Unlike some languages like JavaScript or Lua, which have a clear 'function' declaration, C and Java functions are not obvious to find.
Regex is an approach, used by some syntax highlighter.
Scintilla lexers use a parsing automaton (state machine), going char by char, keeping track of the context (in a comment, in a string...), which can avoid some false positives.
Processing uses a more robust method: it do a real parsing of the code, using Antlr, to perform its pre-processing (like converting #AAF00D to an int). It uses a grammar for Java provided by the Antlr site.
The learning curve might be step (I haven't dug this field myself) but the result is probably reliable.