homebrew-core/Formula/ocrmypdf.rb

class Ocrmypdf < Formula
  include Language::Python::Virtualenv

  desc "Adds an OCR text layer to scanned PDF files"
  homepage "https://github.com/jbarlow83/OCRmyPDF"
  url "https://files.pythonhosted.org/packages/d9/df/47b1590553600dde69b1dd54c065dbc580574dde3e09aa3d2677fe79c58f/ocrmypdf-9.1.0.tar.gz"
  sha256 "11760b4705b8c7770ca57610cd9c9faa0f64e2dc09c0d69c3a0f3afdf189012d"

  bottle do
    cellar :any
    sha256 "f35cd651640d9d5d9b1a750f7583a44c5e6026f80d636471d8facc80cdc762c2" => :catalina
    sha256 "91f96afc22857d8cb5388f211cedadc9cc5f5bbc8f516a4348d76f8e06c5e3e4" => :mojave
    sha256 "397f30a7f5939203ebe14730bae35e0cf07f30735b20e261b05afe1f5ed75351" => :high_sierra
  end

  depends_on "pkg-config" => :build
  depends_on "freetype"
  depends_on "ghostscript"
  depends_on "jbig2enc"
  depends_on "jpeg"
  depends_on "leptonica"
  depends_on "libpng"
  depends_on "libxml2"
  depends_on "pngquant"
  depends_on "pybind11"
  depends_on "python"
  depends_on "qpdf"
  depends_on "tesseract"
  depends_on "unpaper"

  resource "cffi" do
    url "https://files.pythonhosted.org/packages/2d/bf/960e5a422db3ac1a5e612cb35ca436c3fc985ed4b7ed13a1b4879006f450/cffi-1.13.2.tar.gz"
    sha256 "599a1e8ff057ac530c9ad1778293c665cb81a791421f46922d80a86473c13346"
  end

  resource "chardet" do
    url "https://files.pythonhosted.org/packages/fc/bb/a5768c230f9ddb03acc9ef3f0d4a3cf93462473795d18e9535498c8f929d/chardet-3.0.4.tar.gz"
    sha256 "84ab92ed1c4d4f16916e05906b6b75a6c0fb5db821cc65e70cbd64a3e2a5eaae"
  end

  resource "img2pdf" do
    url "https://files.pythonhosted.org/packages/e0/c6/7cd14232a1b10bf884c12daf3626afb76c4f60b52ae0eb23ce1519542ae4/img2pdf-0.3.3.tar.gz"
    sha256 "9d77c17ee65a736abe92ef8cba9cca009c064ea4ed74492c01aea596e41856cf"
  end

  resource "lxml" do
    url "https://files.pythonhosted.org/packages/c4/43/3f1e7d742e2a7925be180b6af5e0f67d38de2f37560365ac1a0b9a04c015/lxml-4.4.1.tar.gz"
    sha256 "c81cb40bff373ab7a7446d6bbca0190bccc5be3448b47b51d729e37799bb5692"
  end

  resource "pdfminer.six" do
    url "https://files.pythonhosted.org/packages/8e/d9/d1ce92de1afd5b7785a1439bfbe95520f5ca9e9ad217c80e5728cb545175/pdfminer.six-20191020.tar.gz"
    sha256 "5aa19ee9de443eb7abd3d6ee6e571b139cb321bf411ec8544f472866b1ef2835"
  end

  resource "pikepdf" do
    url "https://files.pythonhosted.org/packages/5b/53/fefb843e75905d7dc581c163f1290aa5a426a50786d127bcd699f5f06111/pikepdf-1.7.0.tar.gz"
    sha256 "ef4402b5b1669bc2a6234a60632229df9c025414acde591ad4bfe77b61c3441d"
  end

  resource "Pillow" do
    url "https://files.pythonhosted.org/packages/5b/bb/cdc8086db1f15d0664dd22a62c69613cdc00f1dd430b5b19df1bea83f2a3/Pillow-6.2.1.tar.gz"
    sha256 "bf4e972a88f8841d8fdc6db1a75e0f8d763e66e3754b03006cbc3854d89f1cb1"
  end

  resource "pycparser" do
    url "https://files.pythonhosted.org/packages/68/9e/49196946aee219aead1290e00d1e7fdeab8567783e83e1b9ab5585e6206a/pycparser-2.19.tar.gz"
    sha256 "a988718abfad80b6b157acce7bf130a30876d27603738ac39f140993246b25b3"
  end

  resource "pycryptodome" do
    url "https://files.pythonhosted.org/packages/b6/41/d3749c0caa860041eb6b0832c7087253d59cb5af8bb303a8d4b6daa74014/pycryptodome-3.9.0.tar.gz"
    sha256 "dbeb08ad850056747aa7d5f33273b7ce0b9a77910604a1be7b7a6f2ef076213f"
  end

  resource "reportlab" do
    url "https://files.pythonhosted.org/packages/9b/e0/97b334f0e6a805c921ac369a153cea6ea822787a1d8247c7e65b1386c520/reportlab-3.5.32.tar.gz"
    sha256 "83ef44936ef4e9c432d62bc2b72ec8d772b87af319d123e827a72e9b6884c851"
  end

  resource "six" do
    url "https://files.pythonhosted.org/packages/dd/bf/4138e7bfb757de47d1f4b6994648ec67a51efe58fa907c1e11e350cddfca/six-1.12.0.tar.gz"
    sha256 "d16a0141ec1a18405cd4ce8b4613101da75da0e9a7aec5bdd4fa804d0e0eba73"
  end

  resource "sortedcontainers" do
    url "https://files.pythonhosted.org/packages/29/e0/135df2e733790a3d3bcda970fd080617be8cea3bd98f411e76e6847c17ef/sortedcontainers-2.1.0.tar.gz"
    sha256 "974e9a32f56b17c1bac2aebd9dcf197f3eb9cd30553c5852a3187ad162e1a03a"
  end

  resource "tqdm" do
    url "https://files.pythonhosted.org/packages/08/ad/be20d03f9fae7059cd2bd866f47cef00cfaebab304d3699200245bc31583/tqdm-4.38.0.tar.gz"
    sha256 "c1d677f3a85fa291b34bdf8f770f877119b9754b32673699653556f85e2c2f13"
  end

  def install
    venv = virtualenv_create(libexec, "python3")

    resource("Pillow").stage do
      inreplace "setup.py" do |s|
        sdkprefix = MacOS.sdk_path_if_needed ? MacOS.sdk_path : ""
        s.gsub! "openjpeg.h", "probably_not_a_header_called_this_eh.h"
        s.gsub! "ZLIB_ROOT = None", "ZLIB_ROOT = ('#{sdkprefix}/usr/lib', '#{sdkprefix}/usr/include')"
        s.gsub! "JPEG_ROOT = None", "JPEG_ROOT = ('#{Formula["jpeg"].opt_prefix}/lib', '#{Formula["jpeg"].opt_prefix}/include')"
        s.gsub! "FREETYPE_ROOT = None", "FREETYPE_ROOT = ('#{Formula["freetype"].opt_prefix}/lib', '#{Formula["freetype"].opt_prefix}/include')"
      end

      # avoid triggering "helpful" distutils code that doesn't recognize Xcode 7 .tbd stubs
      ENV.append "CFLAGS", "-I#{MacOS.sdk_path}/System/Library/Frameworks/Tk.framework/Versions/8.5/Headers" unless MacOS::CLT.installed?
      venv.pip_install Pathname.pwd
    end

    # Fix "ld: file not found: /usr/lib/system/libsystem_darwin.dylib" for lxml
    ENV["SDKROOT"] = MacOS.sdk_path if MacOS.version == :sierra

    res = resources.map(&:name).to_set - ["Pillow"]
    res.each do |r|
      venv.pip_install resource(r)
    end

    venv.pip_install_and_link buildpath
    bash_completion.install "misc/completion/ocrmypdf.bash" => "ocrmypdf"
    fish_completion.install "misc/completion/ocrmypdf.fish"
  end

  test do
    system "#{bin}/ocrmypdf", "-f", "-q", "--deskew",
                              test_fixtures("test.pdf"), "ocr.pdf"
    assert_predicate testpath/"ocr.pdf", :exist?
  end
end