possible bug when using --xml option

[goekce]

I noticed that if --xml is used after the XPath expression, then it is not applied. Is this a feature or a bug?

$ wget https://raw.githubusercontent.com/DiseaseOntology/HumanDiseaseOntology/main/src/ontology/doid.owl
$ xidel -se "//rdfs:label[text()='malignant hyperthermia']/../rdfs:subClassOf" doid.owl --xml
<?xml version="1.0" encoding="UTF-8"?>
<xml>



            
                
                
            
        
</xml>

$ xidel --xml -se "//rdfs:label[text()='malignant hyperthermia']/../rdfs:subClassOf" doid.owl
<?xml version="1.0" encoding="UTF-8"?>
<xml>
<rdfs:subClassOf xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.obolibrary.org/obo/doid.owl#" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:so="http://purl.obolibrary.org/obo/so#" xmlns:obo="http://purl.obolibrary.org/obo/" xmlns:owl="http://www.w3.org/2002/07/owl#" xmlns:xsd="http://www.w3.org/2001/XMLSchema#" xmlns:doid="http://purl.obolibrary.org/obo/doid#" xmlns:skos="http://www.w3.org/2004/02/skos/core#" xmlns:terms="http://purl.org/dc/terms/" xmlns:oboInOwl="http://www.geneontology.org/formats/oboInOwl#" rdf:resource="http://purl.obolibrary.org/obo/DOID_0050736"/>
<rdfs:subClassOf xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.obolibrary.org/obo/doid.owl#" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:so="http://purl.obolibrary.org/obo/so#" xmlns:obo="http://purl.obolibrary.org/obo/" xmlns:owl="http://www.w3.org/2002/07/owl#" xmlns:xsd="http://www.w3.org/2001/XMLSchema#" xmlns:doid="http://purl.obolibrary.org/obo/doid#" xmlns:skos="http://www.w3.org/2004/02/skos/core#" xmlns:terms="http://purl.org/dc/terms/" xmlns:oboInOwl="http://www.geneontology.org/formats/oboInOwl#" rdf:resource="http://purl.obolibrary.org/obo/DOID_66"/>
<rdfs:subClassOf xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns="http://purl.obolibrary.org/obo/doid.owl#" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:so="http://purl.obolibrary.org/obo/so#" xmlns:obo="http://purl.obolibrary.org/obo/" xmlns:owl="http://www.w3.org/2002/07/owl#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:xsd="http://www.w3.org/2001/XMLSchema#" xmlns:doid="http://purl.obolibrary.org/obo/doid#" xmlns:skos="http://www.w3.org/2004/02/skos/core#" xmlns:terms="http://purl.org/dc/terms/" xmlns:oboInOwl="http://www.geneontology.org/formats/oboInOwl#">
            <owl:Restriction>
                <owl:onProperty rdf:resource="http://purl.obolibrary.org/obo/IDO_0000664"/>
                <owl:someValuesFrom rdf:resource="http://purl.obolibrary.org/obo/GENO_0000147"/>
            </owl:Restriction>
        </rdfs:subClassOf>
</xml>

[Reino17]

$ wget https://[...]/doid.owl

If you're using xidel, then there's no reason to use wget:

xidel -s "https://[...]/doid.owl" -e '$raw' > doid.owl
#or
xidel -s "https://[...]/doid.owl" --download=.

$ xidel -se "//rdfs:label[text()='malignant hyperthermia']/../rdfs:subClassOf" doid.owl --xml

• This use of quotes is for Windows. Since you're on Unix please use -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf'.
• Although not mentioned in xidel --help, 'readme.txt', or anywhere else, it's adviced to put the extraction-query behind the input, especially when you specify multiple inputs (xidel -s <file> -e '[...]' <url> -e '[...]'). And I also specify global options, like -s, always first. So I would enter:

xidel -s --xml doid.owl -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf'

Benito has to determine whether this is a bug or not.
For what it's worth, with --input-format=xml --output-format=xml at the end...

xidel -s doid.owl -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf' --input-format=xml --output-format=xml

...xidel does return the node's content.

[benibela]

Generally options should be on the left side of the thing they modify.

That way you can have multiple queries with different options:

xidel --printed-node-format xml -se "//rdfs:label[text()='malignant hyperthermia']/../rdfs:subClassOf" --printed-node-format text -e "//rdfs:label[text()='malignant hyperthermia']/../@rdf:about"  doid.owl

Or multiple inputs:

xidel --input-format xml doid.owl --input-format html http://example.org -e '$url'

Some options are rearranged if they are in the wrong order. Looks like --xml is missing there

For what it's worth, with --input-format=xml --output-format=xml at the end xidel does return the node's content.

Now that is weird, since --xml should be the same as those two

[goekce]

For what it's worth, with --input-format=xml --output-format=xml at the end...

xidel -s doid.owl -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf' --input-format=xml --output-format=xml

...xidel does return the node's content.

I assume node's content means all the XML nodes and not the text part. I can confirm this behavior.

Although not mentioned in xidel --help, 'readme.txt', or anywhere else, it's adviced to put the extraction-query behind the input, especially when you specify multiple inputs (xidel -s <file> -e '[...]' <url> -e '[...]'). And I also specify global options, like -s, always first. So I would enter:

xidel -s --xml doid.owl -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf'

Thank you, my actual problem is solved then. I must use --xml in front of the corresponding expression.

I leave the issue open for the For what it's worth issue ;)

---

The rest is bit OT, but I am grateful for @Reino17 's elaborate feedback, so:

$ wget https://[...]/doid.owl

If you're using xidel, then there's no reason to use wget:

xidel -s "https://[...]/doid.owl" -e '$raw' > doid.owl
#or
xidel -s "https://[...]/doid.owl" --download=.

doid.owl or other ontologies can get about 100 MB, and xidel does not cache. That is the reason why I use wget. But I did not know about --download=, thanks!

$ xidel -se "//rdfs:label[text()='malignant hyperthermia']/../rdfs:subClassOf" doid.owl --xml

This use of quotes is for Windows. Since you're on Unix please use -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf'.

I use xidel extensively in scripts. Using your style I cannot write something like:

ARG='malignant hyperthermia'
xidel -s doid.owl -e '//rdfs:label[text()=$ARG]/..'

That is the reason why I use this style.

[Reino17]

With Windows quoting style that won't work either:

$ xidel -s doid.owl -e "//rdfs:label[text()=$ARG]/../@rdf:about"
Error:
err:XPST0003: Unknown or unexpected operator: "hyperthermia" (possibly missing comma "," or closing parentheses ")}]" )
in line 1 column 21
//rdfs:label[text()=malignant hyperthermia]/../@rdf:about
                    ^^^^^^^^^^^  error occurs around here

You'd always have to quote the external variable:

-e "//rdfs:label[text()='$ARG']/../@rdf:about"   # Windows quoting style
#or
-e '//rdfs:label[text()="'$ARG'"]/../@rdf:about'   # Unix quoting style

[benibela]

I use xidel extensively in scripts. Using your style I cannot write something like:

ARG='malignant hyperthermia'
xidel -s doid.owl -e '//rdfs:label[text()=$ARG]/..'

There is a special option for that:

export ARG='malignant hyperthermia'
xidel --variable ARG -s doid.owl -e '//rdfs:label[text()=$ARG]/..'

[benibela]

For what it's worth, with --input-format=xml --output-format=xml at the end xidel does return the node's content.
xidel -s doid.owl -e '//rdfs:label[text()="malignant hyperthermia"]/../rdfs:subClassOf' --input-format=xml --output-format=xml

Now that is weird, since --xml should be the same as those two

You had me confused there. --xml is indeed always the same as those two. You got the right output, because you put doid.owl at the beginning