qprintf

0.13.4 • Public • Published

qprintf

Build Status Coverage Status

Qprintf is a very fast standard C printf compatible output formatter, supporting the full set of integer, float, and string conversions with field width, alignment and precision, along with some extensions.

Recognizes all traditional output conversions and modifiers (but not %p). Behavior follows closely that of C printf, including very large number support.

var printf = require('qprintf').printf;
printf("%s (%-7s) %05d!", "Hello", "world", 123);
// => Hello (world  ) 00123!

API

qprintf.printf( format, [arg ...] )

Interpolate the arguments into the format string, and write the result to process.stdout.

qprintf.sprintf( format, [arg1, arg2, ...] )

Interpolate the arguments into the format, and return the result.

qprintf.vsprintf( format, argsArray )

Interplate the arguments array into the format string, and return the result.

Conversions

Traditional conversions:

  • %s - interpolate a string into the output
  • %c - the character (1-char string) represented by the given unicode code point
  • %d - a decimal base 10 integer. The integer conversions truncate toward zero (like php). NOTE: qprintf 0.10.0 and earlier converted %d with the nodejs built-in String(), either as an integer, a float, or an exponential. As of qprintf 0.13.0, %d converts like traditional C/C++ printf, as a truncated integer same as %i.
  • %i - a decimal base 10 integer. The integer conversions truncate toward zero (like php).
  • %x - a hexadecimal integer printed using lowercase [0-9a-f]
  • %X - a hexadecimal integer printed using uppercase [0-9A-F]
  • %o - an octal integer
  • %u - an unsigned integer. The native JavaScript number is converted to a 32-bit two's-complement unsigned integer with >>> and printed as %i.
  • %f - a floating-point value "1.23" with integer part, decimal point, and fraction. This conversion never generates exponential notation; use %g for that.
  • %e - a number in exponential notation, eg "1.23e+02". The precision specified for the %e conversion is the number of digits to show to the right of the decimal point.
  • %E - like %e but printed with a capital E, "1.23E+02"
  • %g - a number in either %f or %e notation, depending on its size. Values between 0.0001 and 10^(precision) are converted as %f, values outside this range as %e. The precision specified for the %g conversion is the number of significant digits to show, not the number of decimal places.
  • %G - like %g but in %f or %E notation
  • %% - the % escape character itself
  • %n - consumes an argument that must be a callback cb(int), and calls it with the number of characters converted so far. Does not add to the output.

Qprintf extensions, additional conversions:

  • %b - a binary integer, eg 13 => "1011"
  • %A - an array formatted with util.inspect. For arrays, the field width is the number of elements to show (default 40), and the precision is the depth to which to inspect each element (default 2).
  • %O - an object formatted with util.inspect to depth:6. For objects, the field width is the number of properties to show (default all), and the precision is the depth to which to inspect each element (default 6). Only enumerable "own" properties are included.

Qprintf supports the conversion flags for field width, precision, alignment, padding, sign, and argument selection.

The conversion specifier is constructed as % [argNum$] [(argName)] [flags] [width][.precision] [modif] conversion, with parts in [ ] square brackets optional. Examples: %d, %10ld, %2$d, %2$-010d, %4.2f, %2$(total)+4.3f.

  • N$ - argNum: interpolate the n-th argument with this conversion
  • (NAME) - argName: interpolate the named property (of the first argument, by default). Named arguments can be be mixed with n-th arguments, but do not work well with positional arguments. Named arguments are a qprintf extension.
  • - - minusFlag: left-align the value in the field
  • 0 - zeroFlag: zero pad the field (default is to pad with spaces)
  • + - plusFlag: always print the sign, + for positive and - for negative
  • - spaceFlag: always print the sign, (space) for positive and - for negative
  • NNN - width: a decimal integer that specifies the field width. If the width NNN is specified as '*' (star), the next argument will be consumed and used as the width.
  • NNN.PP - width.precision: a decimal field width followed by the precision. If the precision PP is specified as '*' (star), the next argument will be consumed and used as the precision.
  • NNN. - width of NNN with precision 0
  • .PP - .precision: a decimal precision with a field width wide enough to fit
  • . - precision 0 with a field width just wide enough to fit
  • l, ll, h, hh, L - conversion modifiers: allowed but ignored data size modifier, "long", "long long", "short", "char" and "long double". Invalid usage eg %Ls is not checked.
  • C - conversion: conversion type specifier character

E.g., %3$-+12d will interpolate the 3rd argument into a field 12 characters wide, preceded by the sign (+ for positive and - for negative), and padded with enough spaces on the right for the formatted value to be at least 12 characters. If the value with sign is wider than the field width of 12, it will not be truncated, rather more than 12 characters will be output.

The precision specifier '.' is supported for strings and numbers. For numbers, .N prints a decimal point followed by exactly N decimal digits. . and .0 omit the decimal point and prints no decimals. For strings, the precision specifies the maximum number of characters of the string included in the output.

Unlike C, %0 zero padding pads with zeroes on the right as well, both numbers and strings.

Unlike C, the %n conversion takes a callback cb(int) and not a pointer to int.

Numeric overflow and Infinity are converted to "Infinity".

Examples

var qprintf = require('qprintf')
var sprintf = qprintf.sprintf

sprintf("%5d", 123)                 => "  123"
sprintf("%5.2f", 1.238)             => " 1.24"
sprintf("%05x", 123)                => "0007b"
sprintf("%10s", "Hello")            => "     Hello"
sprintf("%-10s", "Hello")           => "Hello     "
sprintf("%O", {a:1,b:2})            => "{ a: 1, b: 2 }"
sprintf("%2A", [1,2,3,4])           => "[ 1, 2, ... ]"
sprintf("%.f", 12.345)              => "12"
sprintf("%-*.2f", 7, 12.345)        => "12.35  "
sprintf("%*.*f", 8, 3, 1.25)        => "   1.250"
sprintf("%2$d", 1, 2)               => "2"
sprintf("%(x).2f", {x: 1.234})      => "1.23"
sprintf("%2$(x)-4d", {x:1}, {x:2})  => "2   "
sprintf("%-33.f", 1e30)             => "1000000000000000000222784838656  "

Benchmark

Comparative timings generated by the included benchmark.js. The last column is the speed of each relative to printf.

// var fmt1 = "%s %d %s";
// var z = sprintf(fmt1, "Hello", 123, "world");

package versions:
printf 0.5.1
sprintf-js 1.1.2
qprintf 0.13.3

qtimeit=0.21.0 node=12.9.1 v8=7.6.303.29-node.15 platform=linux kernel=4.9.0-0.bpo.4-amd64 up_threshold=false
arch=ia32 mhz=3998 cpuCount=12 cpu="AMD Ryzen 5 2600X Six-Core Processor"
name                speed           rate
printf            348,816 ops/sec    349 >>
sprintfjs       1,613,889 ops/sec   1614 >>>>>>>>
qprintf         8,019,524 ops/sec   8020 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
util_format     5,799,629 ops/sec   5800 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Change Log

  • 0.13.4 - fix support for node-v0.8, allow more native decimal digits, more accurate extreme numbers, support extreme value %i conversion, convert extreme base-N integers as base10 exponentials
  • 0.13.3 - fix breaking typo in test
  • 0.13.2 - update benchmarks for node-v8, faster str_repeat, try to avoid %f rounding errors
  • 0.13.0 - %g fixes, truncating %d (breaking change), l, ll, h, hh modifiers, . widths, large numbers
  • 0.10.0 - e E g G conversions
  • 0.9.2 -
  • 0.8.0 -
  • 0.4.1 - simple version with the basics, d i x o f A O conversions

Related Work

Todo

  • support js-only operation without losing 100% test coverage
  • factor out the low-level formatting into a separate file
  • invoke formatting via a global singleton, to simplify testing
  • flesh out change log

Package Sidebar

Install

npm i qprintf

Weekly Downloads

16

Version

0.13.4

License

Apache-2.0

Unpacked Size

32.9 kB

Total Files

4

Last publish

Collaborators

  • andrasq