phette23

PNPM 11 doesn't migrate global bins

If you upgrade from pnpm v10 to v11, all your global pnpm binaries stop working because there is a new binary directory for PATH but none of the bins are moved to the new folder. This change is not mentioned in the upgrade guide nor does there appear to be a script or command to resolve it. pnpm setup adds $PNPM_HOME/bin to PATH but doesn't move the binaries in $PNPM_HOME into it. You have to manually reinstall all global binaries after upgrading.

$ uname -a
Darwin name 25.x.y Darwin Kernel Version 25.x.y: ... arm64

phette23

// run in browser console to test website release
fetch("https://libraries-libep.cca.edu/hours/?format=json&library=Simpson", {
  "headers": {
    "accept": "*/*",
  },
  "referrer": "https://library.cca.edu/",
})
  .then(response => response.json())
  .then(data => console.log(data))
  .catch(err => console.error(err))

phette23

<!-- goes in config section, replace existing useldapserver -->
<useldapserver>1</useldapserver>
  <ldapserver id="ldapserver">
    <hostname>ldap:10389</hostname>
    <base>ou=people,dc=planetexpress,dc=com</base>
    <user>cn=admin,dc=planetexpress,dc=com</user>
    <pass>GoodNewsEveryone</pass>
    <anonymous>0</anonymous>
    <replicate>1</replicate>
    <update>1</update>

phette23

Git Bz

• perl git bz see readme for more usage, I have had trouble with the perl version and sometimes switch to python
• koha wiki > git bz configuration

g checkout BUG # always work from a branch, helpful if it contains the BUG number
g bz apply BUG # apply patches from BUG to current branch
g bz edit BUG # lets you obsolete patches
g bz attach BUG HEAD # attach last commit to bug as patch

phette23

Recursive Git Garbage Collection

This can save some disk space, especially if you have many large git repos on your hard drive. Meant to be run from a parent directory where all code/projects are kept. Skips directories that aren't git repos.

phette23

Testing helm-invenio on Minikube

Testing helm-invenio chart locally using Minikube.

The values file uses the front matter starter image ghcr.io/front-matter/invenio-rdm-starter:v12.1.0.0 but that uses gunicorn for a web server instead of uwsgi so the web pod crashes. The invenio demo images are incompatible with minikube on apple silicon. The most complete test would be to use a locally built Invenio image or our images in Artifact Registry with an image pull secret.

Prerequisites

brew install minikube kubectl helm chart-testing

phette23

courses = require('./course_section_data_AP_Spring_2025.json')

// the course codes are not consistent so we can't use those
// some say senior project, senior thesis, senior game project, thesis project, and "fashion studio 5"
srprojects = courses.filter(c => c.status != "Preliminary")
    .filter(c => c.section_title.toLowerCase().match(/senior|thesis/))
    .filter(c => {
        let level = parseInt(c.section_code.match(/-(\d{4})-/)[1])
        return level < 5000 && level >= 4000
    })

phette23

Scrape Moodle Module List

On the Manage Activities admin page there's a count of how many courses use a particular activity and you can click the count to execute a search that retrieves all those courses. There's a button to show all the courses on the same page, but no useful way to do further filtering or extract data. The JS above can be copy-pasted into your browser's JavaScript console to create a CSV of the courses listed on the module list page.

We have parentheticals for semester in our course titles like "20th Century Fashion (2022SP)" so this attempts to extract those, though the regex is primitive and will mess up with courses with a second set of parentheses. This is usually easy to cleanup manually.

phette23

#!/usr/bin/env bash
# create a csv of path, file size, last access time for all files in a directory tree
# usage: file-tree-csv.sh /path/to/directory > output.csv
FILEDIR=/opt/moodledata/filedir
cd $FILEDIR || (echo "Moodle filedir not found, exiting" && exit 1)
echo "path,size,atime"
# shellcheck disable=SC2044
# all Moodle file names & paths only contain hexadecimal chars
for file in $(find . -type f); do
  size=$(stat -c %s $file)

phette23

#!/usr/bin/env python
# run from root of cca/vault_migration project over _all_ VAULT metadata JSON like
# `poetry run python missing-names.py vm/*.json`
import csv
import json
import os
import sys

import xmltodict