Additional control over parsing of repeated elements

[yairlenga]

With 2.12, Jackson can parse xml iwith multiple repeated elements into a tree (ignore duplicate feature). This make it possible to parse, but create challenge with extracting data.

the solution result in document structure that some times will have an array and sometimes a scalar, depending on if the element exists one time or multiple time.

Wanted to suggest borrowing a feature from Perl XML::simple, which make it possible to specify list of tags with the “forceArray” flag [https://metacpan.org/pod/XML::Simple#ForceArray-=%3E-%5B-names-%5D-%23-in-important]

From my experience, this approach make it significantly easier to tree parsing most documents with repeated attributes.

With this feature, the developer is can safely access the nodes that are associated with those tags using the array notation.

as a further improvement, allowing specification of tags in hierarchy will make it

Consider: <doc> <x> 1 </x> </doc>

That will parse to { x: q }
With the propose change, the above will parse into { x: [ 1 ] }

and

<doc> <x> 1 </x> <x> 2 </x> </doc>

That will parse to { x : [ 1, 2 ] }

In the first case the ‘x’ has a value. By specifying at attribute to the mapper forceArray = [ ‘x’], the mapper will always generate array for ‘x’ even if there is only one element, simplifying access to those array.

As an extension, if forceArray is provided and multiple elements are provided fir an element not in the forceArray list, this should be considered a parsing error.