Merge pull request #1148 from dra27/git-precommit-hook

Supply a Git pre-commit hook for tools/check-typo

.gitattributes

@@ -56,16 +56,16 @@ README*                  ocaml-typo=missing-header
 asmcomp/*/emit.mlp       ocaml-typo=tab,long-line,unused-prop
 asmcomp/power/NOTES.md   ocaml-typo=missing-header
 
-config/gnu               ocaml-typo=prune
+/config/gnu              ocaml-typo=prune
 
 emacs/*.el               ocaml-typo=long-line,unused-prop
 emacs/caml.el            ocaml-typo=long-line,unused-prop,missing-header
 emacs/COPYING            ocaml-typo=tab,non-printing,missing-header
 emacs/ocamltags.in       ocaml-typo=non-printing
 
-experimental             ocaml-typo=prune
+/experimental            ocaml-typo=prune
 
-manual                   ocaml-typo=prune
+/manual                  ocaml-typo=prune
 
 ocamldoc/Changes.txt     ocaml-typo=missing-header
 ocamldoc/ocamldoc.sty    ocaml-typo=missing-header
@@ -135,6 +135,7 @@ tools/make-package-macosx text eol=lf
 tools/ocaml-objcopy-macosx text eol=lf
 tools/ocamlmktop.tpl text eol=lf
 tools/ocamlsize text eol=lf
+tools/pre-commit-githook text eol=lf
 
 # These two are cat scripts, so may not actually require this
 config/auto-aux/sharpbang text eol=lf

.travis.yml

@@ -34,6 +34,7 @@ matrix:
   - env: CI_KIND=build XARCH=x64 CONFIG_ARG=-flambda OCAMLRUNPARAM=b,v=0
   - env: CI_KIND=changes
   - env: CI_KIND=manual
+  - env: CI_KIND=check-typo
   - env: CI_KIND=tests
   allow_failures:
   - env: CI_KIND=tests

CONTRIBUTING.md

@@ -68,6 +68,9 @@ columns, not use tab characters (spaces only), and not use non-ASCII
 characters. These typographical rules can be checked with the script
 `tools/check-typo`.
 
+If you are working from a Git clone, you can automate this process by
+copying the file `tools/pre-commit-githook` to `.git/hooks/pre-commit`.
+
 Otherwise, there are no strongly enforced guidelines specific to the
 compiler -- and, as a result, the style may differ in the different
 parts of the compiler. The general [OCaml Programming

tools/check-typo

@@ -91,7 +91,8 @@ export LC_ALL=C
 # Special case for recursive call from the find command (see IGNORE_DIRS).
 case "$1" in
   --check-prune)
-    case `git check-attr ocaml-typo "$2" 2>/dev/null` in
+    case `env $OCAML_CT_GIT_INDEX git check-attr $OCAML_CT_CA_FLAG \
+                                                 ocaml-typo "$2" 2>/dev/null` in
       *prune*) echo "INFO: pruned directory $2 (ocaml-typo=prune)" >&2; exit 0;;
       *) exit 3;;
     esac;;
@@ -113,6 +114,16 @@ while : ; do
   esac
 done
 
+OCAML_CT_CAT=${OCAML_CT_CAT:-cat}
+OCAML_CT_LS_FILES=${OCAML_CT_LS_FILES:-git ls-files}
+OCAML_CT_HEAD=${OCAML_CT_HEAD:-HEAD}
+OCAML_CT_AWK=${OCAML_CT_AWK:-awk}
+if [ -z "${OCAML_CT_GIT_INDEX+x}" ] ; then
+  OCAML_CT_GIT_INDEX=
+else
+  OCAML_CT_GIT_INDEX="GIT_INDEX_FILE=$OCAML_CT_GIT_INDEX"
+fi
+
 IGNORE_DIRS="
   -name .git -prune -o
   -type d -exec $0 --check-prune {} ; -prune -o
@@ -124,7 +135,7 @@ IGNORE_DIRS="
   esac
 ) | (
   while read f; do
-    case `git ls-files "$f" 2>&1` in
+    case `$OCAML_CT_LS_FILES "$f" 2>&1` in
       "") is_svn=false;;
       *) is_svn=true;;
     esac
@@ -147,9 +158,11 @@ IGNORE_DIRS="
       #    the documentation for the --numstat option. Commands designated as
       #    "plumbing" commands in git have stable output intended for parsing)
       EMPTY=`git hash-object -t tree /dev/null`
-      git diff-tree --numstat $EMPTY HEAD -- "$f" | grep -q "^-[[:blank:]]-" \
-        && continue
-      svnrules=`git check-attr ocaml-typo "$f" | sed -e 's/.*: //'`
+      git diff-tree --numstat $EMPTY $OCAML_CT_HEAD -- "$f" \
+        | grep -q "^-[[:blank:]]-" && continue
+      svnrules=`env $OCAML_CT_GIT_INDEX git check-attr $OCAML_CT_CA_FLAG \
+                                                       ocaml-typo "$f" \
+                  | sed -e 's/.*: //'`
       case $svnrules in unspecified) svnrules= ;; esac
     fi
     rules="$userrules"
@@ -169,15 +182,16 @@ IGNORE_DIRS="
         # requires the entire line to match. This specifically detects the
         # presence of lines containing malformed UTF-8. It may be tested using
         # https://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt
-        if LC_ALL=en_US.UTF8 grep -qaxv '.*' "$f" ; then
+        if $OCAML_CT_CAT "$OCAML_CT_PREFIX$f" \
+             | LC_ALL=en_US.UTF8 grep -qaxv '.*' ; then
           echo File "$f" is not correctly encoded in UTF-8
           exit 2
         fi
         ;;
     esac
 
-    (cat "$f" | tr -d '\r'; echo) \
-    | awk -v rules="$rules" -v svnrules="$svnrules" -v file="$f" \
+    ($OCAML_CT_CAT "$OCAML_CT_PREFIX$f" | tr -d '\r'; echo) \
+    | $OCAML_CT_AWK -v rules="$rules" -v svnrules="$svnrules" -v file="$f" \
       '
         function is_err(name) {
           return (("," rules svnrules ",") !~ ("[, ]" name "[, ]"));

tools/ci/inria/extra-checks

@@ -226,3 +226,11 @@ TSAN_OPTIONS="die_after_fork=0" $run_testsuite
 # # Build the system (bytecode only) and test
 # make $jobs world
 # $run_testsuite
+
+#########################################################################
+
+# Ensure that the repo still passes the check-typo script
+if [ ! -x tools/check-typo ] ; then
+  error "tools/check-typo does not appear to be executable?"
+fi
+tools/check-typo

tools/ci/travis/travis-ci.sh

@@ -157,6 +157,80 @@ EOF
     testsuite > /dev/null && exit 1 || echo pass
 }
 
+# Test to see if any part of the directory name has been marked prune
+not_pruned () {
+  DIR=$(dirname "$1")
+  if [ "$DIR" = "." ] ; then
+    return 0
+  else
+    case ",$(git check-attr ocaml-typo "$DIR" | sed -e 's/.*: //')," in
+      ,prune,)
+      return 1
+      ;;
+      *)
+
+      not_pruned $DIR
+      return $?
+    esac
+  fi
+}
+
+CheckTypoTree () {
+  export OCAML_CT_HEAD=$1
+  export OCAML_CT_LS_FILES="git diff-tree --no-commit-id --name-only -r $2 --"
+  export OCAML_CT_CAT="git cat-file --textconv"
+  export OCAML_CT_PREFIX="$1:"
+  GIT_INDEX_FILE=tmp-index git read-tree --reset -i $1
+  git diff-tree --diff-filter=d --no-commit-id --name-only -r $1 \
+    | (while IFS= read -r path
+  do
+    if not_pruned $path ; then
+      echo "Checking $1: $path"
+      if ! tools/check-typo ./$path ; then
+        touch check-typo-failed
+      fi
+    else
+      echo "NOT checking $1: $path (ocaml-typo=prune)"
+    fi
+  done)
+  rm -f tmp-index
+}
+
+CHECK_ALL_COMMITS=0
+
+CheckTypo () {
+  export OCAML_CT_GIT_INDEX="tmp-index"
+  export OCAML_CT_CA_FLAG="--cached"
+  # Work around an apparent bug in Ubuntu 12.4.5
+  # See https://bugs.launchpad.net/ubuntu/+source/gawk/+bug/1647879
+  export OCAML_CT_AWK="awk --re-interval"
+  rm -f check-typo-failed
+  if test -z "$TRAVIS_COMMIT_RANGE"
+  then CheckTypoTree $TRAVIS_COMMIT $TRAVIS_COMMIT
+  else
+    if [ "$TRAVIS_EVENT_TYPE" = "pull_request" ]
+    then TRAVIS_COMMIT_RANGE=$TRAVIS_BRANCH..$TRAVIS_PULL_REQUEST_SHA
+    fi
+    if [ $CHECK_ALL_COMMITS -eq 1 ]
+    then
+      for commit in $(git rev-list $TRAVIS_COMMIT_RANGE --reverse)
+      do
+        CheckTypoTree $commit $commit
+      done
+    else
+      if [ -z "$TRAVIS_PULL_REQUEST_SHA" ]
+      then CheckTypoTree $TRAVIS_COMMIT $TRAVIS_COMMIT
+      else CheckTypoTree $TRAVIS_PULL_REQUEST_SHA $TRAVIS_COMMIT_RANGE
+      fi
+    fi
+  fi
+  echo complete
+  if [ -e check-typo-failed ]
+  then exit 1
+  fi
+}
+
+
 case $CI_KIND in
 build) BuildAndTest;;
 changes)
@@ -169,6 +243,9 @@ tests)
     case $TRAVIS_EVENT_TYPE in
         pull_request) CheckTestsuiteModified;;
     esac;;
+check-typo)
+   set +x
+   CheckTypo;;
 *) echo unknown CI kind
    exit 1
    ;;

tools/pre-commit-githook

@@ -0,0 +1,76 @@
+#!/usr/bin/env bash
+#**************************************************************************
+#*                                                                        *
+#*                                 OCaml                                  *
+#*                                                                        *
+#*                David Allsopp, MetaStack Solutions Ltd.                 *
+#*                                                                        *
+#*   Copyright 2017 MetaStack Solutions Ltd.                              *
+#*                                                                        *
+#*   All rights reserved.  This file is distributed under the terms of    *
+#*   the GNU Lesser General Public License version 2.1, with the          *
+#*   special exception on linking described in the file LICENSE.          *
+#*                                                                        *
+#**************************************************************************
+
+# For what it's worth, allow for empty trees!
+if git rev-parse --verify HEAD >/dev/null 2>&1
+then
+  against=HEAD
+else
+  # Initial commit: diff against an empty tree object
+  against=4b825dc642cb6eb9a060e54bf8d69288fbee4904
+fi
+
+# Redirect output to stderr.
+exec 1>&2
+
+# Check to see if the script's been updated
+if git ls-files --error-unmatch tools/pre-commit-githook >/dev/null 2>&1 ; then
+  if ! diff -q <(git cat-file --textconv HEAD:tools/pre-commit-githook) "$0" \
+       >/dev/null 2>&1 ; then
+    echo "Note: tools/pre-commit-githook differs from .git/hooks/pre-commit"
+    echo "You may wish to upgrade your local githook"
+  fi
+fi
+
+# Git's built-in mechanism for whitespace is neater than ours, so do it first.
+# The strange construction below creates a list of files which have either
+# white-at-eol or white-at-eof included in ocaml-typo in .gitattributes and by
+# prefixing the names with :! causes git diff-index to skip over them.
+FILES=$(git diff-index --cached --name-only $against \
+          | xargs git check-attr --cached ocaml-typo \
+          | sed -ne 's/\(.*\): ocaml-typo:.*[ ,]white-at-eo[fl]\(,\|$\)/:!\1/p')
+if ! git diff-index --check --cached $against -- $FILES ; then
+  exit 1
+fi
+
+# Test to see if any part of the directory name has been marked prune
+not_pruned () {
+  DIR=$(dirname "$1")
+  if [ "$DIR" = "." ] ; then
+    return 0
+  else
+    case ",$(git check-attr ocaml-typo "$DIR" | sed -e 's/.*: //')," in
+      ,prune,)
+      return 1
+      ;;
+      *)
+
+      not_pruned $DIR
+      return $?
+    esac
+  fi
+}
+
+# Now run check-typo over all the files in the index
+ERRORS=0
+export OCAML_CT_PREFIX=:
+export OCAML_CT_CAT="git cat-file --textconv"
+export OCAML_CT_CA_FLAG=--cached
+git diff --diff-filter=d --staged --name-only | (while IFS= read -r path
+do
+  if not_pruned $path && ! tools/check-typo ./$path ; then
+    ERRORS=1
+  fi
+done; exit $ERRORS)