LillyMol 7.0 open source release (#23)

.gitignore

@@ -0,0 +1,21 @@
+**/Linux-gcc-6.2.0
+**/Linux-gcc-7.2.1
+**/Linux-gcc-8.3.0
+**/Linux
+**/__pycache__
+**/.idea
+*_pb.rb
+*.pyc
+lib
+include
+.gdb_history
+third_party
+.venv
+.vscode
+
+./src/WORKSPACE
+
+## vs code workspace 
+./LillyMolPrivate.code-workspace
+*ruby-version
+lib64

Dockerfile

@@ -0,0 +1,56 @@
+FROM gcc:10.2
+
+ENV LILLYMOL_HOME=/LillyMol \
+    BUILD_DIR=Linux
+
+# Install Eigen
+# (No need to install zlib since zlib1g and zlib1g-dev are alreday in the gcc base image)
+RUN apt-get update && \ 
+    apt-get upgrade -y && \
+    apt-get install -y libeigen3-dev && \
+    cd /usr/include && \
+        ln -sf eigen3/Eigen Eigen
+
+RUN apt-get install npm -y && \
+    npm install -g @bazel/bazelisk && \
+    apt-get install cmake -y
+
+RUN apt-get install python3-pip -y && \
+    pip3 install pandas && \
+    apt-get install libblas-dev -y && \
+    apt-get install liblapack-dev -y && \
+    pip3 install scipy
+
+COPY . ./LillyMol
+
+WORKDIR /LillyMol
+
+RUN mkdir bin && \
+    mkdir bin/Linux && \
+    cd src && \
+    ./build_third_party.sh
+
+WORKDIR /LillyMol/src
+
+RUN cp /tmp/WORKSPACE .
+
+WORKDIR /LillyMol/src/build_deps
+
+RUN sed -i 's/\/workspaces\/LillyMolPrivate\/bin\/Linux\//\/LillyMol\/bin\/Linux\//g' install.bzl
+
+WORKDIR /LillyMol
+
+ENV LD_LIBRARY_PATH=/LillyMol/third_party/lib
+
+RUN mkdir /node && \
+    mkdir /node/scratch
+
+WORKDIR /LillyMol/src
+
+RUN ./build_from_src.sh
+
+WORKDIR /LillyMol
+
+RUN rm -f /usr/bin/python && ln -s /usr/bin/python3 /usr/bin/python
+
+# RUN apt-get install python3-pip -Y

Makefile

@@ -14,7 +14,8 @@
 # limitations under the License. 
 ########################################################################
 IWPROGRAMMES ?= $(PWD)
-UNAME ?= Linux-gcc-6.2.0
+UNAME ?= Linux-gcc-8.3.0
+#Linux-gcc-6.2.0
 
 prefix = $(IWPROGRAMMES)
 exec_prefix = $(prefix)
@@ -128,3 +129,19 @@ copy_exe: do-copy_exe
 library: do-library
 
 copy_library: do-copy_library
+
+
+build_docker:
+	docker build  -f Dockerfile -t lillymolprivate .
+
+test_lillymol:
+	docker container exec  lilly_mol  bash -c "cd test/ && ./run_all_test.sh"
+
+start_s3:
+	docker-compose up -d
+	sleep 10
+	echo 'run s3 commannd with: aws s3 --endpoint "http://localhost:4566" <s3 command>'
+stop_s3:
+	docker-compose down
+
+

NEWS.md

@@ -0,0 +1,51 @@
+# June 2023
+
+This release of LillyMol contains a great many updates.
+
+Among the most notable are
+
+# Build with bazel/bazelisk.
+* Protocl Buffers used for substructure searching and reactions.
+* Many additional substructure query functions.
+* Many more utilities included in the release.
+
+The build is now very different.  Previously LillyMol had tried hard
+to ensure a lack of third party dependencies.  This became untenable
+and now there are several external dependencies.  In addition,
+building is now done with `bazel`.  This offers many advantages but
+also makes building more complex than before.  See the README file for
+instructions.
+
+Switching to text format protocol buffers offers many advantages,
+and will be used increasingly within LillyMol.
+Old style query and reaction files will remain supported, but no new
+functionality is being added to them. In order to gain access to new
+functionality, you will need to use the text format protocol buffers.
+
+Smarts provides a language for specifying matched atoms. But many
+substructure query concepts are for groups of atoms, rings, substituents,
+scaffolds, linkers... There are now more substructure abstract concepts
+implemented, see especially the `global_id` concept which now
+links the abstract concepts with matched atoms.
+
+The no matched atoms between directive, `...` is now much more feature
+complete, and allows specification of what atoms can, or cannot, be in
+a region. Arbitrary element names are now more fully supported.
+
+Within the code itself, new code is being implemented with a coding
+style similar to what Google uses. clang-format is being used to 
+transition parts of the code, and address sanitizer is used
+during testing. GoogleTest is used for C++ unit tests - much more work
+remains to be done there, but today there are over 25k lines of unit test.
+
+## Future
+As useful substructure search concepts are needed, they will be added
+to LillyMol. If you have needs that are not addressed, please reach out.
+
+A recent emphasis on LillyMol has been in the area of supporting and
+augmenting AI driven generative designs. That will continue with
+future work focussed on bringing precedented functional groups to
+the generative space, combining the strengths of cheminformatics
+enumeration and replacement based approaches and generative models.
+Several tools supporting this are in this distribution.
+