I will try to explain quickly, what's going on :
Your favourite IDE often provides you auto completion depending on the language you're writing your code. To do that, the plug-in reproduce a part of the compiler job ; it's a kind of parser, so it can suggest you what the compiler is waiting for in the next chars you'll write ( often after a "." or a "(" ).
Haxe needn't that. Better, the compiler is so fast, that it does it by itself !
Building your project with the special parameter "--display@", followed by the offset of the byte char in the class file you want the completion ( including CRLF ), haxe outputs you in the command line the suggestions as an XML tree. This means also, that you can easily write your own completion plug-in for, maybe, your own IDE.
Look at the example here, you'll understand why I love this wonderful compiler !