matchbook

Introduction

matchbook is a very simple piece of code in the form of a CLI tool that allows one to extract the ISBN from a OCR'ed pdf (should it exist) and/or then run an ISBN against online APIs and retrieve that book's info for renaming and classification purposes.

I personally use this to rename and sort the free PDF books I get from the Packt Publishing Free-Learning program (no affiliation whatsoever).

Installation

Just check out the code and install the Python requirements (possibly in a virtual-environment).

Usage

CLI

matchbook uses the begins package to provide a very elementary CLI. Have a look at that package's doco if you like to which can be found under http://begins.readthedocs.io/en/latest/api.html but really a simple python matchbook.py --help will get you going just fine.

ISBN Extraction

matchbook defines the isbn subcommand which performs an ISBN extraction on a PDF file. It does so through the pdfminer package which it uses to extract all text from each PDF page which it regexes for ISBN and ISBN13 content (should it exist).

This can be invoked simply as:

python matchbook.py isbn 9781785285486-INTERNET_OF_THINGS_WITH_ARDUINO_BLUEPRINTS.pdf

which returns 978-1-78528-548-6.

Book API Lookup

matchbook provides some very basic integration with book APIs, specifically the Google Books API and ISBNdb API, in order to retrieve information on a given book via its ISBN (possibly extracted through matchbook):

These APIs can be accessed via the search subcommand.

Google Books API

Matching against the Google Books API is performed simply with a request against https://www.googleapis.com/books/v1/volumes and, at least at the time of writing, does not require any authentication or API keys.

Usage is as simple as:

python matchbook.py search google 9781449372620

where:

• search is the subcommand.
• google is the corpus/API to search under
• the last parameter is the ISBN without any hyphens (9781449372620 above).

Upon successful execution the above command spits out a JSON document with the information available on this particular book (or whatever error Google throws should the ISBN not match).

ISBNdb API

Unlike Google Books, ISBNdb requires an account and API key to use their API. A free account made under http://isbndb.com/ can provide you with an API key (capped to a finite number of requests I believe):

Usage is as follows:

python matchbook.py search isbndb 9781449372620 --api-key=<your-api-key>

which, similarly to the Google Books case, will spit out a JSON string.

Usage in Python scripts

While the command-line interface outlined above was put together haphazardly for my convenience the code is entirely usable as a vanilla Python package.

In a nutshell:

• The pdf.py module provides the generate_pdf_page_text generator function that yields a PDF's text in a page-by-page fashion.
• The extractors.py module provides the ExtractorPdf class which can iterate over a PDF file running each page's content against a regex (which in this case is ISBN regexes but could be any other regex as well).
• The retrievers.py module provides the super basic API wrappers for Google Books and ISBNdb which can be easily complemented with additional classes.