Implement removeWhitespaceElements on XMLDecoder
#222
Conversation
wooj2
changed the title
wooj2
changed the title
Co-authored-by: Max Desiatov <max@desiatov.com>
Co-authored-by: Max Desiatov <max@desiatov.com>
Co-authored-by: Max Desiatov <max@desiatov.com>
Co-authored-by: Max Desiatov <max@desiatov.com>
Co-authored-by: Max Desiatov <max@desiatov.com>
Co-authored-by: Max Desiatov <max@desiatov.com>
Codecov Report
@@ Coverage Diff @@
## main #222 +/- ##
==========================================
+ Coverage 73.50% 73.87% +0.37%
==========================================
Files 46 46
Lines 2404 2431 +27
==========================================
+ Hits 1767 1796 +29
+ Misses 637 635 -2
Continue to review full report at Codecov.
|
|
@wooj2 one last request, I promise 🙂 Can you add a subsection to the "Advanced features" section in Just an example of your use case and a short description of the feature, similar to other subsections that describe advanced features. Thank you! |
MaxDesiatov
changed the title
trimValueWhitespaces on XMLDecoder
wooj2
changed the title
trimValueWhitespaces on XMLDecoderremoveWhitespaceElements on XMLDecoder
No prob at all! Just updated it. Thanks again for all the support |
|
@kneekey23 @wooj2 perfect, thanks for uncovering this edge case! Your contribution is very appreciated 👏 |
This updates the `XMLStackParser` to accept a parameter called `removeWhitespaceElements`.
The purpose of the `XMLStackParser` is to call the XML parser and create a tree of `XMLCoderElement` representing the structure of the parsed XML.
Assuming that XMLStackParser has `trimValueWhitespaces` set to `false`, when attempting to parse a nested data structure like the following:
```xml
<SomeType>
<nestedStringList>
<member>
<member>foo</member>
<member>bar</member>
</member>
<member>
<member>baz</member>
<member>qux</member>
</member>
</nestedStringList>
</SomeType>
```
... then there will multiple `XMLCoderElement`s in the tree which will have `elements` set to elements that are either:
a. Purely whitespaced elements or
b. The child elements
These purely whitespaced elements are problematic for users who are implementing custom `Decoder` logic, as they are interpreted as regular child elements. Therefore, setting `removeWhitespaceElements` to `true` while `trimValueWhitespaces` is set to `false`, will remove these whitespace elements during the construction of the `XMLCoderElement` tree.
An in-depth analysis of the original problem can be found [here](CoreOffice#219).
For historical purposes, a separate approach was implemented. It uses a similar algorithm in a different part of the code. CoreOffice#221
This updates the
XMLStackParserto accept a parameter calledremoveWhitespaceElements.The purpose of the
XMLStackParseris to call the XML parser and create a tree ofXMLCoderElementrepresenting the structure of the parsed XML.Assuming that XMLStackParser has
trimValueWhitespacesset tofalse, when attempting to parse a nested data structure like the following:... then there will multiple
XMLCoderElements in the tree which will haveelementsset to elements that are either:a. Purely whitespaced elements or
b. The child elements
These purely whitespaced elements are problematic for users who are implementing custom
Decoderlogic, as they are interpreted as regular child elements. Therefore, settingremoveWhitespaceElementstotruewhiletrimValueWhitespacesis set tofalse, will remove these whitespace elements during the construction of theXMLCoderElementtree.An in-depth analysis of the original problem can be found here.
For historical purposes, a separate approach was implemented. It uses a similar algorithm in a different part of the code. #221