drygdryg/calc
1
0
Fork 0
mirror of https://github.com/lcn2/calc.git synced 2025-08-16 01:03:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

Improvements to %g pull request

Browse Source 
This code %g is preliminary.

Fixed some code style issues to match the current code style.

Added regression tests and help file updates to printf and strprintf.

Fixed use of magic number -4: using -P instead.

Noted two problem areas with XXX comments in qio.c:

    1) need a qprintfg function
    2) re-write to not modify conf->outdigits
This commit is contained in:
Landon Curt Noll
2018-09-30 16:06:28 -07:00
parent fcfe237375
commit da6ccc146f
6 changed files with 82 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

44
help/printf
View File

@@ -28,7 +28,7 @@ DESCRIPTION
zero or more decimal digits
an optional '. followed by zero or more decimal deigits
an optional 'l'
one of the letters: d, s, c, f, e, r, o, x, b,
one of the letters: d, s, c, f, e, g, r, o, x, b,
If any other character is read, the '%' and any characters
between '%' and the character are ignored and no specifier is
@@ -51,26 +51,27 @@ DESCRIPTION
d, s, c current config("mode")
f real (decimal, floating point)
e exponential
g real or exponential depending on config("display")
r fractional
o octal
x hexadecimal
b binary
If the number of arguments after fmt is less than the
number of format specifiers in fmt, the "missing" arguments
may be taken to be null values - these contribute nothing to the
output; if a positive width w has been specified, the effect is
to produce w spaces, e.g. printf("abc%6dxyz") prints "abc xyz".
If the number of arguments after fmt is less than the number
of format specifiers in fmt, the "missing" arguments may be
taken to be null values - these contribute nothing to the output;
if a positive width w has been specified, the effect is to
produce w spaces, e.g., printf("abc%6dxyz") prints "abc xyz".
If i <= the number of specifiers in fmt, the value of argument x_i
is printed in the format specified by the i-th specifier.
If a positive width w has been specified and normal printing of x_i
does not include a '\n' character, what is printed will if necessary
be padded with spaces so that the length of the printed output
is at least the w. Note that control
characters like '\t', '\b' each count as one character. If
the 'right-pad' flag has been set, the padding is on the right;
otherwise it is on the left.
If i <= the number of specifiers in fmt, the value of argument
x_i is printed in the format specified by the i-th specifier.
If a positive width w has been specified and normal printing
of x_i does not include a '\n' character, what is printed will
if necessary be padded with spaces so that the length of the
printed output is at least the w. Note that control characters
like '\t', '\b' each count as one character. If the 'right-pad'
flag has been set, the padding is on the right; otherwise it
is on the left.
If i > the number of specifiers in fmt, the value of argument x_i
does not contribute to the printing. However, as all arguments
@@ -84,9 +85,9 @@ DESCRIPTION
mode.
In the case of floating-point (f) format the precision determines
the maximum number of decimal places to be
displayed. Other aspects of this printing may be affected by the
configuration parameters "outround", "tilde", "fullzero", "leadzero".
the maximum number of decimal places to be displayed. Other
aspects of this printing may be affected by the configuration
parameters "outround", "tilde", "fullzero", "leadzero".
EXAMPLE
; c = config("epsilon", 1e-6); c = config("display", 6);
@@ -116,6 +117,11 @@ EXAMPLE
[2] = "undefined"
[3] = NULL
; c = config("display", 50);
; printf("%g %g\n%g %g\n", 1e5, 1e49, 1e50, 1e500);
100000 100000000000000000000000000000000000000000000000000
1e50 1e500
LIMITS
The number of arguments of printf() is not to exceed 1024.
@@ -126,7 +132,7 @@ LINK LIBRARY
SEE ALSO
fprintf, strprintf, print
## Copyright (C) 1999-2006 Landon Curt Noll
## Copyright (C) 1999-2006,2018 Landon Curt Noll
##
## Calc is open software; you can redistribute it and/or modify it under
## the terms of the version 2.1 of the GNU Lesser General Public License