Releases: clbanning/mxj
Allow validating XML produced by encoder
The toggle XmlCheckIsValid() was added to force checking that the encoded XML is valid. (Per Issue #88.)
Add patch digit
It turns out that go get ignores releases without the "patch" digit. For it to pull v2.4, we need a release v2.4.0. Ugh!!!
Add XMLEscapeCharsDecoder option
This preserves XML escaped characters in the Map values. Note: xml.Decoder nominally unescapes these characters when returning xml.CharData values.
Update for readme.md changes in pull request.
Merge pull request #79 from TomWright/master Update module package to v2 and update readme
Add go.mod file
Merge pull request #77 from TomWright/master Add go.mod to allow mxj to be imported using go modules
Add DisableTrimWhiteSpace option.
Add DisableTrimWhiteSpace option to not trim \x20 from xml.CharData values. (By default they are.)
Improved XML encoding performance.
Replaced string appends with bytes.Buffer. Performance now scales linearly even to very large XML docs with greater than 480,000 nodes.
Handle parsing errors for mv.Exists arguments.
Previously mv.Exists exited silently on errors encountered when parsing the 'path' argument. Now it reports such errors.
Add MapSeq type and remove probably useless XmlWriter methods.
Minor refactoring - add MapSeq type to reduce confusion and move associated functions and methods from Map to MapSeq.
Breaking change - remove XmlWriter methods that return 'raw' value that is written on the io.Writer. This seems silly and unnecessary. (For those who want it they are just commented out in the source code, so they can be easily restored.)(