Nada bahlas
Dilemma
You have twenty dollars. You meet a homeless person who tells you that, for that money, he could afford food for the next week. However, giving him the 20 dollars will mean that you cannot get to a meeting with a close friend you have not seen for several years, and who needs to speak to you about a personal matter.
Should you give the person the money, or keep it for yourself? Is it more important to help other people who are in need, or to live one's life as fully as possible? No-one will blame you if you do either option, but in either case harm will befall the person you do not expend the twenty dollars upon.
Solution
As with most real-word ethical problems, there is no ‘right' solution to this dilemma, only justifications for taking either path. This is because life is a zero-sum game - that is, in order for someone to gain something, someone else must lose something. There are many ways to perceive the problem.
There is, of course, a third option, in which one gives enough to the homeless person that their immediate needs are met, and still leave enough money to get to the friend, or at least a significant part of the distance, relying on the fact that you have a support network that will provide the additional resources that you require. This is a lateral solution, where you rely on facts not stated within the dilemma itself but reasonable to assume.
Exploration
This dilemma is one of the basic questions of philosophy, and has been explored from every angle throughout history. Every society comes to its own broad answer, and every individual must also come to theirs. There is a natural human tendency towards altruism - it is the existence of social behaviour that led to our success as a species, after all, and gave us language and communication.
As with many functional ethical dilemmas, it does not fare well when considered in the world. In the idealised realms of philosophy, arguments are made based on theory alone. In the world, it is not even clear on what scale one's help would be effective. For example, helping the friend may lead to them helping many other people; but the same could be true of the homeless person, though the help would be of a very different kind and have a very different effect. At the same time, it is unclear whether the money would not be wasted on the friend, whose personal problem may turn out to be trivial, or the homeless person, who may spend it outside of his core needs. As such, there is no utilitarian answer.
Psychology fares no better. According to Maslow's Hierarchy of Needs, the most basic needs of the homeless person can be met with this capital, but a much higher-order need may be met by helping the friend with their personal problem. Economically, this may be a better use of the twenty dollars, but it condemns someone to hunger for the sake of another person's peace of mind.
Specific interpretations of individual ethical systems, rather than theories, may offer particular advice. Christian ethics, for example, would state it is better to help the beggar (if you are gnostic) and your friend (if you are dogmatic, concerned with spiritual welfare). There is the enlightened self-interest of Enlightenment philosophers and Buddhism, wherein one devotes all of one's resources to the common good in the knowledge that the relationship is reciprocate, or the Randian extreme self-interest model where nothing is done except in one's own immediate interest, and there is no social contract.
What all options and explanations share is that require making a value judgement that cannot be fixed by any reference, only by argument and belief.