neuralnetworksdl

Browse the Neural Networks and Deep Learning book by Michael Nielsen from the command line.

neuralnetworksdl is a single pure-Go binary. It reads public data from neuralnetworksanddeeplearning.com over plain HTTP, shapes it into clean records, and prints output that pipes into the rest of your tools. No API key, nothing to run alongside it.

The same package is also a resource-URI driver, so a host program like ant can address neuralnetworksdl as neuralnetworksdl:// URIs.

Install

go install github.com/tamnd/neuralnetworksdl-cli/cmd/neuralnetworksdl@latest

Or grab a prebuilt binary from the releases, or run the container image:

docker run --rm ghcr.io/tamnd/neuralnetworksdl:latest --help

Usage

neuralnetworksdl chapters                       # list all 6 chapters + appendix
neuralnetworksdl chapters -o json               # as JSON, ready for jq
neuralnetworksdl chapters -o csv                # CSV with header row
neuralnetworksdl chapters -n 3                  # first 3 chapters only
neuralnetworksdl chapters --fields number,title # pick columns
neuralnetworksdl page <path>                    # fetch one page as a record
neuralnetworksdl links <path>                   # the pages it links to
neuralnetworksdl --help                         # the whole command tree

Every command shares one output contract: -o table|json|jsonl|csv|tsv|url|raw, --fields to pick columns, --template for a custom line, and -n to limit. The default adapts to where output goes (a table on a terminal, JSONL in a pipe), so the same command reads well by hand and parses cleanly downstream.

Example

$ neuralnetworksdl chapters -o table
RANK  NUMBER  TITLE                                                     URL
1     1       Using neural nets to recognize handwritten digits         http://neuralnetworksanddeeplearning.com/chap1.html
2     2       How the backpropagation algorithm works                   http://neuralnetworksanddeeplearning.com/chap2.html
3     3       Improving the way neural networks learn                   http://neuralnetworksanddeeplearning.com/chap3.html
4     4       A visual proof that neural nets can compute any function  http://neuralnetworksanddeeplearning.com/chap4.html
5     5       Why are deep neural networks hard to train?               http://neuralnetworksanddeeplearning.com/chap5.html
6     6       Deep learning                                             http://neuralnetworksanddeeplearning.com/chap6.html
7     A       Appendix: Is there a simple algorithm for intelligence?   http://neuralnetworksanddeeplearning.com/sai.html

Serve it

The same operations are available over HTTP and as an MCP tool set for agents, with no extra code:

neuralnetworksdl serve --addr :7777    # GET /v1/chapters  returns NDJSON
neuralnetworksdl mcp                   # speak MCP over stdio

Use it as a resource-URI driver

neuralnetworksdl registers a neuralnetworksdl domain the way a program registers a database driver with database/sql. A host enables it with one blank import:

import _ "github.com/tamnd/neuralnetworksdl-cli/neuralnetworksdl"

Then ant (or any program that links the package) dereferences neuralnetworksdl:// URIs without knowing anything about the site:

ant get neuralnetworksdl://page/<path>   # fetch the record
ant cat neuralnetworksdl://page/<path>   # just the body text
ant ls  neuralnetworksdl://page/<path>   # the pages it links to, each addressable
ant url neuralnetworksdl://page/<path>   # the live http URL

Development

cmd/neuralnetworksdl/   thin main: hands cli.NewApp to kit.Run
cli/                 assembles the kit App from the neuralnetworksdl domain
neuralnetworksdl/    the library: HTTP client, data models, and domain.go (the driver)
docs/                tago documentation site

make build      # ./bin/neuralnetworksdl
make test       # go test ./...
make vet        # go vet ./...

Releasing

Push a version tag and GitHub Actions runs GoReleaser, which builds the archives, Linux packages, the multi-arch GHCR image, checksums, SBOMs, and a cosign signature:

git tag v0.1.0
git push --tags

The Homebrew and Scoop steps self-disable until their tokens exist, so the first release works with no extra secrets.

License

Apache-2.0. See LICENSE.