Empty node strings

[Marius8881]

Hi, Is it possible to select certain empty node values, "" , and change them to different types, ex, array ? I do not seem to see the option for this., tried it with tagValueProcessor , but it does not see empty "" .

[github-actions[bot]]

I'm glad you find this repository helpful. I'll try to address your issue ASAP. You can watch the repo for new changes or star it.

[amitguptagwl]

@Marius8881 Can you please give some example? I'm working on fxp v4. I can probably consider new requirements in that.

[Marius8881]

 Example:
    <x>0</x>
    <y>44</y>
    <width>2</width>
    <minWidth>1</minWidth>
    <maxWidth>12</maxWidth>
    <height>4</height>
    <minHeight>4</minHeight>
    <maxHeight>48</maxHeight>
    <overlays/>
    <frame>
      <static>true</static>
      <nodePath/>
      <color>transparent</color>
    </frame>
 

<overlays/> is coming as empty, and I can not see the name of the attribute, because of it, even if its set as <overlays> <overlays/> , same for arrays.

[amitguptagwl]

Do you expect overlays = ' ' in case of <overlays> <overlays/> ?

[Marius8881]

Yes, unfortunately we deal with binary data, and need to use xml, and have issues with the types changing, and we needed to change the type value of a node to what we need, that is why is favorable to see them all, even if empty.

[amitguptagwl]

Try trimValues: false. But it may add "#text": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", to the output.

[amitguptagwl]

closing as no update. reopen it if issue still exists

[fpedroza]

In my example, I'm parsing this:

<companies>
  <company></company>
</companies>

which parses into this:

companies: {
  company: ''
},

vs.

companies: {
  company: {}
},

Is there an option to control this?

[amitguptagwl]

@fpedroza Currently no. But you can create another issue for a feature request. If there are more users who are looking for the same or similar behavior then we can provide some common solution.

[bobspace]

+1 for this issue. I would like to treat an empty tag as a null value, and I cannot process it with the tagValueProcessor because it skips empty nodes.

[dobeerman]

in my case, if specific empty tag exists, I need to replace its value with boolean, but currently, it is impossible due to tagValueProcessor skips empty tags.

[amitguptagwl]

Let's assume 2 situation <a><![CDATA[text]]></a> and <a><tag>text</tag></a>. Should we call tagValueProcessor for <a>?

[amitguptagwl]

I've created a branch "emptyNode_v2" for this change. But not publishing them due to a few pending decisions like when not to call tagValueProcessor.

[dobeerman]

We have absolutely the same case when we need to replace empty string with empty object.

<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<PutDataRequest>
    <Commands>
      <PartialRetransmissions>
      </PartialRetransmissions>
      <RetransmissionsTracking>
          <Retransmission Begin="2021-10-29T20:17:10.786+00:00" End="2021-10-29T22:23:01.220+00:00" />
          <Retransmission Begin="2021-11-02T16:04:04.851+00:00" End="2021-11-02T18:23:18.939+00:00" />
      </RetransmissionsTracking>
    </Commands>
</PutDataRequest>

The returned Commands object as follow

{
    "PartialRetransmissions": "",
    "RetransmissionsTracking": {
      "Retransmission": [
        {
          "@_Begin": "2021-10-29T20:17:10.786+00:00",
          "@_End": "2021-10-29T22:23:01.220+00:00"
        },
        {
          "@_Begin": "2021-11-02T16:04:04.851+00:00",
          "@_End": "2021-11-02T18:23:18.939+00:00"
        }
      ]
    }
  }

We simply want to replace, for instance, "PartialRetransmissions": "", with "PartialRetransmissions": {},, as we expect to have an object here, by validating jsonPath and tagValue. I tried to use tagValueProcessor as I expected it will help us, but, unfortunatelly, the tagValueProcessor not calling by parser at all.

  tagValueProcessor: (tagName, tagValue, jPath, hasAttributes, isLeafNode) => {
    if (alwaysArray.includes(jsonPath) && !tagValue) return {};
    return tagValue;
  },

Am I do something wrong?

[htrap007]

I too have same issue where tagValueProcessor does not execute for empty tags, I want to change the structure of some empty tag instead of default empty string.

eg: <User />

`output { user: "", }

instead { user: {}, }`

can we execute tagValueProcessor for empty tags?

[dan852]

I'm having a similar requirement where I would rather return an empty {} or [] instead of the default "" returned by the library. Wouldn't it be a first simple step to either execute tagValueProcessor for empty strings? Maybe an additional setting to activate this would be good if there are performance implications.

Thanks

[amitguptagwl]

check if updateTag can help.

[m-radzikowski]

I also did not find a solution for this apart from manual post-processing.

I think @fpedroza example illustrates this best - with:

<companies>
  <company></company>
</companies>

there should be an option to parse it to:

{
  "companies": {
    "company": {}
  }
}

Otherwise, company is once an object (if it has any child tags), once a string (like in above XML).

What would work could be isObject() function, with behavior analogous to isArray():

const alwaysObject = ["companies.company"];
const parser = new XMLParser({
	isObject: (name, jPath) => {
		return alwaysObject.includes(jPath);
	},
});

[arshia-gh]

Any updates? I also need this feature. In my case, I need to convert empty nodes to null if they have a certain attributes (nil="true")

[amitguptagwl]

Please check v. It's experimental but should help. I've added basic documentation.