Add a script that counts lines of code.

diff --git a/contrib/utilities/count_lines.sh b/contrib/utilities/count_lines.sh
new file mode 100755 (executable)
index 0000000..7397a33
--- /dev/null
+++ b/contrib/utilities/count_lines.sh
@@ -0,0 +1,89 @@
+#!/bin/sh
+## ---------------------------------------------------------------------
+##
+## Copyright (C) 2018 by the deal.II authors
+##
+## This file is part of the deal.II library.
+##
+## The deal.II library is free software; you can use it, redistribute
+## it, and/or modify it under the terms of the GNU Lesser General
+## Public License as published by the Free Software Foundation; either
+## version 2.1 of the License, or (at your option) any later version.
+## The full text of the license can be found in the file LICENSE.md at
+## the top level directory of deal.II.
+##
+## ---------------------------------------------------------------------
+
+#
+# This is a little script that can be used to count the lines of source
+# code in both the regular source/include/example files as well as the
+# tests directories. This is occasionally useful to assess the growth
+# of the library over time using a crude metric of its size.
+#
+# The script assumes that it is called in the top-level directory
+# of a deal.II git repository. It counts the number of code lines
+# for every 20th commit going back from the current 'master' branch
+# to the beginning of the repository history.
+#
+# It takes, on average, about 1/5th of a second to count lines for one
+# commit, averaged over the lifetime of the deal.II project. At the time
+# when this script was written, the repository had about 41,000 commits,
+# so counting lines for every 20th commit takes approximately 400 seconds,
+# or 6m30s.
+#
+# NOTE: The script has existed in this repository only for a finite time.
+#       So, if you go back in history, as this script does, it will
+#       eventually disappear. As a consequence, you cannot execute it
+#       in its current place -- copy it to the top level of the working
+#       copy you are working in, where it will be an untracked file that
+#       git leaves alone as it cycles through repository commits.
+#
+#
+# The output of this file consists of three numbers per line, showing
+#   date source-lines test-lines
+# displaying the number of lines of code in .h and .cc files (and .cu and
+# .cuh files for CUDA), but excluding files in contributed libraries. The
+# resulting output can be piped into a data file and then be visualized
+# by importing into a spreadsheet, or using the following GNUPLOT script:
+#
+#    set xdata time
+#    set timefmt "%Y-%m-%d"
+#    set xrange ["1997-11-01":"2018-08-01"]
+#    set format x "%Y"
+#    set style data lines
+#    set key top left
+#    
+#    set terminal png
+#    set output "line-count.png"
+#    
+#    plot "< cat line-count.dat | sort" using 1:2 title "Lines of code in source files", \
+#         "" using 1:3 title "Lines of code in tests"
+#
+#
+
+git checkout -q master
+
+commits=`git log | \
+         egrep '^commit ' | \
+         perl -p -e 's/^commit //g;' | \
+         perl -e '$i=0; while (<>) { ++$i; if ($i % 200 == 0) { print; } }'`
+
+for commit in $commits ; do
+  git checkout -q $commit
+  
+  date=`git log --date=short --format="%ad" -n 1`
+
+  files_source=`find . -name '*.h' -or -name '*.cc' -or -name '*.cu' -or -name '*.cuh' -type f | \
+                egrep -i -v '(tests|boost|umfpack|bundled)/'`
+  lines_source=`cat $files_source | wc -l`
+
+  files_tests=`find . -name '*.h' -or -name '*.cc' -type f | \
+               egrep -i -v '(boost|umfpack|bundled)/' | \
+               grep tests/`
+  lines_tests=`cat $files_tests | wc -l`
+  
+  echo $date $lines_source $lines_tests
+done
+
+git checkout -q master
+