Parsing and reencoding document should be identical about dart-xml HOT 3 CLOSED

Thank you for the GIST allowing me to quickly reproduce the observed issue.

In your example, re-encoding creates a different output because ">" is valid XML text and doesn't normally need to be encoded.

Same problem exists in various other places where the XML nodes are unable to keep the exact source of the parsed input, e.g. the delimiter types of attributes, whitespace within elements, and whitespace at document level.

While it is technically possible to solve the reported and the various related problems, I am not sure if it is worth the trouble? Most XML libraries I know don't do it.

Would it help to be able to specify a custom encoder function?

from dart-xml.

Yes, allowing me to specify a custom encoding/decoding function will possibly solve this specific issue, however I am not entirely certain we are on the right track.

Question: is it a valid requirement that I should be able to retrieve contents of an XmlElement as XML without any modification if

1. the XML is valid
2. the textual content may contain such character sequences that are decodable

I saw in a previous issue that you brought up PHP's implementation as reference. I created yet another gist (my PHP skills are terrible).

If the answer to the question is yes, the default behavior of this library should reflect it. Until such time arrives, I would be happy with the custom function you mentioned.

My use-case

The following wall of text is only loosely related to the issue at hand; I wanted to offer you some context as to how I'd like to use this library.
I am using it to parse an XML based format, called XLIFF.
A small snippet from an XLIFF file:

<trans-unit id="some_trans_unit_id" datatype="xml">
    <source><bpt ctype="x-g" equiv-text="[" id="_0">&lt;g id=&quot;_0&quot;&gt;</bpt>3. helyezett: </source>
</trans-unit>

From my application's point of view, it is crucial, that the contents of source can be retrieved as-is, without any kind of encoding/decoding.
If I retrieve the text contents of the source sourceElement.text I get <g id="_0">3. helyezett:. It was devised that such fragments be wrapped in bpt tags and encoded to produce valid XML.

from dart-xml.

With c2fb10d single and double quotes in attributes are preserved.

With b2482cd whitespaces, processing instructions and comments at the document level are preserved.

from dart-xml.