I have the good fortune to be a remote metadata intern for the Law Library of Congress this fall. As part of the project we are generating a large number of text files of bill summaries from the 74th Congress and naming them according to a well-structured schema, e.g., 74_2_s81_19360624.txt contains a summary of S. 81 from the 74th Congress, Second Session, whose last action took place on June 25, 1936.

We are listing these file names in a column in a spreadsheet on Google Sheets. Last night, I realized I mislabeled the session for bills whose last action took place in 1935. I needed to change the filenames in the spreadsheet and rename the files themselves on my hard drive. It took a little while to research how to do this, but it was ultimately far less tedious than doing it manually. Here’s what I did.

Google Sheets

=IF(REGEXMATCH(M14, "1935"), REPLACE(M14,4,1,"1"), M14)

If the text in cell M14 matches the text 1935, replace the fourth character in the string with the character 1, otherwise return the original text. REGEXMATCH may be overkill here, but it’s all I could think of late last night.

Shell

This was a little more involved.

#!/usr/bin/env zsh

# https://askubuntu.com/questions/406313/change-multiple-filenames-by-replacing-a-character
# https://stackoverflow.com/questions/6355011/replacing-one-char-with-many-chars-with-using-tr
for f in *1935*; do mv -v "$f" $(echo "$f" | sed 's/2/1/' ); done

I relied heavily on the two links in the comments above the code, but as I understand it,

for f in *1935*

loops through all the files in the directory that contain 1935 in their name.

mv -v "$f"  $(echo "$f" | sed 's/2/1/' )

does a verbose rename of each of those files, using sed to substitute the first occurrence of the digit 2 with the digit 1.

Expediting the Copy and Paste Workflow

The assignment involves a lot of copying and pasting from a PDF into a text file. Here’s a command I came up with to expedite that process. Before running it, the text from the PDF must be on the paste board.

FILE=74_2_s1476_19350507.txt && pbpaste > $FILE && code $FILE

This assigns a file name to the shell variable FILE, puts the contents of the paste board into that file, and then opens the same file in my text editor. It saves some mousework and keystrokes.

Final Thoughts

I’ve read up on this kind of stuff for years, so it’s nice to finally have a practical application for it. I am eager to learn more, especially about using shell scripting.