How to extract few tabs from a xml file using zgrep or sed

Question

I have a big size file like 5GB with .gz. Inside that file, we have few XML files that contains values that I want to search and extract just in case if those values are there.

For example I want to extract the tags that contains the name NOOSS and also the subcontent of this tags like <pmJobId>, <requestedJobState>, <reportingPeriod>, <jobPriority> from the the .gz file

<Pm xmlns="urnCmwPm">
    <pmId>1</pmId>
    <PmJob>
        <pmJobId>NOOSSCONTROLExample</pmJobId>
        <requestedJobState>ACTIVE</requestedJobState>
        <reportingPeriod>FIVE_MIN</reportingPeriod>
        <jobType>MEASUREMENTJOB</jobType>
        <jobPriority>HIGH</jobPriority>
        <granularityPeriod>FIVE_MIN</granularityPeriod>
        <jobGroup>Sla</jobGroup>
        <reportContentGeneration>CHANGED_ONLY</reportContentGeneration>
        <MeasurementReader>
            <measurementReaderId>mr_2</measurementReaderId>
            <measurementSpecification struct="MeasurementSpecification">
                <measurementTypeRef>Anything</measurementTypeRef>
            </measurementSpecification>
            <thresholdRateOfVariation>PER_SECOND</thresholdRateOfVariation>
        </MeasurementReader>
        <MeasurementReader>
            <measurementReaderId>mr_1</measurementReaderId>
            <measurementSpecification struct="MeasurementSpecification">
                <measurementTypeRef>ManagedElement=1,SystemFunctions=1,Pm=1,PmGroup=OSProcessingLogicalUnit,MeasurementType=CPULoad.Total</measurementTypeRef>
            </measurementSpecification>
            <thresholdRateOfVariation>PER_SECOND</thresholdRateOfVariation>
        </MeasurementReader>
    </PmJob>
</Pm>

I was using cat *gz 1 zgrep -a "PmJobId" but the output only show the <pmJobId> value and not the rest of the information or tags.

Please your help, I'm noobie on this.

Im using CentOS - RedHat Linux.

Thanks

Answer 1

Assuming that the XML document in file.xml is well-formed and correct in all ways (the example in the question has a faulty namespace declaration), then you would be able to extract the part of the document that corresponds to the PmJob nodes that have a pmJobID value that contains the substring NOOSS with the command-line XML parser xmlstarlet.

xmlstarlet sel -t -c '//PmJob[contains(pmJobId,"NOOSS")]' -nl file.xml

This command selects all PmJob nodes with a child node, pmJobId, whose value contains the substring NOOSS. The utility will return a copy of the selected PmJob node and all its child nodes.

Answer 2

Assuming that the XML document is well-formed and valid, you can use the xmllint utility to output the requisite nodes.

$ xmllint --xpath '//PmJob[contains(pmJobId,"NOOSS")]' file.xml
<PmJob>
        <pmJobId>NOOSSCONTROLExample</pmJobId>
        <requestedJobState>ACTIVE</requestedJobState>
        <reportingPeriod>FIVE_MIN</reportingPeriod>
        <jobType>MEASUREMENTJOB</jobType>
        <jobPriority>HIGH</jobPriority>
        <granularityPeriod>FIVE_MIN</granularityPeriod>
        <jobGroup>Sla</jobGroup>
        <reportContentGeneration>CHANGED_ONLY</reportContentGeneration>
        <MeasurementReader>
            <measurementReaderId>mr_2</measurementReaderId>
            <measurementSpecification struct="MeasurementSpecification">
                <measurementTypeRef>Anything</measurementTypeRef>
            </measurementSpecification>
            <thresholdRateOfVariation>PER_SECOND</thresholdRateOfVariation>
        </MeasurementReader>
        <MeasurementReader>
            <measurementReaderId>mr_1</measurementReaderId>
            <measurementSpecification struct="MeasurementSpecification">
                <measurementTypeRef>ManagedElement=1,SystemFunctions=1,Pm=1,PmGroup=OSProcessingLogicalUnit,MeasurementType=CPULoad.Total</measurementTypeRef>
            </measurementSpecification>
            <thresholdRateOfVariation>PER_SECOND</thresholdRateOfVariation>
        </MeasurementReader>
    </PmJob>
$

This utility is installed by default on many Linux distributions.