Most Rewarding Paths

### Description
You are given a directed graph *G* with *N* vertices and *M* edges. Each vertex has a positive reward associated with it that you can collect the **first time** you visit that vertex. What are the maximum total rewards you can collect if you are free to choose your starting vertex, can visit each vertex as many times as you want and can end at any reachable vertex?

#### Input
The first line contains *N* and *M*. The next line contains *N* positive space separated integers and ***ith*** integer is the reward at vertex ***i***.
The next *M* lines contain a pair of integers each, say ***a*** and ***b***, such that there is a directed edge from vertex ***a*** to ***b***.

#### Output
Print the maximum total rewards that can be collected

#### Constraints
* 1<=*N*<=1e5
* 1<=*M*<=2e5
* 1<=***a, b***<=n
* 1<= reward at a vertex <= 1e9

#### Example
**Input**
4 4
4 5 2 7
1 2
2 1
1 3
2 4

**Output**
16

### Details
- Technical Specifications: C,C++,Python
- Type of issue: Single
- Time Limit: 24 hours after the issue is assigned

### Directory Structure
Create a folder called ```rewardingPath``` under the ```Algorithms``` folder and add your code there. Name the file ```rewardingPath.cpp```

#### Note
1. Please claim the issue first by commenting here before starting to work on it.
2. Once you are done with the task and have created a Pull Request, please tag @pranavdv to request a review.

Create Algorithms/rewardingPath/rewardingPath.cpp as specified. Start by reviewing the problem statement and nearby algorithm implementations for repository conventions, then compile and run the provided sample; done means the program handles the stated constraints and produces 16 for the example.

Written by the indexing model from the issue text.