emacs-orgmode@gnu.org archives
 help / color / mirror / code / Atom feed
From: Aaron Ecay <aaronecay@gmail.com>
To: Nicolas Goaziou <mail@nicolasgoaziou.fr>,
	emacs-org list <emacs-orgmode@gnu.org>
Subject: Re: Lexical binding bug in org-list.el?
Date: Sun, 08 Nov 2015 19:55:09 +0000	[thread overview]
Message-ID: <87pozks1uq.fsf@gmail.com> (raw)
In-Reply-To: <878u688rov.fsf@nicolasgoaziou.fr>

[-- Attachment #1: Type: text/plain, Size: 1782 bytes --]

Hi Nicolas,

2015ko azaroak 8an, Nicolas Goaziou-ek idatzi zuen:

[...]

> Anyway, this we're really nitpicking. There's nothing fundamentally
> wrong in either choice.

You’re right, and I’m not trying to be quarrelsome.  At least for me,
it’s helpful to understand the perspective on things like this, so I can
try to model what is normative guidance and what is not.  Thanks for
humoring me.

> 
>>> - change `org-list-parse-list' to provide a simpler output and update
>>> Babel should to use that new output.
> 
> So the new output could be

[...] details elided

> 
> WDYT?

LGTM.  I’ve probably met my quota of org-related fun for the day (see
below...), but implementing this in terms of elements will be my next
org-list related task.

> 
>>> - re-implement `org-list-to-subtree' using directly Elements, or even
>>> string massaging.

It’s obvious to me that string-massaging causes friction in parts of
org’s code, so I’d like to try a different approach here.  The attached
patch makes a stab at doing the reimplementation in terms of elements.
I think it came out rather nicely, but I’d really value hearing your
opinion on it.

It’s very lightly tested so far.  I basically just used the following
snippet as a test case: put it in an org-mode buffer, put your cursor
somewhere inside the list, and M-: (org-list-to-subtree2)

- foo
- *bar*
  blorfle
- [X] baz
  - quux
  - 123

> Having slept over the idea, I think we could simply update
> `org-list-to-generic' to be more robust and be done with it. I can
> implement it in a couple of hours (and debug it in a couple of
> months...).

OK.  Don’t hesitate to ask if there’s some way we can help, of course.

Thanks,

-- 
Aaron Ecay

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-draft-implementation-of-org-list-to-subtree-in-terms.patch --]
[-- Type: text/x-diff, Size: 4710 bytes --]

From 2c41ae7704c133086a772b8651a1c3cd67feab78 Mon Sep 17 00:00:00 2001
From: Aaron Ecay <aaronecay@gmail.com>
Date: Sun, 8 Nov 2015 19:37:22 +0000
Subject: [PATCH] draft implementation of org-list-to-subtree in terms of
 org-element

---
 lisp/org-list.el | 104 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 104 insertions(+)

diff --git a/lisp/org-list.el b/lisp/org-list.el
index 19d5b03..1612e4e 100644
--- a/lisp/org-list.el
+++ b/lisp/org-list.el
@@ -3263,6 +3263,110 @@ syntax.  Return converted list as a string."
   (require 'ox-texinfo)
   (org-export-string-as list 'texinfo t))
 
+(defun org-list--partial-parse-contents (parse)
+  "Get the actual contents of a partial org-mode parse.
+
+Specifically, when parsing a piece of text smaller than a
+headline, `org-element-parse-buffer' wraps its result with a
+dummy `section' element, as well as the standard `org-data'
+wrapper.  This function removes these, returning a list of
+org-elements.
+
+TODO: maybe this needs a more general name."
+  (org-element-contents
+   ;; strip the org-data element
+   (nth 0 (org-element-contents
+	   ;; and the section element
+	   parse))))
+
+(defun org-list--split-first-line (contents)
+  "Remove the first line of text from an org-element item.
+
+CONTENTS are the contents of the item org-element: at least a
+paragraph followed by zero or more other elements.
+
+Returns a cons of the first textual line and a list of
+org-elements representing the structure of the item minus this
+line.
+
+TODO: is the first daughter of an item always a paragraph?"
+  (let ((graf (nth 0 contents)))
+    (unless (eq (org-element-type graf) 'paragraph)
+      (error "`org-list--split-first-line' got confused"))
+    (goto-char (org-element-property :begin graf))
+    (let* ((eol (point-at-eol))
+	   (end (org-element-property :end graf))
+	   (first-line (buffer-substring-no-properties (point) eol)))
+      (if (> (1+ eol) end)
+	  ;; One line paragraph: it becomes the entirety of the
+	  ;; headline, and we remove it from contents
+	  (setq contents (cdr contents))
+	;; Multi-line paragraph: narrow the buffer to lines 2-N, parse
+	;; them, and set them as the contents of the paragraph.
+	(save-restriction
+	  (widen)
+	  (narrow-to-region (1+ eol) end)
+	  (org-element-set-contents graf
+				    (org-list--partial-parse-contents
+				     ;; TODO: We're playing a trick on
+				     ;; the parser here.  AFAICT, the
+				     ;; parse does not rely on the
+				     ;; cache.  But maybe we should
+				     ;; let org-element-use-cache to
+				     ;; nil around this call, in case
+				     ;; that changes in the future.
+				     (org-element-parse-buffer)))))
+      (cons first-line contents))))
+
+(defun org-list--item-to-headline (item level)
+  "Convert an org-element list item to a headline.
+
+The first line of the list item becomes the "
+  (unless (eq (car item) 'item)
+    (error "`org-list--item-to-headline' expects an item argument"))
+  (let* ((r (org-list--split-first-line (org-element-contents item)))
+	 (title (car r))
+	 (other-contents (cdr r)))
+    (list 'headline
+	  `(:level ,level
+		   ,@(when (eq (org-element-property :checkbox item) 'on)
+		       (list :todo-keyword
+			     ;; TODO: how to fish the approporiate
+			     ;; value out of org-todo-keywords?
+			     "TODO"))
+		   :title ,title)
+	  (mapcar (lambda (x) (if (eq (org-element-type x) 'plain-list)
+				  (org-list--to-headlines x (1+ level))
+				x))
+		  other-contents))))
+
+(defun org-list--to-headlines (list level)
+  (unless (eq (car list) 'plain-list)
+    (error "`org-list-to-subtree' expects a plain-list argument"))
+  (mapcar (lambda (x) (org-list--item-to-headline x level))
+	  (org-element-contents list)))
+
+(defun org-list-to-subtree2 ()
+  (let* ((e (org-element-at-point))
+	 (l (org-element-lineage e))
+	 (list (cl-find-if (lambda (x) (eq (org-element-type x) 'plain-list))
+			   (nreverse l)))
+	 (level (org-reduced-level (or (org-current-level) 0)))
+	 (begin (org-element-property :begin list))
+	 (end (org-element-property :end list))
+	 (parse (save-restriction
+		  (widen)
+		  (narrow-to-region begin end)
+		  (org-element-parse-buffer)))
+	 (new-subtree (org-list--to-headlines
+		       (nth 0 (org-list--partial-parse-contents parse))
+		       level)))
+    (goto-char end)
+    ;; Don't eat the blank lines after the list.
+    (skip-chars-backward " \n\t\f")
+    (delete-region begin (point))
+    (insert (org-element-interpret-data new-subtree))))
+
 (defun org-list-to-subtree (list &optional params)
   "Convert LIST into an Org subtree.
 LIST is as returned by `org-list-parse-list'.  PARAMS is a property list
-- 
2.6.2


  reply	other threads:[~2015-11-08 19:55 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-06 19:43 Lexical binding bug in org-list.el? Kaushal Modi
2015-11-06 19:47 ` Kaushal Modi
2015-11-06 20:45   ` Aaron Ecay
2015-11-06 21:13     ` Kaushal Modi
2015-11-07  0:20     ` Nicolas Goaziou
2015-11-07 11:54       ` Aaron Ecay
2015-11-07 16:48         ` Nicolas Goaziou
2015-11-07 21:30           ` Aaron Ecay
2015-11-08 14:57             ` Nicolas Goaziou
2015-11-08 19:55               ` Aaron Ecay [this message]
2015-11-09 15:23                 ` Kaushal Modi
2015-11-11  9:33                 ` Nicolas Goaziou

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.orgmode.org/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87pozks1uq.fsf@gmail.com \
    --to=aaronecay@gmail.com \
    --cc=emacs-orgmode@gnu.org \
    --cc=mail@nicolasgoaziou.fr \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs/org-mode.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).