Strong attacking possibilities is the hard part. White can almost always kill the game if he so wishes.
I would suggest either the Modern Ben-Oni or the Fajarowicz Budapest. A good introductory text to the Modern Ben-Oni is Palliser's, while for the Faja. you'll have to play through games in Moskalenko's or just pick some games you fancy in a database.
The QGD:
It is good and classical, as per the thread title. Your other points:
1. The objective is to blunt white's initiative by simply holding your ground in the center and possibly exchanging a piece or two. After development is completed both sides should be equal.
2. Well, it's not very attacking, I'll grant you that -- the goal is to reach equality first. That's not quite the same as an endgame though, and knowing the QGD you can later branch out to other queen's gambit lines like the Semi-Slav, which combines some ideas of the QGD with more risk and therefore more attacking possibilities.
3. No early exchange of queens, check.
4. The opposite of hypermodern, check.
5. Sadler's "Queen's Gambit Declined" is one of the absolute classics of opening books (so are the other two in the trilogy, The Slav and The Semi-Slav).
I have one friend who shot from ~1900 to ~2250 after the age of 20, and his defence to 1.d4 has always been the QGD.
That said, if you're < 1500, your first opening book should be "Fundamental Chess Openings" by van der Sterren, followed by trying whatever you feel like on that particular day -- your openings really don't have as much influence on your games' results.
at this level the budapest gambit will kill 20%, beat 50% and will be equal the other 30%
Two suggestions.
First: the QGA. There are good books on it. It is solid if white stays in the mainline, but if he doesn't, black can develop a devastating queenside pawn rush. If white deviates from book, your play generally becomes thematic: protect the pawn and move forward on the queenside. The only thorn here is the number of non-queen's gambit alternatives white can hurl at you. But those are generally handled by playing Nf6 and c5 as soon as seems reasonable. Often with a quick Qb6 to take advantage of some kind of early bishop sortie. Rizzitano's book offers both a QGA repertoire and some stuff on all white's offbeat tries. Look up the reviews on Silman's site.
Second, the Stonewall Dutch. Although it doesn't have a great attacker's reputation, there are very few lines better designed to launch an early attack against the castled white king if white doesn't play correctly. If white doesn't play the mainlines, you still stick your pawns on light colored squares, jump your knight to e4, castle, move your queen over to the kingside, and look for ways to push the f-pawn aggressively with that rook behind it. The real bonus here is that you can avoid all of white's Bg5 shenanigans because you play the French. Just play 1.d4 e6, and play 2...f5 against almost anything but 2.e4. It's got a couple good books. Win With the Stonewall and Aagaard's book are both top-flight.
If not the QGA, I'd recommend the Tarrasch QGD. The Dutch is generally good, and it forces you to play actively, or get crushed.
Rizzitano's book is good. Aagard's book Meeting 1d4 is pretty good for the Tarrasch.
But even if you pick a Nimzo/Bogo-Indian repertoire, if you play actively, you'll find good results against lazy d4 players who think they're guaranteed at least a draw and that their choice on move 1 is somehow tha last word on the course of the game.
I'd also suggest the Budapest. But the Fajarowicz variation is not for the faint hearted (sorry, hicetnunc), the regular lines with 3 Ng4 is good, solid and easy to learn.
And you get a lot of quick wins too :)
Try Slav Chebanenko variation.
I'd also suggest the Budapest. But the Fajarowicz variation is not for the faint hearted (sorry, hicetnunc), the regular lines with 3 Ng4 is good, solid and easy to learn.
And you get a lot of quick wins too :)
The OP wants "strong attacking chances", hence my suggestion of Faja. over regular Budapest 
Good enough, but to get to the endgame, you have to get past the opening and middlegame.
If you learn how to complicate things properly, you can pretty much skip a lot of endgame study. 
I would have suggested the semi-slav, but at 1200, or even 1500 level it is probably too difficult to handle. Just forget about having to attack and play the QGD, that way you can play 1...e6 on the first move since your other opening is the french. Also, spend some time on the endgame. As you get stronger you will find that fewer of your opponents fall prey to an attack in the middle game, and it will be necessary to convert your advantage in the endgame.
Endings are inevitable, so LEARN HOW TO PLAY THEM. I am probably less than 2000 strength and I swear that 80-90% of my games against equal opposition reach an endgame, even in blitz.
The Grunfeld is exactly what you have in mind.
Imo KID fits very well
Wow, you've recommended just about every opening under the sun by now! I think you're missing the Slav though. That is also totally what you're looking for, if you're wondering.
In all seriousness, I'd definitely stick with 1...d5 for now (ok, I guess you can play 1...e6 and 2...d5, hoping to trick your foe into a French). No need for you to avoid the Exchange variation; in the resulting pawn formation, black's plan is actually a kingside attack.
Anyway, there might be - there are definitely problems with your play that opening choices won't fix. Rather than buy an opening book, go read Wikipedia on your chosen opening, then practice it against people of your own skill, and maybe buy a book on tactics or strategy. (My System taught me some things, and I still only understood maybe a tenth of what the book is trying to tell me.)
I wonder why people would want to skip endgame study. It's always been the most fun part of the game for me.
That is the most common advice for French players, but rarely you will have a French. Most probably you will have to play the Semi-Slav. This is sharp as you want it but is difficult to play (as Black)
If you go the Slav way, you may get an Exchange variation, which is equivalent with Queens Exchange (drawish).
Take a look at QGD-Albin Counter Gambit (1.d4 d5 2.c4 e5). This is my alternative to my QGA.
The good thing about Albin and KID is the high probability to be chosen by both players so your study will be limited to the same thematic game.
Dutch-Leningrad is equivalent to KID in sharpness. But I like KID much better. KID is more strategical.
Because it is difficult. Try to study Bishop vs Knight for example. With time usually limited in the endgame stage, it is very hard to play the defending side. I just try to study those with high likelyhood to happen or the one that can be used to build up my middlegame plan.
That is the most common advice for French players, but rarely you will have a French. Most probably you will have to play the Semi-Slav.
Simply none of this is true.
There are large amateur databases available freely. Check them out. At the sub-master level, 1.d4 e6 transposes into a French by move two about a third of the time, which is significantly higher than the number of Semi-Slavs you'll see.
You're statistically a lot more likely, at the club level, to face a Colle, Colle-Zukertort, or a London than you are a Semi-Slav. (All of which, incidentally, offer white real chance at maintaining an edge after you've already committed to an early ...e6.)
Depending on the move order you choose after 2.c4, you're about equally likely to face any of the mishmash of mainline QGD's, QGD exchanges, Marshalls, Catalan attempts, etc.
Only very, very rarely will you get to play a Semi-Slav, even if that's your heart's desire. Backdoor routes to questionable Veresovs and Torres occur about as frequently.
There is a strange habit on this site of looking at the numbers on grandmaster databases and thinking for some reason that they reflect the club-player's OTB reality. They do not.
This is why I advocate the Stonewall or the QGA. The Stonewall remains perfectly viable against the white attacks a club player is actually going to face. The QGA doesn't allow white to choose from a sea of choices after he plays 2.c4, but more importantly, it leaves the door open to all the critical lines against the d-pawn specials.
You're probably right. I didn't speak based on statistics, nor GM database, just common sense. It is very unlikely to my opinion that White play 1.d4 then switch to playing against the French (It is actually me who hate the French 
)
I, on the other hand, rely heavily on GM database to build my repertoire. I want a theoretically strong repertoire. There are many benefits from this approach. First, it will never obsolete (as I climb the ladder) so I don't waste my time studying weak sidelines. Second, I have the engine and the database so basically I can find the right answer for every diversion, and know how to punish those who are not prepared.
That is the most common advice for French players, but rarely you will have a French. Most probably you will have to play the Semi-Slav.
Simply none of this is true.
There are large amateur databases available freely. Check them out. At the sub-master level, 1.d4 e6 transposes into a French by move two about a third of the time, which is significantly higher than the number of Semi-Slavs you'll see.
I'm very suprised by this stat. If only because at amateur level, non-e4 players, usually avoid transpositions into 1.e4 systems like the plague. 
I had a look at my database, by filtering games where both players were U2200, and I found ~700 transpositions out of 6200 games, so roughly 11%, no more.
In some databases, there may be a confusion between moves actually played from the current position on the board and # of games with the resulting position in the database.
I'm newly back into chess after a 30+ year absence and would like suggestions for playing black agtainst 1.d4 which meets the following criteria:
1. It must have a clearly defined, specific objective so that if my opponent goes out of book I can assess the strength or weakness of his deviation and how I should react in accorance with my objective.
2. It must lead to a position with strong attacking possibilities. My end game skills are non-existant so nothing that leads to good end game possibilities. I will develop end game skills in time.
3. No early exchange of queens.
4. Nothing hypermodern or crazy like 1...b6, Vulture, etc.
5. Something with a GOOD book that you can recommend. I prefer books that give an overview and explain plans and goals rather tham reams of alabraic notation only.
1.d4 has always been a thorn in my side. Right now I play the Colle with white and the French Defense against 1.e4. Thirty-some years ago my rating was about 1500. Today I'm probably aound 1200.
Any help will be greatly appreciated.
Thank you