* tests/gpgscm/repl.scm (repl): Add an argument 'environment'.
(interactive-repl): Add an optional argument 'environment'.
--
With this change, we can drop
(interactive-repl (current-environment))
anywhere into the code and do some interactive debugging.
Signed-off-by: Justus Winter <justus@g10code.com>
* tests/gpgscm/ffi.c (ffi_init): Rename 'exit' to '_exit'.
* tests/gpgscm/ffi.scm (*interpreter-exit*): New variable.
(throw): New function.
(exit): New function.
--
This allows a proper cleanup of resources.
Signed-off-by: Justus Winter <justus@g10code.com>
* tests/gpgscm/init.scm (catch): Bind all arguments to '*error*' in
the error handler, update and fix comment.
(*error-hook*): Revert to original definition.
* tests/gpgscm/tests.scm (tr:do): Adapt accordingly.
* tests/openpgp/issue2419.scm: Likewise.
Signed-off-by: Justus Winter <justus@g10code.com>
* tests/gpgscm/ffi.c (do_get_isotime): New function.
(ffi_init): Add parameter 'scriptname', bind new function and
scriptname.
* tests/gpgscm/ffi.h (ffi_init): Update prototype.
* tests/gpgscm/main.c (main): Hand in the script name.
* tests/gpgscm/tests.scm (mkdtemp): Use current time and script name
for the names of temporary directories.
Signed-off-by: Justus Winter <justus@g10code.com>
* g10/keylist.c (list_keyblock_print): Do not depend calling
print_fingerprint on opt.keyid_format.
--
Regression-due-to: d757009a24
Debian-bug-id: 838153
Signed-off-by: Werner Koch <wk@gnupg.org>
* tests/fake-pinentries/COPYING: Rename to ...
* COPYING.CC0: this. Add a note on the scope of this license.
* COPYING.LIB: Add a note on the scope of this license.
* AUTHORS (License): Mention CC) license.
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/keyedit.c (check_all_keysigs): Check early for no key
signatures. Use xtrycalloc.
--
GnuPG-bug-id: 2690
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmgnr/cdblib.c: comment used unnecesary hyphenation
* dirmngr/crlcache.h: comment was iso-8859-1
* doc/contrib.text: list contributors using UTF-8 (now we can
acknowledge many more people using their preferred orthography)
At least one other files remains in a non-UTF-8 encoding, which i'm
not sure what to do with:
- build-aux/speedo/w32/inst.nsi is ISO-8859-1, but maybe Windows needs
it that way?
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* g10/tofu.c (tofu_set_policy): If record_binding fails, fail. If the
function fails, rollback the transaction.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
* g10/tofu.c (get_trust): Use the pk_main_keyid accessor function.
(tofu_register_signature): Likewise.
(tofu_register_encryption): Likewise.
(tofu_set_policy): Likewise and also use pk_keyid and keyid_cmp.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
* scd/app-openpgp.c (ECC_FLAG_PUBKEY): New.
(send_key_attr, get_public_key, ecc_writekey, do_auth, do_decipher)
(parse_algorithm_attribute): Check ECC_FLAG_DJB_TWEAK.
(build_ecc_privkey_template): Add ECC_Q and ECC_Q_LEN.
Support offering public key when ECC_FLAG_PUBKEY sets.
(ecc_writekey): Supply ECC_Q and ECC_Q_LEN.
(parse_algorithm_attribute): Parse pubkey-required byte.
--
OpenPGPcard protocol specification version 3.2 supports algorithm
attributes for ECC key which specifies public key data is required for
"keytocard" command. This change supports the feature.
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* g10/trustdb.c (tdb_get_validity_core): If the signer's user id
subpacket is present, only consider matching user ids.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
* g10/mainproc.c (check_sig_and_print): When printing information
about a binding don't include the current signature.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
* tests/fake-pinentries/README.txt and
tests/fake-pinentries/fake-pinentry.{sh,py,pl,php}}: New public
domain (CC0) files to encourage better test suite practices from
downstream developers.
* tests/fake-pinentries/COPYING (new): a copy of
https://creativecommons.org/publicdomain/zero/1.0/legalcode.txt
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* tests/gpgscm/scheme-private.h (struct scheme): Use (void *) for
alloc_seg.
* tests/gpgscm/scheme.c (alloc_cellseg): Use (void *) for cp. Use
(void *) for coercion of address calculation.
--
In old C language, (char *) means an address. In modern C, it's
specifically an address with alignment=1. It's good to use (void *) for
an address, because newer compiler emits warnings. Note: in this
particular case, it is just a warning and the code is safe against
invalid alignment, though.
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* g10/tofu.c (signed_by_utk): New function.
(get_trust): If a key is signed by an ultimately trusted key, then
set any bindings to good.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
* g10/keygen.c (quick_generate_keypair): Write a status error.
(parse_algo_usage_expire): Set a default curve.
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/keygen.c (FUTURE_STD_): New constants.
(parse_expire_string): Handle special keywords.
(parse_algo_usage_expire): Allow "future-default". Simplify call to
parse_expire_string.
(quick_generate_keypair): Always allow an expiration date. Replace
former "test-default" by "future-default".
--
Using an expiration date is pretty common, thus we now allow the
creation of a standard key with expiration date.
Signed-off-by: Werner Koch <wk@gnupg.org>
--
* tools/ccidmon.c (GNUPG_NAME): Define if not defined.
(digitp, hexdigitp): Change to our common semantics.
--
Note that this tool only exists in the repo.
GnuPG-bug-id: 2678
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/keylist.c (list_keyblock_colon): Avoid calling functions which
trigger a status line output before having printed a LF.
--
Status lines like KEY_CONSIDERED and KEYEPXIRED were messing up the
colons output, like here:
pub:[GNUPG:] KEY_CONSIDERED 94A5C9A03C2FE5CA3B095D8E1FDF723CF46[...]
Reported-by: Andreas Stieger <astieger@suse.com>
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/keylist.c (print_fingerprint): use compact format independent of
keyid-format; (print_key_line): always print the fingerprint
--
The choice of fingerprint display should be independent of the
keyid-format.
Currently, the representation of the fingerprint changes depending on
whether the user has specified --keyid-format to anything besides
"none". (this is common, for example, if someone happens to have
"keyid-format long" in their gpg.conf for interoperability with older
versions of gpg)
With this changeset, keyid-format governs only the format of the
displayed keyID, while the fingerprint display is governed only by the
fingerprint options:
[default]::
compact fpr of pubkey only
--with-fingerprint::
human-readable form of fpr of pubkey only
--with-fingerprint --with-fingerprint::
human-readable form of pubkey and subkey
--with-subkey-fingerprint:
compact fpr for pubkey and subkeys
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* dirmngr/dirmngr.c [HAVE_INOTIFY_INIT]: Include sys/inotify.h.
(oDisableCheckOwnSocket): New.
(opts): Add --disable-check-own-socket.
(disable_check_own_socket): New var.
(parse_rereadable_options): Set that var.
(my_inotify_is_name) [HAVE_INOTIFY_INIT]: New.
(handle_connections) [HAVE_INOTIFY_INIT]: New.
Signed-off-by: Werner Koch <wk@gnupg.org>
* g10/tofu.c (ask_about_binding): Use the time that a signature was
seen, not allegedly generated, when generating statistics.
--
Signed-off-by: Neal H. Walfield <neal@g10code.com>
Ineiev