escaping single quote xmllint xpath query in Bash script

Question

I'm trying to Query information from an xmltv document using xmllint inside of a bash script and I'm having trouble with program titles that have single quotes in the title. For example:

XML Example:

<programme start="20150106090000 -0500" stop="20150106093000 -0500" channel="I53.28457646.microsoft.com">
        <title lang="en">Daniel Tiger's Neighborhood</title>
        <sub-title lang="en">Safety Patrol; Safety at the Beach</sub-title>
        <desc lang="en">Prince Tuesday visits the school in his crossing guard uniform and helps the children practice safety rules; Daniel and Katerina drift too far from Mom while playing at the beach.</desc>
        <credits>
            <actor>Jake Beale</actor>
            <actor>Ted Dykstra</actor>
            <actor>Heather Bambrick</actor>
            <actor>Amariah Faulkner</actor>
            <actor>Stuart Ralston</actor>
            <actor>Zachary Bloch</actor>
            <actor>Addison Holley</actor>
            <actor>Nicholas Kaegi</actor>
        </credits>
        <date>20130715</date>
        <category lang="en">Children</category>
        <category lang="en">Educational</category>
        <category lang="en">Episodic</category>
        <category lang="en">Kids</category>
        <category lang="en">Series</category>
        <episode-num system="onscreen">130</episode-num>
        <episode-num system="ms_progid">1.EP015507510029</episode-num>
        <episode-num system="dd_progid">EP01550751.0029</episode-num>
        <video>
            <aspect>16:9</aspect>
            <quality>HDTV</quality>
        </video>
        <audio>
            <stereo>stereo</stereo>
        </audio>
        <previously-shown start="20130715000000" />
        <subtitles type="teletext" />
        <rating system="VCHIP">
            <value>TV-Y</value>
        </rating>
    </programme>

xmllint line:

xmllint --xpath '/tv/programme[title='Daniel Tiger's Neighborhood']' xmltv.xml

I've tried several options to try an escape it including wrapping in double quotes and writing Tiger''s and Tiger's and nothing I try seems to help.

Any help is appreciated. Thanks!

Which information do you need from tag programme with title "Daniel Tiger's Neighborhood"? — Cyrus
– Cyrus, Commented Jan 13, 2015 at 16:48
Using "/tv/programme[title='Daniel Tiger's Neighborhood']" should work from the shell's perspective but that isn't going to help the internal problem. "/tv/programme[title=\"Daniel Tiger's Neighborhood\"]" might work so might "/tv/programme[title='Daniel Tiger\'s Neighborhood']" but it depends on xmllint and xml quoting. — Etan Reisner
– Etan Reisner, Commented Jan 13, 2015 at 17:07
@Cyrus the end game is to get the content of the start="" attribute in the previously-shown tag to get the original air date using show title, description, and start time as my matching criteria. That way I can name the recordings with the original air date for Plex Media Server to pick up and query TheTVDB.com info. — AndyRBlank
– AndyRBlank, Commented Jan 13, 2015 at 20:49
@EtanReisner This worked!: xmllint --xpath "/tv/programme[title=\"Daniel Tiger's Neighborhood\"]" xmltv.xml — AndyRBlank
– AndyRBlank, Commented Jan 13, 2015 at 20:54

Undo · Accepted Answer · 2015-01-13 21:20:27Z

2

Thanks to @EtanReisner This worked!:

xmllint --xpath "/tv/programme[title=\"Daniel Tiger's Neighborhood\"]" xmltv.xml

edited Jan 13, 2015 at 21:20

Undo♦

25.7k38 gold badges114 silver badges132 bronze badges

answered Jan 13, 2015 at 21:11

AndyRBlank

1061 silver badge8 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Marc Bredt · Accepted Answer · 2015-01-13 17:32:51Z

1

After trying a bit i found two ways. The solution is to escape the single quote by \' outside the xpath XPATHPRE''XPATHPOST and concatenate it, while holding attribute/text surrounding with ".

xmllint --xpath '/programme/title[text()="Daniel Tiger'\''s Neighborhood"]' xmltv.xml
xmllint --xpath '/programme/title[text()="'Daniel Tiger\'s Neighborhood'"]' xmltv.xml

answered Jan 13, 2015 at 17:32

Marc Bredt

9651 gold badge5 silver badges14 bronze badges

4 Comments

AndyRBlank Over a year ago

I just tried those lines with this result: andrew@andyrblank:$ xmllint --xpath '/programme/title[text()="Daniel Tiger'\''s Neighborhood"]' xmltv.xml XPath set is empty andrew@andyrblank:$ xmllint --xpath '/programme/title[text()="'Daniel Tiger\'s Neighborhood'"]' xmltv.xml warning: failed to load external entity "Tiger's" warning: failed to load external entity "Neighborhood"]" XPath error : Unfinished literal xmlXPathEval: evaluation failed XPath evaluation failure

Etan Reisner Over a year ago

That second suggested command is not correct. It would need to be ..."'"Daniel Tiger's Neighboorhood"'"... I believe.

Marc Bredt Over a year ago

for my minimal example both solutions worked fine. so here is what i got: minimal XML <foo> <bar title="ke'ks">pu'ps</bar> </foo> and the minimal script

#!/bin/bash xmllint --xpath '/foo/bar[text()="pu'\''ps" and @title="ke'\''ks"]' foo.xml xmllint --xpath '/foo/bar[text()="'pu\'ps'" and @title="'ke\'ks'"]' foo.xml

. Both commands end up in displaying <bar title="ke'ks">pu'ps</bar>. @Etan Reisner your solution works for me as well :)

Etan Reisner Over a year ago

Lack of spaces in your sample causes the problem there. You don't need to quote to keep the word together for your case but you do for @AndyRBlank's case.

Collectives™ on Stack Overflow

escaping single quote xmllint xpath query in Bash script

2 Answers 2

Comments

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related