Improve handling extractors output

[norberttech]

Currently, when dealing with just extracted rows, we need to properly handle them, this is how it looks now:

(new Flow())
    ->read(
        From::array(
            [['id' => 1, 'array' => ['a' => 1, 'b' => 2, 'c' => 3]]],
        )
    )
    ->withEntry('row', ref('row')->unpack())
    ->renameAll('row.', '')
    ->drop('row')
    ->withEntry('array', ref('array')->arrayMerge(lit(['d' => 4])))
    ->write(To::memory($memory = new ArrayMemory()))
    ->run();

Following lines are repeated almost* always

    ->withEntry('row', ref('row')->unpack())
    ->renameAll('row.', '')
    ->drop('row')

We should look into this and introduce an expression that will do all of those 3 things, something like:

ref('row')->rowUnpack() 

One might ask, why extractors are exposing the entire row under ArrayEntry.
The reason comes from Config::shouldPutInputIntoRows() option.

When this option is set to true, extractors will provide also additional data from the input, for example
file based extractors with that option will return something like this:

[
   'input_file_uri' => 'string',
   'row' => [...]
]

Or Http extractor will put there headers, request URL etc.

This is very handy when for example our datasets are stored like this:

• /datasets/sales/2023/january.json
• /datasets/sales/2023/february.json
• /datasets/reports/report_type/123456.xml
• /datasets/reports/report_type/789010.xml

Thanks to input_file_uri we can also parse input file path and get the month name or report id from it.