Add XML conversion to bodies, and fix XML data conversion

[ySnoopyDogy]

The new XML validation functions just convert the data to JSON before validating the schema. With that, we can just convert the data on Request and Response validations just by looking at the content type.

Because in XML everything is a string, there can be problems to validating against a schema that includes other primitive types. Now, we convert the strings based on the schema types. If the property is a boolean, we convert to a boolean to pass the validation. If the value is not a valid boolean, we do nothing to let the schema validation to raise an error.

Also, now we are validating if the payload includes prefixes and namespaces defined in the schema (the XMLNS feature of XML).

Added a new option to ValidationOptions: AllowXMLBodyValidation to enable XML validations in request/response bodies (the default is false). Created a function WithXmlBodyValidation to enable the new validator option.

No breaking changes with the new code, as the updated function was internal, and the XML body validation setting is false by default.

[codecov]

Codecov Report

❌ Patch coverage is 99.25651% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 97.73%. Comparing base (26d43b7) to head (25009e6).

Files with missing lines	Patch %	Lines
schema_validation/validate_xml.go	98.65%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #214      +/-   ##
==========================================
+ Coverage   97.63%   97.73%   +0.10%     
==========================================
  Files          56       57       +1     
  Lines        5323     5527     +204     
==========================================
+ Hits         5197     5402     +205     
+ Misses        126      125       -1     

Flag	Coverage Δ
unittests	97.73% <99.25%> (+0.10%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[ySnoopyDogy]

Going to implement a smarter conversion, like the one used in Query Parameter validation, looking to the type defined in the schema. This idea can be used to convert the body to JSON with correct defined types, so we can validate application/xml, x-www-form-urlencoded and also multipart/form-data

[daveshanley]

A few missing lines of coverage, any chance of a bump?

[ySnoopyDogy]

The coverage lines from validate_xml I think are impossible lines to hit. I tried forcing errors in the xml2json package, but they seem just to not generate errors at all. We create a reader from a string, so we need to force an invalid string that would cause errors in the LIB.

Looking the code from the xj.Convert, the Decode and Encode functions seem to not error. The lib just ignores every error. The most likely error that I found would be an error in the xmlDec.Token from the Decode function, but if it errors, the xml2json lib just breaks out the loop and stop converting the data. I don't think we can force errors for that.

The other line is the validation for the marshal error, but the xml lib only returns valid json. If the lib receives wrong XML data, they just return an empty buffer, that the line 53 ensures to return early. For that reason, the line 59 is practically unreachable, as the xml2json will only generate valid JSON with string values, so we can't even force NaN conversions.

For that, I would say there is no way of covering that lines, the IFs are just safeguards for impossible errors. What could we do in this situation about that two uncovered lines?