Validateeach.sh

From DLXS Documentation

validateeach.sh

   1  #!/bin/sh
    2  
    3  #DATADIR=$DLXSROOT/prep/s/samplefa/data/
    4  #XMLDECL=$DLXSROOT/misc/sgml/xml.dcl
    5  #SINGLEDOCTYPE=$DLXSROOT/prep/s/samplefa/samplefa.text.inp
    6  BINDIR=$DLXSROOT/bin/s/samplefa/
    7  
    8  # Argument processing
    9  #
   10  USAGE="USAGE: $0 -d DATA_DIRECTORY -x XML_DECLARATION -t DOCTYPE_DECLARATION"
   11  
   12  while getopts d:x:t: OPT
   13  do
   14            case $OPT in
   15                 d)   if [ -d ${OPTARG} ]
   16                      then
   17                          DATADIR=${OPTARG}
   18                      else
   19                          echo "$0: Data directory: $OPTARG is not a directory." >&2
   20                          exit 1
   21                      fi 
   22                      ;;
   23                 x)   if [ -e ${OPTARG} ]
   24                      then
   25                          XMLDECL=${OPTARG}
   26                      else
   27                          echo "$0: xmldecl file: $OPTARG does not exist." >&2
   28                          exit 1
   29                      fi 
   30                      ;;
   31                 t)   if [ -e ${OPTARG} ]
   32                      then
   33                          SINGLEDOCTYPE=${OPTARG}
   34                      else
   35                          echo "$0: Doctype  file $OPTARG does not exist." >&2
   36                          exit 1
   37                      fi 
   38                      ;;
   39  
   40                 \?)  echo "$USAGE" 1>&2
   41                      exit 1
   42                      ;;
   43            esac
   44  done
   45  #---------------------------------------------------------------------
   46  
   47  
   48  for file in   ${DATADIR}*.xml 
   49  do
   50  
   51      echo "working on $file"
   52      ${BINDIR}fixdoctype.pl $file  > $file.tmp
   53      onsgmls  -wxml -w no-explicit-sgml-decl -s -f $file.err  ${XMLDECL} ${SINGLEDOCTYPE} $file.tmp
   54      rm $file.tmp
   55  done
   56  
   57  
   58  find $DATADIR -type f -size 0 -prune -exec rm {} \;
   59  

• Lines 12-44 process the command line arguments passed to the script
• Lines 48-55 process each file in the $DATADIR that end with ".xml"
• Line 52 Creates a copy of the EAD with a ".tmp" extension after it removes the doctype declaration by running the perl script "fixdoctype.pl" This program correctly deals with multiline document declarations.
• Line 53 runs onsgmls to validate the copy using the doctype declaration in $SINGLEDOCTYPE
  • -wxml tells onsgmls to warn about constructs that are not allowed in xml
  • -w no-explicit-sgml-decl tells onsgmls not to warn us that the SGML declaration was not implied
  • -s suppresses output. That means instead of getting a parse tree, we just get any errors
  • -f $file.err tells onsgmls to write errors to a file with the ".err" extension
  • $XMLDECL is an SGML Declaration for valid XML documents.
  • $SINGLEDOCTYPE is a replacement for the deleted document type declaration. It points to a copy of the EAD2002 dtd. If you have custom entities such as a logo in your finding aids, you may want to declare them in your customized version of samplefa.txt.inp and samplefa.xml.inp.
• line 54 removes the temporary files
• line 58 removes 0 byte files (any *err files that are completely empty)

samplefa.text.inp

 <!DOCTYPE ead SYSTEM "/l1/misc/sgml/ead.dtd" []>

See http://openjade.sourceforge.net/doc/nsgmls.htm for more info on onsgmls