Opening choice for complex middlegame

Hi Friends, I would like to know which openings(for white, for black against e4,d4) can be used for a player of my level. I am asking this question on behalf of my real life friend(for the purpose of OTB rated tournaments).

Few requirements which he said to me are,

1. The middlegame should be as complex as possible.

2. Whether winning or losing, the game should end in middlegame itself.

3. My friend strongly feels that his strength lies in opposite side castling while his weakness is entering into an endgame.

4. Containing Bishop/knight sacrifice around the opponent's castled king along with our pawnstorm.

Any unsound openings are also welcome. Thank you

First off, why look for complex middlegames if you struggle with the endgame? It would stand to reason that since the endgame is the key to the middlegame then if you struggle with that then simple middle games must be best. Nonetheless I'll answer anyway:

I would forget about the bishop/knight sacrifice stuff as sacrifices only work in specific situations. My recommendation for white is definitely 1.d4 as you can play aggressively against all of black's options, whereas it is more difficult to do so with 1.e4 despite the common conception that 1.e4 is for aggressive players. Learn the Queen's Gambit for white and also responses to all of the Indian Defenses - you could look into the recommendations of GM Simon Williams who is well known for being a very aggressive player and who advocates 1.d4. As black, you can play the Modern Benoni  against 1.d4 (if white avoids it then the Kasparov Gambit is an option) and against 1.e4 obviously the Sicilian defense - look into different lines in the Sicilian and see which suits you best

Already the first requirement is a total BS. U should be a positional genius (at least, as good as Nimzovich) in order to survive in a complex middlegame vs a strong opponent. 

The Dutch, French and Caro-Kann always give me original and interesting positions.

The second requirement is utter bullsh*t and often unavoidable.

Sure, some openings are more complex than others.  For example, The Marshall Attack in the Ruy Lopez (7...O-O 8.c3 d5) and the Modern Benoni are far more complex than the Berlin Defense and the Orthodox Variation of the Queens Gambit Declined, but to avoid endgame is utter bullsh*t.

The objective in chess is to play the best move.  Part of that is understanding when to trade and what to trade.  Many positions see cases where trading down to an endgame wins while keeping pieces on the board draws at best and has a high chance of losing.  Other positions see cases where trading down to an endgame loses outright, but keeping the pieces on is close to winning.

To just write off trading down because you hate endgame shows a clear bias and a blatant case of showing that you do not understand the basics and the root goal in the game of chess.  Having a no endgame bias is just as bad as a must reach the endgame ASAP bias, or a bishop bias (over knights) or a knight bias (over bishops), or any other bias.

If endgames are your problem, study the endgame.  If you are under 1400, you have no business studying openings anyway.  If you are over 1400, you ought to know Lucena's Position, Philidor's Draw, the Short Side Defense, and the Passive Defense like the back of your hand in R and P versus R endgames, and when each can be used.

Lucena's Position works for all pawns except the Rook pawn.  Philidor's Draw works against all pawns, the Short Side Defense against the Bishop pawn, and the Passive Defense works against a Knight or Rook pawn.  If you are 1400 and do not kniw this like the back of your hand, first thing you need to do is study endgames.

You should also know how to win Q vs pawn on 6th rank.