rclxml

2012-05-19 09:23:24 +02:00 · 2012-05-19 09:23:24 +02:00 · cbe7fd21cb
commit cbe7fd21cb
parent aee22e093a
1 changed files with 119 additions and 0 deletions
--- a/src/filters/rclxml
+++ b/src/filters/rclxml
@ -0,0 +1,119 @@
+#!/bin/sh
+
+#================================================================
+# Extract text from a generic XML file (Justus Piater)
+#================================================================
+
+# set variables
+LANG=C ; export LANG
+LC_ALL=C ; export LC_ALL
+progname="rclxml"
+filetype=xml
+
+
+#RECFILTCOMMONCODE
+##############################################################################
+# !! Leave the previous line unmodified!! Code imported from the
+# recfiltcommon file
+
+# Utility code common to all shell filters. This could be sourced at run
+# time, but it's slightly more efficient to include the code in the
+# filters at build time (with a sed script).
+
+# Describe error in a way that can be interpreted by our caller
+senderror()
+{
+    echo RECFILTERROR $*
+    # Also alert on stderr just in case
+    echo ":2:$progname::: $*" 1>&2
+    exit 1
+}
+
+iscmd()
+{
+    cmd=$1
+    case $cmd in
+    */*)
+	if test -x $cmd -a ! -d $cmd ; then return 0; else return 1; fi ;;
+    *)
+      oldifs=$IFS; IFS=":"; set -- $PATH; IFS=$oldifs
+      for d in $*;do test -x $d/$cmd -a ! -d $d/$cmd && return 0;done
+      return 1 ;;
+    esac
+}
+
+checkcmds()
+{
+    for cmd in $*;do
+      if iscmd $cmd 
+      then 
+        a=1
+      else 
+        senderror HELPERNOTFOUND $cmd
+      fi
+    done
+}
+
+# show help message
+if test $# -ne 1 -o "$1" = "--help" 
+then
+  echo "Convert a $filetype file to HTML text for Recoll indexing."
+  echo "Usage: $progname [infile]"
+  exit 1
+fi
+
+infile="$1"
+
+# check the input file existence (may be '-' for stdin)
+if test "X$infile" != X- -a ! -f "$infile"
+then
+  senderror INPUTNOSUCHFILE "$infile"
+fi
+
+# protect access to our temp files and directories
+umask 77
+
+##############################################################################
+# !! Leave the following line unmodified !
+#ENDRECFILTCOMMONCODE
+
+checkcmds xsltproc
+
+xsltproc --novalid --nonet - "$infile" <<EOF
+<?xml version="1.0"?>
+<xsl:stylesheet version="1.0"
+		xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+
+  <xsl:output method="html" encoding="UTF-8"/>
+
+  <xsl:template match="/">
+    <html>
+      <head>
+	<xsl:if test="//*[local-name() = 'title']">
+	  <title>
+	    <xsl:value-of select="//*[local-name() = 'title'][1]"/>
+	  </title>
+	</xsl:if>
+      </head>
+      <body>
+	<xsl:apply-templates/>
+      </body>
+    </html>
+  </xsl:template>
+
+  <xsl:template match="text()">
+    <xsl:if test="string-length(normalize-space(.)) &gt; 0">
+      <p><xsl:value-of select="."/></p>
+      <xsl:text>
+      </xsl:text>
+    </xsl:if>
+  </xsl:template>
+
+  <xsl:template match="*">
+    <xsl:apply-templates/>
+  </xsl:template>
+
+</xsl:stylesheet>
+EOF
+
+exit 0