mutt-kz June 2012

mutt-kz@lists.fedoraproject.org

9 participants
10 discussions

A bunch of fixes from debian
by David Sterba 26 Jul '12

26 Jul '12

Hi, I've set up my mutt-kz git tree, please find a few fixes taken from debian's mutt-patched package, all of them from category 'upstream', git://repo.or.cz/mutt-kz-devel.git debian-bugfixes Shortlog: debian/upstream/228671-pipe-mime.patch debian/upstream/383769-score-match.patch debian/upstream/528233-readonly-open.patch debian/upstream/531430-imapuser.patch debian/upstream/537061-dont-recode-saved-attachments.patch debian/upstream/537694-segv-imap-headers.patch debian/upstream/537818-emptycharset.patch debian/upstream/542817-smimekeys-tmpdir.patch debian/upstream/543467-thread-segfault.patch debian/upstream/547980-smime_keys-chaining.patch debian/upstream/548577-gpgme-1.2.patch debian/upstream/553321-ansi-escape-segfault.patch debian/upstream/568295-references.patch debian/upstream/603288-split-fetches.patch debian/upstream/608706-fix-spelling-errors.patch debian/upstream/611410-no-implicit_autoview-for-text-html.patch debian/upstream/619216-gnutls-CN-validation.patch debian/upstream/624058-gnutls-deprecated-set-priority.patch For completeness, these are already applied in current master: 578087-header-strchr.patch 584138-mx_update_context-segfault.patch 611412-bts-regexp.patch 620854-pop3-segfault.patch 624085-gnutls-deprecated-verify-peers.patch Next I'm preparing more sidebar fixes. david

2 2

[PATCH 1/2] Test for the latest notmuch API change
by David Riebenbauer 27 Jun '12

27 Jun '12

In notmuch 0.13 the library's API has changed, so we explicitly test for that with autoconf. --- configure.ac | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/configure.ac b/configure.ac index b6464da..3142ad5 100644 --- a/configure.ac +++ b/configure.ac @@ -177,6 +177,18 @@ AC_ARG_ENABLE(notmuch, AC_HELP_STRING([--enable-notmuch], [Enable NOTMUCH suppor NOTMUCH_LIBS="-lnotmuch" OPS="$OPS \$(srcdir)/OPS.NOTMUCH" need_notmuch="yes" + + AC_MSG_CHECKING([for notmuch api version 3]) + AC_COMPILE_IFELSE( [AC_LANG_PROGRAM( + [[#include <notmuch.h>]], + [[notmuch_database_open("/path", NOTMUCH_DATABASE_MODE_READ_ONLY, (notmuch_database_t**)NULL);]] + )], + [notmuch_api_3=yes + AC_DEFINE([NOTMUCH_API_3], 1, [Define to 1 if you have the notmuch api version 3.]) + ], + [notmuch_api_3=no] + ) + AC_MSG_RESULT([$notmuch_api_3]) fi ]) AM_CONDITIONAL(BUILD_NOTMUCH, test x$need_notmuch = xyes) -- 1.7.10

2 3

[PATCH] fix recent notmuch api change
by Chris Mason 27 Jun '12

27 Jun '12

Hi everyone, I'm not sure how to test for notmuch api versions, but this makes things work with the latest notmuch. Of course, it'll also break things for older notmuch... diff --git a/mutt_notmuch.c b/mutt_notmuch.c index 3a05a62..6a27a35 100644 --- a/mutt_notmuch.c +++ b/mutt_notmuch.c @@ -347,13 +347,14 @@ static notmuch_database_t *do_database_open(const char *filename, { notmuch_database_t *db = NULL; unsigned int ct = 0; + int status = 0; dprint(1, (debugfile, "nm: db open '%s' %s (timeout %d)\n", filename, writable ? "[WRITE]" : "[READ]", NotmuchOpenTimeout)); do { - db = notmuch_database_open(filename, + status = notmuch_database_open(filename, writable ? NOTMUCH_DATABASE_MODE_READ_WRITE : - NOTMUCH_DATABASE_MODE_READ_ONLY); + NOTMUCH_DATABASE_MODE_READ_ONLY, &db); if (db || !NotmuchOpenTimeout || ct / 2 > NotmuchOpenTimeout) break; @@ -364,11 +365,13 @@ static notmuch_database_t *do_database_open(const char *filename, } while (1); if (verbose) { - if (!db) + if (status) mutt_error (_("Cannot open notmuch database: %s"), filename); else if (ct > 1) mutt_clear_error(); } + if (status) + return NULL; return db; }

2 1

[PATCH] sidebar: add optional refresh interval
by David Sterba 27 Jun '12

27 Jun '12

New option sidebar_refresh: do not refresh sidebar in less than $sidebar_refresh seconds (0 disables refreshing). This helps when there are 'mbox' mailboxes and scanning for new mails blocks UI and consumes CPU, although it might not be desired to refresh. This de-syncs message counts listed in sidebar, can be updated by entering the mailbox. Signed-off-by: David Sterba <dsterba(a)suse.cz> --- I'm concerned about CPU consumption mainly because I read mail on a shared host and I usually know what mboxes I want to read and do not want to wait when toggling sidebar to easily pick mbox. buffy.c | 13 ++++++++++--- globals.h | 2 ++ init.h | 6 ++++++ sidebar.c | 15 +++++++++++++++ sidebar.h | 2 ++ 5 files changed, 35 insertions(+), 3 deletions(-) diff --git a/buffy.c b/buffy.c index 989ea10..fc8efa6 100644 --- a/buffy.c +++ b/buffy.c @@ -26,6 +26,7 @@ #include "mx.h" #include "mutt_curses.h" +#include "sidebar.h" #ifdef USE_IMAP #include "imap.h" @@ -543,22 +544,28 @@ static void buffy_check(BUFFY *tmp, struct stat *contex_sb) { case M_MBOX: case M_MMDF: - if (option(OPTSIDEBAR)) + if (sidebar_should_refresh()) { buffy_mbox_update (tmp); + sidebar_updated(); + } if (buffy_mbox_hasnew (tmp, &sb) > 0) BuffyCount++; break; case M_MAILDIR: - if (option(OPTSIDEBAR)) + if (sidebar_should_refresh()) { buffy_maildir_update (tmp); + sidebar_updated(); + } if (buffy_maildir_hasnew (tmp) > 0) BuffyCount++; break; case M_MH: - if (option(OPTSIDEBAR)) + if (sidebar_should_refresh()) { mh_buffy_update (tmp->path, &tmp->msgcount, &tmp->msg_unread, &tmp->msg_flagged); + sidebar_updated(); + } if ((tmp->new = mh_buffy (tmp->path)) > 0) BuffyCount++; break; diff --git a/globals.h b/globals.h --- a/globals.h +++ b/globals.h @@ -212,6 +212,8 @@ WHERE short ScoreThresholdFlag; WHERE struct buffy_t *CurBuffy INITVAL(0); WHERE short DrawFullLine INITVAL(0); WHERE short SidebarWidth; +WHERE short SidebarRefresh; +WHERE short SidebarLastRefresh; #ifdef USE_IMAP WHERE short ImapKeepalive; WHERE short ImapPipelineDepth; diff --git a/init.h b/init.h --- a/init.h +++ b/init.h @@ -1995,6 +1995,12 @@ struct option_t MuttVars[] = { { "sidebar_width", DT_NUM, R_BOTH, UL &SidebarWidth, 0 }, /* ** .pp + ** Do not refresh sidebar in less than $sidebar_refresh seconds, + ** (0 disables refreshing). + */ + { "sidebar_refresh", DT_NUM, R_BOTH, UL &SidebarRefresh, 60 }, + /* + ** .pp ** The width of the sidebar. */ { "pgp_use_gpg_agent", DT_BOOL, R_NONE, OPTUSEGPGAGENT, 0}, diff --git a/sidebar.c b/sidebar.c --- a/sidebar.c +++ b/sidebar.c @@ -199,6 +199,7 @@ int draw_sidebar(int menu) { saveSidebarWidth = SidebarWidth; if(!option(OPTSIDEBAR)) SidebarWidth = 0; initialized = true; + SidebarLastRefresh = time(NULL); } /* save or restore the value SidebarWidth */ @@ -434,3 +435,16 @@ void toggle_sidebar(int menu) set_curbuffy(""); /* default is the first mailbox */ draw_sidebar(menu); } + +int sidebar_should_refresh() +{ + if (option(OPTSIDEBAR) && SidebarRefresh > 0) { + if (time(NULL) - SidebarLastRefresh >= SidebarRefresh) + return 1; + } + return 0; +} +void sidebar_updated() +{ + SidebarLastRefresh = time(NULL); +} diff --git a/sidebar.h b/sidebar.h index 066b33b..94e08be 100644 --- a/sidebar.h +++ b/sidebar.h @@ -33,5 +33,7 @@ void scroll_sidebar(int, int); void set_curbuffy(char*); void set_buffystats(CONTEXT*); void toggle_sidebar(int menu); +int sidebar_should_refresh(); +void sidebar_updated(); #endif /* SIDEBAR_H */ -- 1.7.6.233.gd79bc

2 1

[PATCH] buildsys: add compiler flags enabling debugging symbols
by David Sterba 27 Jun '12

27 Jun '12

Automatically enabled when configured with --enable-debug Signed-off-by: David Sterba <dsterba(a)suse.cz> --- configure.ac | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/configure.ac b/configure.ac index b6464da..167034a 100644 --- a/configure.ac +++ b/configure.ac @@ -773,6 +773,7 @@ dnl -- end socket -- AC_ARG_ENABLE(debug, AC_HELP_STRING([--enable-debug], [Enable debugging support]), [ if test x$enableval = xyes ; then AC_DEFINE(DEBUG,1,[ Define to enable debugging info. ]) + CFLAGS="$CFLAGS -g" fi ]) -- 1.7.6.233.gd79bc

2 1

Yay! I can build now!
by Darryl L. Pierce 25 Jun '12

25 Jun '12

I've been pretty busy recently and haven't had time to play with the code or even build it. Now that I've tried, it's building for me and I'm ready to start putting some time into the code. Do we have a backlog of tasks to accomplish? And are we at a point where we can release a stable build to package for Fedora? -- Darryl L. Pierce, Sr. Software Engineer @ Red Hat, Inc. Delivering value year after year. Red Hat ranks #1 in value among software vendors. http://www.redhat.com/promo/vendor/

1 0

displaying full threads (even archived messages)
by Chris Mason 15 Jun '12

15 Jun '12

Hi everyone, The TODO list mentions a features I really wanted, which is making a way to display an entire thread, including archived messages. I'm not sure what the best way to wire this up is, but here is some sample code that replaces the current search with something that returns the entire thread. sup worked this way, but it was very difficult to use with big threads because it was almost impossible to find individual messages that you cared about. mutt has all the normal limiting/searching commands, so having a huge thread returned isn't a big deal. This patch is about 10 minutes old, treat it gently. It needs a conditional to switch between whole-thread matching and strict message matching, but I'll let people that know mutt internals better suggest how to do that. diff --git a/mutt_notmuch.c b/mutt_notmuch.c --- a/mutt_notmuch.c +++ b/mutt_notmuch.c @@ -699,9 +702,95 @@ done: FREE(&newpath); } +/* + * add all the replies to a given messages into the display. + * Careful, this calls itself recursively to make sure we get + * everything. + */ +static void append_replies(CONTEXT *ctx, notmuch_message_t *top) +{ + notmuch_messages_t *msgs; + for (msgs = notmuch_message_get_replies(top); + notmuch_messages_valid(msgs); + notmuch_messages_move_to_next(msgs)) { + notmuch_message_t *m = notmuch_messages_get(msgs); + append_message(ctx, m); + /* recurse through all the replies to this message too */ + append_replies(ctx, m); + notmuch_message_destroy(m); + } +} + +/* + * add each top level reply in the thread, and then add each + * reply to the top level replies + */ +static void append_thread(CONTEXT *ctx, notmuch_thread_t *thread) +{ + notmuch_messages_t *msgs; + for (msgs = notmuch_thread_get_toplevel_messages(thread); + notmuch_messages_valid(msgs); + notmuch_messages_move_to_next(msgs)) { + notmuch_message_t *m = notmuch_messages_get(msgs); + append_message(ctx, m); + append_replies(ctx, m); + notmuch_message_destroy(m); + } +} + +/* + * The default read_query function, this one finds all the + * matching threads and then iterates through them + */ int nm_read_query(CONTEXT *ctx) { notmuch_query_t *q; + notmuch_threads_t *threads; + struct nm_ctxdata *data; + int limit, rc = -1; + + if (init_context(ctx) != 0) + return -1; + + data = get_ctxdata(ctx); + if (!data) + return -1; + + dprint(1, (debugfile, "nm: reading messages...\n")); + + q = get_query(data, FALSE); + if (q) { + limit = get_limit(data); + + for (threads = notmuch_query_search_threads(q); + notmuch_threads_valid(threads) && + (limit == 0 || ctx->msgcount < limit); + notmuch_threads_move_to_next(threads)) { + + notmuch_thread_t *thread = notmuch_threads_get(threads); + append_thread(ctx, thread); + notmuch_thread_destroy(thread); + } + + notmuch_query_destroy(q); + rc = 0; + } + + if (!is_longrun(data)) + release_db(data); + + ctx->mtime = time(NULL); + + mx_update_context(ctx, ctx->msgcount); + + dprint(1, (debugfile, "nm: reading messages... done [rc=%d, count=%d]\n", + rc, ctx->msgcount)); + return rc; +} + +int nm_read_query_messages(CONTEXT *ctx) +{ + notmuch_query_t *q; notmuch_messages_t *msgs; struct nm_ctxdata *data; int limit, rc = -1;

2 1

Re: displaying full threads (even archived messages)
by Wesley Hearn 13 Jun '12

13 Jun '12

Chris Mason <chris.mason(a)fusionio.com> wrote: Hi everyone, The TODO list mentions a features I really wanted, which is making a way to display an entire thread, including archived messages. I'm not sure what the best way to wire this up is, but here is some sample code that replaces the current search with something that returns the entire thread. sup worked this way, but it was very difficult to use with big threads because it was almost impossible to find individual messages that you cared about. mutt has all the normal limiting/searching commands, so having a huge thread returned isn't a big deal. This patch is about 10 minutes old, treat it gently. It needs a conditional to switch between whole-thread matching and strict message matching, but I'll let people that know mutt internals better suggest how to do that. diff --git a/mutt_notmuch.c b/mutt_notmuch.c --- a/mutt_notmuch.c +++ b/mutt_notmuch.c @@ -699,9 +702,95 @@ done: FREE(&newpath); } +/* + * add all the replies to a given messages into the display. + * Careful, this calls itself recursively to make sure we get + * everything. + */ +static void append_replies(CONTEXT *ctx, notmuch_message_t *top) +{ + notmuch_messages_t *msgs; + for (msgs = notmuch_message_get_replies(top); + notmuch_messages_valid(msgs); + notmuch_messages_move_to_next(msgs)) { + notmuch_message_t *m = notmuch_messages_get(msgs); + append_message(ctx, m); + /* recurse through all the replies to this message too */ + append_replies(ctx, m); + notmuch_message_destroy(m); + } +} + +/* + * add each top level reply in the thread, and then add each + * reply to the top level replies + */ +static void append_thread(CONTEXT *ctx, notmuch_thread_t *thread) +{ + notmuch_messages_t *msgs; + for (msgs = notmuch_thread_get_toplevel_messages(thread); + notmuch_messages_valid(msgs); + notmuch_messages_move_to_next(msgs)) { + notmuch_message_t *m = notmuch_messages_get(msgs); + append_message(ctx, m); + append_replies(ctx, m); + notmuch_message_destroy(m); + } +} + +/* + * The default read_query function, this one finds all the + * matching threads and then iterates through them + */ int nm_read_query(CONTEXT *ctx) { notmuch_query_t *q; + notmuch_threads_t *threads; + struct nm_ctxdata *data; + int limit, rc = -1; + + if (init_context(ctx) != 0) + return -1; + + data = get_ctxdata(ctx); + if (!data) + return -1; + + dprint(1, (debugfile, "nm: reading messages...\n")); + + q = get_query(data, FALSE); + if (q) { + limit = get_limit(data); + + for (threads = notmuch_query_search_threads(q); + notmuch_threads_valid(threads) && + (limit == 0 || ctx->msgcount < limit); + notmuch_threads_move_to_next(threads)) { + + notmuch_thread_t *thread = notmuch_threads_get(threads); + append_thread(ctx, thread); + notmuch_thread_destroy(thread); + } + + notmuch_query_destroy(q); + rc = 0; + } + + if (!is_longrun(data)) + release_db(data); + + ctx->mtime = time(NULL); + + mx_update_context(ctx, ctx->msgcount); + + dprint(1, (debugfile, "nm: reading messages... done [rc=%d, count=%d]\n", + rc, ctx->msgcount)); + return rc; +} + +int nm_read_query_messages(CONTEXT *ctx) +{ + notmuch_query_t *q; notmuch_messages_t *msgs; struct nm_ctxdata *data; int limit, rc = -1; _______________________________________________ mutt-kz mailing list mutt-kz(a)lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/mutt-kz

1 0

whole thread archiving
by Chris Mason 13 Jun '12

13 Jun '12

Hi everyone, This muttrc line: macro index A "<modify-labels-then-hide>+archive -inbox\n<sync-mailbox>" Archives a message. Is there a way to do this for a whole thread? I know I can tag the thread and then archive the tagged messages, but I'm hoping for muttrc solution. -chris

3 2

Build Issues
by Andrew Davidson 11 Jun '12

11 Jun '12

I'm seeing some issues building mutt-kz. Here's a log excerpt: gcc -DPKGDATADIR=\"/usr/local/share/mutt\" -DSYSCONFDIR=\"/usr/local/etc\" -DBINDIR=\"/usr/local/bin\" -DMUTTLOCALEDIR=\"/usr/local/share/locale\" -DHAVE_CONFIG_H=1 -I. -I. -I. -Iintl -I./intl -Wall -pedantic -Wno-long-long -g -O2 -MT safe_asprintf.o -MD -MP -MF .deps/safe_asprintf.Tpo -c -o safe_asprintf.o safe_asprintf.c mv -f .deps/safe_asprintf.Tpo .deps/safe_asprintf.Po gcc -DPKGDATADIR=\"/usr/local/share/mutt\" -DSYSCONFDIR=\"/usr/local/etc\" -DBINDIR=\"/usr/local/bin\" -DMUTTLOCALEDIR=\"/usr/local/share/locale\" -DHAVE_CONFIG_H=1 -I. -I. -I. -Iintl -I./intl -Wall -pedantic -Wno-long-long -g -O2 -MT mutt_notmuch.o -MD -MP -MF .deps/mutt_notmuch.Tpo -c -o mutt_notmuch.o mutt_notmuch.c mutt_notmuch.c: In function 'do_database_open': mutt_notmuch.c:356:6: error: too few arguments to function 'notmuch_database_open' /usr/local/include/notmuch.h:200:1: note: declared here make[2]: *** [mutt_notmuch.o] Error 1 make[2]: Leaving directory `/home/amdavidson/src/mutt-kz' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/amdavidson/src/mutt-kz' make: *** [all] Error 2 Any tips? Thanks, -- Andrew Davidson

3 3

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

mutt-kz June 2012