How to get rawXmlString

[martijn10kb]

Heey,
I need to get an XML string of something I have build with this module.
I am at the point where I have an array of elements, and need to convert that to xml so I can then insert that into an already existing docx (no worries I know this module doesn't do that..) just need to xml string.

We currently have a docx document as template in which we need some rich text elements to be inserted.
Now we already have a HTML to "array with 'docx' elements" script. Just need a way to convert that array to a raw XML string so I can do a search and replace in my existing docx with that resulting XML string.

[dolanmiu]

Currently that is not exposed, but that is a good idea, perhaps it can be part of Packer

However, I am actually making a solution right now to allow for documents to be edited, by exactly the method you are describing. No ETA yet, but it will be in the next release.

[martijn10kb]

Yes, I will try to do that tomorrow for you in one of your sample docx.. ..I do hope I can reproduce it in that document..

[martijn10kb]

@dolanmiu Hmm can't really isolate it. No matter how I modify your document it still succeeds. And no matter how I edit my document it still fails...
This document is now completely empty but it still fails if I run it as input document in demo 85.
Could you maybe have a look it at? Maybe you've better luck/skills with docx's internals..

e738cf8d-3070-4387-b043-b34ffa676039 (2).docx

[martijn10kb]

@dolanmiu I just came to learn that it may have been fixed in 8.0.1 trough the PR from Ewout :)

[dolanmiu]

Great! good to know.

Let me know if there are any other issues

[martijn10kb]

@dolanmiu Alright this works really cool!
For the next step. It would be really great if we had some kind of function support. Like i can have in word the following..

{{ generateTableFromData('employeeCollection', (a) => a.isNotFired, ['first_name', 'last_name', 'age']  ) }}
{{ generateTableFromData('documentCollection', (a) => a.type === 'foo', ['path', 'title', 'description']  ) }}

And then in your code you have something like this:

const { data } = graphqlQuery()

await patchDocument(data, {
  patches: [
    ...,
    generateTableFromData: (collection: string, filter: Fn, columns: Array<string>) => {
      return new table(children: [
        new row(isHeader: true, children: column.map(i => t(i)) ),
        ...data[collection].filter(filter).map((item) =>{
          return new row(children: columns.map( ...generate column per row..)
        }),
      ])
    },
    ...,
  ]
});

Although this would be really great I think.. ..it still might not 100% work for our needs I think..
Maybe @ewoutkleinsmann has another idea?

[DharyaSoni]

Please Suggest, How can I show rich text as html tag formatted string in Documents.
Eg. "

Text to Show

".

[martijn10kb]

See the demo case 85.
Besides that you will need your own rich text parser to docx mapper.
I cant share the code we used 1 on 1, but we use himalaya to parse the rich text to json. Then we have a recursive function that generates all the elements like TextRun, ExternalHyperlink, Paragraph, etc.

[dolanmiu]

Interesting approach!

[DharyaSoni]

See the demo case 85. Besides that you will need your own rich text parser to docx mapper. I cant share the code we used 1 on 1, but we use himalaya to parse the rich text to json. Then we have a recursive function that generates all the elements like TextRun, ExternalHyperlink, Paragraph, etc.

Could you please share that recursive method that generate all elements like text run, and all

[martijn10kb]

import { parse } from 'himalaya';

export default html = (html: string, root = true) => {
  const elements = parse(html);
  return mapToDocx(elements, root);
};

const mapToDocx = (elements: Element[], styles = {}, root: boolean): any[] => {
  return elements.flatMap((element) => toDocx(element, root));
};

export const toDocx = (
  { type, tagName, attributes, content, children }: Element,
  extraStyles: object,
  root: boolean,
) => {

  ...here have your code..
  fi. if type === 'text' return TextRun({...extraStyles, children: content})
  if tagname is bold or b return mapToDocx(children, { bold: true }, root);
  if tagname img return imageRun
  etcetc..
  
  return [];
};