[Ethereum] call mapping as arguments of function

it works if you change to int,address (which doesn't have a dynamic size as the string)or bytesX (fixed byte array)

  string[] public myArray;

  function   myContract (int[] myArray)  {

  }

so to understand why there is 2 answers :

1-in the doc :

generic arrays of fixed and dynamic size are supported in calldata and storage with the following features: Index access, copying (from calldata to storage, inside storage, both including implicit type conversion), enlarging and shrinking and deleting. Not supported are memory-based arrays (i.e. usage in non-external functions or local variables)

2-

Solidity doesn't have a built-in way yet to deserialize string arrays. It can cheat when the contract is passing an array it created itself to itself, which is why it works with private functions. But if you want to take string arrays from the outside, you're going to have to handle the serialization and deserialization from a string (or bytes) yourself. (E.g., concatenating fixed-length items, or using null characters as separators, or using commas, etc.)

source: Is it impossible to use an array of strings as the argument to solidity function?

You need to break the interface down into smaller chunks. In practice, you'll want to extend the internal storage pattern a little, I think.

Iterable mapping comes to mind. You keep lists of keys to mappings in arrays. Something like this:

  struct VoterStruct {
    address voterAddress;
    uint256 tokensBought;
    mapping (bytes32 => uint256) tokensUsed;
    bytes32[] tokenIndex; // a list of mapping keys that exist for the voter
  }

  mapping (address => VoterStruct) public voterStructs;

Then you can expose functions that allow iterating over lists while the contract returns one set of values at a time. Very sketchy to inspire some ideas:

// client can discover how may tokens are in the list for one voter
function getVoterTokenCount(address voter) returns(uint tokenCount) {
    return voterStructs[voter].tokenIndex.length;}

// client can discover the tokenId for a voter in a given row
function getVoterTokenAtIndex(address voter, uint index) returns(bytes32 tokenId) {
    return voterStructs[voter].tokenIndex[index];}

// client can discover token used by voter and tokenId
function getVoterTokenUsed(address voter, bytes32 tokenId) returns(uint tokensUsed) {
    return voterStructs[voter].tokensUsed[tokenId];}

// append keys to the list as you go
function insertVoterToken(address voter, bytes32 tokenId) returns(bool success) {
    voterStructs[voter].tokenIndex.push(tokenId);
    return true;}

Hope I didn't flub the syntax. It might be possible to sketch out the contract a little more. Would need a general description of what it's going to do. When I use this pattern, I always add some rules to ensure there are no duplicate keys, and so on.

Anyway, possibly the hints will inspire some ideas.

Hope it helps.