TextCleaner

TextCleaner is a macOS application that utilizes the imgclean CLI tool to enhance scanned images by turning noisy, colored scans into clean, black-and-white images. It is particularly useful for preparing documents for Optical Character Recognition (OCR) or for improving the readability of scanned text.

I have started this project as a learning exercise to familiarize myself with macOS application development using XCode and Swift. The idea is simple: provide a GUI wrapper around the imgclean CLI tool to make it accessible to users who may not be comfortable using command-line tools. Unfortunately, I do not have an Apple Developer license, so I cannot distribute a signed version of this application. This means that for now, you will have to build the application yourself if you wish to use it.

Furthermore, Apple's notarization requirements have made packaging this app even more challenging. imgclean relies on several third-party libraries, which means that they need to be included with the application bundle as well. However, these libraries must be properly signed by the same developer ID as the main application to satisfy Apple's security checks. Since the libraries are neither built nor signed by me, I cannot include them as-is. Therefore, a re-signing step is necessary, but the script collect_dylibs.sh provided in this repository automates this process for you.

Building

1. Pray that everything will work as intended and that Apple's shenanigans won't get in your way.
2. Clone the project to your machine.
   • imgclean is a git submodule and should be automatically cloned when you clone the repository.
   • If not, run git submodule update --init --recursive
3. Using the terminal, navigate to TextCleaner/imgclean/project and build imgclean by following the instructions in the imgclean README. I recommend using the config flags -DCMAKE_BUILD_TYPE=Release -DMEASURE_PERFORMANCE=OFF for this project.
4. Collect all dylibs that imgclean needs by invoking the collect_dylibs.sh script in the root folder of this project:

   ./collect_dylibs.sh
   

   The resulting dylibs will be placed in the imgclean-frameworks folder.
5. Open the project in XCode.
6. In this step, you need to ensure that all the dylibs collected in step 3 are included in the application bundle:
   • Go to the Build Phases tab of the TextCleaner target.
   • Look for a Copy Files phase with the destination set to Frameworks.
     • If it does not exist, create one by clicking the + button and selecting New Copy Files Phase, then set the destination to Frameworks.
     • If it exists, make sure to remove any old dylibs that may be there from previous builds.
   • Add all the dylibs in the imgclean-frameworks folder to this phase. Make sure to select Create folder references when adding them and ensure that Code Sign on Copy is checked for each dylib.
7. Now, the imgclean binary needs to be included in the application bundle as well:
   • Still in the Build Phases tab, look for another Copy Files phase with the destination set to Executables.
     • If it does not exist, create one by clicking the + button and selecting New Copy Files Phase, then set the destination to Executables.
     • If it exists, make sure to remove any old imgclean binaries that may be there from previous builds.
   • Add the imgclean binary (located in TextCleaner/imgclean/project/build/) to this phase. Ensure that Code Sign on Copy is checked.
8. You should now be able to build and run the application from XCode. Have fun!