Updated to latest, string unix timestamp is not understood by 'get'

## Issue Description

```python
import arrow

arrow.get('1657831049')
```
results in
```
arrow.parser.ParserError: Could not match input '1657831049' to any of the following formats: YYYY-MM-DD, YYYY-M-DD, YYYY-M-D, YYYY/MM/DD, YYYY/M/DD, YYYY/M/D, YYYY.MM.DD, YYYY.M.DD, YYYY.M.D, YYYYMMDD, YYYY-DDDD, YYYYDDDD, YYYY-MM, YYYY/MM, YYYY.MM, YYYY, W.
```

worked beforehand but with the latest version it does not.

We would like to accept a lot of time formats and some of them are strings some of them are unix timestamps, so forcing a int/float conversion is unacceptable. It would be nice to use string unix timestamps.

## System Info

- 🖥 **OS name and version**: ubuntu
- 🐍 **Python version**: 3.8
- 🏹 **Arrow version**: 1.2.2

Start by reproducing the issue with arrow.get('1657831049') on Arrow 1.2.2, then trace the parsing path used by arrow.get and review its existing parser tests. Done means string Unix timestamps are accepted while the listed date formats continue to parse correctly.

Written by the indexing model from the issue text.