>>>>> "CR" == Colin Rafferty <[EMAIL PROTECTED]> writes:
CR> Marco Walther writes:
>> I've seen some minor problems with bbdb and I hope I've fixed some of
>> them;-)
CR> I'm having a hard time reading the diff. If you do `diff -u', it is
CR> much easier to tell what was changed.
Yes, but it looses the portability. Only GNU diff supports this
option/format and that's not always the default diff/patch you get:-(
But here it comes;-)
-- Marco
--- ChangeLog~ Fri Jul 23 00:35:07 1999
+++ ChangeLog Tue Apr 25 18:49:19 2000
@@ -1,5 +1,15 @@
+2000-04-25 Marco Walther <[EMAIL PROTECTED]>
+
+ * bbdb-com.el (bbdb-display-completion-list): It's not a good idea
+ to discard the call-back and user data!!
+ (bbdb-complete-name): The 'pred does not work as expected with the
+ hash table. And I've a general question whether it's a good idea
+ to have one completion entry per record?! I've a lot of records
+ with the name and at least two (work & home) email addresses and I
+ would like to see all of them as choices;-)
+
1999-01-24 Matt Simmons <[EMAIL PROTECTED]>
* lisp/bbdb.el: BBDB 2.00.06 released
1999-01-24 Colin Rafferty <[EMAIL PROTECTED]>
--- bbdb-com.el~ Sat Nov 27 12:49:43 1999
+++ bbdb-com.el Tue Apr 25 18:48:27 2000
@@ -1678,13 +1678,18 @@
(set-window-configuration
bbdb-complete-name-saved-window-config))
(setq bbdb-complete-name-saved-window-config nil))))
(defun bbdb-display-completion-list (list &optional callback data)
- "Wrapper for display-completion-list.
-CALLBACK and DATA are discarded."
- (display-completion-list list))
+ "Wrapper for display-completion-list."
+;; MARCO CALLBACK and DATA are discarded.
+;; That's a bad idea if you have not only net addresses in the
+;; completion list!
+;; Or the second case is: typed `Marco' and completion choosen
+;; `marco walther' and you will end up with `Marcomarco walther' in
+;; the buffer:-(
+ (display-completion-list list :activate-callback callback :user-data data))
(defun bbdb-complete-clicked-name (event extent user-data)
"Find the record for a name clicked in a completion buffer.
Currently only used by XEmacs."
(let ((buffer (nth 0 user-data))
@@ -1731,14 +1736,20 @@
(if (and yeah-yeah-this-one
(not (eq rec yeah-yeah-this-one)))
(setq only-one-p nil))
(setq all-the-completions
(cons sym all-the-completions))
- (if (eq rec yeah-yeah-this-one)
- nil
+;; MARCO: This does not work as expected with the hash table!!
+;; And I've a general question whether it's a good idea to have one
+;; completion entry per record?! I've a lot of records with the name
+;; and at least two (work & home) email addresses and I would like to
+;; see all of them as choices;-)
+;; (if (eq rec yeah-yeah-this-one)
+;; nil
(and net (setq yeah-yeah-this-one rec))
- net))))))
+ net)))))
+;; MARCO )
(completion (try-completion pattern ht pred)))
;; If there were multiple completions for this record, the one that was
;; picked is random (hash order.) So canonicalize that to be the one
;; closest to the front of the list.
(if (and (stringp completion)
