Grep for pattern in first occurrence of XML element - bash

Question

I have a file with multiple occurrences of an XML element. I want to grep for a pattern only in the first element. I want to use grep because I need to use this as the condition of an if check in a bash script. NOTE that unfortunately, I am not guaranteed that the XML element(s) are contained in an enclosing tag (this file is generated by another program out of my control).
Example of a match for "mango"
<element>
apple
banana
orange
mango
</element>
<element>
apple
banana
orange
mango
</element>
Example of a non-match for "mango"
In the following XML snippet, I want my search to fail b/c mango doesn't exist in the first element.
<element>
apple
banana
orange
</element>
<element>
apple
banana
orange
mango
</element>

Answer 1

Here's how I solved this, but I had to use a pipe combining grep with sed. This solution only worked for me because the first <element> is on the first line of the file.
sed -n '0,/<\/element>/p' /path/to/file | grep -q mango
Uses sed to print the first line of the file up to the first closing tag for element.
Uses grep to exit true or false if it matches mango.

Answer 2

For the handling of XML data I would always recommend XML tools. Only that tools can handle the specifics of XML in a save way. For the commandline is a tool called xsltproc available. This is a simple to use XSLT processor and it can do the job better than sed. The only drawback you need an additional xslt stylesheet.
Example stylesheet: test.xslt
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
version="1.0">
<xsl:output method="text"/>
<xsl:template match="element[position()=1]">
<xsl:value-of select="."/>
</xsl:template>
<xsl:template match="*|#*|text()|comment()|processing-instruction()">
<xsl:apply-templates select="*|#*|text()|comment()|processing-instruction()"/>
</xsl:template>
</xsl:stylesheet>
With the stylesheet and xsltproc you can do a command like this
xsltproc test.xslt test.xml | grep mango

Answer 3

This may be quite a lengthy solution, however it works.
./check.sh mango
This calls a simple awk script for each file, referenced by the FILES variable
note:
The xml files I saved as: xml1, xml2
For the example above, it produces the following output:
mango found in xml1
mango not found in xml2
is-here.awk:
BEGIN {
tagOpened="not yet"
tagsPresent=0
}
/<[[:alnum:]]+>/ {
if (tagsPresent <= 1) # remove this condition to check ALL occurencies
{
tagOpened="true"
tagsPresent++
}
}
/<[/][[:alnum:]]+>/ {
tagOpened="false"
}
// {
if (match($1, value) && tagOpened=="true" && length($1)==length(value))
{
found++
}
}
END {
if (found == tagsPresent)
{
print "present"
}
else
{
print "not"
}
}
check.sh
#! /bin/bash
function check()
{
local file=$1
local pattern=$2
local result=$(cat $file | gawk -f is-here.awk -v value=$pattern)
echo $result
}
FILES="xml1 xml2"
for file in $FILES
do
result=$(check $file $1)
if [ "$result" == "present" ]
then
echo "$1 found in $file"
else
echo "$1 not found in $file"
fi
done