adding start to script for spack export.

There are a ton of symbols from system includes that are deemed missing,
but we do not have a way to know that they are from system (and not a dependency).

Signed-off-by: vsoch <vsoch@users.noreply.github.com>

abi-python/spack/Dockerfile.clingo

@@ -0,0 +1,19 @@
+FROM ghcr.io/autamus/clingo:latest
+# docker build -f Dockerfile.clingo -t clingo .
+ENV DEBIAN_FRONTEND=noninteractive
+ENV PATH=/opt/conda/bin:$PATH
+RUN apt-get update && apt-get install -y curl
+RUN /bin/bash -c "curl -L https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh > miniconda.sh && \
+    bash miniconda.sh -b -p /opt/conda && \
+    rm miniconda.sh"
+RUN /bin/bash -c "conda install -y -c conda-forge mamba && \
+    mamba create -q -y -c potassco -n clingo-env clingo && \
+    source activate clingo-env && \
+    conda clean --all -y && \
+    pip install pyelftools && \
+    conda install clingo ipython six"
+RUN echo "source activate clingo-env" > ~/.bashrc
+ENV PATH /opt/conda/envs/clingo-env/bin:${PATH}
+ENTRYPOINT /bin/bash
+WORKDIR /code
+ADD . /code

abi-python/spack/README.md

@@ -0,0 +1,44 @@
+# A Logic Program to work with Spack
+
+The file [abi-facts.lp](abi-facts.lp) is generated with [this branch](https://github.com/vsoch/spack/tree/add/clingo-analyzer) of spack, which has an added analyzer for clingo:
+
+```bash
+$ spack install zlib+debug
+$ spack install tcl+debug
+$ spack analyze run  -a clingo tcl
+```
+
+Note above that both tcl and it's dependency zlib were installed with debug information,
+otherwise this would not work. Also note that in the command above, we are stating
+that "tcl" is the main library we are interested in to assess compatibility with
+it's dependencies. In the case of generating facts for a solver (likely for a splice)
+the problem will be slightly different - we will still be interested in some primary package,
+but we will be including facts (atoms) for the dependency version we want to splice in.
+
+The above command generates that file in the analyzers folder, in the clingo subdirectory. We are copying it here so it's easy to develop with.
+Next, let's build a container to develop in:
+
+```bash
+$ docker build -f Dockerfile.clingo -t clingo-docker .
+```
+
+And shell into it.
+
+```bash
+$ docker run -it --rm -v $PWD/:/code clingo-docker bash
+```
+
+Unlike our first attempt in the [python](../python) folder, instead of having
+a main binary and two libraries (one known to work and one we are testing),
+for this logic program we are throwing all symbols in the space into a bag,
+including the compiler, and we target the binaries of the main package as the 
+ones we are interested in to assess if they still work. This means that we assume
+that a dependency of this main library (that is already installed) is working,
+and we don't add all nested dependency symbols into the mix. If any symbol is
+missing for a main binary/library that we are interested in, we will throw an issue.
+
+We can then develop and run our program:
+
+```bash
+(clingo-env) root@12069473da65:/code/python# clingo --out-ifs=\\n abi-facts.lp is_compatible.lp 
+```