending on the implementations. The
|
||
|
## following commands work, but may be overkill.
|
||
|
# dd iflag=skip_bytes skip=3 if=file.srt of=temp.srt
|
||
|
# dd bs=1 skip=3 if=file.srt of=temp.srt
|
||
|
# tail -c +32 file.srt > temp.srt
|
||
|
fi
|
||
|
|
||
|
if [ -n "$(echo $CODING | grep 'CRLF')" ]; then
|
||
|
echo "CRLF: [$i]"
|
||
|
sed -i 's/\r//g' "$i"
|
||
|
fi
|
||
|
|
||
|
done <<EOF
|
||
|
$(find . -type f -size -50M -print)
|
||
|
EOF
|