We keep this class for backward compatibility for a while, but we encourage to use GenericSSLSocketFactory directly.
HSM enables to convert double and triple length DES key encrypted using X9.17 methods to more secure key using the variant method. In theory is possible do it using export a key under a ZMK and import it again under LMK with more secure encryption scheme, but very often exporting and importing X9.17 keys is locked. Translate Key Scheme to variant methods is enabled by default.
In addition: * fix several PMD EmptyCatchBlock violations * key scheme and Variant checking and throwing IllegalArgumentException
In addition: * fix several PMD EmptyCatchBlock violations * key scheme and variant checking and throwing IllegalArgumentException * add lacking JUnit tests
Fix swapped arguments
Chhil suggested to add: In section 8.1.1. QBean descriptor add the timeout element and describe it. In section 8.1.3. Channel timeout, keep-alive, connectiontimeout, referencing the timeouts xml element and other elements would be beneficial to new users. This is based on a recent question on the list requiring explanation of the timeout. [ci skip]
@demsey comment (baae23b#commitcomment-10330412) I was looking for why my BankNet packager stopped working (after jPOS upgrade). And i found this commit. After reverting them all is working again;) BankNet (ASCII) packager is very similar to existining in jPOS sources europay.xml. Except that subfields in DE48 use IFMC_LLCHAR instead IFEP_LLCHAR. Change to LLLL in that commit, affect only unpacking. So packing and unpacking are asymetric and isn't possible to unpack earlier packed message (at europay.xml and banknet). Upon closer examination, I found that this is my ugly solution (jPOS-28) for receiving messages (BankNet - ASCII version) without a field TCC (DE48.0) and is required for processing mastercard 0800 mesages. Before that commit it works. I have checked that europay don't work corectly before it and change to LLLL don't improve that. If possible please revert that change. I thing that we should e.g.: introduce MasterCardSubFieldPackager (just exists MasterCardEBCDICSubFieldPackager) and do it properly by overriding some methots