High performance library for creating, modiyfing and parsing PDF files in C++
Welcome to PDF-Writer, Also known as PDFHummus.
PDFHummus is a Fast and Free C++ Library for Creating, Parsing an Manipulating PDF Files and Streams.
Documentation is available here.
Project site is here.
There is also a NodeJS module named MuhammaraJS wrapping PDFHummus PDF library and making it available for that language. It is the current and still supported version of a now deprecated HummusJS project of mine, which julianhille maintains.
This is a C++ Project using CMake as project builder.
To build/develop You will need:
Given that this is a Library and not an executable, you can ignore this cmake setup and just use the code as is by copying the folders into your own project. There are, however, better ways to include the code in your own project. The project cmake setup defines PDFHummus as a package and allows you to import the project directly from this repo (remotely) or by pre-installing the package. The instructions below contains information about building the project locally, testing and installing it, as well as explaining how to use CMake FetchContent functionality in order to import the project automatically from this repo
For documentation about how to use the library API you should head for the Wiki pages here.
There are 8 folders to this project:
Once you installed pre-reqs, you can now build the project.
To build you project start by creating a project file in a “build” folder off of the cmake configuration, like this:
mkdir build
cd build
cmake ..
The project defines some optional flags to allow you to control some aspects of building PDFHummus.
PDFHUMMUS_NO_DCT - defines whether to exclude DCT functionality (essentially - not include LibJpeg) from the library. defaults to FALSE. when setting TRUE the library will not require the existance of LibJpeg however will not be able to decode DCT streams from PDF files. (note that this has no baring on the ability to include JPEG images. That ability does not require LibJpeg given the innate ability of PDF files to include DCT encoded streams).PDFHUMMUS_NO_TIFF - defines whether to exclude TIFF functionality (essentially - not include LibTiff) from the library. defaults to FALSE. when setting TRUE the library will not require the existance of LibTiff however will not be able to embed TIFF images.PDFHuMMUS_NO_PNG - defines whether to exclude PNG functionality (essentially - not include LibPng) from the library. defaults to FALSE. when setting TRUE the library will not require the existance of LibPng however will not be able to embed PNG images.USE_BUNDLED - defines whether to use bundled dependencies when building the project or use system libraries. defaults to TRUE. when defined as FALSE, the project configuration will look for installed versions of LibJpeg, Zlib, LibTiff, FreeType, LibAesgm, LibPng and use them instead of the bundled ones (i.e. those contained in the project). Note that for optional dependencies - LibJpeg, LibTiff, LibPng - if not installed the coniguration will succeed but will automatically set the optional building flags (the 3 just described) according to the libraries avialability. As for required dependencies - FreeType, LibAesgm, Zlib - the configuration will fail if those dependencies are not found. see USE_UNBUNDLED_FALLBACK_BUNDLED for an alternative method to deal with dependencies not being found.USE_UNBUNDLED_FALLBACK_BUNDLED - Defines an alternative behavior when using USE_BUNDLED=OFF and a certain dependency is not installed on the system. If set to TRUE then for a dependency that’s not found it will fallback on the bundled version of this dependency. This is essentially attempting to find installed library and if not avialable use a bundled one to ensure that the build will succeed.BUILD_FUZZING_HARNESS - Enable a fuzz testing target for PDFParser. This makes it possible to runn fuzz tests and debug potential vulnarbilities. read more about this in here. By default it’s OFF, set it up with ON.You can set any of those options when calling the cmake command. For example to use system libraries replaced the earlier sequence with:
cd build
cmake .. -DUSE_BUNDLED=FALSE
Once you got the project file, you can now build the project. If you created an IDE file, you can use the IDE file to build the project.
Alternatively you can do so from the command line, again using cmake.
The following builds the project from its root folder:
cmake --build build [--config release]
This will build the project inside the build folder. what’s in brackets is optional and will specify a release onfiguration build. You will be able to look up the result library files per how you normally do when building with the relevant build environment. For example, for windows, the build/PDFWriter/Release folder will have the result PDFWriter file.
This project uses ctest for running tests. ctest is part of cmake and should be installed as part of cmake installation.
The tests run various checks on PDFHummus…and I admit quite a lot of them are not great as unitests as they may just create PDF files without verifying they are good…one uses ones eyes to inspect the test files to do that…or revert to being OK with no exceptions being thrown, which is also good. They are decent as sample code to learn how to do things though 😬.
To run the project tests (after having created the project files in ./build) go:
ctest --test-dir build [-C release]
This should scan the folders for tests and run them.
Consider appending -j22 to the command in order to run tests in parallel to speed things up.
You should be able to see result output files from the tests under ./build/Testing/Output.
Note that ctest does NOT build the project. It may fail if there’s no previous build, or will not pick up on your changes if you made them
since the last build. For This purpose there’s an extra target created in the project to make sure the project and test code is built (as well as recreating the output folder to clear previous runs output):
cmake --build build --target pdfWriterCheck [--config release]
If you want, you can use the install verb of cmake to install a built product (the library files and includes). Use the prefix param to specify where you want the result to be installed to
cmake --install build --prefix ./etc/install [--config release]
This will install all the library files in ./etc/install. You should see an “include” folder and a “lib” folder with include files and library files respectively.
If you want to use PDFHummus there are several methods:
Not much to say about the first option. 2nd option just means to follow the installation instructions and then pointing to the resultant lib and include folders to build your project.
3rd option is probably the best, especially if you already have cmake in your project. This project has package definition for PDFHummus package, which means you can have cmake look for this package and include it in your project with find_package. Then link to the PDFHummus::PDFWriter target and you are done. Another option is to do this + allow for fetching the project content from the repo with FetchContent. Here’s an example from the PDF TextExtraction project of mine:
include(FetchContent)
FetchContent_Declare(
PDFHummus
GIT_REPOSITORY https://github.com/galkahana/PDF-Writer.git
GIT_TAG v4.6.2
FIND_PACKAGE_ARGS
)
FetchContent_MakeAvailable(PDFHummus)
target_link_libraries (TextExtraction PDFHummus::PDFWriter)
This will either download the project and build it or use an installed version (provided that one exists and has a matching version).
Change the GIT_TAG value to what version you’d like to install. You can use tags, branches, commit hashs. anything goes.
Includes are included haha.
You may consider an alternative form that uses URL instead of GIT_REPOSITORY, like this:
include(FetchContent)
FetchContent_Declare(
PDFHummus
URL https://github.com/galkahana/PDF-Writer/archive/refs/tags/v4.6.2.tar.gz
URL_HASH SHA256=0a36815ccc9d207028567f90039785c824b211169ba5da68de84d0c15455ab62
DOWNLOAD_EXTRACT_TIMESTAMP FALSE
FIND_PACKAGE_ARGS
)
FetchContent_MakeAvailable(PDFHummus)
This has the benefit of fetching the archive URL rather than cmake runnig git clone on the specified target.
PDFWriter archives since version v4.6.2 do not include PDFWriterTesting folder and its materials, making it a singificantly smaller download.
You can find the archive urls in the Releases area for this repository.
Note that when installing PDFHummus with the bundled libraries built (this is the default behvaior which can be changed by setting USE_BUNDLED variable to FALSE) there are additional targets that PDFHummus includes:
You can use those targets in additon or instead of using PDFWriter if this makes sense to your project (like if you are extracting images, having LibJpeg or LibPng around can be useful).
The project contains definitions for cpack, cmake packaging mechanism. It might be useful for when you want to build PDFHummus and then install it someplace else.
The following will create a zip file with all libs and includes:
cd build
cpack .
If you are developing this project using vscode here’s some suggestions to help you:
This should help you enable testing and debugging the tests in vscode.
I wrote a post about how to compile and use the library for the iPhone and iPad environments. you can read it here.
It should be quite simple to construct project files in the various building environments (say VS and Xcode) if you want them. Here are some pointers: