Duplicate headers are not handled correctly

Duplicate headers are not handled correctly with following csv data:
```
c;c;c;c_1
1;2;3;4
```

Current output from 5.4.0 is:
```
c: 1
c_1: 4
c_2: 3
```
Based on (outdated) documentation the expected output should be:
```
c: 3
c_1: 4
```

or if _[1..n] is appended to duplicate column name:
```
c: 1
c_1_1: 2
c_2: 3
c_1: 4
```
