char * vs. unsigned char * warnings. The GNU coding standards used to
say that these mismatches are okay and better than a bunch of casts.
Obviously this has changed now.
* import.c (parse_p12, popen_protect_tool): New.
* base64.c (gpgsm_create_reader): New arg ALLOW_MULTI_PEM.
Changed all callers.
(base64_reader_cb): Handle it here.
(gpgsm_reader_eof_seen): New.
(base64_reader_cb): Set a flag for EOF.
(simple_reader_cb): Ditto.
* server.c (cmd_genkey): New.
* certreqgen.c: New. The parameter handling code has been taken
from gnupg/g10/keygen.c version 1.0.6.
* call-agent.c (gpgsm_agent_genkey): New.
* certdump.c (print_sexp): Renamed to gpgsm_dump_serial, made
global.
(print_time): Renamed to gpgsm_dump_time, made global.
(gpgsm_dump_serial): Take a real S-Expression as argument and
print the first item.
* keylist.c (list_cert_colon): Ditto.
* keydb.c (keydb_search_issuer_sn): Ditto.
* decrypt.c (print_integer_sexp): Removed and made callers
use gpgsm_dump_serial.
* verify.c (print_time): Removed, made callers use gpgsm_dump_time.
* sign.c (gpgsm_sign): Ditto.
* base64.c (base64_reader_cb): Reset the linelen when we need to
skip the line and adjusted test; I somehow forgot about DeMorgan.
* server.c (cmd_encrypt,cmd_decrypt,cmd_sign,cmd_verify)
(cmd_import): Close the FDs on success.
(close_message_fd): New.
(input_notify): Setting autodetect_encoding to 0 after initializing
it to 0 is pretty pointless. Easy to fix.
of direct creation of ksba_reader/writer.
* gpgsm.c (main): Set ctrl.auto_encoding unless --no-armor is used.
This way we can feed PEM encoded stuff to --verify.
