92 lines
3.7 KiB
Ruby
92 lines
3.7 KiB
Ruby
class Tesseract < Formula
|
|
desc "OCR (Optical Character Recognition) engine"
|
|
homepage "https://github.com/tesseract-ocr/"
|
|
url "https://github.com/tesseract-ocr/tesseract/archive/5.3.0.tar.gz"
|
|
sha256 "7e70870f8341e5ea228af2836ce79a36eefa11b01b56177b4a8997f330c014b8"
|
|
license "Apache-2.0"
|
|
head "https://github.com/tesseract-ocr/tesseract.git", branch: "main"
|
|
|
|
livecheck do
|
|
url :stable
|
|
regex(/^v?(\d+(?:\.\d+)+)$/i)
|
|
end
|
|
|
|
bottle do
|
|
sha256 cellar: :any, arm64_ventura: "7a036201708ae3eb711d3a11bb14a2a8f3f9c33dbc2a18d575a38a2a100866dc"
|
|
sha256 cellar: :any, arm64_monterey: "ca1f6cfc7e239c18e8b27127bfde0f7c8ac66070c05c598b21f900f900dbf253"
|
|
sha256 cellar: :any, arm64_big_sur: "fe3ed8985df97bdc41671a601fa4076e11662539a51cec523f9f3610b8ae877f"
|
|
sha256 cellar: :any, ventura: "3dfa521b91ee4f9d1eb601594085eef3f399e2b3a05a745e641cb124e5238026"
|
|
sha256 cellar: :any, monterey: "83d2c0f4da40cbd12a1ea08fb62f0b7418610ee0d1f434b8cb1d865c8005c28c"
|
|
sha256 cellar: :any, big_sur: "ebe07d9cbfc7bfa5cbbbbd9781b67a09fdd707f077036915289b12c03f5a672b"
|
|
sha256 cellar: :any_skip_relocation, x86_64_linux: "6b5f137e8137383a86711b21b6e11e6c19bed3a0ae59fed5bb1f221b5f8deb07"
|
|
end
|
|
|
|
depends_on "autoconf" => :build
|
|
depends_on "automake" => :build
|
|
depends_on "libtool" => :build
|
|
depends_on "pkg-config" => :build
|
|
depends_on "cairo"
|
|
depends_on "icu4c"
|
|
depends_on "leptonica"
|
|
depends_on "libarchive"
|
|
depends_on "pango"
|
|
|
|
fails_with gcc: "5"
|
|
|
|
resource "eng" do
|
|
url "https://github.com/tesseract-ocr/tessdata_fast/raw/4.1.0/eng.traineddata"
|
|
sha256 "7d4322bd2a7749724879683fc3912cb542f19906c83bcc1a52132556427170b2"
|
|
end
|
|
|
|
resource "osd" do
|
|
url "https://github.com/tesseract-ocr/tessdata_fast/raw/4.1.0/osd.traineddata"
|
|
sha256 "9cf5d576fcc47564f11265841e5ca839001e7e6f38ff7f7aacf46d15a96b00ff"
|
|
end
|
|
|
|
resource "snum" do
|
|
url "https://github.com/USCDataScience/counterfeit-electronics-tesseract/raw/319a6eeacff181dad5c02f3e7a3aff804eaadeca/Training%20Tesseract/snum.traineddata"
|
|
sha256 "36f772980ff17c66a767f584a0d80bf2302a1afa585c01a226c1863afcea1392"
|
|
end
|
|
|
|
resource "test_resource" do
|
|
url "https://raw.githubusercontent.com/tesseract-ocr/test/6dd816cdaf3e76153271daf773e562e24c928bf5/testing/eurotext.tif"
|
|
sha256 "7b9bd14aba7d5e30df686fbb6f71782a97f48f81b32dc201a1b75afe6de747d6"
|
|
end
|
|
|
|
def install
|
|
# explicitly state leptonica header location, as the makefile defaults to /usr/local/include,
|
|
# which doesn't work for non-default homebrew location
|
|
ENV["LIBLEPT_HEADERSDIR"] = HOMEBREW_PREFIX/"include"
|
|
|
|
ENV.cxx11
|
|
|
|
system "./autogen.sh"
|
|
system "./configure", "--prefix=#{prefix}",
|
|
"--disable-dependency-tracking",
|
|
"--datarootdir=#{HOMEBREW_PREFIX}/share"
|
|
|
|
system "make", "training"
|
|
|
|
# make install in the local share folder to avoid permission errors
|
|
system "make", "install", "training-install", "datarootdir=#{share}"
|
|
|
|
resource("snum").stage { mv "snum.traineddata", share/"tessdata" }
|
|
resource("eng").stage { mv "eng.traineddata", share/"tessdata" }
|
|
resource("osd").stage { mv "osd.traineddata", share/"tessdata" }
|
|
end
|
|
|
|
def caveats
|
|
<<~EOS
|
|
This formula contains only the "eng", "osd", and "snum" language data files.
|
|
If you need any other supported languages, run `brew install tesseract-lang`.
|
|
EOS
|
|
end
|
|
|
|
test do
|
|
resource("test_resource").stage do
|
|
system bin/"tesseract", "./eurotext.tif", "./output", "-l", "eng"
|
|
assert_match "The (quick) [brown] {fox} jumps!\n", File.read("output.txt")
|
|
end
|
|
end
|
|
end
|