根据文件中的时间戳对文件进行排序

Question

I have a bunch of files which contain an ascii header with a time stamp WITHIN the file, followed by a large chunck of binary data. I would like to list the files sorted by this time stamp, at the command line (bash, etc).

The file headers look similar to the following:

encoding: raw
endian: big
dimension: 4
sizes: 128 128 1 4
spacings: 1.0 1.0 1.0 NaN
position: -3164,-13678
date_time: 06.02.12.18:59
user_name: Operator1
sample_name: 
dwell_time: 4.000
count_time: 65.536
duration: 202.000
raster: 79912
pixel_width: 624.3125
pixel_height: 624.3125

....binary data....

I would like to sort based on the "date_time" time stamp, which uses the format dd.mm.yy.hh:mm

The sort --key option looks promising but all my attempts have failed. Any help is much appreciated. Thanks.

Answer 1

Assuming these files are images, so you can use a tool like exiftool to rename them based on their creation dates, then sort them by name.

If you can't rename them, just dump the names with the creation date to STDOUT and sort that, eg:

exiftool -p '$dateTimeOriginal $filename' -q -f DIRECTORY/WHERE/IMAGES/ARE | sort -n 

If you only want the filenames in the output, append a | cut -f 2 -d " " | cut -f 2 -d " " to the end.

If it's file format which is not recognized by exiftool this might or might not work:

for f in YOURFILES* ; do
    filedate=`grep --binary-file=text -i -o 'date_time: ...........:..' $f | head -1`
    echo "$filedate $f"
done | sort -n

Note : this won't work when there are spaces in the filenames (and I'm leaving that to you to resolve). And if you want to output only the sorted filenames, append | awk '{print $NF}' | awk '{print $NF}' after the sort -n .