Parsing XML and replacing specific elements or attribute values via shell script

Question

For the below xml ,I need to replace <studentStatus> for <studentName>CLASSA</studentName> to <studentStatus>failed</studentStatus>.

<studentFile>
    <student>
        <studentName>CLASSA</studentName>
        <studentStatus>Success</studentStatus>
        <studentActions>
            <studentAction>
                <studentType>Juniour</studentType>
                <studentStatus>Completed</studentStatus>
                <studentMsg/>
            </studentAction>
            <studentAction>
                <studentType>HighSchool</studentType>
                <studentStatus>Completed</studentStatus>
                <studentMsg/>
            </studentAction>
        </studentActions>
    </student>
    <student>
        <studentName>CLASSB</studentName>
        <studentStatus>Success</studentStatus>
        <studentActions>
            <studentAction>
                <studentType>Senior</studentType>
                <studentStatus>Completed</studentStatus>
            </studentAction>
            <studentAction>
                <studentType>Middle</studentType>
                <studentStatus>Completed</studentStatus>
            </studentAction>                         
        </studentActions>
    </student>
</studentFile>

What I got so far,

xmllint -xpath "/studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType="Juniour"]/studentStatus" myxml.xml

now i got the status of the student as Completed , now this value should be changed to Failed . Only for <studentType>Juniour</studentType>. How should I edit the xml inorder to get it as ,

<studentFile>
    <student>
        <studentName>CLASSA</studentName>
        <studentStatus>Success</studentStatus>
        <studentActions>
            <studentAction>
                <studentType>Juniour</studentType>
                <studentStatus>Failed</studentStatus>
                <studentMsg/>
            </studentAction>
            <studentAction>
                <studentType>HighSchool</studentType>
                <studentStatus>Completed</studentStatus>
                <studentMsg/>
            </studentAction>
        </studentActions>
    </student>
    <student>
        <studentName>CLASSB</studentName>
        <studentStatus>Success</studentStatus>
        <studentActions>
            <studentAction>
                <studentType>Senior</studentType>
                <studentStatus>Completed</studentStatus>
            </studentAction>
            <studentAction>
                <studentType>Middle</studentType>
                <studentStatus>Completed</studentStatus>
            </studentAction>                         
        </studentActions>
    </student>
</studentFile>

Can this be done using sed. I know there are tools like xsltproc but not sure if this is installed in all nodes in our cluster .

Any help will be appreciated. Thanks in advance!

"*Can this be done using sed.*" See: http://stackoverflow.com/a/1732454/3016153 (applies to XML just as well). — michael.hor257k, May 14 '17 at 18:29
Pretty sure python comes with a xml module on most systems, could try that. — 123, May 14 '17 at 18:39
@Cyrus edited now . I checked for xmlstarlet. `xmlstarlet` `-bash: xmlstarlet: command not found` — Neethu Lalitha, May 14 '17 at 18:43
@Neethu, and you have no privileges to install anything? xmlstarlet would be a good tool for further xml processing. — RomanPerekhrest, May 14 '17 at 19:27
@RomanPerekhrest yeah from the below answers , using xmlstarlet is the easiest way. Let me check with the admin teams . Do you know any sed way to updating the xml? — Neethu Lalitha, May 14 '17 at 19:31
@Neethu, xmlstarlet would be a good tool for further xml processing. And if xmlstarlet installation will be refused, we may look forward Python solution (you'll need to tell your Python version in that case) — RomanPerekhrest, May 14 '17 at 19:34

Cyrus · Accepted Answer · 2017-05-15T19:18:57.963

22

Update value with xmllint in file.xml:

xmllint --shell file.xml << EOF
cd /studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus
set failed
save
EOF

or without here document:

echo -e "cd /studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus\nset failed\nsave" | xmllint --shell file.xml

Update: With bash and XML in a variable:

xml=$(xmllint --shell <(echo "$xml") << EOF
cd /studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus
set failed
save -
EOF
)

or without here document:

xml=$(echo -e "cd /studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus\nset failed\nsave -" | xmllint --shell <(echo "$xml"))

edited May 15 '17 at 19:18

answered May 14 '17 at 19:47

Cyrus

84,225
14
89
153

if my xml is in a variable , how can this command be modified? – Neethu Lalitha May 15 '17 at 03:43
Hi @Cyrus , I tried adding xml=$(xmllint --shell <(echo "$xml") << EOF cd /studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus set failed save - EOF ) to my script but it says: unexpected EOF while looking for matching `)' , should i enclose them in quotes ? – Neethu Lalitha May 15 '17 at 18:56
1

Hi Cyrus, For an xml tag that looks like < studentStatus/> , this command is not working . xml=$(echo -e "cd /studentFile/student[studentName='CLASSA']/studentActions/st‌udentAction[studentT‌ype='Juniour']/stude‌ntStatus\nset failed\nsave -" | xmllint --shell <(echo "$xml")) ,it says `/studentFile/student[studentName='CLASSA']/studentActions/st‌udentAction[studentT‌ype='Juniour']/stude‌ntStatus is an empty Node Set`, Can this command be modified in any way? Any help is appreciated! – Neethu Lalitha May 17 '17 at 03:31
Unfortunately, I can not reproduce this. – Cyrus May 17 '17 at 05:24
1

Just found a workaround for this, i stored the new xml node schema in a variable , update it using xmllint and then replaced it via sed. Thanks Cyrus for the help – Neethu Lalitha May 19 '17 at 12:37

score 3 · Answer 2 · answered May 14 '17 at 19:20

xlmlint, as the name implies, is for parsing and validating XML, not editing it. If you can install xmlstarlet on your cluster, you can do the following:

xmlstarlet ed --update "/studentFile/student[studentName='CLASSA']/studentActions/studentAction[studentType='Juniour']/studentStatus" --value "Failed" *file*

RomanPerekhrest · Answer 3 · 2017-05-14T19:26:51.147

2

In case if xmlstarlet (a command line toolkit to query/edit/check/transform XML documents) is accessible:

xmlstarlet ed -u "//studentAction/studentStatus[preceding-sibling::studentType[1][text() = 'Juniour'] \
           and ancestor::student/studentName[text() = 'CLASSA']]" -v failed students.xml

The above will output the initial XML document with needed replacement

The command details:

ed -u - edit/update mode

//studentAction/studentStatus - xpath expression to select studentStatus element which has:

preceding-sibling::studentType[1][text() = 'Juniour'] - preceding sibling element studentType with value Juniour
ancestor::student/studentName[text() = 'CLASSA'] - nearest element studentName with value CLASSA

edited May 14 '17 at 19:26

answered May 14 '17 at 19:20

RomanPerekhrest

88,541
4
65
105

Not sure the point of this answer, they say in the comments it isn't... – 123 May 14 '17 at 19:25
1

@123, that's pretty sad. Missed that comment while preparing the anser – RomanPerekhrest May 14 '17 at 19:28
The comment they just added makes it seem like it probably will be a useful answer, so not too sad – 123 May 14 '17 at 19:33
@123, the sad fact is that someone's unable to install such useful tools – RomanPerekhrest May 14 '17 at 19:38

score 0 · Answer 4 · answered Aug 23 '21 at 20:05

0

You can try my Xembly command-line tool:

$ xembly --xml file.xml 'XPATH "/studentFile/student[studentName=\'CLASSA\']/studentActions/studentAction[studentType=\'Juniour\']/studentStatus"; SET "failed";'

The full syntax of Xembly is here.

answered Aug 23 '21 at 20:05

yegor256

102,010
123
446
597

score 0 · Answer 5 · answered Sep 15 '22 at 16:12

Update your xml file through xmllint with --shell.

$ xmllint --shell pom.xml
$ cd //*[local-name()='project']/*[local-name()='parent']/*[local-name()='version']
$ set 1.0.0
$ save
$ quit

You can combile bash with:

xml=$(xmllint --shell pom.xml <<EOF
cd //*[local-name()='project']/*[local-name()='parent']/*[local-name()='version']
set 1.0.0
save -
EOF
)

Parsing XML and replacing specific elements or attribute values via shell script

5 Answers5

Linked