scripts.md

June 14, 2026 · View on GitHub

Scripts

The following scripts are under src/cldwalker/bb_clis/bin/:

gh-pr-for-commit

Prints github url of PR associated with a commit. It assumes a current directory's repository but any repository can be specified. See https://github.com/mislav/hub-api-utils/blob/master/bin/hub-pr-with-commit for an alternative implementation.

Setup

Usage

# Prints url and opens url for osx
$ ./gh-pr-for-commit -r atom/atom 0f521f1e8afbcaf73479ea93dd4c87d9187903cb
"https://github.com/atom/atom/pull/20350"

# Open url of current github repository
$ ./gh-pr-for-commit SHA

gh-repo

For the current github repo, open different repo urls e.g. commit or branch. Inspired by this ruby version.

To open a commit, gh-repo -c SHA.

bb-table

Usage

Prints an ascii table given an EDN collection on stdin or as a file:

$ echo '[{:a 4 :b 2} {:a 2 :c 3}]' | bb-table

| :a | :b | :c |
|----+----+----|
|  4 |  2 |    |
|  2 |    |  3 |

$ bb-table -f something.edn
...

bb-project-clj

Prints a project.clj defproject form as a map. Useful for manipulating this data on the commandline

# Pretty prints a project's dependencies
$ bb-project-clj -d 1 | bb -I '(-> *input* first :dependencies clojure.pprint/pprint)'

bb-replace

Replaces a substring in a file using a regex to match it. Much less powerful than sed but more user friendly as it supports configuring and naming regexs. bb-replace reads configs found in ~/.bb-replace.edn and ./bb-replace.edn. See script for documentation on config format.

# Use the default name replacements provided
$ cp .bb-replace.edn ~/.bb-replace.edn

# Navigate to a lein project and update project's version
$ bb-replace lein-version 1.2.1

# Navigate to a nodejs project and update project's version
$ bb-replace json-version 2.1.1

# A one-off regex can be used. This updates a map entry to false
$ bb-replace -f project.clj -F '\$1 %s' "(:pseudo-names)\s+\w+" false

bb-try

Try a Clojure library easily with bb. Inspired by https://github.com/avescodes/lein-try.

# Adds latest version of this library to classpath and starts repl
$ bb-try camel-snake-kebab

# Any additional arguments are passed on to bb
$ bb-try camel-snake-kebab "(require '[camel-snake-kebab.core :as csk]) (csk/->SCREAMING_SNAKE_CASE :babashka-classpath)"
:BABASHKA_CLASSPATH

Currently it only fetches the latest version of a library but I'm thinking of making a version optional.

bb-ns-dep-tree

Print the ns dependency tree for a given ns or file. For example, if we want to print the dependencies of datascript.datafy:

# On a local checkout of datascript
$ bb-ns-dep-tree src/datascript/datafy.cljc
datascript.datafy
├── clojure.core.protocols
├── datascript.pull-api
│   ├── datascript.db ...
│   └── datascript.pull-parser
│       └── datascript.db ...
├── datascript.db
│   ├── clojure.walk
│   ├── clojure.data
│   ├── me.tonsky.persistent-sorted-set
│   └── me.tonsky.persistent-sorted-set.arrays
└── datascript.impl.entity
    ├── clojure.core
    └── datascript.db ...

# We can also print the cljs dependencies of the same ns
$ bb-ns-dep-tree -l cljs datascript.datafy
datascript.datafy
├── clojure.core.protocols
├── datascript.pull-api
│   ├── datascript.db ...
│   └── datascript.pull-parser
│       └── datascript.db ...
├── datascript.db
│   ├── goog.array
│   ├── clojure.walk
│   ├── clojure.data
│   ├── me.tonsky.persistent-sorted-set
│   ├── me.tonsky.persistent-sorted-set.arrays
│   └── datascript.db ...
└── datascript.impl.entity
    ├── cljs.core
    └── datascript.db ...

bb-update-lein-dependency

Updates lein dependency of specified directories and optionally commits and pushes the change. For example, if I'm in the dependency's directory and I want to update two dependent projects to use its latest SHA, commit and git push:

bb-update-lein-dependency -c -d ../proj1 -d ../proj2 my-dep $(git rev-parse HEAD).

bb-logseq-convert

Given a url, returns auto populated properties as a logseq block. The properties are derived from the url's rdf data and what is converted to the logseq block is highly configurable. My config of these scripts are in this directory.

Logseq CLIs

There are several logseq commands to try. These are all for the DB version. Here are just a few:

  • logseq-bookmark - Quickly adds bookmarks with my preferred ontology
  • logseq-graph-backup - Backs up one or more graphs using export of build EDN. Has a number of options to ensure export data is idempotent and bug free.
  • logseq-validate-* - CLIs with this prefix are validations that most of my graphs have to pass. I run these with a pre-commit hook
  • logseq-search-* - CLIs that search, let users pick specific nodes from the search results and then apply a subsequent action on them