lightsystemdarkgreen
@ 369f607 · fetchingraw

AGENTS.md

Guidance for AI coding agents (and humans) working in the ack3 repository.

What this project is

ack is a grep-like code-searching tool optimized for programmers searching large source trees. It is written in Perl, distributed on CPAN as ack (modules under App::Ack::), and also built as a single-file “standalone” script. The author and maintainer is Andy Lester. Project home: https://beyondgrep.com/. The license is the Artistic License 2.0.

The current version is in lib/App/Ack.pm ($VERSION) and in the ack script itself. Current release line: v3.x.

Repository layout

Generated files (not committed; already in .gitignore): Makefile, MYMETA.*, blib/, ack-standalone, pm_to_blib, Build, *~.

Hard design constraints

These are settled design decisions (see DESIGN.md). Do not violate them without an explicit decision from the maintainer.

Building and testing

Prerequisite check:

perl -MFile::Next -E'say "ack is ready to build!"'

Build and test:

perl Makefile.PL
make
make test            # runs test_classic and test_standalone

Individual targets:

make test_classic     # tests against lib/ via blib/
make test_standalone  # tests against the built ack-standalone
make test_xt          # author tests in xt/
make fulltest         # classic + standalone + xt
make pytest           # runs the YAML-driven tests via t/run_yaml_test.py
make ack-standalone   # build the single-file version
make critic           # perlcritic with perlcriticrc

Running one test file (requires make first so blib/ is current):

prove -b t/context.t

Quick manual run of the built code from the repo root: ./tack <args> (again, run make first).

Test notes:

Coding standards

Changes, versioning, and releases

Git workflow

Things to avoid

Where to get more context