While this is something more internal, its still important. old oclHashcat-* version (and CPU version as well) use a very static hash parser code. In the past I thought this is a good idea since this way the loading of the hashes, especially when load multi-million hashlists, is much faster.
But then there is a compatibility problem. For example the nsldap hashes are by reference set to an 8-byte salt. But some sites change this and if they do, we get the well known line-length exception. This fix this it required new parser code.
In theory this enables also a new parameter like --base64-salt or --base64-charset like we already have for the HEX encoding. Not sure of this of any use... Suggestions?
But then there is a compatibility problem. For example the nsldap hashes are by reference set to an 8-byte salt. But some sites change this and if they do, we get the well known line-length exception. This fix this it required new parser code.
In theory this enables also a new parameter like --base64-salt or --base64-charset like we already have for the HEX encoding. Not sure of this of any use... Suggestions?