use setup (#17)

* use setup

* added files back

* bump version numbers

* Update install.rst

* Update setup.py

* Update .travis.yml

Co-authored-by: Neil Wu <[email protected]>

.travis.yml

@@ -43,15 +43,16 @@ install:
       fi
     fi
   # Compile
-  - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && cd $DOCKER_WORKING_DIR && make"
+  - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && cd $DOCKER_WORKING_DIR && make && pip install ."
+
   # the version of testflo is too new, we downgrade to the last version that supports py2
   - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && pip install testflo==1.3.5 "
 
 script:
   # We need to source the mdolab bashrc before running anything
   - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && cd $DOCKER_WORKING_DIR && testflo . -v -n 1"
   # check that new CGNS files have been generated
-  - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && cd $DOCKER_WORKING_DIR/python/examples && ls -l */*.cgns"
+  - docker exec -it app /bin/bash -c ". \$HOME/.bashrc_mdolab && cd $DOCKER_WORKING_DIR/examples && ls -l */*.cgns"
 
 after_script:
   - docker rm -f app

doc/install.rst

@@ -10,99 +10,28 @@ Prerequisites
 "heavy lifting". The following external components are required for
 pyHyp:
 
+- CGNS Libarary
+- PETSc
 
-1. `CGNS Library <http://cgns.sourceforge.net>`_
-
-The CGNS library is used to provide output CGNS functionality for
-pyHyp. The latest CGNS version 3.1 is recommended, but the older
-version 2.5.x may also be used. The After downloading, untar::
-
-   $ tar -xzf cgnslib_3.1.4-2.tar.gz 
-
-.. NOTE:: cmake and ccmake are required for building the CGNS
-   library.
-
-Enter cgnslib_3.1.4 and type::
-
-   $ cmake . 
-
-By default, the CGNS library does not include the Fortran bindings
-that are required for `pyHyp`. This needs to be enabled using the
-cmake configure utility, `ccmake`.::
-
-   $ ccamke .
-
-A "GUI" appears and toggle ENABLE_FORTRAN by pressing [enter]. Type
-'c' to reconfigure and 'g' to generate and exit. Then build the
-library using::
-
-   $ make
-
-If you are compiling on machine without root access, you're done.
-If you are on a local machine with root access, it is usually
-easiest to 'install' into the default directory using::
-
-    $ sudo make install
-
-.. NOTE:: It is also suggested that the BUILD_CGNSTOOLS option be
-   used in the configuration, but this is optional. 
-
-2. `PETSc <http://www.mcs.anl.gov/petsc/index.html>`_ 
-
-Download the latest tarball and extract::
-
-   $ tar -xzf petsc-3.4.1.tar.gz
-
-PETSc must be first configured. There are a wide variety of
-options. The only ones that is strictly necessary are
---with-shared-libraries and --with-fortran-interfaces. However, it is
-recommend the following options are used since some are required for
-pyWarp. To compile without debugging use: --with-debugging=no. It is
-HIGHLY recommended to use debugging until ready to perform production
-runs. In that case, you should compile with a separate architecture
-for the optimized build. In this case use '--PETSC_ARCH=real-opt' for
-example::
-
-       $ ./configure --with-shared-libraries --download-superlu_dist=yes --download-parmetis=yes --download-metis=yes --with-fortran-interfaces=1 --with-debuggig=yes --with-scalar-type=real --PETSC_ARCH=real-debug 
-
-After the configuration step, PETSc must be built. This is
-accomplished with the command provided at the end of the configure
-script. It will look something like below::
-
-   $ make PETSC_DIR=/home/gaetan/Downloads/petsc-3.4.1 PETSC_ARCH=real-debug all
-
-The last step is to add PETSC_DIR and PETSC_ARCH entries to your
-.bashrc file. This is essential. It should look something like
-this: (Make sure the CORRECT directory is used!)::
-
-    export PETSC_ARCH=real-debug
-    export PETSC_DIR=/home/user/packages/petsc-3.4.1
-   
-Make sure the .bashrc file is sourced before pyHyp is compiled using::
-
-   $ source ~/.bashrc
-
-Or simply open a new terminal before compiling pyHyp.
-
-See the documentation of each of these packages for further
-information. 
+See the MDO Lab installation guide `here <http://mdolab.engin.umich.edu/docs/installInstructions/install3rdPartyPackages.html>`_ for the supported versions and installation instructions.
 
 .. NOTE:: A working MPI is not strictly required. However, in most
    cases PETSc should be configured with MPI.
 
 Compilation 
 ------------ 
+``pyHyp`` follows the standard MDO Lab build procedure.
+To start, first clone the repo. For stability we recommend checking out a tagged release.
 
-`pyHyp` follows the standard MDO Lab build procedure.
-To start, find a configuration file close to your current setup in::
+Next, find a configuration file close to your current setup in::
 
     $ config/defaults
 
 and copy it to ''config/config.mk''. For example::
 
     $ cp config/defaults/config.LINUX_GFORTRAN.mk config/config.mk
 
-If you are a beginner user installing the packages on a linux desktop, 
+If you are a beginner user installing the packages on a Linux desktop, 
 you should use the ``config.LINUX_GFORTRAN.mk`` versions of the configuration 
 files. The ``config.LINUX_INTEL.mk`` versions are usually used on clusters.
 
@@ -116,10 +45,10 @@ the screen (near the end)::
    Testing if module hyp can be imported...
    Module hyp was successfully imported.
 
-Finally, add the directory containing the pyhyp folder to the
-$PYTHONPATH variable in your bashrc file::
+Finally, install the Python interface with::
+
+    pip install .
 
-    export $PYTHONPATH:/path/to/folder/containing/pyhyp/
 
 .. _pyhyp_theory:
 
@@ -136,4 +65,4 @@ After you run some of the files, you will get a message like this::
   *** Your MPI job will now abort.
   [MDO-John:7977] Local abort after MPI_FINALIZE completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
   
-Despite its scary look, this is a non-issue and means that the script successfully finished.
+Despite its scary look, this is a non-issue and means that the script successfully finished.