2021-01-10 15:00:47 +01:00
|
|
|
;;; notmuch-wash.el --- cleaning up message bodies -*- lexical-binding: t -*-
|
2010-03-22 17:49:16 +01:00
|
|
|
;;
|
|
|
|
;; Copyright © Carl Worth
|
2010-04-26 15:45:30 +02:00
|
|
|
;; Copyright © David Edmondson
|
2010-03-22 17:49:16 +01:00
|
|
|
;;
|
|
|
|
;; This file is part of Notmuch.
|
|
|
|
;;
|
|
|
|
;; Notmuch is free software: you can redistribute it and/or modify it
|
|
|
|
;; under the terms of the GNU General Public License as published by
|
|
|
|
;; the Free Software Foundation, either version 3 of the License, or
|
|
|
|
;; (at your option) any later version.
|
|
|
|
;;
|
|
|
|
;; Notmuch is distributed in the hope that it will be useful, but
|
|
|
|
;; WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
;; General Public License for more details.
|
|
|
|
;;
|
|
|
|
;; You should have received a copy of the GNU General Public License
|
2016-06-02 18:26:14 +02:00
|
|
|
;; along with Notmuch. If not, see <https://www.gnu.org/licenses/>.
|
2010-03-22 17:49:16 +01:00
|
|
|
;;
|
|
|
|
;; Authors: Carl Worth <cworth@cworth.org>
|
2010-04-26 15:45:30 +02:00
|
|
|
;; David Edmondson <dme@dme.org>
|
|
|
|
|
2016-04-13 09:58:47 +02:00
|
|
|
;;; Code:
|
|
|
|
|
2010-04-26 15:45:30 +02:00
|
|
|
(require 'coolj)
|
2018-06-14 02:32:36 +02:00
|
|
|
(require 'notmuch-lib)
|
2020-08-08 13:49:36 +02:00
|
|
|
|
|
|
|
(declare-function notmuch-show-insert-bodypart "notmuch-show"
|
|
|
|
(msg part depth &optional hide))
|
2016-08-06 17:29:33 +02:00
|
|
|
(defvar notmuch-show-indent-messages-width)
|
2010-04-27 12:20:58 +02:00
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;; Options
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:20 +01:00
|
|
|
(defgroup notmuch-wash nil
|
|
|
|
"Cleaning up messages for display."
|
|
|
|
:group 'notmuch)
|
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-signature-regexp "^\\(-- ?\\|_+\\)$"
|
|
|
|
"Pattern to match a line that separates content from signature."
|
|
|
|
:type 'regexp
|
|
|
|
:group 'notmuch-wash)
|
|
|
|
|
|
|
|
(defcustom notmuch-wash-citation-regexp "\\(^[[:space:]]*>.*\n\\)+"
|
|
|
|
"Pattern to match citation lines."
|
|
|
|
:type 'regexp
|
|
|
|
:group 'notmuch-wash)
|
|
|
|
|
|
|
|
(defcustom notmuch-wash-original-regexp "^\\(--+\s?[oO]riginal [mM]essage\s?--+\\)$"
|
|
|
|
"Pattern to match a line that separates original message from
|
|
|
|
reply in top-posted message."
|
|
|
|
:type 'regexp
|
|
|
|
:group 'notmuch-wash)
|
|
|
|
|
|
|
|
(defcustom notmuch-wash-button-signature-hidden-format
|
2011-05-22 20:57:53 +02:00
|
|
|
"[ %d-line signature. Click/Enter to show. ]"
|
2010-03-22 17:49:16 +01:00
|
|
|
"String used to construct button text for hidden signatures.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-button-signature-visible-format
|
2011-05-22 20:57:53 +02:00
|
|
|
"[ %d-line signature. Click/Enter to hide. ]"
|
|
|
|
"String used to construct button text for visible signatures.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2011-05-22 20:57:53 +02:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-button-citation-hidden-format
|
2011-05-22 20:57:53 +02:00
|
|
|
"[ %d more citation lines. Click/Enter to show. ]"
|
2010-03-22 17:49:16 +01:00
|
|
|
"String used to construct button text for hidden citations.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-button-citation-visible-format
|
2011-05-22 20:57:53 +02:00
|
|
|
"[ %d more citation lines. Click/Enter to hide. ]"
|
|
|
|
"String used to construct button text for visible citations.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2011-05-22 20:57:53 +02:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-button-original-hidden-format
|
2010-02-24 19:30:06 +01:00
|
|
|
"[ %d-line hidden original message. Click/Enter to show. ]"
|
|
|
|
"String used to construct button text for hidden citations.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2010-02-24 19:30:06 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-button-original-visible-format
|
2010-02-24 19:30:06 +01:00
|
|
|
"[ %d-line original message. Click/Enter to hide. ]"
|
|
|
|
"String used to construct button text for visible citations.
|
2014-10-31 09:05:21 +01:00
|
|
|
Can use up to one integer format parameter, i.e. %d."
|
|
|
|
:type 'string
|
|
|
|
:group 'notmuch-wash)
|
2010-02-24 19:30:06 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-signature-lines-max 12
|
|
|
|
"Maximum length of signature that will be hidden by default."
|
|
|
|
:type 'integer
|
|
|
|
:group 'notmuch-wash)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-citation-lines-prefix 3
|
2010-03-22 17:49:16 +01:00
|
|
|
"Always show at least this many lines from the start of a citation.
|
|
|
|
|
|
|
|
If there is one more line than the sum of
|
|
|
|
`notmuch-wash-citation-lines-prefix' and
|
|
|
|
`notmuch-wash-citation-lines-suffix', show that, otherwise
|
2014-10-31 09:05:21 +01:00
|
|
|
collapse the remaining lines into a button."
|
|
|
|
:type 'integer
|
|
|
|
:group 'notmuch-wash)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-citation-lines-suffix 3
|
2010-03-22 17:49:16 +01:00
|
|
|
"Always show at least this many lines from the end of a citation.
|
|
|
|
|
|
|
|
If there is one more line than the sum of
|
|
|
|
`notmuch-wash-citation-lines-prefix' and
|
|
|
|
`notmuch-wash-citation-lines-suffix', show that, otherwise
|
2014-10-31 09:05:21 +01:00
|
|
|
collapse the remaining lines into a button."
|
|
|
|
:type 'integer
|
|
|
|
:group 'notmuch-wash)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2014-10-31 09:05:21 +01:00
|
|
|
(defcustom notmuch-wash-wrap-lines-length nil
|
2012-02-17 19:34:44 +01:00
|
|
|
"Wrap line after at most this many characters.
|
|
|
|
|
|
|
|
If this is nil, lines in messages will be wrapped to fit in the
|
|
|
|
current window. If this is a number, lines will be wrapped after
|
2016-08-06 17:29:34 +02:00
|
|
|
this many characters (ignoring indentation due to thread depth)
|
|
|
|
or at the window width (whichever one is lower)."
|
2014-10-31 09:05:21 +01:00
|
|
|
:type '(choice (const :tag "window width" nil)
|
|
|
|
(integer :tag "number of characters"))
|
|
|
|
:group 'notmuch-wash)
|
2012-02-17 19:34:44 +01:00
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;; Faces
|
|
|
|
|
2014-10-31 09:05:20 +01:00
|
|
|
(defface notmuch-wash-toggle-button
|
|
|
|
'((t (:inherit font-lock-comment-face)))
|
|
|
|
"Face used for buttons toggling the visibility of washed away
|
|
|
|
message parts."
|
|
|
|
:group 'notmuch-wash
|
|
|
|
:group 'notmuch-faces)
|
|
|
|
|
|
|
|
(defface notmuch-wash-cited-text
|
|
|
|
'((t (:inherit message-cited-text)))
|
|
|
|
"Face used for cited text."
|
|
|
|
:group 'notmuch-wash
|
|
|
|
:group 'notmuch-faces)
|
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;; Buttons
|
|
|
|
|
2010-03-22 17:49:16 +01:00
|
|
|
(defun notmuch-wash-toggle-invisible-action (cite-button)
|
emacs: Eliminate buffer invisibility specs from show and wash
Previously, all visibility in show buffers for headers, message
bodies, and washed text was specified by generating one or more
symbols for each region and creating overlays with their 'invisible
property set to carefully crafted combinations of these symbols.
Visibility was controlled not by modifying the overlays directly, but
by adding and removing the generated symbols from a gigantic buffer
invisibilty spec.
This has myriad negative consequences. It's slow because Emacs'
display engine has to traverse the buffer invisibility list for every
overlay and, since every overlay has its own symbol, this makes
rendering O(N^2) in the number of overlays. It composes poorly
because symbol-type 'invisible properties are taken from the highest
priority overlay over a given character (which is often ambiguous!),
rather than being gathered from all overlays over a character. As a
result, we have to include symbols related to message hiding in the
wash code lest the wash overlays un-hide parts of hidden messages. It
also requires various workarounds for isearch to properly open
overlays, to set up buffer-invisibility-spec for
remove-from-invisibility-spec to work right, and to explicitly refresh
the display after updating the buffer invisibility spec.
None of this is necessary.
This patch converts show and wash to use simple boolean 'invisible
properties and to not use the buffer invisibility spec. Rather than
adding and removing generated symbols from the invisibility spec, the
code now directly toggles the 'invisible property of the appropriate
overlay. This speeds up rendering because the display engine only has
to check the boolean values of the overlays over a character. It
composes nicely because text will be invisible if *any* overlay over
it has 'invisible t, which means we can overlap invisibility overlays
with abandon. We no longer need any of the workarounds mentioned
above. And it fixes a minor bug for free: now, when isearch opens a
washed region, the button text will update to say "Click/Enter to
hide" rather than remaining unchanged.
2012-12-18 07:40:10 +01:00
|
|
|
;; Toggle overlay visibility
|
|
|
|
(let ((overlay (button-get cite-button 'overlay)))
|
|
|
|
(overlay-put overlay 'invisible (not (overlay-get overlay 'invisible))))
|
|
|
|
;; Update button text
|
2011-05-22 20:57:53 +02:00
|
|
|
(let* ((new-start (button-start cite-button))
|
|
|
|
(overlay (button-get cite-button 'overlay))
|
|
|
|
(button-label (notmuch-wash-button-label overlay))
|
2011-05-25 01:02:43 +02:00
|
|
|
(old-point (point))
|
2013-05-30 03:13:44 +02:00
|
|
|
(properties (text-properties-at (point)))
|
2011-05-22 20:57:53 +02:00
|
|
|
(inhibit-read-only t))
|
2011-05-25 01:02:43 +02:00
|
|
|
(goto-char new-start)
|
|
|
|
(insert button-label)
|
2013-05-30 03:13:44 +02:00
|
|
|
(set-text-properties new-start (point) properties)
|
2011-05-25 01:02:43 +02:00
|
|
|
(let ((old-end (button-end cite-button)))
|
|
|
|
(move-overlay cite-button new-start (point))
|
|
|
|
(delete-region (point) old-end))
|
emacs: Eliminate buffer invisibility specs from show and wash
Previously, all visibility in show buffers for headers, message
bodies, and washed text was specified by generating one or more
symbols for each region and creating overlays with their 'invisible
property set to carefully crafted combinations of these symbols.
Visibility was controlled not by modifying the overlays directly, but
by adding and removing the generated symbols from a gigantic buffer
invisibilty spec.
This has myriad negative consequences. It's slow because Emacs'
display engine has to traverse the buffer invisibility list for every
overlay and, since every overlay has its own symbol, this makes
rendering O(N^2) in the number of overlays. It composes poorly
because symbol-type 'invisible properties are taken from the highest
priority overlay over a given character (which is often ambiguous!),
rather than being gathered from all overlays over a character. As a
result, we have to include symbols related to message hiding in the
wash code lest the wash overlays un-hide parts of hidden messages. It
also requires various workarounds for isearch to properly open
overlays, to set up buffer-invisibility-spec for
remove-from-invisibility-spec to work right, and to explicitly refresh
the display after updating the buffer invisibility spec.
None of this is necessary.
This patch converts show and wash to use simple boolean 'invisible
properties and to not use the buffer invisibility spec. Rather than
adding and removing generated symbols from the invisibility spec, the
code now directly toggles the 'invisible property of the appropriate
overlay. This speeds up rendering because the display engine only has
to check the boolean values of the overlays over a character. It
composes nicely because text will be invisible if *any* overlay over
it has 'invisible t, which means we can overlap invisibility overlays
with abandon. We no longer need any of the workarounds mentioned
above. And it fixes a minor bug for free: now, when isearch opens a
washed region, the button text will update to say "Click/Enter to
hide" rather than remaining unchanged.
2012-12-18 07:40:10 +01:00
|
|
|
(goto-char (min old-point (1- (button-end cite-button))))))
|
2010-03-22 17:49:16 +01:00
|
|
|
|
|
|
|
(define-button-type 'notmuch-wash-button-invisibility-toggle-type
|
|
|
|
'action 'notmuch-wash-toggle-invisible-action
|
|
|
|
'follow-link t
|
2014-10-31 09:05:20 +01:00
|
|
|
'face 'notmuch-wash-toggle-button
|
2013-01-07 22:07:20 +01:00
|
|
|
:supertype 'notmuch-button-type)
|
2010-03-22 17:49:16 +01:00
|
|
|
|
|
|
|
(define-button-type 'notmuch-wash-button-citation-toggle-type
|
|
|
|
'help-echo "mouse-1, RET: Show citation"
|
|
|
|
:supertype 'notmuch-wash-button-invisibility-toggle-type)
|
|
|
|
|
|
|
|
(define-button-type 'notmuch-wash-button-signature-toggle-type
|
|
|
|
'help-echo "mouse-1, RET: Show signature"
|
|
|
|
:supertype 'notmuch-wash-button-invisibility-toggle-type)
|
|
|
|
|
2010-02-24 19:30:06 +01:00
|
|
|
(define-button-type 'notmuch-wash-button-original-toggle-type
|
|
|
|
'help-echo "mouse-1, RET: Show original message"
|
|
|
|
:supertype 'notmuch-wash-button-invisibility-toggle-type)
|
|
|
|
|
2010-04-23 19:39:33 +02:00
|
|
|
(defun notmuch-wash-region-isearch-show (overlay)
|
emacs: Eliminate buffer invisibility specs from show and wash
Previously, all visibility in show buffers for headers, message
bodies, and washed text was specified by generating one or more
symbols for each region and creating overlays with their 'invisible
property set to carefully crafted combinations of these symbols.
Visibility was controlled not by modifying the overlays directly, but
by adding and removing the generated symbols from a gigantic buffer
invisibilty spec.
This has myriad negative consequences. It's slow because Emacs'
display engine has to traverse the buffer invisibility list for every
overlay and, since every overlay has its own symbol, this makes
rendering O(N^2) in the number of overlays. It composes poorly
because symbol-type 'invisible properties are taken from the highest
priority overlay over a given character (which is often ambiguous!),
rather than being gathered from all overlays over a character. As a
result, we have to include symbols related to message hiding in the
wash code lest the wash overlays un-hide parts of hidden messages. It
also requires various workarounds for isearch to properly open
overlays, to set up buffer-invisibility-spec for
remove-from-invisibility-spec to work right, and to explicitly refresh
the display after updating the buffer invisibility spec.
None of this is necessary.
This patch converts show and wash to use simple boolean 'invisible
properties and to not use the buffer invisibility spec. Rather than
adding and removing generated symbols from the invisibility spec, the
code now directly toggles the 'invisible property of the appropriate
overlay. This speeds up rendering because the display engine only has
to check the boolean values of the overlays over a character. It
composes nicely because text will be invisible if *any* overlay over
it has 'invisible t, which means we can overlap invisibility overlays
with abandon. We no longer need any of the workarounds mentioned
above. And it fixes a minor bug for free: now, when isearch opens a
washed region, the button text will update to say "Click/Enter to
hide" rather than remaining unchanged.
2012-12-18 07:40:10 +01:00
|
|
|
(notmuch-wash-toggle-invisible-action
|
|
|
|
(overlay-get overlay 'notmuch-wash-button)))
|
2010-04-23 19:39:33 +02:00
|
|
|
|
2011-05-22 20:57:53 +02:00
|
|
|
(defun notmuch-wash-button-label (overlay)
|
|
|
|
(let* ((type (overlay-get overlay 'type))
|
|
|
|
(invis-spec (overlay-get overlay 'invisible))
|
|
|
|
(state (if (invisible-p invis-spec) "hidden" "visible"))
|
2020-08-08 13:49:36 +02:00
|
|
|
(label-format (symbol-value
|
|
|
|
(intern-soft
|
|
|
|
(format "notmuch-wash-button-%s-%s-format"
|
|
|
|
type state))))
|
|
|
|
(lines-count (count-lines (overlay-start overlay)
|
|
|
|
(overlay-end overlay))))
|
2011-05-22 20:57:53 +02:00
|
|
|
(format label-format lines-count)))
|
|
|
|
|
2021-01-10 15:00:51 +01:00
|
|
|
(defun notmuch-wash-region-to-button (beg end type &optional prefix)
|
2020-05-24 21:17:56 +02:00
|
|
|
"Auxiliary function to do the actual making of overlays and buttons.
|
2010-03-22 17:49:16 +01:00
|
|
|
|
|
|
|
BEG and END are buffer locations. TYPE should a string, either
|
2012-02-04 08:36:37 +01:00
|
|
|
\"citation\" or \"signature\". Optional PREFIX is some arbitrary
|
|
|
|
text to insert before the button, probably for indentation. Note
|
|
|
|
that PREFIX should not include a newline."
|
2010-03-22 17:49:16 +01:00
|
|
|
;; This uses some slightly tricky conversions between strings and
|
|
|
|
;; symbols because of the way the button code works. Note that
|
|
|
|
;; replacing intern-soft with make-symbol will cause this to fail,
|
|
|
|
;; since the newly created symbol has no plist.
|
|
|
|
(let ((overlay (make-overlay beg end))
|
|
|
|
(button-type (intern-soft (concat "notmuch-wash-button-"
|
|
|
|
type "-toggle-type"))))
|
emacs: Eliminate buffer invisibility specs from show and wash
Previously, all visibility in show buffers for headers, message
bodies, and washed text was specified by generating one or more
symbols for each region and creating overlays with their 'invisible
property set to carefully crafted combinations of these symbols.
Visibility was controlled not by modifying the overlays directly, but
by adding and removing the generated symbols from a gigantic buffer
invisibilty spec.
This has myriad negative consequences. It's slow because Emacs'
display engine has to traverse the buffer invisibility list for every
overlay and, since every overlay has its own symbol, this makes
rendering O(N^2) in the number of overlays. It composes poorly
because symbol-type 'invisible properties are taken from the highest
priority overlay over a given character (which is often ambiguous!),
rather than being gathered from all overlays over a character. As a
result, we have to include symbols related to message hiding in the
wash code lest the wash overlays un-hide parts of hidden messages. It
also requires various workarounds for isearch to properly open
overlays, to set up buffer-invisibility-spec for
remove-from-invisibility-spec to work right, and to explicitly refresh
the display after updating the buffer invisibility spec.
None of this is necessary.
This patch converts show and wash to use simple boolean 'invisible
properties and to not use the buffer invisibility spec. Rather than
adding and removing generated symbols from the invisibility spec, the
code now directly toggles the 'invisible property of the appropriate
overlay. This speeds up rendering because the display engine only has
to check the boolean values of the overlays over a character. It
composes nicely because text will be invisible if *any* overlay over
it has 'invisible t, which means we can overlap invisibility overlays
with abandon. We no longer need any of the workarounds mentioned
above. And it fixes a minor bug for free: now, when isearch opens a
washed region, the button text will update to say "Click/Enter to
hide" rather than remaining unchanged.
2012-12-18 07:40:10 +01:00
|
|
|
(overlay-put overlay 'invisible t)
|
2010-04-23 19:39:33 +02:00
|
|
|
(overlay-put overlay 'isearch-open-invisible #'notmuch-wash-region-isearch-show)
|
2011-05-22 20:57:53 +02:00
|
|
|
(overlay-put overlay 'type type)
|
2010-03-22 17:49:16 +01:00
|
|
|
(goto-char (1+ end))
|
|
|
|
(save-excursion
|
2012-02-04 08:36:37 +01:00
|
|
|
(goto-char beg)
|
2020-08-08 13:49:43 +02:00
|
|
|
(when prefix
|
|
|
|
(insert-before-markers prefix))
|
2012-02-04 08:36:37 +01:00
|
|
|
(let ((button-beg (point)))
|
|
|
|
(insert-before-markers (notmuch-wash-button-label overlay) "\n")
|
emacs: Eliminate buffer invisibility specs from show and wash
Previously, all visibility in show buffers for headers, message
bodies, and washed text was specified by generating one or more
symbols for each region and creating overlays with their 'invisible
property set to carefully crafted combinations of these symbols.
Visibility was controlled not by modifying the overlays directly, but
by adding and removing the generated symbols from a gigantic buffer
invisibilty spec.
This has myriad negative consequences. It's slow because Emacs'
display engine has to traverse the buffer invisibility list for every
overlay and, since every overlay has its own symbol, this makes
rendering O(N^2) in the number of overlays. It composes poorly
because symbol-type 'invisible properties are taken from the highest
priority overlay over a given character (which is often ambiguous!),
rather than being gathered from all overlays over a character. As a
result, we have to include symbols related to message hiding in the
wash code lest the wash overlays un-hide parts of hidden messages. It
also requires various workarounds for isearch to properly open
overlays, to set up buffer-invisibility-spec for
remove-from-invisibility-spec to work right, and to explicitly refresh
the display after updating the buffer invisibility spec.
None of this is necessary.
This patch converts show and wash to use simple boolean 'invisible
properties and to not use the buffer invisibility spec. Rather than
adding and removing generated symbols from the invisibility spec, the
code now directly toggles the 'invisible property of the appropriate
overlay. This speeds up rendering because the display engine only has
to check the boolean values of the overlays over a character. It
composes nicely because text will be invisible if *any* overlay over
it has 'invisible t, which means we can overlap invisibility overlays
with abandon. We no longer need any of the workarounds mentioned
above. And it fixes a minor bug for free: now, when isearch opens a
washed region, the button text will update to say "Click/Enter to
hide" rather than remaining unchanged.
2012-12-18 07:40:10 +01:00
|
|
|
(let ((button (make-button button-beg (1- (point))
|
|
|
|
'overlay overlay
|
|
|
|
:type button-type)))
|
|
|
|
(overlay-put overlay 'notmuch-wash-button button))))))
|
2010-03-22 17:49:16 +01:00
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;; Hook functions
|
|
|
|
|
2021-01-10 15:00:52 +01:00
|
|
|
(defun notmuch-wash-excerpt-citations (_msg _depth)
|
2010-04-26 15:45:30 +02:00
|
|
|
"Excerpt citations and up to one signature."
|
2010-03-22 17:49:16 +01:00
|
|
|
(goto-char (point-min))
|
|
|
|
(beginning-of-line)
|
2020-08-08 13:49:43 +02:00
|
|
|
(when (and (< (point) (point-max))
|
|
|
|
(re-search-forward notmuch-wash-original-regexp nil t))
|
2021-01-10 15:00:51 +01:00
|
|
|
(notmuch-wash-region-to-button (match-beginning 0)
|
2021-01-10 15:00:48 +01:00
|
|
|
(point-max)
|
|
|
|
"original"))
|
2010-03-22 17:49:16 +01:00
|
|
|
(while (and (< (point) (point-max))
|
|
|
|
(re-search-forward notmuch-wash-citation-regexp nil t))
|
|
|
|
(let* ((cite-start (match-beginning 0))
|
|
|
|
(cite-end (match-end 0))
|
|
|
|
(cite-lines (count-lines cite-start cite-end)))
|
2020-08-08 13:49:36 +02:00
|
|
|
(overlay-put (make-overlay cite-start cite-end)
|
|
|
|
'face 'notmuch-wash-cited-text)
|
2010-03-22 17:49:16 +01:00
|
|
|
(when (> cite-lines (+ notmuch-wash-citation-lines-prefix
|
|
|
|
notmuch-wash-citation-lines-suffix
|
|
|
|
1))
|
|
|
|
(goto-char cite-start)
|
|
|
|
(forward-line notmuch-wash-citation-lines-prefix)
|
|
|
|
(let ((hidden-start (point-marker)))
|
|
|
|
(goto-char cite-end)
|
|
|
|
(forward-line (- notmuch-wash-citation-lines-suffix))
|
|
|
|
(notmuch-wash-region-to-button
|
2021-01-10 15:00:51 +01:00
|
|
|
hidden-start (point-marker)
|
2012-02-04 08:36:37 +01:00
|
|
|
"citation")))))
|
2020-08-08 13:49:43 +02:00
|
|
|
(when (and (not (eobp))
|
|
|
|
(re-search-forward notmuch-wash-signature-regexp nil t))
|
2021-01-10 15:00:48 +01:00
|
|
|
(let ((sig-start (match-beginning 0)))
|
|
|
|
(when (<= (count-lines sig-start (point-max))
|
|
|
|
notmuch-wash-signature-lines-max)
|
2020-08-08 13:49:43 +02:00
|
|
|
(let ((sig-start-marker (make-marker))
|
|
|
|
(sig-end-marker (make-marker)))
|
|
|
|
(set-marker sig-start-marker sig-start)
|
|
|
|
(set-marker sig-end-marker (point-max))
|
|
|
|
(overlay-put (make-overlay sig-start-marker sig-end-marker)
|
|
|
|
'face 'message-cited-text)
|
|
|
|
(notmuch-wash-region-to-button
|
2021-01-10 15:00:51 +01:00
|
|
|
sig-start-marker sig-end-marker
|
2020-08-08 13:49:43 +02:00
|
|
|
"signature"))))))
|
2010-03-22 17:49:16 +01:00
|
|
|
|
2021-01-10 15:00:52 +01:00
|
|
|
(defun notmuch-wash-elide-blank-lines (_msg _depth)
|
2010-04-26 15:45:30 +02:00
|
|
|
"Elide leading, trailing and successive blank lines."
|
|
|
|
;; Algorithm derived from `article-strip-multiple-blank-lines' in
|
|
|
|
;; `gnus-art.el'.
|
|
|
|
;; Make all blank lines empty.
|
|
|
|
(goto-char (point-min))
|
|
|
|
(while (re-search-forward "^[[:space:]\t]+$" nil t)
|
|
|
|
(replace-match "" nil t))
|
|
|
|
;; Replace multiple empty lines with a single empty line.
|
|
|
|
(goto-char (point-min))
|
|
|
|
(while (re-search-forward "^\n\\(\n+\\)" nil t)
|
|
|
|
(delete-region (match-beginning 1) (match-end 1)))
|
|
|
|
;; Remove a leading blank line.
|
|
|
|
(goto-char (point-min))
|
2020-08-08 13:49:43 +02:00
|
|
|
(when (looking-at "\n")
|
|
|
|
(delete-region (match-beginning 0) (match-end 0)))
|
2010-04-26 15:45:30 +02:00
|
|
|
;; Remove a trailing blank line.
|
|
|
|
(goto-char (point-max))
|
2020-08-08 13:49:43 +02:00
|
|
|
(when (looking-at "\n")
|
|
|
|
(delete-region (match-beginning 0) (match-end 0))))
|
2010-04-26 15:45:30 +02:00
|
|
|
|
2021-01-10 15:00:52 +01:00
|
|
|
(defun notmuch-wash-tidy-citations (_msg _depth)
|
2010-04-26 15:45:30 +02:00
|
|
|
"Improve the display of cited regions of a message.
|
|
|
|
|
2010-11-17 14:32:33 +01:00
|
|
|
Perform several transformations on the message body:
|
2010-04-26 15:45:30 +02:00
|
|
|
|
|
|
|
- Remove lines of repeated citation leaders with no other
|
|
|
|
content,
|
|
|
|
- Remove citation leaders standing alone before a block of cited
|
|
|
|
text,
|
|
|
|
- Remove citation trailers standing alone after a block of cited
|
|
|
|
text."
|
|
|
|
;; Remove lines of repeated citation leaders with no other content.
|
|
|
|
(goto-char (point-min))
|
|
|
|
(while (re-search-forward "\\(^>[> ]*\n\\)\\{2,\\}" nil t)
|
|
|
|
(replace-match "\\1"))
|
2020-08-08 13:49:37 +02:00
|
|
|
;; Remove citation leaders standing alone before a block of cited text.
|
2010-04-26 15:45:30 +02:00
|
|
|
(goto-char (point-min))
|
|
|
|
(while (re-search-forward "\\(\n\\|^[^>].*\\)\n\\(^>[> ]*\n\\)" nil t)
|
|
|
|
(replace-match "\\1\n"))
|
2020-08-08 13:49:37 +02:00
|
|
|
;; Remove citation trailers standing alone after a block of cited text.
|
2010-04-26 15:45:30 +02:00
|
|
|
(goto-char (point-min))
|
|
|
|
(while (re-search-forward "\\(^>[> ]*\n\\)\\(^$\\|^[^>].*\\)" nil t)
|
2010-11-17 14:32:33 +01:00
|
|
|
(replace-match "\\2")))
|
2010-04-26 15:45:30 +02:00
|
|
|
|
2021-01-10 15:00:52 +01:00
|
|
|
(defun notmuch-wash-wrap-long-lines (_msg depth)
|
2012-02-17 19:34:44 +01:00
|
|
|
"Wrap long lines in the message.
|
|
|
|
|
|
|
|
If `notmuch-wash-wrap-lines-length' is a number, this will wrap
|
|
|
|
the message lines to the minimum of the width of the window or
|
|
|
|
its value. Otherwise, this function will wrap long lines in the
|
|
|
|
message at the window width. When doing so, citation leaders in
|
|
|
|
the wrapped text are maintained."
|
|
|
|
(let* ((coolj-wrap-follows-window-size nil)
|
2016-08-06 17:29:33 +02:00
|
|
|
(indent (* depth notmuch-show-indent-messages-width))
|
2012-02-17 19:34:44 +01:00
|
|
|
(limit (if (numberp notmuch-wash-wrap-lines-length)
|
2016-08-06 17:29:34 +02:00
|
|
|
(min (+ notmuch-wash-wrap-lines-length indent)
|
2012-02-17 19:34:44 +01:00
|
|
|
(window-width))
|
|
|
|
(window-width)))
|
|
|
|
(fill-column (- limit
|
2016-08-06 17:29:33 +02:00
|
|
|
indent
|
2012-02-17 19:34:44 +01:00
|
|
|
;; 2 to avoid poor interaction with
|
|
|
|
;; `word-wrap'.
|
|
|
|
2)))
|
2010-04-26 15:45:30 +02:00
|
|
|
(coolj-wrap-region (point-min) (point-max))))
|
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;;; Convert Inline Patches
|
2010-04-26 15:45:30 +02:00
|
|
|
|
2010-04-27 12:20:58 +02:00
|
|
|
(require 'diff-mode)
|
|
|
|
|
|
|
|
(defvar diff-file-header-re) ; From `diff-mode.el'.
|
|
|
|
|
2011-12-27 17:04:39 +01:00
|
|
|
(defun notmuch-wash-subject-to-filename (subject &optional maxlen)
|
|
|
|
"Convert a mail SUBJECT into a filename.
|
|
|
|
|
|
|
|
The resulting filename is similar to the names generated by \"git
|
|
|
|
format-patch\", without the leading patch sequence number
|
|
|
|
\"0001-\" and \".patch\" extension. Any leading \"[PREFIX]\"
|
|
|
|
style strings are removed prior to conversion.
|
|
|
|
|
|
|
|
Optional argument MAXLEN is the maximum length of the resulting
|
|
|
|
filename, before trimming any trailing . and - characters."
|
|
|
|
(let* ((s (replace-regexp-in-string "^ *\\(\\[[^]]*\\] *\\)*" "" subject))
|
|
|
|
(s (replace-regexp-in-string "[^A-Za-z0-9._]+" "-" s))
|
|
|
|
(s (replace-regexp-in-string "\\.+" "." s))
|
|
|
|
(s (if maxlen (substring s 0 (min (length s) maxlen)) s))
|
|
|
|
(s (replace-regexp-in-string "[.-]*$" "" s)))
|
|
|
|
s))
|
|
|
|
|
|
|
|
(defun notmuch-wash-subject-to-patch-sequence-number (subject)
|
|
|
|
"Convert a patch mail SUBJECT into a patch sequence number.
|
|
|
|
|
|
|
|
Return the patch sequence number N from the last \"[PATCH N/M]\"
|
|
|
|
style prefix in SUBJECT, or nil if such a prefix can't be found."
|
2020-08-08 13:49:41 +02:00
|
|
|
(and (string-match
|
|
|
|
"^ *\\(\\[[^]]*\\] *\\)*\\[[^]]*?\\([0-9]+\\)/[0-9]+[^]]*\\].*"
|
|
|
|
subject)
|
|
|
|
(string-to-number (substring subject (match-beginning 2) (match-end 2)))))
|
2011-12-27 17:04:39 +01:00
|
|
|
|
|
|
|
(defun notmuch-wash-subject-to-patch-filename (subject)
|
|
|
|
"Convert a patch mail SUBJECT into a filename.
|
|
|
|
|
|
|
|
The resulting filename is similar to the names generated by \"git
|
|
|
|
format-patch\". If the patch mail was generated and sent using
|
|
|
|
\"git format-patch/send-email\", this should re-create the
|
|
|
|
original filename the sender had."
|
|
|
|
(format "%04d-%s.patch"
|
|
|
|
(or (notmuch-wash-subject-to-patch-sequence-number subject) 1)
|
|
|
|
(notmuch-wash-subject-to-filename subject 52)))
|
|
|
|
|
2011-05-26 00:10:12 +02:00
|
|
|
(defun notmuch-wash-convert-inline-patch-to-part (msg depth)
|
2010-04-27 12:20:58 +02:00
|
|
|
"Convert an inline patch into a fake 'text/x-diff' attachment.
|
|
|
|
|
|
|
|
Given that this function guesses whether a buffer includes a
|
|
|
|
patch and then guesses the extent of the patch, there is scope
|
|
|
|
for error."
|
|
|
|
(goto-char (point-min))
|
2012-02-01 14:50:00 +01:00
|
|
|
(when (re-search-forward diff-file-header-re nil t)
|
|
|
|
(beginning-of-line -1)
|
|
|
|
(let ((patch-start (point))
|
|
|
|
(patch-end (point-max))
|
|
|
|
part)
|
|
|
|
(goto-char patch-start)
|
2020-08-08 13:49:43 +02:00
|
|
|
(when (or
|
|
|
|
;; Patch ends with signature.
|
|
|
|
(re-search-forward notmuch-wash-signature-regexp nil t)
|
|
|
|
;; Patch ends with bugtraq comment.
|
|
|
|
(re-search-forward "^\\*\\*\\* " nil t))
|
|
|
|
(setq patch-end (match-beginning 0)))
|
2012-02-01 14:50:00 +01:00
|
|
|
(save-restriction
|
|
|
|
(narrow-to-region patch-start patch-end)
|
2013-05-04 15:01:15 +02:00
|
|
|
(setq part (plist-put part :content-type "inline patch"))
|
2012-02-01 14:50:00 +01:00
|
|
|
(setq part (plist-put part :content (buffer-string)))
|
|
|
|
(setq part (plist-put part :id -1))
|
|
|
|
(setq part (plist-put part :filename
|
|
|
|
(notmuch-wash-subject-to-patch-filename
|
|
|
|
(plist-get
|
|
|
|
(plist-get msg :headers) :Subject))))
|
|
|
|
(delete-region (point-min) (point-max))
|
|
|
|
(notmuch-show-insert-bodypart nil part depth)))))
|
2010-04-27 12:20:58 +02:00
|
|
|
|
emacs: make headings outline-minor-mode compatible
`outline-minor-mode' treats comments that begin with three or more
semicolons as headings. That makes it very convenient to navigate
code and to show/hide parts of a file.
Elips libraries typically have four top-level sections, e.g.:
;;; notmuch.el --- run notmuch within emacs...
;;; Commentary:...
;;; Code:...
;;; notmuch.el ends here
In this package many libraries lack a "Commentary:" section, which is
not optimal but okay for most libraries, except major entry points.
Depending on how one chooses to look at it, the "... ends here" line
is not really a heading that begins a section, because it should never
have a "section" body (after all it marks eof).
If the file is rather short, then I left "Code:" as the only section
that contains code. Otherwise I split the file into multiple sibling
sections. The "Code:" section continues to contain `require' and
`declare-function' forms and other such "front matter".
If and only if I have split the code into multiple sections anyway,
then I also added an additional section named just "_" before the
`provide' form and shortly before the "...end here" line. This
section could also be called "Back matter", but I feel it would be
distracting to be that explicit about it. (The IMO unnecessary but
unfortunately still obligatory "... ends here" line is already
distracting enough as far as I am concerned.)
Before this commit some libraries already uses section headings, some
of them consistently. When a library already had some headings, then
this commit often sticks to that style, even at the cost inconsistent
styling across all libraries.
A very limited number of variable and function definitions have to be
moved around because they would otherwise end up in sections they do
not belong into.
Sections, including but not limited to their heading, can and should
be further improved in the future.
2021-01-10 15:00:46 +01:00
|
|
|
;;; _
|
2010-04-27 12:20:58 +02:00
|
|
|
|
2010-03-22 17:49:16 +01:00
|
|
|
(provide 'notmuch-wash)
|
2016-04-13 09:58:47 +02:00
|
|
|
|
|
|
|
;;; notmuch-wash.el ends here
|