I think that you are looking for the name_search function. This function is used for the search in many2one fields. This is for example the Product in the Order Line. You can override this function to fulfill you requirements.

As an example look at the name_search function of the Product (product.product).

def name_search(self, cr, user, name='', args=None, operator='ilike', context=None, limit=100):
    if not args:
        args = []
    if name:
        ids = self.search(cr, user, [('default_code','=',name)]+ args, limit=limit, context=context)
        if not ids:
            ids = self.search(cr, user, [('ean13','=',name)]+ args, limit=limit, context=context)
        if not ids:
            # Do not merge the 2 next lines into one single search, SQL search performance would be abysmal
            # on a database with thousands of matching products, due to the huge merge+unique needed for the
            # OR operator (and given the fact that the 'name' lookup results come from the ir.translation table
            # Performing a quick memory merge of ids in Python will give much better performance
            ids = set()
            ids.update(self.search(cr, user, args + [('default_code',operator,name)], limit=limit, context=context))
            if not limit or len(ids) < limit:
                # we may underrun the limit because of dupes in the results, that's fine
                ids.update(self.search(cr, user, args + [('name',operator,name)], limit=(limit and (limit-len(ids)) or False) , context=context))
            ids = list(ids)
        if not ids:
            ptrn = re.compile('(\[(.*?)\])')
            res = ptrn.search(name)
            if res:
                ids = self.search(cr, user, [('default_code','=', res.group(2))] + args, limit=limit, context=context)
    else:
        ids = self.search(cr, user, args, limit=limit, context=context)
    result = self.name_get(cr, user, ids, context=context)
    return result