emre şahin's digital garden 🍃 - regex

bits 21

Emre Şahin — 2025-09-28T18:00:14+00:00

I wanted to get the number of word changes between commits in my blog. I asked for a git-diff-based Nushell pipeline from Gemini.

Gemini proposed this:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

What I ended up doing:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

I notice LLMs tend to err on the side of complexity. I don’t know if this is due to their training to spit out as many tokens as possible, but using regular expressions when a simple string replacement may suffice is a good sign that they may be adding more complexity than necessary.

bits 20

Emre Şahin — 2025-09-28T17:52:31+00:00

I wanted to get the number of word changes between commits in my blog. I asked for a git-diff-based Nushell pipeline from Gemini.

Gemini proposed this:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

What I ended up doing:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

bits 19

Emre Şahin — 2025-09-28T17:42:11+00:00

I wanted to get the number of word changes between commits in my blog. I asked for a git-diff-based Nushell pipeline from Gemini.

Gemini proposed this:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

What I ended up doing:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

bits 18

Emre Şahin — 2025-09-28T17:22:48+00:00

I wanted to get the number of word changes between commits in my blog. I asked for a git-diff-based Nushell pipeline from Gemini.

Gemini proposed this:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

What I ended up doing:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

bits 17

Emre Şahin — 2025-09-28T17:20:37+00:00

I wanted to calculate the number of words added between commits in my blog posts. I asked Gemini for a Nushell pipeline based on git diff:

It proposed the following:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

However, I ended up using this simplified version:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

I’ve noticed that LLMs often lean toward overly complex solutions. I’m not sure if this is due to being trained to generate as many tokens as possible, but using regular expressions where a simple string replacement would suffice is often a sign of unnecessary complexity.

bits 16

Emre Şahin — 2025-09-28T17:15:52+00:00

I wanted to get the number of word changes between commits in my blogs. I asked for a git diff based nushell pipeline from Gemini:

Gemini proposed this:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -r '\{\+|\+\}' '' | str join " " | str words | length

What I ended up doing:

git diff --word-diff=plain -- "*.md" | rg -o '\{\+(.*?)\+\}' | str replace -a "{+" "" | str replace -a "+}" "" | str replace -a ' ' "\n" | lines | length

I notice LLMs tend to err on the side of complexity. I don’t know if this is due to their training to spit as many tokens as possible, but using regular expressions when a simple single replace may suffice is a good sign that they may be adding more complexity than necessary.

TIL: sed

Emre Şahin — 2020-07-15T15:51:00+00:00

When I try to use sed to find and edit multiple files, I always remember that perl -pe is better suited for this task. Today, this happened again. I tried to find and replace lines starting with # Bla bla with title: Bla bla, and it was easier to use perl -pe 's|^#+ (.*)|title: $1|g' than identifying what kind of regular expressions sed uses.

It’s possible to get a section from a Markdown file with a command like:

sed -n -e '/^#/,/^#/p'

It is also possible to use line numbers instead of regexes. The p at the end is the print command, which can be replaced by, for example, d to delete the lines.