Giuseppe's patch works like a charm. Although I think we should consider adding this as an option.
e.g:
--ignore-error-codes 1048,...
although that may be a bit obscure.
maybe, since this is the only error that has cropped up as possibly legitimate to ignore:
--allow-null-to-non-null (or something clearer)
Perhaps the tool could suggest using that option if it detects the error or parses the modify clause correctly.
Giuseppe's patch works like a charm. Although I think we should consider adding this as an option.
e.g:
--ignore- error-codes 1048,...
although that may be a bit obscure.
maybe, since this is the only error that has cropped up as possibly legitimate to ignore:
--allow- null-to- non-null (or something clearer)
Perhaps the tool could suggest using that option if it detects the error or parses the modify clause correctly.