Davide Angelocola

@andimarek please have a look #2216

@andimarek of course! It looks very interesting... :-) Would be possible to publish this branch in maven central?

@andimarek there is a breaking change in a `@PublicApi` class: ``` symbol: method getObjectType() location: variable field of type graphql.schema.SelectedField ``` Not sure if this replacement is good enough for...

another breaking change is: ``` symbol: method getFieldDefinition() location: variable field of type graphql.schema.SelectedField ```

@andimarek using yourkit with cpu tracing profilter. I cannot see anymore the "collectField" hotspot. This is really nice result! 🥇 However, I can't push this version up to the integration...

probably should be designed together with #264

and with #273

TSV and CSV could be covered by http://commons.apache.org/proper/commons-csv/ (a 44KB library with no additional dependencies). JUnit5 is using https://github.com/uniVocity/univocity-parsers but jar file is quite big (400KB).

JSON could be covered by: - [GSON](https://github.com/google/gson/) 240KB, no additional dependencies - [JSR 374](https://javaee.github.io/jsonp/) implementation is also needed (glassfish provides one) - [joy](https://github.com/leadpony/joy) another implementation for JSR 374 Please note...

XML could be covered without any additional dependency. Sample: - `xpath file.xml expression | ...`, to extract data from any document