Reproducible software package and source for publishable paper for introducing the elaphrocentre into extragalactic physics
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

623 lines
24 KiB

# High-level script to manage the project.
# Run `./project --help' for a description of how to use it.
# Copyright (C) 2019-2021 Mohammad Akhlaghi <>
# Copyright (C) 2020-2021 Boud Roukema
# Copyright (C) 2021 Raul Infante-Sainz <>
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <>.
# Basic settings
# --------------
# Stop the script if there are any errors.
set -e
# Default option values
jobs=0 # 0 is for the default for the `' script.
# Parse the options
# -----------------
# Separate command-line arguments from options. Then put the option value
# into the respective variable.
# Each option has two lines because we want to process both these formats:
# `--name=value' and `--name value'. The former (with `=') is a single
# command-line argument, so we just need to shift the counter by one. The
# latter (without `=') is two arguments, so we'll need two shifts.
# Note on the case strings: for every option, we need three lines: one when
# the option name and value are separate. Another when there is an equal
# between them, and finally one where the value is immediately after the
# short-format. This exact order is important. Otherwise, there will be a
# conflict between them.
print_help() {
# Print the output.
cat <<EOF
Usage: $scriptname configure [OPTIONS]
$scriptname shell [OPTIONS]
$scriptname make [OPTIONS]
Top-level script to manage the reproducible project. The high-level
operation is defined by the (mandatory) second argument:
configure - Configure project for this machine (e.g., build software).
make - Run the project (do analysis and build outputs).
shell - Execute the project's shell for interactive testing.
RECOMMENDATION: If this is the first time you are configuring this
template, please don't use the options and let the script explain each
parameter in full detail by simply running './project configure'.
Project 'make' special features.
./project make Build the project on one thread
./project make -jN Built the project in parallel on N threads.
./project make clean Clean all files generated by 'make' (not software).
./project make texclean Clean all files built by (La)TeX.
./project make distclean Clean everything (including compiled software).
./project make dist Produce a LaTeX-ready-to-build distribution tarball
('tar.gz') of the project. This is ready to be
uploaded to servers like ''.
./project make dist-lzip Similar to 'dist', but compress to '.tar.lz'.
./project make dist-zip Similar to 'dist', but compress to '.zip'.
./project make dist-software Build a .tar.gz tarball containing all
software source tarballs necessary for the
With the options below you can modify the default behavior.
Configure options:
-b, --build-dir=STR Top directory to build the project in.
-e, --existing-conf Use (possibly existing) local configuration.
--host-cc Use host system's C compiler, don't build GCC.
-i, --input-dir=STR Directory containing input datasets (optional).
-s, --software-dir=STR Directory containing necessary software tarballs.
--check-config During configuration, show what is being built.
--clean-texdir Remove possibly existing build-time subdirectories
under the project's 'tex/' directory (can happen
when source is from arXiv for example).
--all-highlevel Build all high-level software (for development).
Configure and Make options:
-d, --debug[=FLAGS] In configure: use -j1, no -k, and no Zenodo check.
In make: 'FLAGS' will be directly passed to 'make'.
-g, --group=STR Build and run with write permissions for a group.
-j, --jobs=INT Number of threads to build/run the software.
-?, --help Print this help list.
Make (analysis) options:
-P, --help-pipeline Print help about the galaxy formation pipeline.
--highlight-new Highlight revised text in bold red.
--remove-some-fits Remove some fits as requested by the referee.
-d, --debug=FLAGS Print various types of debugging information.
-p, --prepare-redo Re-do preparation (only done automatically once).
Make (final PDF) options:
--highlight-notes Show '\tonote' regions as red text in PDF.
Mandatory or optional arguments to long options are also mandatory or optional
for any corresponding short options.
Maneage URL:
Report bugs to
print_help_pipeline() {
cat <<EOF
Usage: $scriptname configure [OPTIONS]
$scriptname make [OPTIONS]
Main help option.
./project --help Get help on command line options.
To reconfigure/compile/install an executable 'foo':
rm .build/software/installed/version-info/proglib/foo
./project configure --existing-conf
Galaxy formation pipeline options for creating or removing output files. Most of
the pipeline steps run the prerequisite steps if needed, except for 'run-simulation':
Separate steps:
./project make init-conditions Run the initial conditions.
./project make run-simulation Run the simulations (not run automatically).
./project make detect-haloes Detect dark matter haloes.
./project make create-mergertree Create a merger history tree.
./project make create-galaxies Create galaxies from SFR histories.
./project make detect-voids Detect voids in dark matter distribution.
./project make analyse-plot Analyse and plot.
Combined steps:
./project make create-galaxies-voids Run all steps after the N-body simulation.
./project make full-calculations Full run (= simulation + galaxies + voids; and prerequisites).
./project make full-pipeline Full run + analysis/plotting.
Separate removal steps:
./project make clean-init-conditions Remove initial conditions output files.
./project make clean-simulation Remove simulation output files.
./project make clean-haloes Remove dark matter halo output files.
./project make clean-mergertree Remove merger history tree.
./project make clean-galaxies Remove galaxies.
./project make clean-voids Remove voids.
./project make clean-paper Remove LaTeX-related auxiliary (.aux, ...) files.
Combined removal steps:
./project make clean-post-simulation Remove all steps after the N-body simulation.
./project make clean-analyse-plot Remove analysis/plotting files.
./project make clean-pipeline Remove all calculations and plotting.
./project make clean Remove *all* significant files generated by
'analysis/make/*.mk' make rules (not software).
on_off_option_error() {
if [ "x$2" = x ]; then
echo "$scriptname: '$1' doesn't take any values."
echo "$scriptname: '$1' (or '$2') doesn't take any values."
exit 1
check_v() {
if [ x"$2" = x ]; then
echo "$scriptname: option '$1' requires an argument."
echo "Try '$scriptname --help' for more information."
exit 1;
func_operation_set() {
if [ x$operation = x ]; then
echo "Only one operation ('configure', 'make' or 'shell') may be given."
exit 1
while [ $# -gt 0 ]
case $1 in
# Main operation.
configure) func_operation_set $1; shift;;
make) func_operation_set $1; shift;;
shell) func_operation_set $1; shift;;
# Configure options:
-b|--build-dir) build_dir="$2"; check_v "$1" "$build_dir"; shift;shift;;
-b=*|--build-dir=*) build_dir="${1#*=}"; check_v "$1" "$build_dir"; shift;;
-b*) build_dir=$(echo "$1" | sed -e's/-b//'); check_v "$1" "$build_dir"; shift;;
-e|--existing-conf) existing_conf=1; shift;;
-e*|--existing-conf=*) on_off_option_error --existing-conf -e;;
--host-cc) host_cc=1; shift;;
--host-cc=*) on_off_option_error --host-cc;;
-i|--input-dir) input_dir="$2"; check_v "$1" "$input_dir"; shift;shift;;
-i=*|--input-dir=*) input_dir="${1#*=}"; check_v "$1" "$input_dir"; shift;;
-i*) input_dir=$(echo "$1" | sed -e's/-i//'); check_v "$1" "$input_dir"; shift;;
-s|--software-dir) software_dir="$2"; check_v "$1" "$software_dir"; shift;shift;;
-s=*|--software-dir=*) software_dir="${1#*=}"; check_v "$1" "$software_dir"; shift;;
-s*) software_dir=$(echo "$1" | sed -e's/-s//'); check_v "$1" "$software_dir"; shift;;
--check-config) check_config=1; shift;;
--check-config=*) on_off_option_error --check-config;;
--clean-texdir) clean_texdir=1; shift;;
--clean-texdir=*) on_off_option_error --clean-texdir;;
--all-highlevel) all_highlevel=1; shift;;
--all-highlevel=*) on_off_option_error --all-highlevel;;
# Configure and Make options:
-g|--group) group="$2"; check_v group "$group"; shift;shift;;
-g=*|--group=*) group="${1#*=}"; check_v group "$group"; shift;;
-g*) group=$(echo "$1" | sed -e's/-g//'); check_v group "$group"; shift;;
-j|--jobs) jobs="$2"; check_v jobs "$jobs"; shift;shift;;
-j=*|--jobs=*) jobs="${1#*=}"; check_v jobs "$jobs"; shift;;
-j*) jobs=$(echo "$1" | sed -e's/-j//'); check_v jobs "$jobs"; shift;;
-'?'|--help) print_help; exit 0;;
-'?'*|--help=*) on_off_option_error --help -?;;
# Make options
# ------------
# Note that Make's `debug' can take values, but when called without any
# value, it is like giving it a value of `a'):
-P|--help-pipeline) print_help_pipeline; exit 0;;
--highlight-new) highlightnew=1; shift;;
--highlight-new=*) on_off_option_error --highlight-new;;
--remove-some-fits) removesomefits=1; shift;;
--remove-some-fits=*) on_off_option_error --remove-some-fits;;
--highlight-notes) highlightnotes=1; shift;;
--highlight-notes=*) on_off_option_error --highlight-notes;;
-d|--debug) if [ x"$2" = x ]; then debug=a; shift;
else debug="$2"; check_v debug "$debug"; shift;shift; fi;;
-d=*|--debug=*) debug="${1#*=}"; check_v debug "$debug"; shift;;
-d*) debug=$(echo "$1" | sed -e's/-d//'); check_v debug "$debug"; shift;;
-p|--prepare-redo) prepare_redo=1; shift;;
-p=*|--prepare-redo=*) on_off_option_error --prepare-redo; shift;;
# Unrecognized option:
-*) echo "$scriptname: unknown option '$1'"; exit 1;;
# Not an option, an argument (so its a Make target).
*) make_targets="$make_targets $1"; shift;;
# Check configuration status
# --------------------------
if ! [ x$check_config = x ]; then
# Find the color option to pass to `ls'. Note that `--color' (for GNU
# Coreutils `ls') should be checked first because it also has `-G', but
# for something else.
if ls --color 2> /dev/null > /dev/null; then coloropt="--color=auto"
elif ls -G 2> /dev/null > /dev/null; then coloropt="-G"
else coloropt=""
# Print a notice to let the user know what is happening.
cat <<EOF
This is an infinite loop that will print what software are being built at
every moment. It is actually a listing ('ls' command) of the temporary
directory where software source code are unpacked while they are being
built. If the project isn't being configured, the output (every second)
will either be empty (only a date) or a with an error about a non-existant
directory. This feature is thus only useful when the project's software are
being built.
# Run the infinite loop. To be able to find the last built programs, we
# need to put all the Python programs and high-level program into one
# directory.
while true; do
# Make sure the '.build' directory has already been created.
echo "========================"
if [ -d .build ]; then
echo "$(date) [[press CTRL-C to stop]]";
echo "--- Currently being built:"
if [ -d .build/software/build-tmp ]; then
ls $coloropt .build/software/build-tmp || junk=1;
# Make the temporary directory, delete its contents, then put new
# links of all built software.
if ! [ -d $checkdir ]; then mkdir $checkdir; fi
rm -f $checkdir/*
# Check if any programs exist in the given directory yet.
check=$(ls .local/version-info/python/)
if ! [ "x$check" = x ]; then
ln -s .local/version-info/python/* $checkdir/
check=$(ls .local/version-info/proglib/)
if ! [ "x$check" = x ]; then
ln -s "$(pwd)"/.local/version-info/proglib/* $checkdir/
# If something was actually found, then print them.
if [ $printresults = 1 ]; then
echo "--- Last 5 packages that were built:"
# Then sort all the links based on the most recent dates of the
# files they link to (with '-L').
ls -Llt $checkdir \
| awk '/^-/ && c++<5 {printf "[at %s] %s\n", $(NF-1), $NF}'
cat <<EOF
The connection to the build directory (.build) is not yet created.
If you have just ran './project configure', it should be created and will
be used in a few seconds to report the build status of various software.
If not, please run './project configure' (in another terminal) and you will
see the results shortly afterwards.
echo "========================"
# Wait for the next round of checks.
sleep 1
exit 0
# Basic group settings
# --------------------
if ! [ x$group = x ]; then
# Check if group is usable.
if ! sg "$group" "echo Group \'$group\' exists"; then
echo "$scriptname: '$group' is not a usable group name on this system.";
echo "(TIP: you can use the 'groups' command to see your groups)"
exit 1
# Set the group option for running Make.
# Error when configuration isn't run
configuration_necessary() {
cat <<EOF
The project is either (1) not configured on this system, or (2) the
configuration wasn't successful.
(1) If it hasn't been configured at all, use the command below to configure
it (set a build directory and let it build its necessary software in it).
$ ./project configure
(2) If it has been configured, but the configuration failed in a step, you
can re-configure it using your previous settings with the command
below. All successful steps will be skipped, allowing a fast completion.
$ ./project configure -e
If there was a problem, please let us know by filling this online form:
exit 1
# Error when configuration isn't run
configuration_necessary() {
cat <<EOF
The project is either (1) not configured on this system, or (2) the
configuration wasn't successful.
(1) If it hasn't been configured at all, use the command below to configure
it (set a build directory and let it build its necessary software in it).
$ ./project configure
(2) If it has been configured, but the configuration failed in a step, you
can re-configure it using your previous settings with the command
below. All successful steps will be skipped, allowing a fast completion.
$ ./project configure -e
If there was a problem, please let us know by filling this online form:
exit 1
# Run operations in controlled environment
# ----------------------------------------
controlled_env() {
# Get the full address of the build directory:
bdir=`.local/bin/realpath .build`
# Remove all existing environment variables (with `env -i') and only
# use some pre-defined environment variables, then build the project.
envmake=".local/bin/env -i HOME=$bdir sys_rm=$(which rm) $gopt"
envmake="$envmake highlightnew=$highlightnew"
envmake="$envmake removesomefits=$removesomefits"
envmake="$envmake highlightnotes=$highlightnotes .local/bin/make"
envmake="$envmake --no-builtin-rules --no-builtin-variables -f $1"
if ! [ x"$debug" = x ]; then envmake="$envmake --debug=$debug"; fi
# Set the number of jobs. Note that for the `' script the
# default value has to be 0, so the default is the maximum number of
# threads. But here, the default value is 1.
if ! [ x"$jobs" = x0 ]; then envmake="$envmake -j$jobs"; fi
# Run the project
if [ x"$group" = x ]; then
$envmake $make_targets
# Set the group and permission flags.
sg "$group" "umask $perms && $envmake $make_targets"
# Do requested operation
# ----------------------
case $operation in
# Build the project's software.
# Set executable flags
# --------------------
# In some scenarios (for example when using a tarball from arXiv),
# it may happen that the host server has removed the executable
# flags of all the files. In `' we instruct the readers on
# setting the executable flag of this script. But we don't want the
# user to have to worry about any other file that needs an
# executable flag.
# Basically, all the project shell scripts need executable flags so
# to make sure they have them, we are activating the executable
# flags by default here every time './project configure' is run. If
# any other file in your project needs such flags, add them here.
chmod u+x reproduce/software/shell/* reproduce/software/config/*.sh \
# If the user requested, clean the TeX directory from the extra
# (to-be-built) directories that may already be there (and will not
# allow the configuration to complete).
if [ x"$clean_texdir" = x1 ]; then
rm -rf tex/build tex/tikz
# Variables to pass to the configuration script.
export jobs=$jobs
export debug=$debug
export host_cc=$host_cc
export need_gfortran=$need_gfortran
export build_dir=$build_dir
export input_dir=$input_dir
export scriptname=$scriptname
export maneage_group_name=$group
export software_dir=$software_dir
export existing_conf=$existing_conf
export all_highlevel=$all_highlevel
# Run the configuration script
if [ x"$group" = x ]; then
# Set the group and permission flags.
sg "$group" "umask $perms && $configscript"
# Set the group writing permission for everything in the
# installed software directory. The common build process sets
# the writing permissions of the installed programs/libraries
# to `755'. So group members can't write over a file. This
# creates problems when another group member wants to update
# the software for example. We thus need to manually add the
# group writing flag to all installed software files.
echo "Enabling group writing permission on all installed software..."
.local/bin/chmod -R g+w .local/;
# Batch execution of the project.
# Make sure the configure script has been completed properly
# (`configuration-done.txt' exists).
if ! [ -f .build/software/configuration-done.txt ]; then
# Run data preparation phase (optionally build Makefiles with
# special values for optimizing the main `'). But note
# that data preparation is only done automatically the first time
# the project is built (when `.build/software/'
# doesn't yet exist). After that, if the user wants to re-do the
# preparation they have to use the `--prepare-redo' option.
if ! [ -f .build/software/ ] \
|| [ x"$prepare_redo" = x1 ]; then
controlled_env reproduce/analysis/make/
# Run the actual project.
controlled_env reproduce/analysis/make/
# Make sure the configure script has been completed properly
# (`configuration-done.txt' exists).
if ! [ -f .build/software/configuration-done.txt ]; then
# Run the project's own shell without inheriting any environment
# from the host. The 'TERM' environment variable is necessary for
# tools like some text editors.
bdir=`.local/bin/realpath .build`
.local/bin/env -i \
HOME="$bdir" \
PATH="$instdir"/bin \
LDFLAGS=-L"$instdir"/lib \
SHELL="$instdir"/bin/bash \
CPPFLAGS=-I"$instdir"/include \
LD_LIBRARY_PATH="$instdir"/lib \
OMPI_MCA_plm_rsh_agent=/bin/false \
PYTHONPATH="$instdir"/lib/python/site-packages \
PYTHONPATH3="$instdir"/lib/python/site-packages \
PS1="[\[\033[32m\](maneage)\[\033[00m\] \u@\h \W]$ " \
# Operation not specified.
cat <<EOF
No operation defined!
Please run with '--help' for more information.
(TIP: available operations are: 'configure', 'make', or 'shell').
exit 1