Checking file encoding (UTF-8) in svn pre-commit

pre-commit

#!/bin/bash 

#  SVN pre-commit hook for "svn-read-only"

##############################################################
#
#  FEATURES: 
#
#  1) Introduces the "svn-read-only" property for SVN.
#
#  If the "svn-read-only" property is set on a FILE, then no
#  change to this file (except for a property change) can be
#  committed into the repository until after the property is
#  deleted.
#
#  If the "svn-read-only" property is set on a DIRECTORY,
#  then no change to an item inside the directory or inside
#  the subdirectories can be committed.
#
#  2) Checks the SPELLING of the property being set 
#  against the complete list of known properties. 
#
#  This is to prevent user from unintentional spelling
#  errors while setting properties. Trying to set the
#  properties "svn-rea-nly" for example, will trigger an
#  error message. This feature is optional and is controlled
#  by variable SPELLCHK in the script below.
#
#  PLATFORMS CURRENTLY SUPPORTED: Tested on Linux and Cygwin
#  under Windows.  Should work on a Macs as well, but not
#  tested. No support for native Windows so far.
#
#  SETTING UP: Copy this file into the hooks directory
#  (located right inside your svn repository) and modify the
#  value of SVNLOOK below.  If using a Linux platform, make
#  this file executable with chmod.
#
##############################################################
#
#  EXAMPLE OF USE: Assuming your "Project-1.0.0" directory
#  is located at /home/user/projects/tags/Project-1.0.0 on
#  your machine. Open the bash terminal and type:
#
#       cd /home/user/projects/tags
#
#  Ensure that your working copy does not have any local
#  modifications by typing "svn st". If it does have
#  modifications, they should either be committed or
#  reverted before proceeding. Now, set the property and
#  commit the property change by
#
#       svn ps svn-read-only on Project-1.0.0
#       svn ci -m "Making the project read only" Project-1.0.0
#
#  You can still make any local changes within the
#  "Project-1.0.0" directory at this point. They however can
#  not be committed. Your "Project-1.0.0" directory is now
#  safe from any unintentional commits. In order to undo
#  this behavior, type
#   
#       svn pd svn-read-only Project-1.0.0
#       svn ci -m "Making the project modifiable again" Project-1.0.0
#
##############################################################
#
#  AUTHOR: This program is designed and written by Alexander
#  Shirokov in Toronto Canada on a personal initiative. Its
#  first copy was made available on December 5, 2011 at
#  http://www.cita.utoronto.ca/~shirokov/soft/svn-hooks/svn-read-only
#
#  LICENSE & DISCLAIMER: The program is available for anyone
#  who wishes to use it, without any restrictions. There is
#  no explicit or implied warranty for its use.
#
##############################################################

REPOS="$1"
TXN="$2"

# Please make sure to set SVNLOOK to an existing path
SVNLOOK=/usr/bin/svnlook
ICONV=/usr/bin/iconv

#chenyang20141112
SVN_USER="`$SVNLOOK author -t "$TXN" "$REPOS"`"

#
# check glue stable code
#

# Set SPELLCHK to "yes" or "no".
SPELLCHK=yes

# Do not modify this, unless you want to change the name of the property.
#SVNREADONLY=svn-read-only
SVNREADONLY="glue:stable"

# If you use any additional svn properties, please add them
# to the list below.
KNOWNPROPERTIES=(
    $SVNREADONLY
    svn:externals \
    glue:stable
)

USER1="chenyang"
USER2="chenzhongming"
USER3="fengmingming"
USER4="wanglingzhao"

typeset -i blockcnt=0
typeset -i contentchangecnt=0

if [ ! -f "$SVNLOOK" ]; then
    echo "Error: file missing: $SVNLOOK"  1>&2
    echo "To correct this error, please fix the SVNLOOK in your pre-commit hook."  1>&2
    exit 3
fi


while IFS= read -r line
do
    status="${line:0:2}"
    filepath="${line:4}"

    # valid property change is always allowed
    if [ "$status" != "_U" ]; then
	contentchangecnt+=1
    fi

    while IFS= read -r property
    do
	# Strip property of whitespaces to get property
	property=`echo "$property" | awk '{$1=$1;print}'`

	# Check for nocommit property
	if [ "$property" == "$SVNREADONLY" ] && [ "$status" != "_U" ]; then
	    #chenyang20141112
	    msg="[`date`] Property \"$property\" is set on path being changed: \"$filepath\""
	    echo "$msg" >> /tmp/svn-pre-commit
	    if [ "X$SVN_USER" != "X$USER1" -a "X$SVN_USER" != "X$USER2" -a "X$SVN_USER" != "X$USER3" -a "X$SVN_USER" != "X$USER4" ]; then
	        echo "$msg"  1>&2
		blockcnt+=1
	    fi
	fi


	# Alternatively, check spelling.
	if [ "$SPELLCHK" != "yes" ]; then
	    typeset -i block=1
	    for svnproperty in ${KNOWNPROPERTIES[*]}; do
		if [ "$property" == "$svnproperty" ]; then
		    block=0
	            break
		fi
	    done

	    if [ "$block" != "0" ]; then
		echo "Invalid property svn:* : \"$property\" on path $filepath"  1>&2
		blockcnt+=$block
	    fi
	    unset block
	fi



    done < <("$SVNLOOK" proplist  -t "$TXN"  "$REPOS" "$filepath")

done  < <("$SVNLOOK" changed -t "$TXN" "$REPOS")

if [ $blockcnt != 0 ]; then
    echo "[ERROR] Commit blocked by changed item check" 1>&2
    exit 1
fi

# If the content is changed on an item that does not bear the
# SVNREADONLY, investigate upper directories for this property.
if [ "$contentchangecnt" != "0" ]; then
  
  while IFS= read -r line
  do
      dirpath="$line"
      upperdir="$dirpath"
  
      while [ 1 ]; do
	  
  	  while IFS= read -r property
  	  do
    	      # Strip the line of leading and trailing blankspaces
  	      property=`echo "$property" | awk '{$1=$1;print}'`
  
  	      if [ "$property" == "$SVNREADONLY" ]; then
	          #chenyang20150205
                  msg="[`date`] Property \"$property\" is set on the superdirectory \"$upperdir\" of an item changed \"$dirpath\""
	          echo $msg >> /tmp/svn-pre-commit
	          if [ "X$SVN_USER" != "X$USER1" -a "X$SVN_USER" != "X$USER2" -a "X$SVN_USER" != "X$USER3" -a "X$SVN_USER" != "X$USER4" ]; then
  		      echo "$msg"  1>&2
  		      blockcnt+=1
                  fi
  	      fi
  	  done < <("$SVNLOOK" proplist -t "$TXN" "$REPOS" "${upperdir}")
  
  	  if [  "$upperdir" == "." ] || [  "$upperdir" == "/" ] || [  "$upperdir" == "" ] ; then
  	      break 
  	  fi
  	  upperdir=`dirname $upperdir`
      done
  done  < <($SVNLOOK dirs-changed -t "$TXN" "$REPOS" )

  if [ $blockcnt != 0 ]; then
      echo "[ERROR] Commit blocked by superdirectory check" 1>&2
      exit 2
  fi
fi

#
# Make sure that the log message contains some text.  
#

SVNLOOKOK=1  
$SVNLOOK log -t "$TXN" "$REPOS" | grep "[a-zA-Z0-9]" > /dev/null || SVNLOOKOK=0
if [ $SVNLOOKOK = 0 ]; then  
 echo "[ERROR] Empty log messages are not allowed. Please provide a proper log message." >&2  
 exit 1  
fi  

#
# Comments should have more than 5 characters  
#

LOGMSG=$($SVNLOOK log -t "$TXN" "$REPOS" | grep [a-zA-Z0-9] | wc -c)  
  
if [ "$LOGMSG" -lt 6 ]; then  
 echo -e "[ERROR] Please provide a meaningful comment when committing changes." 1>&2  
 exit 1  
fi 

#
# Make sure that all files to be committed are encoded in UTF-8.  
#

while read changeline;   
do  
  
  # Get just the file (not the add / update / etc. status).  
  file=${changeline:4}  
  
  # Only check source files.  
  if [[ $file == *.java || $file == *.html || $file == *.esp ]] ; then  
    $SVNLOOK cat -t "$TXN" "$REPOS" "$file" | $ICONV -f UTF-8 -t UTF-8 -o /dev/null  
    if [ "${PIPESTATUS[1]}" != 0 ] ; then  
      echo "[ERROR] Only UTF-8 files can be committed ("$file")" 1>&2  
      exit 1  
    fi
    # ticket:14508 has Zero-width space?
    # hex dump
    #---------------------------------------------
    #e2808b  68747470 3a 2f2f 777777 
    #        http     :  //   www    .xuaran001.com
    #---------------------------------------------
    #$SVNLOOK cat -t "$TXN" "$REPOS" "$file" | grep "$(printf %b '\u200b')"
    $SVNLOOK cat -t "$TXN" "$REPOS" "$file" | xxd -p - | tr -d '\n' | grep "e2808b68747470"
    if [ "${PIPESTATUS[3]}" = 0 ] ; then  
      echo "[ERROR100] Find zero-width space ("$file")" 1>&2  
      #exit 1
    fi
  fi  
done < <($SVNLOOK changed -t "$TXN" "$REPOS")  

# All checks passed, so allow the commit.
exit 0