Enhancment Request/Suggestion
[1/4] from: joel:neely:fedex at: 19-Sep-2000 17:25
Within the recent benchmark discussion, Gabriele made a comment
which reminded me of a suggestion I have been meaning to make
for some time.
Almost every real application I write has a need for at least a
rudimentary level of output formatting. Examples include:
- truncating strings and/or padding strings with trailing blanks
to a fixed length,
- converting numerical values to string values, padding with
leading blanks to a fixed length,
- formatting decimal values as strings, complete with the usual
sorts of leading-blank, trailing-zero, fractional-rounding
string thrashing,
- combining all of the above with interpolation into literal
string data to form an entire message or output line.
I understand/assume that the underlying REBOL infrastructure is
written in C.
Therefore...
Please consider exposing an interface to printf-style formatting
of data values into text. Let me quickly say that I don't view
printf as anything NEAR the acme of readability or elegance, but
it has a few offsetting benefits to compensate for its offputting
appearance:
1) It's already in C, which I assume means it could be exposed
in REBOL more quickly than a completely new formatting scheme
could be conceived, designed, and developed. I certainly
understand that RT has MANY priorities and demands on your
time and personnel; I view this need as basic enough to
justify considering a bit of a quick-and-dirty solution (and
printf ain't hospital sanitary...)
2) It could serve as the underlying mechanism supporting nicer
interfaces that could be build above as mezzanine functions.
3) A large portion of professional programmers have been exposed
(in the epidemiological sense ;-) to C, and therefore would
require almost no training or documentation to be able to
use printf-style format strings.
4) It could provide a "standard" means of bridging the need until
RT has the time to build a more elegant, REBOL-style solution.
(Here I use "standard" in the sense of "available to everyone
using REBOL", without the need to write one's own or go find
some code written by someone else.)
-jn-
[2/4] from: sterling::rebol::com at: 19-Sep-2000 16:25
A quick comment:
I understand what you're asking for and I personally think that it is
important as well but remember that "quick-and-dirty today" often
equals "major legacy problem" tomorrow. I bet a lot of bloat in
existing software can be attributed to this type of thing. Those are
some footsteps that we would rather not follow and I think you don't
want us headed that way either.
I believe Carl already has in mind how he wants to solve these types
of formatting issues within REBOL. I don't know when it'll happen
though.
Take care,
Sterling
[3/4] from: rebol:techscribe at: 19-Sep-2000 16:37
Hi Joel,
have you looked at
http://www.rebol.org/utility/format.html
>From the intro:
Purpose: {
A few functions for formatting.
FORMAT aims to be a Reboloid replacement for sprintf.
It converts data into strings, with rounding of decimal
values, choice of scientific or fixed-point notation,
padding to a given length, and right and left justification.
FORMAT also handles blocks of data, with corresponding blocks
of formatting options to be applied to successive items.
QT is useful for printing out long blocks of words or numerical
data, and QC formats and prints out tabular data stored in
nested blocks.
FULL-FORM returns a string representation of a decimal that will
load back to exactly the original value. All values that correspond
to an exact integer up to 2 ** 53 are represented as an integer.
}
At 05:25 PM 9/19/00 -0500, you wrote:
>Within the recent benchmark discussion, Gabriele made a comment
>which reminded me of a suggestion I have been meaning to make
<<quoted lines omitted: 37>>
> some code written by someone else.)
>-jn-
;- Elan [ : - ) ]
author of REBOL: THE OFFICIAL GUIDE
REBOL Press: The Official Source for REBOL Books
http://www.REBOLpress.com
visit me at http://www.TechScribe.com
[4/4] from: jhagman:infa:abo:fi at: 20-Sep-2000 5:43
Quoting [sterling--rebol--com] ([sterling--rebol--com]):
> I believe Carl already has in mind how he wants to solve these types
> of formatting issues within REBOL. I don't know when it'll happen
> though.
I hope this happens sooner rather than later, because it seems weird
that a language is used for scripting, and is aimed to be able to
replace perl, has no standard means of formatting text. Formatting
text is what most perl programs do. If REBOL really wants to guide
perl programmers to the right direction a standard means of text
formatting is a must.
I thought once that Carl had commented this on the list, but I was
unable to find the article. I used slow Rebol.org then so maybe I just
had not the patience.
--
Jussi Hagman CS in Åbo Akademi University
Studentbyn 4 D 33 [juhagman--abo--fi]
20540 Åbo [jhagman--infa--abo--fi]
Finland
Notes
- Quoted lines have been omitted from some messages.
View the message alone to see the lines that have been omitted