[PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Mark Walters 

On Mon, 12 Mar 2012 13:03:12 -0700, Jameson Graef Rollins  wrote:
> On Mon, 12 Mar 2012 11:31:52 +, Mark Walters  gmail.com> wrote:
> > There is a bug in the exclude code (found by jrollins in the
> > --with-excluded series) but also present in master.  None of the
> > current tests were finding it so the first patch adds two tests.
> 
> Hey, Mark.  Thanks so much for looking in to this.  It look like this
> patch fixes the issue for me too!  Very excited to see this all coming
> together.
> 
> > The bug (and test failure) do not appear in all configuations: on my
> > main test machine (an oldish debian testing 32bit userspace with a
> > 64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
> > debian testing 64bit userspace and xapian 1.2.8) one of the new tests
> > fails.
> 
> It was failing for me too, but looking closer at the test I actually
> found a bug: you accidentally used the old style --no-exclude option,
> instead of the new --with-excludes.  When you fix the call all the tests
> pass fine.

That's great.

> > The second patch fixes the behaviour for me but I don't see why it
> > should make a difference: searches for A and not B should give the
> > same results as A and not (A and B). It could be a bug in xapian, it
> > could be that I am not allowed to reuse queries as I do (is query1 =
> > query1 and query2 allowed?) or it could be some memory use bug on my
> > part.
> 
> I can't explain it either, but there's certainly a lot about xapian that
> I don't understand.  Maybe one of the xapian gurus will have some ideas
> (Olly?  Austin?).
> 
> Anyway, thanks again for pushing on all of this, Mark.
> 
> jamie.
> 
> PS. Not a big deal, but it would have been nice for this patch set to
> have been sent in-reply-to the original series it fixes, just to keep
> everything together.

Just to emphasise the bug is already present in current master (just
better hidden because of the defaults). Hence this pair of patches
(unlike the first one I sent privately) are to current master rather
than to the exclude the series (though they apply there to to modulo the
minor change you mention).

Best wishes

Mark

[PATCH v6 06/10] schemata: Add documentation for JSON reply format.

2012-03-12 Thread Austin Clements 
Quoth Adam Wolfe Gordon on Mar 11 at 10:09 pm:
> On Sun, Mar 11, 2012 at 18:36, Austin Clements  wrote:
> > Oops. ?Looks like I left references to various old JSON functions in
> > the schemata file. ?I'll submit a patch to fix those up, but you might
> > as well use the correct function names in the new documentation.
> 
> Erk, I missed this review as I was making up the latest version of the
> series. I guess I'll have to send another, but I'll wait for reviews
> on the other patches first.

No worries.  I'm happy to fold these in to fixing up other stale
references in that file.  This shouldn't hold up your series.

[PATCH 1/2] test: add tests for message only search

2012-03-12 Thread Jameson Graef Rollins 
On Mon, 12 Mar 2012 11:31:53 +, Mark Walters  
wrote:
> +test_begin_subtest "Exclude \"deleted\" messages from message search 
> (no-exclude)"
> +output=$(notmuch search --no-exclude --output=messages subject:deleted | 
> notmuch_search_sanitize)

This should be 's/no-exclude/with-excludes/'.  With this change all
tests pass when the subsequent patch is applied.

jamie.
-- next part --
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: not available
URL: 
<http://notmuchmail.org/pipermail/notmuch/attachments/20120312/16108061/attachment.pgp>

[PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Jameson Graef Rollins 
On Mon, 12 Mar 2012 11:31:52 +, Mark Walters  
wrote:
> There is a bug in the exclude code (found by jrollins in the
> --with-excluded series) but also present in master.  None of the
> current tests were finding it so the first patch adds two tests.

Hey, Mark.  Thanks so much for looking in to this.  It look like this
patch fixes the issue for me too!  Very excited to see this all coming
together.

> The bug (and test failure) do not appear in all configuations: on my
> main test machine (an oldish debian testing 32bit userspace with a
> 64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
> debian testing 64bit userspace and xapian 1.2.8) one of the new tests
> fails.

It was failing for me too, but looking closer at the test I actually
found a bug: you accidentally used the old style --no-exclude option,
instead of the new --with-excludes.  When you fix the call all the tests
pass fine.

> The second patch fixes the behaviour for me but I don't see why it
> should make a difference: searches for A and not B should give the
> same results as A and not (A and B). It could be a bug in xapian, it
> could be that I am not allowed to reuse queries as I do (is query1 =
> query1 and query2 allowed?) or it could be some memory use bug on my
> part.

I can't explain it either, but there's certainly a lot about xapian that
I don't understand.  Maybe one of the xapian gurus will have some ideas
(Olly?  Austin?).

Anyway, thanks again for pushing on all of this, Mark.

jamie.

PS. Not a big deal, but it would have been nice for this patch set to
have been sent in-reply-to the original series it fixes, just to keep
everything together.
-- next part --
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: not available
URL: 
<http://notmuchmail.org/pipermail/notmuch/attachments/20120312/5a0d4dfe/attachment.pgp>

[PATCH 2/2] lib: fix an exclude bug

2012-03-12 Thread Mark Walters 
One test using the new exclude code was failing in some
configurations.  This patch makes it work for me. It may be a "fix"
but I do not see why it fixes it.
---
 lib/query.cc |5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/lib/query.cc b/lib/query.cc
index ab18fbc..2b73d72 100644
--- a/lib/query.cc
+++ b/lib/query.cc
@@ -213,13 +213,14 @@ notmuch_query_search_messages (notmuch_query_t *query)

if (query->exclude_terms) {
exclude_query = _notmuch_exclude_tags (query, final_query);
-   exclude_query = Xapian::Query (Xapian::Query::OP_AND,
-  exclude_query, final_query);

if (query->omit_excluded_messages)
final_query = Xapian::Query (Xapian::Query::OP_AND_NOT,
 final_query, exclude_query);
else {
+   exclude_query = Xapian::Query (Xapian::Query::OP_AND,
+  exclude_query, final_query);
+
enquire.set_weighting_scheme (Xapian::BoolWeight());
enquire.set_query (exclude_query);

-- 
1.7.9.1

[PATCH 1/2] test: add tests for message only search

2012-03-12 Thread Mark Walters 
This adds two tests for --output=messages searches one of which fails
in some configurations.
---
 test/search |   11 +++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/test/search b/test/search
index 081f60c..276271b 100755
--- a/test/search
+++ b/test/search
@@ -132,13 +132,24 @@ test_expect_equal "$output" "thread:XXX   2000-01-01 
[1/1] Notmuch Test Suite; u
 test_begin_subtest "Exclude \"deleted\" messages from search"
 notmuch config set search.exclude_tags = deleted
 generate_message '[subject]="Not deleted"'
+not_deleted_id=$gen_msg_id
 generate_message '[subject]="Deleted"'
 notmuch new > /dev/null
 notmuch tag +deleted id:$gen_msg_id
+deleted_id=$gen_msg_id
 output=$(notmuch search subject:deleted | notmuch_search_sanitize)
 test_expect_equal "$output" "thread:XXX   2001-01-05 [1/1] Notmuch Test Suite; 
Not deleted (inbox unread)
 thread:XXX   2001-01-05 [0/1] Notmuch Test Suite; Deleted (deleted inbox 
unread)"

+test_begin_subtest "Exclude \"deleted\" messages from message search"
+output=$(notmuch search --output=messages subject:deleted | 
notmuch_search_sanitize)
+test_expect_equal "$output" "id:$not_deleted_id"
+
+test_begin_subtest "Exclude \"deleted\" messages from message search 
(no-exclude)"
+output=$(notmuch search --no-exclude --output=messages subject:deleted | 
notmuch_search_sanitize)
+test_expect_equal "$output" "id:$not_deleted_id
+id:$deleted_id"
+
 test_begin_subtest "Exclude \"deleted\" messages from search, overridden"
 output=$(notmuch search subject:deleted and tag:deleted | 
notmuch_search_sanitize)
 test_expect_equal "$output" "thread:XXX   2001-01-05 [1/1] Notmuch Test Suite; 
Deleted (deleted inbox unread)"
-- 
1.7.9.1

[PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Mark Walters 
There is a bug in the exclude code (found by jrollins in the
--with-excluded series) but also present in master.  None of the
current tests were finding it so the first patch adds two tests.

The bug (and test failure) do not appear in all configuations: on my
main test machine (an oldish debian testing 32bit userspace with a
64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
debian testing 64bit userspace and xapian 1.2.8) one of the new tests
fails.

The second patch fixes the behaviour for me but I don't see why it
should make a difference: searches for A and not B should give the
same results as A and not (A and B). It could be a bug in xapian, it
could be that I am not allowed to reuse queries as I do (is query1 =
query1 and query2 allowed?) or it could be some memory use bug on my
part.

Anyway the "fix" is small which should help narrow down the actual
cause.

Best wishes

Mark 

Mark Walters (2):
  test: add tests for message only search
  lib: fix an exclude bug

 lib/query.cc |5 +++--
 test/search  |   11 +++
 2 files changed, 14 insertions(+), 2 deletions(-)

-- 
1.7.9.1

[PATCH] man: update SEE ALSO references in man pages

2012-03-12 Thread Jani Nikula 
Drop references to notmuch-part(1). Reference all man pages. Fix man
page section of notmuch-dump(1).
---
 man/man1/notmuch-config.1   |5 ++---
 man/man1/notmuch-count.1|8 
 man/man1/notmuch-dump.1 |8 
 man/man1/notmuch-new.1  |8 
 man/man1/notmuch-reply.1|8 
 man/man1/notmuch-restore.1  |8 
 man/man1/notmuch-search.1   |8 
 man/man1/notmuch-show.1 |8 
 man/man1/notmuch-tag.1  |8 
 man/man1/notmuch.1  |7 +++
 man/man5/notmuch-hooks.5|8 
 man/man7/notmuch-search-terms.7 |5 ++---
 12 files changed, 43 insertions(+), 46 deletions(-)

diff --git a/man/man1/notmuch-config.1 b/man/man1/notmuch-config.1
index e62577c..d22988a 100644
--- a/man/man1/notmuch-config.1
+++ b/man/man1/notmuch-config.1
@@ -137,9 +137,8 @@ Specifies the location of the notmuch configuration file. 
Notmuch will
 use ${HOME}/.notmuch\-config if this variable is not set.
 .SH SEE ALSO

-\fBnotmuch\fR(1), \fBnotmuch-count\fR(1),
-\fBnotmuch-dump\fR(1), \fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1),
-\fBnotmuch-part\fR(1), \fBnotmuch-reply\fR(1),
+\fBnotmuch\fR(1), \fBnotmuch-count\fR(1), \fBnotmuch-dump\fR(1),
+\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-reply\fR(1),
 \fBnotmuch-restore\fR(1), \fBnotmuch-search\fR(1),
 \fBnotmuch-search-terms\fR(7), \fBnotmuch-show\fR(1),
 \fBnotmuch-tag\fR(1)
diff --git a/man/man1/notmuch-count.1 b/man/man1/notmuch-count.1
index 805a8ae..f593065 100644
--- a/man/man1/notmuch-count.1
+++ b/man/man1/notmuch-count.1
@@ -51,7 +51,7 @@ Do not exclude the messages matching search.exclude_tags in 
the config file.
 .SH SEE ALSO

 \fBnotmuch\fR(1), \fBnotmuch-config\fR(1), \fBnotmuch-dump\fR(1),
-\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-part\fR(1),
-\fBnotmuch-reply\fR(1), \fBnotmuch-restore\fR(1),
-\fBnotmuch-search\fR(1), \fBnotmuch-search-terms\fR(7),
-\fBnotmuch-show\fR(1), \fBnotmuch-tag\fR(1)
+\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-reply\fR(1),
+\fBnotmuch-restore\fR(1), \fBnotmuch-search\fR(1),
+\fBnotmuch-search-terms\fR(7), \fBnotmuch-show\fR(1),
+\fBnotmuch-tag\fR(1)
diff --git a/man/man1/notmuch-dump.1 b/man/man1/notmuch-dump.1
index bd7e274..4e24392 100644
--- a/man/man1/notmuch-dump.1
+++ b/man/man1/notmuch-dump.1
@@ -31,7 +31,7 @@ for details of the supported syntax for .
 .SH SEE ALSO

 \fBnotmuch\fR(1), \fBnotmuch-config\fR(1), \fBnotmuch-count\fR(1),
-\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-part\fR(1),
-\fBnotmuch-reply\fR(1), \fBnotmuch-restore\fR(1),
-\fBnotmuch-search\fR(1), \fBnotmuch-search-terms\fR(7),
-\fBnotmuch-show\fR(1), \fBnotmuch-tag\fR(1)
+\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-reply\fR(1),
+\fBnotmuch-restore\fR(1), \fBnotmuch-search\fR(1),
+\fBnotmuch-search-terms\fR(7), \fBnotmuch-show\fR(1),
+\fBnotmuch-tag\fR(1)
diff --git a/man/man1/notmuch-new.1 b/man/man1/notmuch-new.1
index ccee738..19cef3d 100644
--- a/man/man1/notmuch-new.1
+++ b/man/man1/notmuch-new.1
@@ -64,7 +64,7 @@ Prevents hooks from being run.
 .SH SEE ALSO

 \fBnotmuch\fR(1), \fBnotmuch-config\fR(1), \fBnotmuch-count\fR(1),
-\fBnotmuch-dump\fR(5), \fBnotmuch-hooks\fR(5), \fBnotmuch-part\fR(1),
-\fBnotmuch-reply\fR(1), \fBnotmuch-restore\fR(1),
-\fBnotmuch-search\fR(1), \fBnotmuch-search-terms\fR(7),
-\fBnotmuch-show\fR(1), \fBnotmuch-tag\fR(1)
+\fBnotmuch-dump\fR(1), \fBnotmuch-hooks\fR(5), \fBnotmuch-reply\fR(1),
+\fBnotmuch-restore\fR(1), \fBnotmuch-search\fR(1),
+\fBnotmuch-search-terms\fR(7), \fBnotmuch-show\fR(1),
+\fBnotmuch-tag\fR(1)
diff --git a/man/man1/notmuch-reply.1 b/man/man1/notmuch-reply.1
index 7ed7f0f..0fb68b4 100644
--- a/man/man1/notmuch-reply.1
+++ b/man/man1/notmuch-reply.1
@@ -80,7 +80,7 @@ on issue found in multiple patches.
 .SH SEE ALSO

 \fBnotmuch\fR(1), \fBnotmuch-config\fR(1), \fBnotmuch-count\fR(1),
-\fBnotmuch-dump\fR(5), \fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1),
-\fBnotmuch-part\fR(1), \fBnotmuch-restore\fR(1),
-\fBnotmuch-search\fR(1), \fBnotmuch-search-terms\fR(7),
-\fBnotmuch-show\fR(1), \fBnotmuch-tag\fR(1)
+\fBnotmuch-dump\fR(1), \fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1),
+\fBnotmuch-restore\fR(1), \fBnotmuch-search\fR(1),
+\fBnotmuch-search-terms\fR(7), \fBnotmuch-show\fR(1),
+\fBnotmuch-tag\fR(1)
diff --git a/man/man1/notmuch-restore.1 b/man/man1/notmuch-restore.1
index 8bd59e4..bcb765c 100644
--- a/man/man1/notmuch-restore.1
+++ b/man/man1/notmuch-restore.1
@@ -39,7 +39,7 @@ details.
 .SH SEE ALSO

 \fBnotmuch\fR(1), \fBnotmuch-config\fR(1), \fBnotmuch-count\fR(1),
-\fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1), \fBnotmuch-part\fR(1),
-\fBnotmuch-reply\fR(1), \fBnotmuch-dump\fR(1),
-\fBnotmuch-search\fR(1), \fBnotmuch-search-terms\fR(7),
-\fBnotmuch-show\fR(1), \fBnotmuch-tag\fR(1)
+\fBnotmuch-dump\fR(1), \fBnotmuch-hooks\fR(5), \fBnotmuch-new\fR(1),
+\fBnotmuch-reply\fR(1),

[PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Mark Walters 
There is a bug in the exclude code (found by jrollins in the
--with-excluded series) but also present in master.  None of the
current tests were finding it so the first patch adds two tests.

The bug (and test failure) do not appear in all configuations: on my
main test machine (an oldish debian testing 32bit userspace with a
64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
debian testing 64bit userspace and xapian 1.2.8) one of the new tests
fails.

The second patch fixes the behaviour for me but I don't see why it
should make a difference: searches for A and not B should give the
same results as A and not (A and B). It could be a bug in xapian, it
could be that I am not allowed to reuse queries as I do (is query1 =
query1 and query2 allowed?) or it could be some memory use bug on my
part.

Anyway the fix is small which should help narrow down the actual
cause.

Best wishes

Mark 

Mark Walters (2):
  test: add tests for message only search
  lib: fix an exclude bug

 lib/query.cc |5 +++--
 test/search  |   11 +++
 2 files changed, 14 insertions(+), 2 deletions(-)

-- 
1.7.9.1

___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch

[PATCH 1/2] test: add tests for message only search

2012-03-12 Thread Mark Walters 
This adds two tests for --output=messages searches one of which fails
in some configurations.
---
 test/search |   11 +++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/test/search b/test/search
index 081f60c..276271b 100755
--- a/test/search
+++ b/test/search
@@ -132,13 +132,24 @@ test_expect_equal $output thread:XXX   2000-01-01 
[1/1] Notmuch Test Suite; u
 test_begin_subtest Exclude \deleted\ messages from search
 notmuch config set search.exclude_tags = deleted
 generate_message '[subject]=Not deleted'
+not_deleted_id=$gen_msg_id
 generate_message '[subject]=Deleted'
 notmuch new  /dev/null
 notmuch tag +deleted id:$gen_msg_id
+deleted_id=$gen_msg_id
 output=$(notmuch search subject:deleted | notmuch_search_sanitize)
 test_expect_equal $output thread:XXX   2001-01-05 [1/1] Notmuch Test Suite; 
Not deleted (inbox unread)
 thread:XXX   2001-01-05 [0/1] Notmuch Test Suite; Deleted (deleted inbox 
unread)
 
+test_begin_subtest Exclude \deleted\ messages from message search
+output=$(notmuch search --output=messages subject:deleted | 
notmuch_search_sanitize)
+test_expect_equal $output id:$not_deleted_id
+
+test_begin_subtest Exclude \deleted\ messages from message search 
(no-exclude)
+output=$(notmuch search --no-exclude --output=messages subject:deleted | 
notmuch_search_sanitize)
+test_expect_equal $output id:$not_deleted_id
+id:$deleted_id
+
 test_begin_subtest Exclude \deleted\ messages from search, overridden
 output=$(notmuch search subject:deleted and tag:deleted | 
notmuch_search_sanitize)
 test_expect_equal $output thread:XXX   2001-01-05 [1/1] Notmuch Test Suite; 
Deleted (deleted inbox unread)
-- 
1.7.9.1

___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch

Re: [PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Jameson Graef Rollins 
On Mon, 12 Mar 2012 11:31:52 +, Mark Walters markwalters1...@gmail.com 
wrote:
 There is a bug in the exclude code (found by jrollins in the
 --with-excluded series) but also present in master.  None of the
 current tests were finding it so the first patch adds two tests.

Hey, Mark.  Thanks so much for looking in to this.  It look like this
patch fixes the issue for me too!  Very excited to see this all coming
together.

 The bug (and test failure) do not appear in all configuations: on my
 main test machine (an oldish debian testing 32bit userspace with a
 64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
 debian testing 64bit userspace and xapian 1.2.8) one of the new tests
 fails.

It was failing for me too, but looking closer at the test I actually
found a bug: you accidentally used the old style --no-exclude option,
instead of the new --with-excludes.  When you fix the call all the tests
pass fine.

 The second patch fixes the behaviour for me but I don't see why it
 should make a difference: searches for A and not B should give the
 same results as A and not (A and B). It could be a bug in xapian, it
 could be that I am not allowed to reuse queries as I do (is query1 =
 query1 and query2 allowed?) or it could be some memory use bug on my
 part.

I can't explain it either, but there's certainly a lot about xapian that
I don't understand.  Maybe one of the xapian gurus will have some ideas
(Olly?  Austin?).

Anyway, thanks again for pushing on all of this, Mark.

jamie.

PS. Not a big deal, but it would have been nice for this patch set to
have been sent in-reply-to the original series it fixes, just to keep
everything together.


pgpBG5NEdaeRi.pgp
Description: PGP signature
___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch

Re: [PATCH 1/2] test: add tests for message only search

2012-03-12 Thread Jameson Graef Rollins 
On Mon, 12 Mar 2012 11:31:53 +, Mark Walters markwalters1...@gmail.com 
wrote:
 +test_begin_subtest Exclude \deleted\ messages from message search 
 (no-exclude)
 +output=$(notmuch search --no-exclude --output=messages subject:deleted | 
 notmuch_search_sanitize)

This should be 's/no-exclude/with-excludes/'.  With this change all
tests pass when the subsequent patch is applied.

jamie.


pgpt2YSpUKffG.pgp
Description: PGP signature
___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch

Re: [PATCH v6 06/10] schemata: Add documentation for JSON reply format.

2012-03-12 Thread Austin Clements 
Quoth Adam Wolfe Gordon on Mar 11 at 10:09 pm:
 On Sun, Mar 11, 2012 at 18:36, Austin Clements amdra...@mit.edu wrote:
  Oops.  Looks like I left references to various old JSON functions in
  the schemata file.  I'll submit a patch to fix those up, but you might
  as well use the correct function names in the new documentation.
 
 Erk, I missed this review as I was making up the latest version of the
 series. I guess I'll have to send another, but I'll wait for reviews
 on the other patches first.

No worries.  I'm happy to fold these in to fixing up other stale
references in that file.  This shouldn't hold up your series.
___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch

Re: [PATCH 0/2] A bug in the exclude code

2012-03-12 Thread Mark Walters 

On Mon, 12 Mar 2012 13:03:12 -0700, Jameson Graef Rollins 
jroll...@finestructure.net wrote:
 On Mon, 12 Mar 2012 11:31:52 +, Mark Walters markwalters1...@gmail.com 
 wrote:
  There is a bug in the exclude code (found by jrollins in the
  --with-excluded series) but also present in master.  None of the
  current tests were finding it so the first patch adds two tests.
 
 Hey, Mark.  Thanks so much for looking in to this.  It look like this
 patch fixes the issue for me too!  Very excited to see this all coming
 together.
 
  The bug (and test failure) do not appear in all configuations: on my
  main test machine (an oldish debian testing 32bit userspace with a
  64bit kernel and xapian 1.2.7) all tests pass. On my laptop (a recent
  debian testing 64bit userspace and xapian 1.2.8) one of the new tests
  fails.
 
 It was failing for me too, but looking closer at the test I actually
 found a bug: you accidentally used the old style --no-exclude option,
 instead of the new --with-excludes.  When you fix the call all the tests
 pass fine.

That's great.

  The second patch fixes the behaviour for me but I don't see why it
  should make a difference: searches for A and not B should give the
  same results as A and not (A and B). It could be a bug in xapian, it
  could be that I am not allowed to reuse queries as I do (is query1 =
  query1 and query2 allowed?) or it could be some memory use bug on my
  part.
 
 I can't explain it either, but there's certainly a lot about xapian that
 I don't understand.  Maybe one of the xapian gurus will have some ideas
 (Olly?  Austin?).
 
 Anyway, thanks again for pushing on all of this, Mark.
 
 jamie.
 
 PS. Not a big deal, but it would have been nice for this patch set to
 have been sent in-reply-to the original series it fixes, just to keep
 everything together.

Just to emphasise the bug is already present in current master (just
better hidden because of the defaults). Hence this pair of patches
(unlike the first one I sent privately) are to current master rather
than to the exclude the series (though they apply there to to modulo the
minor change you mention).

Best wishes

Mark

___
notmuch mailing list
notmuch@notmuchmail.org
http://notmuchmail.org/mailman/listinfo/notmuch