Apple Terminal script to copy files based on a .csv list of filenames?

Question

Been trying for hours to get this to work - but I'm a complete novice (e.g. started using terminal today) and struggling to get this working.

I have a csv file that lists filenames row by row. I want to find them and copy them to a new directory. The files are listed in one directory, but each file is in its own subdirectory.

This is what I have tried, but it is not copying anything. But I can see it cycling through the searches:

    cat /Users/Will/Downloads/three-new.csv | while IFS=, read col1
do
    find /Volumes/LaCie\ 8tb/BPOTY/2021\ Photos -path $col1 -exec cp {} /Volumes/LaCie\ 8tb/BPOTY/2021\ book \;
done

I have also tried this, but it just copies all files from the directory (stripping the subdirectories) and doesn't find the actual filenames.

cat /Users/Will/Downloads/three-new.csv | while IFS=, read col1
do
    find /Volumes/LaCie\ 8tb/BPOTY/2021\ Photos $col1 -exec cp {} /Volumes/LaCie\ 8tb/BPOTY/2021\ book \;
done

The file can be seen here: https://www.dropbox.com/s/6uqwzky43r25ors/three.csv?dl=0

Have run a command as requested and the CSV outputs:

00000000: 636f 6c31 0d0a 494d 475f 3437 3538 2d64  col1..IMG_4758-d
00000010: 6e2e 6a70 670d 0a43 4f45 5f32 3330 352d  n.jpg..COE_2305-
00000020: 4564 6974 2e74 6966 2e7a 6970 0d0a 5f44  Edit.tif.zip.._D
00000030: 5343 3435 3933 2d68 6972 6573 2d72 6573  SC4593-hires-res
00000040: 6f6c 7574 696f 6e2d 5065 7465 722d 6365  olution-Peter-ce
00000050: 6368 2d54 6865 2d6c 6f6f 6b2e 6a70 67    ch-The-look.jpg

Answer 1

Since your are doing this under macOS and the file has CRLF endings, first lets remove the carriage returns from the three.csv file, in Terminal:

tr -d '\r' < three.csv > new.csv

As the three.csv file only has one column and there are actually no commas, nor are the entries quoted, this is really not a CSV file and can be treated in a more straight forward manner. To that end, use the following shell script syntax to accomplish your goal.

#!/bin/bash

while IFS= read -r line; do
    cd '/path/to/parent_dir' || exit
    find . -type f -iname "$line" -exec cp -av {} '/path/to/destnation_dir' \;
done < '/path/to/new.csv'

---

Notes:

• The cd '/path/to/parent_dir' || exit isn't absolutely necessary, it's just my preferred method in many use cases. The . after find can be replaced with '/path/to/parent_directory'.
• Pathnames with spaces need to either be escaped with a backslash character or the pathname be quoted, single, or double if containing variables, however, not both. As a general rule I almost alway quote non-variable strings with single-quotes. (I did use double-quotes last night on non-variable strings but that was a typo.)
• If executing directly in Terminal then the #!/bin/bash shebang isn't necessary and was included if using it as a standalone shell script.
• The last line of the three.csv file, and thus the new.csv file does not have a line feed character. Edit the file by placing the cursor at the end of line three and press enter, then save the file. Otherwise the last line is not processed.
• As mentioned in the comments of the OP by Gordon Davisson, quote your variables and use ShellCheck to check your bash shell script code.