Chris Hanson [Thu, 27 Apr 2000 02:16:47 +0000 (02:16 +0000)]
Eliminate ability to edit the contents of a message. IMAP doesn't
permit this, so we won't either. RMAIL is unusual among mail clients
in permitting this.
Additionally, eliminate generic procedure HEADER-FIELDS, and stop
treating the "summary-line" header specially.
Chris Hanson [Sat, 22 Apr 2000 05:04:31 +0000 (05:04 +0000)]
Modify DECODING-PARSER to accept a parser to parse the decoded text,
rather that a matcher and a keyword. Implement ENCAPSULATING-PARSER,
and abstraction mechanism for this language. Implement PARSE-STRING
and PARSE-SUBSTRING to handle top-level parsing.
Chris Hanson [Thu, 20 Apr 2000 04:09:19 +0000 (04:09 +0000)]
When deciding whether a keystroke is interesting enough to translate
manually, ignore NUMLOCK or SCROLLLOCK. Previously if either of these
were depressed, it would force manual translation. Also: raise trace
level for some common but uninteresting message types.
Chris Hanson [Thu, 13 Apr 2000 16:58:40 +0000 (16:58 +0000)]
Change all one-arg rexp combinators to accept any number of args and
treat it as an implicit sequence. Change sequence combinator to
collapse nested sequences.
Chris Hanson [Wed, 12 Apr 2000 03:08:15 +0000 (03:08 +0000)]
Reconcile the character sets and regexps used here with the
descriptions in RFC 1738. Also export them since the IMAP URL code
needs access to some of them.
Chris Hanson [Tue, 11 Apr 2000 16:01:42 +0000 (16:01 +0000)]
Change keygen-type generation so that salt is supplied in the call to
MHASH-KEYGEN, if needed. This is desirable because the salt is
usually unique for each passphrase.
Chris Hanson [Tue, 11 Apr 2000 16:00:12 +0000 (16:00 +0000)]
Change keygen-type generation so that salt is supplied in the call to
MHASH-KEYGEN, if needed. This is desirable because the salt is
usually unique for each passphrase.
Chris Hanson [Tue, 11 Apr 2000 15:24:54 +0000 (15:24 +0000)]
Change MHASH-KEYGEN to accept all of the key-generation parameters as
a compound structure, and implement MAKE-MHASH-KEYGEN-TYPE to build
that structure.
Chris Hanson [Tue, 11 Apr 2000 15:19:18 +0000 (15:19 +0000)]
Change MHASH-KEYGEN to accept all of the key-generation parameters as
a compound structure, and implement MAKE-MHASH-KEYGEN-TYPE to build
that structure.
Chris Hanson [Tue, 11 Apr 2000 03:46:57 +0000 (03:46 +0000)]
Define RANDOM-BYTE-VECTOR to generate random bytes. This is useful
for cryptographic applications that don't require an RNG with
cryptographic properties, e.g. to generate an init vector or to salt a
password.
Chris Hanson [Mon, 10 Apr 2000 19:09:18 +0000 (19:09 +0000)]
Add new implementation of GC finalizers, a cleaner replacement for the
old protection list abstraction. Unlike protection lists, GC
finalizers keep themselves clean, eliminating the need for the
programmer to interact with GC daemons and events.
Chris Hanson [Mon, 10 Apr 2000 18:32:39 +0000 (18:32 +0000)]
Add new implementation of GC finalizers, a cleaner replacement for the
old protection list abstraction. Unlike protection lists, GC
finalizers keep themselves clean, eliminating the need for the
programmer to interact with GC daemons and events.
Chris Hanson [Mon, 10 Apr 2000 02:27:47 +0000 (02:27 +0000)]
Add code to detect case where the first visible line in a window is a
partially-visible continued line and is subsequently deleted.
Previously this caused the redisplay to go into an uninterruptible
infinite loop.
Chris Hanson [Mon, 10 Apr 2000 02:27:41 +0000 (02:27 +0000)]
Add code to detect case where the first visible line in a window is a
partially-visible continued line and is subsequently deleted.
Previously this caused the redisplay to go into an uninterruptible
infinite loop.
This is yet another change in the series of revisions 1.25 and 1.30.
Revision 1.30 only partially fixed this bug. This revision introduces
a new fix that should be more robust.