5:26 p.m.
It’s a trivial file, and I don’t think it merits an entire package to
itself, but there’s no real “miscellaneous-Mule” package it to put it
in. If I don’t get a suggestion I will make a separate package for it, but I
feel pretty strongly that wouldn’t be ideal.
;;; url-coding.el --- Support for the web's URL encoding.
;; Copyright (C) 2004 Aidan Kehoe
;; Keywords: HTTP, urlencode
;; This file is not part of XEmacs, but its licence terms apply to it
;; nevertheless.
;; XEmacs 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 2, or (at your option)
;; any later version.
;; XEmacs 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 XEmacs; see the file COPYING. If not, write to the Free
;; Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
;; 02111-1307, USA.
;; This is broken for non-ASCII characters, but we'll ignore that for the
;; moment.
(eval-and-compile
;; This is a table that maps from an octet's value to the ASCII encoding of its
;; first four bits, when those four bits are treated as a single hexadecimal
;; digit.
(defvar url-coding-high-order-nybble-as-ascii
(let ((val (make-vector 256 0))
(i 0))
(while (< i (length val))
(aset val i (char-to-int (aref (format "%02X" i) 0)))
(setq i (1+ i)))
val)
"Table to find an ASCII version of an octet's most significant 4 bits.")
;; Almost the same thing, but this time it's a map from the octet's value to
;; the ASCII encoding of its low-order four bits, when those bits are treated
;; as a hexadecimal digit.
;;
(defvar url-coding-low-order-nybble-as-ascii
(let ((val (make-vector 256 0))
(i 0))
(while (< i (length val))
(aset val i (char-to-int (aref (format "%02X" i) 1)))
(setq i (1+ i)))
val)
"Table to find an ASCII version of an octet's least significant 4 bits.")
(defvar url-coding-escape-character-code (char-to-int ?%)
"The code point for the percentage sign, in ASCII.")
(defvar url-coding-escaped-space-code (char-to-int ?+)
"The URL-encoded value of the space character, that is, +.")
(defvar url-coding-hex-digit-table
(let ((i 0)
(val (make-vector 16 0)))
(while (< i 16)
(aset val i (char-to-int (aref (format "%X" i) 0)))
(setq i (1+ i)))
val)
"A map from a hexadecimal digit's numeric value to its encoding in
ASCII.")
(defvar url-coding-latin-1-as-hex-table
(let ((val (make-vector 256 0))
(i 0))
(while (< i (length val))
;; Get a hex val for this ASCII character.
(aset val i (string-to-int (format "%c" i) 16))
(setq i (1+ i)))
val)
"A map from Latin 1 code points to their values as hexadecimal digits.")
(defvar url-coding-should-preserve-table
(let ((preserve
(list ?- ?_ ?. ?a ?b ?c ?d ?e ?f ?g ?h ?i ?j ?k ?l ?m ?n ?o
?p ?q ?r ?s ?t ?u ?v ?w ?x ?y ?z ?A ?B ?C ?D ?E ?F ?G
?H ?I ?J ?K ?L ?M ?N ?O ?P ?Q ?R ?S ?T ?U ?V ?W ?X ?Y
?Z ?0 ?1 ?2 ?3 ?4 ?5 ?6 ?7 ?8 ?9))
(i 0)
(res (make-vector 256 0)))
(while (< i 256)
(when (member (int-char i) preserve)
(aset res i 1))
(setq i (1+ i)))
res)
"A 256-entry array of binary values, indicating whether to preserve an
octet as ASCII.")
) ;; end of eval-and-compile.
(define-ccl-program ccl-decode-urlcoding
`(1
((read r0)
(loop
(if (r0 == ,url-coding-escape-character-code)
((read r2 r3)
;; Assign the value at offset r2 in the url-coding-hex-digit-table
;; to r3.
(r2 = r2 ,url-coding-latin-1-as-hex-table)
(r3 = r3 ,url-coding-latin-1-as-hex-table)
(r2 <<= 4)
(r3 |= r2)
(write r3))
(if (r0 == ,url-coding-escaped-space-code)
(write #x20)
(write r0)))
(read r0)
(repeat))))
"CCL program to take URL-encoded ISO 8859-1 text and transform it to our
internal encoding.")
(define-ccl-program ccl-encode-urlcoding
`(3
((read r0)
(loop
(r1 = r0 ,url-coding-should-preserve-table)
;; If we should preserve the value, just write the octet directly.
(if r1
(write r0)
;; else, write a percentage sign, and the hex value of the octet, in
;; an ASCII-friendly format.
((write ,url-coding-escape-character-code)
(write r0 ,url-coding-high-order-nybble-as-ascii)
(write r0 ,url-coding-low-order-nybble-as-ascii)))
(read r0)
(repeat))))
"CCL program to encode octets (almost) according to RFC 1738")
(make-coding-system
'url-coding 'ccl
"The coding used by application/x-www-form-urlencoded HTTP applications.
This coding form doesn't specify anything about non-ASCII characters, so
make sure you've transformed to a seven-bit coding system first."
'(decode ccl-decode-urlcoding
encode ccl-encode-urlcoding
mnemonic "URLenc"))
--
“I, for instance, am gung-ho about open source because my family is being
held hostage in Rob Malda’s basement. But who fact-checks me, or Enderle,
when we say something in public? No-one!” -- Danny O’Brien
4:29 p.m.
* Aidan Kehoe <kehoea(a)parhasard.net> writes:
It’s a trivial file, and I don’t think it merits an entire package
to itself, but there’s no real
“miscellaneous-Mule” package it to
put it in. If I don’t get a suggestion I will make a separate
package for it, but I feel pretty strongly that wouldn’t be ideal.
;;; url-coding.el --- Support for the web's URL encoding.
Sounds like its home is net-utils. Does it need Mule? Or can a
non-Mule XEmacs use it too?
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| I am Dyslexic of Borg. |
| Fusistance is retile. Your arse will be laminated. |
|------------------------------------<steve(a)sxemacs.org>---|
5:23 p.m.
Ar an seachtú lá is fiche de mí Márta, scríobh Steve Youngs:
Sounds like its home is net-utils. Does it need Mule? Or can a
non-Mule XEmacs use it too?
It needs define-ccl-program and make-coding-system, neither of which are
available without Mule.
That said, its source code is entirely ASCII, so it can certainly be loaded
by a non-Mule XEmacs, if only to give an error saying "Sorry, this package
needs an XEmacs compiled with Mule support."
define-ccl-program is a macro, though, so putting it into net-utils is only
an option if the distributed packages are _always_ compiled by an XEmacs
with Mule support, something I’m not sure is the case.
--
“I, for instance, am gung-ho about open source because my family is being
held hostage in Rob Malda’s basement. But who fact-checks me, or Enderle,
when we say something in public? No-one!” -- Danny O’Brien
7:13 p.m.
Aidan Kehoe <kehoea(a)parhasard.net> writes:
Hello Aidan,
define-ccl-program is a macro, though, so putting it into net-utils
is only
an option if the distributed packages are _always_ compiled by an XEmacs
with Mule support, something I’m not sure is the case.
Perhaps, we should create a new package mule-utils?
norbert.
7:47 p.m.
Ar an séú lá is fiche de mí Márta, scríobh Norbert Koch:
> define-ccl-program is a macro, though, so putting it into
net-utils is
> only an option if the distributed packages are _always_ compiled by an
> XEmacs with Mule support, something I’m not sure is the case.
Perhaps, we should create a new package mule-utils?
Certainly, if you think it appropriate :-) . I would suggest, though, naming
the package mule-misc instead, because mule-utils implies, in my mind,
functions that are related to internationalisation, which this package
isn’t.
--
“I, for instance, am gung-ho about open source because my family is being
held hostage in Rob Malda’s basement. But who fact-checks me, or Enderle,
when we say something in public? No-one!” -- Danny O’Brien
8 p.m.
Aidan Kehoe <kehoea(a)parhasard.net> writes:
Certainly, if you think it appropriate :-) . I would suggest, though,
naming
the package mule-misc instead, because mule-utils implies, in my mind,
functions that are related to internationalisation, which this package
isn’t.
Sure, I'm all for it. -utils just sprang into mind from the net-utils
reference ;-)
norbert.
9:21 p.m.
* Aidan Kehoe <kehoea(a)parhasard.net> writes:
Ar an seachtú lá is fiche de mí Márta, scríobh Steve Youngs:
> Sounds like its home is net-utils. Does it need Mule? Or can a
> non-Mule XEmacs use it too?
It needs define-ccl-program and make-coding-system, neither of
which are available without Mule.
OK.
That said, its source code is entirely ASCII, so it can certainly
be loaded by a non-Mule XEmacs, if only to give an error saying
"Sorry, this package needs an XEmacs compiled with Mule support."
Easy enough to do.
define-ccl-program is a macro, though, so putting it into net-utils
is only an option if the distributed packages are _always_ compiled
by an XEmacs with Mule support, something I’m not sure is the case.
For XEmacs.org distributed packages, yes, it is the case. For the
case where people are building their own packages, obviously not. But
even that can be worked around in net-util's Makefile.
I've now actually taken a look at url-coding.el (beforehand I only
went as far as the first line) and can honestly say that I have no
idea what it does. :-) So I'll let you decide where to put it. If
you feel net-utils is appropriate, go ahead and add it.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| I am Dyslexic of Borg. |
| Fusistance is retile. Your arse will be laminated. |
|------------------------------------<steve(a)sxemacs.org>---|
11:27 p.m.
Ar an seachtú lá is fiche de mí Márta, scríobh Steve Youngs:
I've now actually taken a look at url-coding.el (beforehand I
only
went as far as the first line) and can honestly say that I have no
idea what it does. :-)
Really--it’s that obscure? Well, to be a bit clearer, it defines a coding
system such that you can do something like:
(setq full-url
(format "http://www.google.com/search?q=%s"
(encode-coding-string "this is our query string &c."
'url-coding)))
=> "http://www.google.com/search?q=this%20is%20our%20query%20string%20%26c."
to generate valid URLs. You can also go in the other direction, which can be
useful when your browser has saved a file to disk without transforming %20
to space:
(decode-coding-string
"Awkward%20file%20name--it%27s%20got%20punctuation%20%2B%20everything.jpg"
'url-coding)
=> "Awkward file name--it's got punctuation + everything.jpg"
The HTTP spec says any byte outside of an extremely limited range must be
encoded as %AB, where AB is the hexadecimal numeric value of that byte, when
that byte is to be passed as a part query string or as part of form
submission data.
It is a toy--I wrote it as part of an effort to better document the CCL
mini-language for writing character encoding transformations. w3 certainly
already has a better version of it, written outside of the CCL framework,
and thus portable to non-Mule. But making it available as part of the
packages does make the documentation better, because it makes it easier for
people to try it out.
So I'll let you decide where to put it. If you feel net-utils
is
appropriate, go ahead and add it.
And it’s looking like we have a new package, mule-misc :-) .
--
“I, for instance, am gung-ho about open source because my family is being
held hostage in Rob Malda’s basement. But who fact-checks me, or Enderle,
when we say something in public? No-one!” -- Danny O’Brien
6:45 p.m.
On Sat, 2005-03-26 at 17:26 +0000, Aidan Kehoe wrote:
It’s a trivial file, and I don’t think it merits an entire package
to
itself, but there’s no real “miscellaneous-Mule” package it to put it
in. If I don’t get a suggestion I will make a separate package for it, but I
feel pretty strongly that wouldn’t be ideal.
;;; url-coding.el --- Support for the web's URL encoding.
At least some similar functionality is included in the "url" elisp
package: https://savannah.gnu.org/projects/url/
The above web page says the package as a separate entity is obsolete and
has been incorporated into the GNU Emacs.
Anyway, it might save some compatibility headache later if we'd setup a
separate "url" XEmacs package, eg. by prying out the latest and greatest
files that comprised the original "url" package from GNU Emacs CVS, and
include url-encoding.el in it as an additional file, or merge it with
the existing files in the package if appropriate.
5:32 a.m.
>>>> "Ville" == Ville Skytt <Ville>
writes:
Ville> On Sat, 2005-03-26 at 17:26 +0000, Aidan Kehoe wrote:
> Itÿs a trivial file, and I donÿt think it merits an entire
> package to itself, but thereÿs no real ´miscellaneous-Mule¡
> package it to put it in. If I donÿt get a suggestion I will
> make a separate package for it, but I feel pretty strongly that
> wouldnÿt be ideal.
> ;;; url-coding.el --- Support for the web's URL encoding.
Ville> At least some similar functionality is included in the
Ville> "url" elisp package: https://savannah.gnu.org/projects/url/
Ville> Anyway, it might save some compatibility headache later if
Ville> we'd setup a separate "url" XEmacs package, eg. by prying
Ville> out the latest and greatest files that comprised the
Ville> original "url" package from GNU Emacs CVS,
I think Ville has this right. Really, there's no reason to do this as
an opaque coding system, except that it's cool that we can do it.
--
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
6960
days inactive
6969
days old
9 comments
5 participants
participants (5)
-
Aidan Kehoe -
Norbert Koch -
Stephen J. Turnbull -
Steve Youngs -
Ville Skyttä