What's the general feeling about this approach to producing tests for
the emacs UI? (That is, code the test in a .el file and call the
relevant function(s) from the test harness.)
It makes it simpler to develop and maintain the test (because you can do
more work with traditional emacs support for
Hi!
When I try to run the attached test.py after adding the attached email
(4EFC743A.3060609_april.org) to notmuch db I got a segmentation fault
(gdb bt attached).
This is what I think a relevant part of the gdb:
(gdb) frame 1
#1 0x75f2759c in g_mime_iconv_open (to=0x761ef0
On Wed, 28 Dec 2011 21:34:56 -0500, Austin Clements amdra...@mit.edu wrote:
It makes no sense to run test-lib.sh, so it makes no sense to give it
an interpreter. This is particularly annoying for Emacs users who
have executable-insert set, since the presence of the #! line will
cause Emacs to
On Wed, 28 Dec 2011 21:34:43 -0500, Austin Clements amdra...@mit.edu wrote:
January 5, 2001 was a Tuesday, not a Friday. Jameson fixed this exact
problem for the multipart test in ec2b0a98cc, but not for
generate_message itself.
Pushed,
d
___
Hi!
When I try to run the attached test.py after adding the attached email
(4EFC743A.3060609_april.org) to notmuch db I got a segmentation fault
(gdb bt attached).
This is what I think a relevant part of it:
(gdb) frame 1
#1 0x75f2759c in g_mime_iconv_open (to=0x761ef0 UTF-8,
`mail-header-parse-address' expects un-decoded mailbox parts, which is
not what we have at this point. Replace it with simple string
deconstruction.
---
Added removal of backslashes from the mailbox part.
emacs/notmuch-show.el | 52
1 files
On Sun, 25 Dec 2011 08:14:52 +0400, Dmitry Kurochkin
dmitry.kuroch...@gmail.com wrote:
Before the change, there were two ways to do search in Emacs UI:
search widget in notmuch-hello buffer and `notmuch-search' function
bound to s. Internally, notmuch-hello search widget uses
On Fri, 21 Oct 2011 22:50:48 +0200, Daniel Schoepe dan...@schoepe.org wrote:
On Tue, 23 Feb 2010 11:32:51 -0500, James Vasile ja...@hackervisions.org
wrote:
Added notmuch-enqueue-asynch to replace calls to
notmuch-call-notmuch-process. Calls to notmuch are then queued and
executed
On Wed, 28 Dec 2011 12:14:29 -0800, Jameson Graef Rollins
jroll...@finestructure.net wrote:
It appears to be an oversight that encrypted parts were indexed
previously. The terms generated from encrypted parts are meaningless
and do nothing but add bloat to the database. It is not worth
Hi!
When I try to run the attached test.py after adding the attached email
(4EFC743A.3060609_april.org) to notmuch db I got a segmentation fault
(gdb bt attached).
This is what I think a relevant part of it:
(gdb) frame 1
#1 0x75f2759c in g_mime_iconv_open (to=0x761ef0 UTF-8,
---
Added backslash test. UTF8 round-trip still isn't right.
test/emacs | 20
.../notmuch-address-simplification | 10 ++
2 files changed, 30 insertions(+), 0 deletions(-)
create mode 100644
---
test/emacs |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/test/emacs b/test/emacs
index 69738cd..32d49ae 100755
--- a/test/emacs
+++ b/test/emacs
@@ -78,7 +78,7 @@ thread=$(notmuch search --output=threads
subject:message-with-invalid-from)
test_emacs (notmuch-show
After setting up a new user recently[1] and discussion with j4ni, it
occurred to me that some of my recent patches are interrelated in a way
that I didn't previously realise.
Getting natural feeling `notmuch-show-rewind' behaviour (latest patches
are [2]) is made more difficult if
Add various functions to print notmuch messages and tie them together
with a simple frontend. No keybinding is currently made to encourage
paper saving.
---
Resurrecting an old patch.
emacs/Makefile.local |3 +-
emacs/notmuch-print.el | 75
Konrad, can you provide a specific example of when this is useful? I
don't understand what it is for.
pgpLhgKrmkbXn.pgp
Description: PGP signature
___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch
`notmuch-hello-mode' inherits the keymap for widgets, which confused
`notmuch-substitute-command-keys'. Fix the confusion.
Simplify `notmuch-substitute-command-keys' a little to make it easier
to read.
---
emacs/notmuch.el | 17 -
1 files changed, 12 insertions(+), 5
On Sun, 25 Dec 2011 15:55:43 -0400, David Bremner da...@tethera.net wrote:
Well, it's -19C outside, so guess that is appropriate.
It's warmed up to -10C, with a brief interlude of +10C, so we must be
getting ready to unfreeze. Well, release, but that doesn't really work
as a pun.
Speaking of
On Thu, 29 Dec 2011 22:57:27 -0200, Kazuo Teramoto kaz@gmail.com wrote:
notmuch CLI show the message correct. I know nothing about gmime or
notmuch code, but can this be the case of the python bindings not
calling g_mime_init() correctly?
That seems likely. We worked around a similar
On Fri, 30 Dec 2011 10:43:37 -0400, David Bremner da...@tethera.net wrote:
On Fri, 30 Dec 2011 10:44:28 +, David Edmondson d...@dme.org wrote:
On Sun, 25 Dec 2011 08:14:52 +0400, Dmitry Kurochkin
dmitry.kuroch...@gmail.com
Were there actual complaints about the usability of the
Tester may have set LD_LIBRARY_PATH to find libraries required
by notmuch. Therefore add $TEST_DIRECTORY/../lib to the beginning
of current list of library paths in $LD_LIBRARY_PATH before running
symbol-test.
---
test/symbol-hiding |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
nottoomuch-addresses.sh -- email address substring matcher -- completion
helper version 1.6 is available.
This version fixes some 'encoded-text' (=?encoding?type?text?=)
handling.
Header version was not changed so rebuilding of address cache
is not enforced. However running
On Fri, 30 Dec 2011 10:43:37 -0400, David Bremner da...@tethera.net wrote:
Personally I would like to be able to customize the hello screen to
remove the search box. I find it confusing that that 's' jumps to the
search box there, and everywhere else brings up the mini-buffer.
+1. I rarely
---
NEWS | 10 ++
1 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/NEWS b/NEWS
index eaed960..3688944 100644
--- a/NEWS
+++ b/NEWS
@@ -20,6 +20,16 @@ Automatic tag query optimization
exclude messages whose tags won't change. In the past, we've
suggested that people
The C side looks good to me (with the updated commit message). I
can't speak for the nroff side.
Quoth David Bremner on Dec 29 at 9:11 am:
From: David Bremner brem...@debian.org
Since we lose the fancy scanning of the commands for which ones take
search terms, the search terms page
On Fri, 30 Dec 2011 10:52:17 +, David Edmondson d...@dme.org wrote:
I discovered this patch a while ago and it almost applies cleanly (the
conflicts are easy to fix though) and it has made the emacs UI for
notmuch _much_ more responsive and enjoyable for me.
This is definitely true for
Oh, and I moved the code to
http://commonmeasure.org/~jkr/notmuch_addresses.git
Best,
Jesse
On Fri, 30 Dec 2011 14:04:43 -0500, Jesse Rosenthal jrosent...@jhu.edu wrote:
Pushed.
Thanks,
Jesse
On Wed, 21 Dec 2011 13:49:17 +, David Edmondson d...@dme.org wrote:
---
The gmime docs don't says if is ok to call g_mime_init multiple times,
but the code have a check for it in a form like this:
static unsigned int initialized = 0;
g_mime_init (guint32 flags)
{
if (initialized++)
return;
so the init code is run only once and notmuch
g_mime_init already check for multiple initializations.
---
lib/index.cc |4
lib/message-file.c |4
2 files changed, 0 insertions(+), 8 deletions(-)
diff --git a/lib/index.cc b/lib/index.cc
index d8f8b2b..6764929 100644
--- a/lib/index.cc
+++ b/lib/index.cc
@@ -419,12
We need to call g_mime_init to correct initialize the structures needed
by gmime before using it.
---
lib/database.cc |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/lib/database.cc b/lib/database.cc
index d11dfaf..07ca3fd 100644
--- a/lib/database.cc
+++
hi!
while I cannot confirm that this patch solves the stated issue (cannot
reproduce),
It seeminly doesn't break things either and I'm all in favour of doing things
right™.
On a related note: It also doesn't fix the recent segfaults of the bindings
that occur when you try to write to a locked
On 2011-12-30T23:16:47, Patrick Totzke wrote:
It seeminly doesn't break things either and I'm all in favour of doing
things right™.
And talking about doing things right, we probably need to call
g_mime_shutdown() too, one for every time we call g_mime_init(). I will
send a patch after I do some
On Fri, 30 Dec 2011 13:42:24 -0500, Aaron Ecay aarone...@gmail.com wrote:
---
NEWS | 10 ++
1 files changed, 10 insertions(+), 0 deletions(-)
pushed, although I had to mess with it a fair amount. Maybe it was
against master, rather than release?
d
On Fri, 30 Dec 2011 19:58:09 -0200, Kazuo Teramoto kaz@gmail.com wrote:
g_mime_init already check for multiple initializations.
---
lib/index.cc |4
lib/message-file.c |4
2 files changed, 0 insertions(+), 8 deletions(-)
Is this needed to fix the bug? It seems
On Fri, 30 Dec 2011 19:58:10 -0200, Kazuo Teramoto kaz@gmail.com wrote:
We need to call g_mime_init to correct initialize the structures needed
by gmime before using it.
---
lib/database.cc |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
I can confirm this patch (alone)
On 2011-12-31T00:55:59, David Bremner wrote:
On Fri, 30 Dec 2011 19:58:09 -0200, Kazuo Teramoto kaz@gmail.com wrote:
g_mime_init already check for multiple initializations.
---
lib/index.cc |4
lib/message-file.c |4
2 files changed, 0 insertions(+), 8
I'm resending the PATCH 2/2 as a standalone, leaving the matter of
removing the checks for the future. I added a more detailed commit too.
This is the only patch needed to fix the segfault I reported.
Kazuo Teramoto (1):
lib: call g_mime_init() from notmuch_database_open()
lib/database.cc |
As reported in
id:caebopgyuhnz4bptdutntpuhcp3eycrcrkxhyojr43rumw67...@mail.gmail.com
sometimes gmime try to access a NULL pointer, e.g. g_mime_iconv_open()
try to access iconv_cache that is NULL if g_mime_init() is not called.
This cause notmuch to segfault when calling gmime functions.
Calling
On Fri, 30 Dec 2011 22:54:30 -0400, David Bremner da...@tethera.net wrote:
pushed, although I had to mess with it a fair amount. Maybe it was
against master, rather than release?
It was. I didn’t realize that the patch should be against the release
branch; sorry about that... :-|
--
Aaron
Shouldn't we remove the g_mime_init's from this code entirely if we're
going to do it in notmuch_database_open?
Quoth Kazuo Teramoto on Dec 30 at 7:58 pm:
g_mime_init already check for multiple initializations.
---
lib/index.cc |4
lib/message-file.c |4
2 files
`mail-header-parse-address' expects un-decoded mailbox parts, which is
not what we have at this point. Replace it with simple string
deconstruction.
---
Added removal of backslashes from the mailbox part.
emacs/notmuch-show.el | 52
1 files
bytes
Desc: not available
URL:
<http://notmuchmail.org/pipermail/notmuch/attachments/20111230/c5d8f6f8/attachment.pgp>
197 bytes
Desc: not available
URL:
<http://notmuchmail.org/pipermail/notmuch/attachments/20111230/42d48223/attachment.pgp>
---
Added backslash test. UTF8 round-trip still isn't right.
test/emacs | 20
.../notmuch-address-simplification | 10 ++
2 files changed, 30 insertions(+), 0 deletions(-)
create mode 100644
---
test/emacs |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/test/emacs b/test/emacs
index 69738cd..32d49ae 100755
--- a/test/emacs
+++ b/test/emacs
@@ -78,7 +78,7 @@ thread=$(notmuch search --output=threads
subject:message-with-invalid-from)
test_emacs
bytes
Desc: not available
URL:
<http://notmuchmail.org/pipermail/notmuch/attachments/20111230/0aa5d8bd/attachment-0001.pgp>
Add various functions to print notmuch messages and tie them together
with a simple frontend. No keybinding is currently made to encourage
paper saving.
---
Resurrecting an old patch.
emacs/Makefile.local |3 +-
emacs/notmuch-print.el | 75
org/pipermail/notmuch/attachments/20111230/a0ed97a3/attachment.pgp>
`notmuch-hello-mode' inherits the keymap for widgets, which confused
`notmuch-substitute-command-keys'. Fix the confusion.
Simplify `notmuch-substitute-command-keys' a little to make it easier
to read.
---
emacs/notmuch.el | 17 -
1 files changed, 12 insertions(+), 5
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 315 bytes
Desc: not available
URL:
<http://notmuchmail.org/pipermail/notmuch/attachments/20111230/6c0548e2/attachment.pgp>
On Fri, 30 Dec 2011 10:44:28 +, David Edmondson wrote:
> On Sun, 25 Dec 2011 08:14:52 +0400, Dmitry Kurochkin gmail.com>
> Were there actual complaints about the usability of the current
> approach?
>
> I much preferred your alternative approach (merging the history while
> leaving the UI
On Thu, 29 Dec 2011 22:57:27 -0200, Kazuo Teramoto wrote:
> notmuch CLI show the message correct. I know nothing about gmime or
> notmuch code, but can this be the case of the python bindings not
> calling g_mime_init() correctly?
That seems likely. We worked around a similar problem by calling
Tester may have set LD_LIBRARY_PATH to find libraries required
by notmuch. Therefore add $TEST_DIRECTORY/../lib to the beginning
of current list of library paths in $LD_LIBRARY_PATH before running
symbol-test.
---
test/symbol-hiding |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
nottoomuch-addresses.sh -- email address substring matcher -- completion
helper version 1.6 is available.
This version fixes some 'encoded-text' (=?encoding?type?text?=)
handling.
Header version was not changed so rebuilding of address cache
is not enforced. However running
On Fri, 30 Dec 2011 10:43:37 -0400, David Bremner wrote:
> Personally I would like to be able to customize the hello screen to
> remove the search box. I find it confusing that that 's' jumps to the
> search box there, and everywhere else brings up the mini-buffer.
+1. I rarely use
---
NEWS | 10 ++
1 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/NEWS b/NEWS
index eaed960..3688944 100644
--- a/NEWS
+++ b/NEWS
@@ -20,6 +20,16 @@ Automatic tag query optimization
exclude messages whose tags won't change. In the past, we've
suggested that people
The C side looks good to me (with the updated commit message). I
can't speak for the nroff side.
Quoth David Bremner on Dec 29 at 9:11 am:
> From: David Bremner
>
> Since we lose the fancy scanning of the commands for which ones take
> search terms, the search terms page SYNOPSIS should be
Pushed.
Thanks,
Jesse
On Wed, 21 Dec 2011 13:49:17 +, David Edmondson wrote:
> ---
> notmuch_addresses.py |7 +--
> 1 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/notmuch_addresses.py b/notmuch_addresses.py
> index bf45151..74a743c 100755
> ---
On Fri, 30 Dec 2011 10:52:17 +, David Edmondson wrote:
> > I discovered this patch a while ago and it almost applies cleanly (the
> > conflicts are easy to fix though) and it has made the emacs UI for
> > notmuch _much_ more responsive and enjoyable for me.
This is definitely true for me as
Oh, and I moved the code to
http://commonmeasure.org/~jkr/notmuch_addresses.git
Best,
Jesse
On Fri, 30 Dec 2011 14:04:43 -0500, Jesse Rosenthal
wrote:
> Pushed.
>
> Thanks,
> Jesse
>
> On Wed, 21 Dec 2011 13:49:17 +, David Edmondson wrote:
> > ---
> > notmuch_addresses.py |7
The gmime docs don't says if is ok to call g_mime_init multiple times,
but the code have a check for it in a form like this:
static unsigned int initialized = 0;
g_mime_init (guint32 flags)
{
if (initialized++)
return;
so the init code is run only once and notmuch
g_mime_init already check for multiple initializations.
---
lib/index.cc |4
lib/message-file.c |4
2 files changed, 0 insertions(+), 8 deletions(-)
diff --git a/lib/index.cc b/lib/index.cc
index d8f8b2b..6764929 100644
--- a/lib/index.cc
+++ b/lib/index.cc
@@ -419,12
We need to call g_mime_init to correct initialize the structures needed
by gmime before using it.
---
lib/database.cc |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/lib/database.cc b/lib/database.cc
index d11dfaf..07ca3fd 100644
--- a/lib/database.cc
+++
On Fri, 30 Dec 2011 13:42:24 -0500, Aaron Ecay wrote:
> ---
> NEWS | 10 ++
> 1 files changed, 10 insertions(+), 0 deletions(-)
pushed, although I had to mess with it a fair amount. Maybe it was
against master, rather than release?
d
On Fri, 30 Dec 2011 19:58:09 -0200, Kazuo Teramoto wrote:
> g_mime_init already check for multiple initializations.
> ---
> lib/index.cc |4
> lib/message-file.c |4
> 2 files changed, 0 insertions(+), 8 deletions(-)
Is this needed to fix the bug? It seems less
On Fri, 30 Dec 2011 19:58:10 -0200, Kazuo Teramoto wrote:
> We need to call g_mime_init to correct initialize the structures needed
> by gmime before using it.
> ---
> lib/database.cc |5 +
> 1 files changed, 5 insertions(+), 0 deletions(-)
I can confirm this patch (alone) fixes the
On Fri, 30 Dec 2011 22:54:30 -0400, David Bremner wrote:
> pushed, although I had to mess with it a fair amount. Maybe it was
> against master, rather than release?
It was. I didn?t realize that the patch should be against the release
branch; sorry about that... :-|
--
Aaron Ecay
Shouldn't we remove the g_mime_init's from this code entirely if we're
going to do it in notmuch_database_open?
Quoth Kazuo Teramoto on Dec 30 at 7:58 pm:
> g_mime_init already check for multiple initializations.
> ---
> lib/index.cc |4
> lib/message-file.c |4
> 2
67 matches
Mail list logo
