Remove yaml-cpp as a dependency of the CalculiX adapter (#547)

Co-authored-by: Gerasimos Chourdakis <[email protected]>

Author: IshaanDesai

content/docs/adapters/calculix/adapter-calculix-get-adapter.md

@@ -42,7 +42,7 @@ We started offering Debian packages for the CalculiX adapter since v2.19.0. Plea
 
 2. Edit the `Makefile` to set the paths to dependencies.
    - If you have the CalculiX source in `~/CalculiX/` and the dependencies in your global paths, you don't need to change anything.
-   - Otherwise, set `CCX` and, if built from source, the include and lib flags for SPOOLES, ARPACK, and yaml-cpp.
+   - Otherwise, set `CCX` and, if built from source, the include and lib flags for the dependencies.
    - Read below if you are [using GCC 10 or later](#compiling-with-gcc-10-or-newer).
 3. Clean any previous builds with `make clean`.
 4. Build with `make` (e.g. `make -j 4` for parallel).
@@ -56,6 +56,10 @@ Since version 2.17 of CalculiX, it is possible to link the PaStiX solver for inc
 
 The adapter is built using GNU Make. The `Makefile` contains a few variables on top, which need to be adapted to your system:
 
+{% version %}
+The variables `YAML_INCLUDE` and `YAML_LIBS` are only relevant up to the adapter v2.20.1.
+{% endversion %}
+
  1. `CCX`: Location of the original CalculiX solver (CCX) source code ("src" directory)
     - Example: `$(HOME)/CalculiX/ccx_{{site.calculix_version}}/src`
  2. `SPOOLES_INCLUDE`: Include flags for SPOOLES

content/docs/adapters/calculix/adapter-calculix-get-calculix.md

@@ -18,18 +18,22 @@ You don't need to build the "vanilla" CalculiX before building the adapter. But
 
 CalculiX itself depends on [SPOOLES2.2](http://www.netlib.org/linalg/spooles/spooles.2.2.html) and [ARPACK](https://en.wikipedia.org/wiki/ARPACK).
 
-Additionally, our adapter also depends on [yaml-cpp](https://github.com/jbeder/yaml-cpp).
+{% version %}
+Up to version v2.20.1, the adapter also requires [yaml-cpp](https://github.com/jbeder/yaml-cpp) to be installed.
+In later versions, a header-only variant of this library is included with the adapter source code.
+You can get this on Ubuntu by installing the `libyaml-cpp-dev` package with APT, or on Arch/Manjaro by installing the `yaml-cpp` package with pacman.
+{% endversion %}
 
 These can be found in many distributions as binary packages. For example, in Ubuntu, do:
 
 ```bash
-sudo apt install libarpack2-dev libspooles-dev libyaml-cpp-dev
+sudo apt install libarpack2-dev libspooles-dev
 ```
 
-For example, in Arch or Manjaro, install `arpack` and `yaml-cpp`, and compile `spooles` using an AUR helper (e.g. `yay`):
+For example, in Arch or Manjaro, install `arpack`, and compile `spooles` using an AUR helper (e.g. `yay`):
 
 ```bash
-sudo pacman -S arpack yaml-cpp
+sudo pacman -S arpack
 yay spooles
 ```
 
@@ -116,6 +120,10 @@ Now we are ready to build the library with `make lib`
 
 ### Building yaml-cpp from source
 
+{% version %}
+Only required up to the adapter v2.20.1.
+{% endversion %}
+
 <details markdown="1"><summary>If you cannot get a binary for yaml-cpp, try these instructions.</summary>
 
 Get the latest version of yaml-cpp and build it as a shared library. For example:

content/docs/adapters/calculix/adapter-calculix-pastix-build.md

@@ -20,6 +20,10 @@ This page was built following CalculiX' documentation with adaptations for preCI
 Make sure you have a working installation of preCICE. Also run these installation commands, (after a call to `sudo apt update` and `sudo apt upgrade`) :
 `sudo apt install build-essential cmake git gfortran flex bison zlib1g-dev nvidia-cuda-toolkit-gcc libspooles-dev libyaml-cpp-dev`
 
+{% version %}
+Up to the adapter v2.20.1, the dependency package `libyaml-cpp-dev` was also required.
+{% endversion %}
+
 {% note %}
 You can get these elsewhere or build them from source. In particular, it is probably wise to have a more up to date CUDA installation than the one provided on the Ubuntu repository. Again, feedback is appreciated!
 {% endnote %}

content/docs/adapters/calculix/adapter-calculix-supermuc.md

@@ -22,7 +22,7 @@ In installing ARPACK, the HOME directory needs to be specified in the "ARmake.in
 
 Any problems with the installation of SPOOLES and ARPACK can be searched in the installation [instructions](http://www.dhondt.de/INST_CCX_2_8_MAC_02_10_2015.pdf).
 
-To install yaml-cpp, run in the source directory:
+To install yaml-cpp (up to adapter v2.20.1), run in the source directory:
 
 ```bash
 mkdir build