Port Empirical tidy enforcement

Author: mmore500

.github/workflows/CI.yml

@@ -49,6 +49,8 @@ jobs:
   tidy:
     runs-on: ubuntu-latest
     steps:
+    - name: Install apt dependencies
+      run: sudo apt-get install -y rename
     - uses: actions/checkout@v1
     - name: Set up Python 3.8
       uses: actions/setup-python@v2
@@ -63,6 +65,8 @@ jobs:
     - uses: editorconfig-checker/action-editorconfig-checker@main
     - name: Test editorconfig tidyness
       run: editorconfig-checker
+    - name: Test tidy/ tidyness
+      run: ./tidy/test_tidy.sh
 
   deploy:
     needs: [coverage, test, tidy]

tidy/impl/alphabetize_includes.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+TARGETS=$(find . -type f \( -name "*.hpp" -o -name "*.cpp" \) ! -path "./third-party/*" ! -path "./node_modules/*")
+
+for filename in ${TARGETS}
+do
+
+  printf "."
+
+  # adapted from https://stackoverflow.com/a/6970681
+  include_linenos=$(awk '/^#include .*/ {print FNR}' "${filename}")
+
+  # adapted from https://stackoverflow.com/a/26809816
+  # and https://superuser.com/a/284192
+  echo "${include_linenos}" | tr ' ' '\n' | awk 'NR==1{first=$1;last=$1;next} $1 == last+1 {last=$1;next} {print first,last;first=$1;last=first} END{print first,last}' | while read line ; do
+
+    # adapted from https://unix.stackexchange.com/a/11064
+    read start stop <<< "${line}"
+
+    # adapted from https://stackoverflow.com/a/46018238
+    # and https://stackoverflow.com/a/32723119
+    # and https://unix.stackexchange.com/a/87748
+    export LC_ALL=C
+    echo "x" | ex -s -c "${start},${stop}!sort --stable --ignore-nonprinting --ignore-case --dictionary-order" $filename
+    # see also https://www.gnu.org/software/coreutils/manual/html_node/sort-invocation.html
+
+  done
+done
+
+echo

tidy/impl/ensure_end_of_file_newlines.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+# greq -qI tests if file is binary
+# adapted from https://stackoverflow.com/a/29689345
+TARGETS=$(find . -type f ! -path "./third-party/*" ! -path "./.git/*" ! -path "*/assets/*" ! -path "*.tar.gz" ! -path "*.jpg" ! -path "*.png" ! -path "*Makefile" ! -path "*Maketemplate*" ! -path "./node_modules/*" -exec grep -qI . {} ';' -print)
+
+for filename in ${TARGETS}
+do
+  sed -i -e '$a\' "${filename}"
+done

tidy/impl/generate_boilerplate.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+set -e
+
+./tidy/impl/generate_boilerplate_headerguards.sh

tidy/impl/generate_boilerplate_headerguards.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+for filename in $(cd include && find * -name '*.hpp' -type f); do
+
+  GUARD=$( sed "s/[^[:alnum:]]/_/g" <<< "${filename}" | tr [a-z] [A-Z] )
+  NDEF_LINE="#ifndef ${GUARD}_INCLUDE"
+  DEF_LINE="#define ${GUARD}_INCLUDE"
+  ENDIF_LINE="#endif // ${NDEF_LINE}"
+
+  sed -i '1s/^.*$/'"#pragma once"'/' "include/${filename}"
+  sed -i '2s/^.*$/'"${NDEF_LINE}"'/' "include/${filename}"
+  sed -i '3s/^.*$/'"${DEF_LINE}"'/' "include/${filename}"
+  sed -i '$ s,^.*$,'"${ENDIF_LINE}"',g' "include/${filename}"
+
+done

tidy/impl/generate_license_notices.sh

@@ -0,0 +1,49 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+# license notice boilerplate for source files that aren't in include/
+
+for filename in $(find . -name '*.cpp' -type f ! -path "./third-party/*") $(find . -name '*.hpp' -type f ! -path "./third-party/*" ! -path "./include/*"); do
+
+  # grow file up to at least 8 lines
+  FILE_NUM_LINES="$(cat "${filename}" | wc -l)"
+  for __ in $(seq "${FILE_NUM_LINES}" 1 6); do
+    echo >> "${filename}"
+  done
+
+  # stamp in expected boilerplate line-by-line
+  # just like file docstrings, but don't require a brief
+  # stamp in expected boilerplate line-by-line
+  sed -i '1s|^.*$|/**|' "${filename}"
+  sed -i '2s|^.*$| *  @note This file is part of Empirical, https://github.com/devosoft/Empirical|' "${filename}"
+  sed -i '3s|^.*$| *  @copyright Copyright (C) Michigan State University, MIT Software license; see doc/LICENSE.md|' "${filename}"
+  # only match if a @date isn't currently in place
+  # if we see @date, "break" (b) sed script for that line
+  # adapted from https://stackoverflow.com/a/5334825
+  # and https://stackoverflow.com/a/12178023
+  # and https://stackoverflow.com/a/9053163
+  sed -i "/^ \*  @date /b; 4s/^.*\$/ *  @date $(date +'%Y')/" "${filename}"
+  sed -i '5s/^.*$/ */' "${filename}"
+  sed -i "6s/^.*\$/ *  @file $(basename "${filename}")/" "${filename}"
+  # only match empty lines
+  # add extra * to replace later with */ when constructing fresh
+  sed -i '7s/^$/ */' "${filename}"
+
+  # close boilerplate file docstring
+  # must accomodate possible additional content in docstring
+  # so, search backwards from first blank line to place close */
+  FILE_NUM_LINES="$(cat "${filename}" | wc -l)"
+  ONE_PAST_LAST_LINE_NO="$((${FILE_NUM_LINES}+1))"
+  # adapted from https://stackoverflow.com/a/66474842
+  FIRST_BLANK_LINE_NO="$(awk '! NF { print NR; exit }' "${filename}")"
+  # two cases: file does or does not have a blank line
+  FIRST_EMPTY_LINE_NO="$(grep -q '^$' "${filename}" && echo "${FIRST_BLANK_LINE_NO}" || echo "${ONE_PAST_LAST_LINE_NO}")"
+  DOCSTRING_CLOSE_LINE_NO="$((${FIRST_EMPTY_LINE_NO}-1))"
+  # stamp */ into line preceding first empty line
+  sed -i "${DOCSTRING_CLOSE_LINE_NO}s|^.*\$| */|" "${filename}"
+
+
+done

tidy/impl/modernize_filename_suffixes.sh

@@ -0,0 +1,12 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+# enforce availability of dependencies
+. ./tidy/util/enforce_dependency.sh rename
+
+find ! -path "./third-party/*" ! -path "./.git/*" -type f | rename 's/\.h$/.hpp/'
+find ! -path "./third-party/*" ! -path "./.git/*" -type f | rename 's/\.c$/.cpp/'
+find ! -path "./third-party/*" ! -path "./.git/*" -type f | rename 's/\.cc$/.cpp/'

tidy/impl/partition_includes.sh

@@ -0,0 +1,24 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+TARGETS=$(find . -type f \( -name "*.hpp" -o -name "*.cpp" \) ! -path "./third-party/*" ! -path "./node_modules/*")
+
+for filename in ${TARGETS}
+do
+
+  # insert a blank line between #include <'s and #include "'s
+  # adapted from https://www.unix.com/shell-programming-and-scripting/186893-how-insert-line-between-two-consecutive-lines-match-special-pattern.html
+  awk -v ADD='' -v ANGLEINCLUDE='^#include <.+>$' -v QUOTEINCLUDE='^#include ".+"$' '/^.+$/ L { print L; if(((L ~ ANGLEINCLUDE) && ($0 ~ QUOTEINCLUDE)) || ((L ~ QUOTEINCLUDE) && ($0 ~ ANGLEINCLUDE)) ) print ADD }; { L=$0 } END { print L }' "${filename}" | awk 'NR!=1 {print}' > "${filename}.bak"
+  mv "${filename}.bak" "${filename}"
+
+  # insert a blank line between #include "'s and #include "../'s
+  # adapted from https://www.unix.com/shell-programming-and-scripting/186893-how-insert-line-between-two-consecutive-lines-match-special-pattern.html
+  awk -v ADD='' -v PWDINCLUDE='^#include "[0-9a-zA-Z_].*"$' -v RELATIVEINCLUDE='^#include "../[0-9a-zA-Z_].*"$' '/^.+$/ L { print L; if(((L ~ PWDINCLUDE) && ($0 ~ RELATIVEINCLUDE)) || ((L ~ RELATIVEINCLUDE) && ($0 ~ PWDINCLUDE)) ) print ADD }; { L=$0 } END { print L }' "${filename}" | awk 'NR!=1 {print}' > "${filename}.bak"
+  mv "${filename}.bak" "${filename}"
+
+done
+
+echo

tidy/impl/replace_tabs.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+# greq -qI tests if file is binary
+# adapted from https://stackoverflow.com/a/29689345
+TARGETS=$(find . -type f ! -path "./third-party/*" ! -path "./.git/*" ! -path "*/assets/*" ! -path "*Makefile" ! -path "*Maketemplate*" ! -path "./node_modules/*" -exec grep -qI . {} ';' -print)
+
+for filename in ${TARGETS}
+do
+  sed -i 's/\t/  /g' "${filename}"
+done

tidy/impl/strip_filename_whitespace.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+set -e
+
+# enforce use of GNU version of coreutils
+. ./tidy/util/enforce_gnu_utils.sh
+
+# enforce availability of dependencies
+. ./tidy/util/enforce_dependency.sh rename
+
+find ! -path "./third-party/*" ! -path "./node_modules/*" -type d | grep '\s' | rename 's/\s/_/g'    # do the directories first
+find ! -path "./third-party/*" ! -path "./node_modules/*" -type f | grep '\s' | rename 's/\s/_/g'    # do the directories first