unofficial mirror of emacs-devel@gnu.org 
 help / color / mirror / code / Atom feed
* [ELPA] Proposed package: vcard
@ 2020-02-01  4:21 Eric Abrahamsen
  2020-02-01 14:22 ` Stefan Monnier
  0 siblings, 1 reply; 6+ messages in thread
From: Eric Abrahamsen @ 2020-02-01  4:21 UTC (permalink / raw)
  To: emacs-devel

[-- Attachment #1: Type: text/plain, Size: 3416 bytes --]

Hi all,

I wanted to add full vCard support to EBDB, including parsing vCard
files. There are some vCard-related packages in the repos already, but
they're all tied to some other specific library (BBDB, Org, etc), and I
thought it would be good to have a library that does a sort of
vCard-to-vanilla-data-structure process, so I wrote a new one. Yes, I am
aware of the relevant XKCD comic.

I've attached the two main files in the package: one for parsing vCard
text, the other providing an extremely basic vcard-mode for looking at
*.vcf files. The missing piece is a vcard-write library, for going from
data structures back to vCard files, but I wanted to wait on that until
I've got some feedback.

vCard-to-vanilla-data-structure means that this vCard data:

BEGIN:VCARD
VERSION:3.0
N:Doe;John;;;
FN:John Doe
TITLE:Imaginary test person
EMAIL;type=INTERNET;type=WORK;type=pref:johnDoe@example.org
item1.URL;type=pref:http\://www.example/com/doe
item1.X-ABLabel:_$!<HomePage>!$_
BDAY:--0415
CATEGORIES:Work,Test group
END:VCARD

Turns into this elisp data:

((version 3.0)
 (n ("Doe" "John" "" "" "") nil)
 (fn "John Doe" nil)
 (title "Imaginary test person" nil)
 (email "johnDoe@example.org"
	((type . "pref")
	 (type . "work")
	 (type . "internet")))
 (url "http://www.example/com/doe"
      ((type . "pref")
       (group . "item1")))
 (x-ablabel "HomePage"
	    ((group . "item1")))
 (bday (nil nil nil 15 4 nil nil -1 nil) nil)
 (categories ("Work" "Test group") nil))

It's possible to register your own consumer functions on a per-property
and a per-card basis. I considered using structs instead of plain lists,
but vCards are essentially just bags of properties, so lists seemed
fine. The library comes with some accessors to help with eg. getting all
the properties with the same "group" parameter key.

Potential issues:

1. All date/time properties are parsed with the new iso8601 library. The
   package includes a copy of the library for use in older Emacs, but
   only loads it if the built-in library isn't found. I hope this is
   okay.
2. vCard files are line-wrapped using a CRLF-(SPC|TAB) sequence. The
   parser unwraps long lines. The RFC says:

   Note: It is possible for very simple implementations to generate
   improperly folded lines in the middle of a UTF-8 multi-octet
   sequence.  For this reason, implementations SHOULD unfold lines in
   such a way as to properly restore the original sequence.

   The dumb solution is to find *.vcf files literally, and delete all
   runs of \015\012\040 or \015\012\011. But then the file should be
   _revisited_ and decoded using whatever coding system and other
   parameters that Emacs had originally guessed heuristically for the
   file. I haven't come up with a clean way of doing that (apart from
   maybe saving a temporary intermediate file).
3. The vcard-mode font-locking doesn't work correctly. The value of
   `font-lock-defaults' is set to:

   (defvar vcard-font-lock-keywords
    '("BEGIN:VCARD" "END:VCARD"
     ("^[^;:]+" . 'vcard-property-face)
     (";\\([^=]+\\)=" (1 'vcard-parameter-key-face))
     ("=\\([^;:]+\\)[;:]" (1 'vcard-parameter-value-face))))

   But this _sometimes_ doesn't highlight the parameter keys correctly,
   and I don't know why. I've never messed with font locking before, and
   would appreciate pointers. I think it might have to do with
   overlapping matches.

All comments welcome!

Eric

[-- Attachment #2: vcard-mode.el --]
[-- Type: text/plain, Size: 2043 bytes --]

;;; vcard-mode.el --- Major mode for viewing vCard files  -*- lexical-binding: t; -*-

;; Copyright (C) 2019  Free Software Foundation, Inc.

;; Version: 0
;; Package-Requires: ((emacs "25.1"))

;; Author: Eric Abrahamsen <eric@ericabrahamsen.net>
;; Maintainer: Eric Abrahamsen <eric@ericabrahamsen.net>

;; This program 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.

;; This program 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
;; along with this program.  If not, see <https://www.gnu.org/licenses/>.

;;; Commentary:

;; This file contains `vcard-mode', for viewing vcard files.

;;; Code:

(defface vcard-property-face
  '((t :inherit font-lock-function-name-face))
  "Face for highlighting property names."
  :group 'vcard)

(defface vcard-parameter-key-face
  '((t :inherit font-lock-comment-face))
  "Face for highlighting parameter keys."
  :group 'vcard)

(defface vcard-parameter-value-face
  '((t :inherit font-lock-type-face))
  "Face for highlighting parameter values."
  :group 'vcard)

(defvar vcard-font-lock-keywords
  '("BEGIN:VCARD" "END:VCARD"
    ("^[^;:]+" . 'vcard-property-face)
    (";\\([^=]+\\)=" (1 'vcard-parameter-key-face))
    ("=\\([^;:]+\\)[;:]" (1 'vcard-parameter-value-face))))

;;;###autoload
(define-derived-mode vcard-mode text-mode "vCard"
  "Major mode for viewing vCard files."
  (turn-off-auto-fill)
  (set (make-local-variable 'paragraph-start) "BEGIN:VCARD")
  (setq font-lock-defaults '(vcard-font-lock-keywords)))

;;;###autoload
(add-to-list 'auto-mode-alist '("\\.[Vv][Cc][Ff]\\'" . vcard-mode))

(provide 'vcard-mode)
;;; vcard-mode.el ends here

[-- Attachment #3: vcard-parse.el --]
[-- Type: text/plain, Size: 14638 bytes --]

;;; vcard-parse.el --- Library for parsing vCards      -*- lexical-binding: t; -*-

;; Copyright (C) 2019  Free Software Foundation, Inc.

;; Author: Eric Abrahamsen <eric@ericabrahamsen.net>
;; Maintainer: Eric Abrahamsen <eric@ericabrahamsen.net>

;; This program 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.

;; This program 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
;; along with this program.  If not, see <https://www.gnu.org/licenses/>.

;;; Commentary:

;; This package provides a consumer-agnostic parser for vCard files,
;; aka Virtual Contact Files.  Its entry points parse a file or buffer
;; containing one or more contacts in vCard format, and return the
;; data as a structure meant for use by other programs.  It can parse
;; versions 2.1, 3.0, and 4.0 of the vCard standard, RFC 6350 (see
;; https://tools.ietf.org/html/rfc6350).

;; Parsed vCards are returned as lists containing contact properties.
;; Each property is a list containing the property name, downcased and
;; interned as a symbol, the property value, cast to the most
;; appropriate type, and a further alist of property parameters,
;; values also cast to type where applicable.  For example, this email
;; property:

;; EMAIL;TYPE=work:simon.perreault@viagenie.ca

;; Will be parsed into:

;; (email "simon.perreault@viagenie.ca" ((type . "work")))

;; A contact is a structure containing a list of properties.  As much
;; as possible, the internal implementation of the structure should be
;; ignored, and the properties of a single contact accessed only
;; through the provided getters.  The getters are:

;; `vcard-contact-properties': Return a list of all properties.

;; `vcard-contact-property-types': Return a list of all the different
;; property types this contact has, as symbols.

;; `vcard-contact-property-type': Return all properties of the given
;; type, for this contact.  The return value, if non-nil, is either a
;; single property, or a list of (possibly just one) properties,
;; depending on the cardinality of the property type (see the RFC).

;; `vcard-contact-property-groups': Return a list of all the property
;; groups for the given contact.  A single property's group is found
;; under the 'group key in its parameter list.

;; `vcard-contact-property-group': Return all the properties of the
;; given group, for this contact, or nil.

;; For reference, these are the property types specified for vCard
;; version 4.0:

;; "SOURCE" "KIND" "FN" "N" "NICKNAME" "PHOTO" "BDAY" "ANNIVERSARY"
;; "GENDER" "ADR" "TEL" "EMAIL" "IMPP" "LANG" "TZ" "GEO" "TITLE"
;; "ROLE" "LOGO" "ORG" "MEMBER" "RELATED" "CATEGORIES" "NOTE" "PRODID"
;; "REV" "SOUND" "UID" "CLIENTPIDMAP" "URL" "KEY" "FBURL" "CALADRURI"
;; "CALURI" "XML" iana-token x-name

;; Value types:

;; Booleans, integers, and floats are all cast as expected.  If
;; `vcard-parse-datetime-values' is non-nil, the code will do the best
;; it can to turn a datetime value into a list of integers a-la
;; `parse-time-string'.  This is done either with the built-in
;; `iso8601' library that exists in newer Emacs, or with a local copy
;; that ships with this package, if the built-in version isn't found.

;; While different vCard versions provide slightly different options,
;; the parsing process attempts to normalize property values as much
;; as possible.  Version 4.0 might have more properties available (the
;; KIND property, for instance), but for the most part the parsed data
;; will look the same.

;; TODO:

;; - Go the other direction: produce vCard files from structures.

;;; Code:

(require 'cl-lib)

(defgroup vcard nil
  "Customization options for the vcard library."
  :group 'mail)

(defgroup vcard-parse nil
  "Customization options for vcard parsing."
  :group 'vcard)

(defcustom vcard-parse-select-fields nil
  "A list of field types to select.
If this variable is non-nil, only the fields listed will be
parsed, all others will be discarded.  Note that the 'version and
'fn properties are always returned.

Most useful when let-bound around one of the parsing functions."
  :type '(repeat symbol))

(defcustom vcard-parse-omit-fields nil
  "A list of field types to omit.
If this variable is non-nil, the fields listed will be discarded.

Most useful when let-bound around one of the parsing functions."
  :type '(repeat symbol))

(defcustom vcard-parse-datetime-values t
  "When non-nil, attempt to parse date/time property values.
If successful, the property value will be (usually) converted to
a list of integers, though if the \"type\" parameter of the
property is \"text\", the value will be returned as a string.  It
is also possible that parsing may fail, in which case the
original string value will also be returned."
  :type 'boolean)

(defcustom vcard-parse-card-consumer-function nil
  "Custom function for consuming a single contact card.
It is called with a list of properties, as produced by the
built-in code, or by the return value of
`vcard-parse-property-consumer-function'."
  :type 'function)

(defcustom vcard-parse-property-consumer-function nil
  "Custom function for consuming a single property.
The function is called with four arguments: the property type as
a symbol, the property value (all un-escaping, decoding,
splitting, etc already complete), the property parameters as an
alist with symbol keys, and the vcard version as a float."
  :type 'function)

(defvar vcard-parse-overriding-version nil
  "vCard version, as a float, used when no VERSION property is present.
vCard versions are sometimes specified outside of the cards
themselves -- as part of the file media type, for instance.  In
these cases, this variable can be let-bound around the parsing
process to specify the version.

If a card contains its own VERSION property, that property value
cannot be overridden.")

(defvar vcard-compound-properties '(n adr gender org)
  "A list of vcard properties with multi-part values.
Properties are symbols.  Values have several parts, separated by
semicolons.")

(defvar vcard-datetime-properties '(bday anniversary rev)
  "A list of vcard properties representing date or time values.
The parsing process will make some attempt at converting these
values into lisp timestamps.")

;; Maybe load our local version of iso8601.
(eval-when-compile
  (unless (fboundp 'iso8601-parse)
    (require 'vcard-iso8601)))

;;;###autoload
(defun vcard-parse-file (file)
  "Parse FILE containing vCard data into an alist."
  (interactive "f")
  (with-temp-buffer
    (insert-file-contents file)
    (vcard-parse-buffer)))

;;;###autoload
(defun vcard-parse-buffer ()
  "Parse current buffer, containing vCard data.
Returns a list of contact objects."
  (interactive)
  (let ((card-consumer (when (functionp vcard-parse-card-consumer-function)
			 vcard-parse-card-consumer-function))
	(prop-consumer (if (functionp vcard-parse-property-consumer-function)
			   vcard-parse-property-consumer-function
			 #'list))
	(warning-series t)
	card out)
    ;; vCard 4.0 files *must* be utf-8 encoded + CRLF.  But we're only
    ;; parsing this file, we're not responsible for how it's saved to
    ;; disk.  Don't enable this for now.

    ;; (when (and (null (eq buffer-file-coding-system 'utf-8-unix))
    ;; 	       (or (eql
    ;; 		    vard-parse-overriding-version 4.0)
    ;; 		   (save-excursion
    ;; 		     (re-search-forward "VERSION:4\\.0" (point-max) t))))
    ;;   (set-buffer-file-coding-system 'utf-8-unix))
    (goto-char (point-min))
    ;; Unfolding consists of removing any instances of
    ;; newline-plus-space-or-horizontal-tab.  Technically there should
    ;; always be a non-space character following the space, but we
    ;; don't really care.

    ;; From the RFC:

    ;; Note: It is possible for very simple implementations to
    ;; generate improperly folded lines in the middle of a UTF-8
    ;; multi-octet sequence.  For this reason, implementations SHOULD
    ;; unfold lines in such a way as to properly restore the original
    ;; sequence.

    ;; How would we do that?  We could operate on
    ;; `find-file-literally', but then what?

    ;; CR = \015
    ;; LF = \012
    ;; SPC = \040
    ;; TAB = \011
    (while (re-search-forward "\n[ \t]" (point-max) t)
      (replace-match ""))

    (goto-char (point-min))

    ;; This routine assumes no blank lines in the whole file, which is
    ;; the way it's supposed to be, but we could be a little kinder
    ;; with a `skip-syntax-forward' check.

    (while (re-search-forward "^BEGIN:VCARD\n" (line-end-position 2) t)
      (when (setq card (ignore-errors
			 ;; `vcard-parse-card' moves point past the
			 ;; card.
			 (vcard-parse-card prop-consumer card-consumer)))
	(push card out)))

    (nreverse out)))

(defun vcard-parse-card (&optional prop-consumer card-consumer)
  "Collect properties from a single vCard and return them as an alist.
Point is at bol on the first property.  Collect properties until
the \"END:VCARD\" tag is reached, then move past that tag.

PROP-CONSUMER, if given, should be a function accepting three
arguments -- a property symbol, property value list, and property
parameter list -- and returning a property object.  CARD-CONSUMER
should be a function accepting one argument -- a list of
properties -- and returning a card/contact object."
  (let ((prop-consumer (or prop-consumer #'list))
	(version
	 ;; First line should be the VERSION property.
	 (or (when (re-search-forward
		    "VERSION:\\([[:digit:].]+\\)\n"
		    (line-end-position 2) t)
	       (string-to-number (match-string 1)))
	     vcard-parse-overriding-version
	     (error "Can't determine vCard version")))
	card)
    (push (list 'version version) card)
    (while (and (null (looking-at-p "^END:VCARD$"))
		(re-search-forward
		 "^\\(?:\\(?1:[-[:alnum:]]+\\)\\.\\)?\\(?2:[-[:alnum:]]+\\)"
		 (line-end-position) t))
      (let ((prop (intern (downcase (match-string 2))))
	    anchor sep params value)
	(when (or (eql prop 'fn)
		  (and (or (null vcard-parse-omit-fields)
			   (null (memql prop vcard-parse-omit-fields)))
		       (or (null vcard-parse-select-fields)
			   (memql prop vcard-parse-select-fields))))
	  ;; Pick up the group.
	  (when-let ((group (match-string-no-properties 1)))
	    (push (cons 'group group) params))
	  ;; Pick up parameters.
	  (while (re-search-forward ";\\([^=]+\\)=\\([^;:]+\\)"
				    (line-end-position) t)
	    (push (cons (intern (match-string-no-properties 1))
			(downcase (match-string-no-properties 2)))
		  params))
	  (skip-chars-forward ":")
	  ;; Break value on unescaped commas or semicolons, as
	  ;; appropriate.  Properties may either be compound
	  ;; (eg. addresses), with parts separated by semicolons, or
	  ;; multi-value (eg. categories), with instances separated by
	  ;; commas, but *not both*.
	  (setq sep (if (memq prop vcard-compound-properties) ";" ",")
		anchor (point))
	  (while (re-search-forward sep (line-end-position) t)
	    ;; 92 = backslash.  Having ?\ in the buffer confuses
	    ;; paredit.
	    (unless (eql (char-before (1- (point))) 92)
	      (push (buffer-substring-no-properties anchor (1- (point))) value)
	      (setq anchor (point))))
	  (push (buffer-substring-no-properties
		 anchor (line-end-position))
		value)
	  ;; Unescape all remaining colons, semicolons, commas,
	  ;; backslashes and newlines.
	  (setq value
		(mapcar (lambda (v)
			  (replace-regexp-in-string
			   "\\\\\\([\n:;\\,]\\)" "\\1" v))
			value))
	  ;; Possibly do some parsing of the value(s).
	  (let ((case-fold-search t))
	    (setq value
		  (mapcar
		   (lambda (v)
		     (cond
		      ((string-match-p "false" v)
		       nil)
		      ((string-match-p "true" v)
		       t)
		      ;; What the hell is this, anyway?
		      ((and (eql prop 'x-ablabel)
			    (string-match "_$!<\\([^>]+\\)>!$_" v))
		       (match-string 1 v))
		      ((memql prop vcard-datetime-properties)
		       (if vcard-parse-datetime-values
			   (let ((val-type (cdr-safe (assoc 'value params))))
			     (cond
			      ((and (stringp val-type)
				    (string-equal val-type "text"))
			       v)
			      ((and (stringp val-type)
				    (string-equal val-type "timestamp"))
			       (parse-time-string v))
			      (t
			       (condition-case nil
				   (iso8601-parse v)
				 (error
				  (lwarn
				   '(vcard) :error
				   "Unable to parse date value: \"%s\"" v))))))
			 v))
		      ((string-match-p "\\`[[:digit:].]+\\'" v)
		       (string-to-number v))
		      (t v)))
		   value)))
	  ;; Do we want to normalize this?  This way consumers have to
	  ;; explicitly check if it's a string or a list.
	  (setq value
		(if (= 1 (length value))
		    (car value)
		  (nreverse value)))
	  (push (funcall prop-consumer prop value params)
		card))
	(forward-line)))
    (if card-consumer
	(funcall card-consumer (nreverse card))
      (nreverse card))))

(cl-defmethod vcard-contact-properties ((contact list))
  "Return a list of all properties in CONTACT."
  contact)

(cl-defmethod vcard-contact-property-types ((contact list))
  "Return a list of all property types in CONTACT.
Each type is a symbol representing a downcased property name."
  (let (types)
    (dolist (p (vcard-contact-properties contact) types)
      (cl-pushnew (car p) types))))

(cl-defmethod vcard-contact-property-type ((contact list)
					   (type symbol))
  "Return all properties of TYPE from CONTACT.
TYPE is a symbol, e.g. 'email."
  (let (props)
    (dolist (p (vcard-contact-properties contact) props)
      (when (eql type (car p))
	(push p props)))))

(cl-defmethod vcard-contact-property-groups ((contact list))
  "Return a list of all properties groups in CONTACT.
Each group is a string."
  (let (groups)
    (dolist (p (vcard-contact-properties contact) (nreverse groups))
      (when-let ((g (cdr-safe (assoc 'group (nth 2 p)))))
	(cl-pushnew g groups :test #'equal)))))

(cl-defmethod vcard-contact-property-group ((contact list)
					    (group string))
  "Return all properties belonging to GROUP in CONTACT.
GROUP is a string."
  (let (props)
    (dolist (p (vcard-contact-properties contact) props)
      (when (string-equal (cdr (assq 'group (nth 2 p)))
			  group)
	(push p props)))))

(provide 'vcard-parse)
;;; vcard-parse.el ends here

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [ELPA] Proposed package: vcard
  2020-02-01  4:21 [ELPA] Proposed package: vcard Eric Abrahamsen
@ 2020-02-01 14:22 ` Stefan Monnier
  2020-02-02 17:50   ` Eric Abrahamsen
  0 siblings, 1 reply; 6+ messages in thread
From: Stefan Monnier @ 2020-02-01 14:22 UTC (permalink / raw)
  To: Eric Abrahamsen; +Cc: emacs-devel

> I wanted to add full vCard support to EBDB, including parsing vCard
> files. There are some vCard-related packages in the repos already, but
> they're all tied to some other specific library (BBDB, Org, etc), and I
> thought it would be good to have a library that does a sort of
> vCard-to-vanilla-data-structure process, so I wrote a new one. Yes, I am
> aware of the relevant XKCD comic.

The solution to the XKCD's objection is to rewrite at least one of the
existing solutions on top of your library.  Bonus points if your new one
is obtained by extracting code from that existing other solution rather
than by writing it all from scratch.

>    Note: It is possible for very simple implementations to generate
>    improperly folded lines in the middle of a UTF-8 multi-octet
>    sequence.  For this reason, implementations SHOULD unfold lines in
>    such a way as to properly restore the original sequence.
>
>    The dumb solution is to find *.vcf files literally, and delete all
>    runs of \015\012\040 or \015\012\011. But then the file should be
>    _revisited_ and decoded using whatever coding system and other
>    parameters that Emacs had originally guessed heuristically for the
>    file.

Hmm... so you have to make sure it's first opened as binary, then
do the unfolding, then do the normal decoding?  

>    I haven't come up with a clean way of doing that (apart from
>    maybe saving a temporary intermediate file).

I think the right answer likely involves `after-insert-file-set-coding`,
and `set-auto-coding`, but there's probably more to it.

> 3. The vcard-mode font-locking doesn't work correctly. The value of
>    `font-lock-defaults' is set to:
>
>    (defvar vcard-font-lock-keywords
>     '("BEGIN:VCARD" "END:VCARD"
>      ("^[^;:]+" . 'vcard-property-face)
>      (";\\([^=]+\\)=" (1 'vcard-parameter-key-face))
>      ("=\\([^;:]+\\)[;:]" (1 'vcard-parameter-value-face))))
>
>    But this _sometimes_ doesn't highlight the parameter keys correctly,
>    and I don't know why. I've never messed with font locking before, and
>    would appreciate pointers. I think it might have to do with
>    overlapping matches.

Give us an example,


        Stefan




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [ELPA] Proposed package: vcard
  2020-02-01 14:22 ` Stefan Monnier
@ 2020-02-02 17:50   ` Eric Abrahamsen
  2020-02-03 20:05     ` Eric Abrahamsen
  0 siblings, 1 reply; 6+ messages in thread
From: Eric Abrahamsen @ 2020-02-02 17:50 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: emacs-devel

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> I wanted to add full vCard support to EBDB, including parsing vCard
>> files. There are some vCard-related packages in the repos already, but
>> they're all tied to some other specific library (BBDB, Org, etc), and I
>> thought it would be good to have a library that does a sort of
>> vCard-to-vanilla-data-structure process, so I wrote a new one. Yes, I am
>> aware of the relevant XKCD comic.
>
> The solution to the XKCD's objection is to rewrite at least one of the
> existing solutions on top of your library.  Bonus points if your new one
> is obtained by extracting code from that existing other solution rather
> than by writing it all from scratch.

Okay, not a bad idea. I'll see if any of them are amenable.

>>    Note: It is possible for very simple implementations to generate
>>    improperly folded lines in the middle of a UTF-8 multi-octet
>>    sequence.  For this reason, implementations SHOULD unfold lines in
>>    such a way as to properly restore the original sequence.
>>
>>    The dumb solution is to find *.vcf files literally, and delete all
>>    runs of \015\012\040 or \015\012\011. But then the file should be
>>    _revisited_ and decoded using whatever coding system and other
>>    parameters that Emacs had originally guessed heuristically for the
>>    file.
>
> Hmm... so you have to make sure it's first opened as binary, then
> do the unfolding, then do the normal decoding?  

Right, sort of interrupt then resume the normal process somehow.

>> I haven't come up with a clean way of doing that (apart from
>>    maybe saving a temporary intermediate file).
>
> I think the right answer likely involves `after-insert-file-set-coding`,
> and `set-auto-coding`, but there's probably more to it.

Okay, I've started looking at this stuff. It might take me a while.

>> 3. The vcard-mode font-locking doesn't work correctly. The value of
>>    `font-lock-defaults' is set to:
>>
>>    (defvar vcard-font-lock-keywords
>>     '("BEGIN:VCARD" "END:VCARD"
>>      ("^[^;:]+" . 'vcard-property-face)
>>      (";\\([^=]+\\)=" (1 'vcard-parameter-key-face))
>>      ("=\\([^;:]+\\)[;:]" (1 'vcard-parameter-value-face))))
>>
>>    But this _sometimes_ doesn't highlight the parameter keys correctly,
>>    and I don't know why. I've never messed with font locking before, and
>>    would appreciate pointers. I think it might have to do with
>>    overlapping matches.
>
> Give us an example,

Here's the full example vCard:

BEGIN:VCARD
VERSION:3.0
N:Doe;John;;;
FN:John Doe
ORG:Example.com Inc.;
TITLE:Imaginary test person
EMAIL;type=INTERNET;type=WORK;type=pref:johnDoe@example.org
TEL;type=WORK;type=pref:+1 617 555 1212
TEL;type=WORK:+1 (617) 555-1234
TEL;type=CELL:+1 781 555 1212
TEL;type=HOME:+1 202 555 1212
item1.ADR;type=WORK:;;2 Enterprise Avenue;Worktown;NY;01111;USA
item1.X-ABADR:us
item2.ADR;type=HOME;type=pref:;;3 Acacia Avenue;Hometown;MA;02222;USA
item2.X-ABADR:us
NOTE:John Doe has a long and varied history\, being documented on more police files that anyone else. Reports of his death are alas numerous.
item3.URL;type=pref:http\://www.example/com/doe
item3.X-ABLabel:_$!<HomePage>!$_
item4.URL:http\://www.example.com/Joe/foaf.df
item4.X-ABLabel:FOAF
item5.X-ABRELATEDNAMES;type=pref:Jane Doe
item5.X-ABLabel:_$!<Friend>!$_
CATEGORIES:Work,Test group
X-ABUID:5AD380FD-B2DE-4261-BA99-DE1D1DB52FBE\:ABPerson
END:VCARD

When I put that into vcard-mode, the first "type" parameter key on the
EMAIL property and the item2.ADR property aren't given a face (but
subsequent "type" keys are).

Thanks,
Eric



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [ELPA] Proposed package: vcard
  2020-02-02 17:50   ` Eric Abrahamsen
@ 2020-02-03 20:05     ` Eric Abrahamsen
  2020-02-03 20:52       ` Stefan Monnier
  0 siblings, 1 reply; 6+ messages in thread
From: Eric Abrahamsen @ 2020-02-03 20:05 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: emacs-devel

Eric Abrahamsen <eric@ericabrahamsen.net> writes:

> When I put that into vcard-mode, the first "type" parameter key on the
> EMAIL property and the item2.ADR property aren't given a face (but
> subsequent "type" keys are).

And if I put point anywhere on the "type" key string, and delete/add a
character, it is then font-locked correctly.



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [ELPA] Proposed package: vcard
  2020-02-03 20:05     ` Eric Abrahamsen
@ 2020-02-03 20:52       ` Stefan Monnier
  2020-02-03 21:26         ` Eric Abrahamsen
  0 siblings, 1 reply; 6+ messages in thread
From: Stefan Monnier @ 2020-02-03 20:52 UTC (permalink / raw)
  To: Eric Abrahamsen; +Cc: emacs-devel

> And if I put point anywhere on the "type" key string, and delete/add a
> character, it is then font-locked correctly.

Ah, that rings a bell: some of your regexps can accidentally match over
multiple lines because of [^..] patterns that can match \n


        Stefan




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [ELPA] Proposed package: vcard
  2020-02-03 20:52       ` Stefan Monnier
@ 2020-02-03 21:26         ` Eric Abrahamsen
  0 siblings, 0 replies; 6+ messages in thread
From: Eric Abrahamsen @ 2020-02-03 21:26 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: emacs-devel

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> And if I put point anywhere on the "type" key string, and delete/add a
>> character, it is then font-locked correctly.
>
> Ah, that rings a bell: some of your regexps can accidentally match over
> multiple lines because of [^..] patterns that can match \n

Hey hey! Adding "\n" to the don't-match range did the trick. Thank you.

I'm going to push version 0 now, just to have it out there. I'll try for
a few more days to refine the line unfolding/decoding problem. It's a
pretty edge case, though, so if that stalls I'd rather get 0.1 out there
and start working with existing libraries, then come back to this.

Thanks,
Eric



^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2020-02-03 21:26 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-02-01  4:21 [ELPA] Proposed package: vcard Eric Abrahamsen
2020-02-01 14:22 ` Stefan Monnier
2020-02-02 17:50   ` Eric Abrahamsen
2020-02-03 20:05     ` Eric Abrahamsen
2020-02-03 20:52       ` Stefan Monnier
2020-02-03 21:26         ` Eric Abrahamsen

Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.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).