根据单词之间的空格查找文本的宽度（JAVA）

Question

I've been working on a problem of optimization, that requires me to find the width of a text that gives me the longest river. A river is defined as a sequence of spaces that are only separated by 1 or less characters in different lines. For example, I have the following text:

Here's the image, the red "x" means the rivers.

-> A river is a sequence of spaces in different lines that are separated by one character to the left, to the right or none. We want the longest of them all.

The text has to be aligned to the left but not to the right, and a word cannot be divided if it does not fit.

I've been trying to work with the length of each word, instead of all the sequence of characters in order to get a better execution, but I'm not being able to solve the problem in this way. I can only think of testing all the possibilities (the minimum width is the length of the longest word), and getting the better one, but I don't think that's efficient in time and in space.

I don't want you to solve the problem, I just want to know if you think it could be solved in another way. Thanks!

EDIT: For example, given a text, I want to return the width and the length of the river.

Answer 1

Well, I see no point is saying YES, YOU CAN withoutsome hints, so here you have...

I believe you should start by defining what you know as rules deriving from the base question:

Assumptions:

1. There will not be at any location within the string two or more consecutive spaces,
2. there will be no word-wrapping and hence the minimal width will be determined by the length of the longest word,
3. Blanks at the end of a line are not to be considered as part of a river (I'm not sure this would be a correct rule as you can imagine situations where you have a 6 lines river and one line out of these 6 happens to finish earlier because the next word is too long).

You would still need to find what is the length of the longest word but this only to set the minimal width to begin from.

So, you do know that:

1. You have a long string that need to be wrapped into at least three records, such that the third record has at least two words (you need this to enable the potential creation of a 3-lines river, which is the minimum, right?),
2. As such, you can calculate the maximal length of any line, meaning, length of the whole string - length of the last two words - 1 (the 1 is the blank between the last two words) divided by 2 .

With these assumptions and facts , you can simply loop deploying the text into a matrix (each cell holding one character) and examining, line by line, of you find rivers complying with the rules.

Answer 2

You cant make this alghoritm more effective as I do not see any pattern that you can use to "I dont have to try this line width, it does not give me better result than this line width".

So basically you have to try to find all solution for all line widths and take the best. The only optimization is when you start from the smallest line widths to bigger ones and you find river longer than number of rows that will come - then you dont have to continue.

For example - if you find longest river with 4 value and your line width reached 4 rows, you can stop it as you cant find longer river.