Skip to content

Latest commit

 

History

History
411 lines (282 loc) · 14.9 KB

documentation.md

File metadata and controls

411 lines (282 loc) · 14.9 KB

String Kit

A string manipulation toolbox, featuring a string formatter (inspired by sprintf), a variable inspector (output featuring ANSI colors and HTML) and various escape functions (shell argument, regexp, html, etc).

  • License: MIT
  • Current status: beta
  • Platform: Node.js only (browser support is planned)

Install

Use Node Package Manager:

npm install string-kit

Reference

.format( formatString , ... )

  • formatString String a string containing some sprintf()-like formating
  • ... mixed a variable list of arguments to insert into the formatString

This function is inspired by the C's sprintf() function.

Basicly, if formatString includes format specifiers (subsequences beginning with %), the additional arguments following formatString are formatted and inserted in the resulting string replacing their respective specifiers.

Also it diverges from C in quite a few places.

New: Since v0.3.x we can add styles markup (color, bold, italic, and so on...) using the ^ caret. See the format markup documentation.

Basic usage:

var format = require( 'string-kit' ).format ;
console.log( format( 'Hello %s %s, how are you?' , 'Joe' , 'Doe' ) ) ;
// Output: 'Hello Joe Doe, how are you?'

Specifiers:

  • %% write a single %
  • %s to string, with sanitizer
  • %S to string, with ^ interpretation
  • %r to raw string, without any sanitizer
  • %n to natural: output the most natural representation for this type
  • %N to even more natural: avoid type hinting marks like bracket for array
  • %f to (float) number
  • %k to number with metric system prefixes (like k, M, G, and so on...)
  • %e to exponential “E notation” (e.g. 13 -> 1.23e+2)
  • %K to scientific notation (e.g. 123 -> 1.23 × 10²)
  • %i to integer
  • %d alias of %i (C's printf() compatibility)
  • %u to unsigned integer
  • %U to unsigned positive integer (>0)
  • %P to (absolute) percent (e.g.: 1.25 -> 125% ; 0.75 -> 75%)
  • %p to relative percent (e.g.: 1.25 -> +25% ; 0.75 -> -25%)
  • %t to time duration, convert ms into H:min:s
  • %m to degrees/minutes/seconds notation
  • %h to unsigned hexadecimal
  • %x to unsigned hexadecimal, force pairs of symbols (e.g. 'f' -> '0f')
  • %o to unsigned octal
  • %b to unsigned binary
  • %X string/binary/buffer to hexadecimal: convert a string into hex charcodes, force pair of symbols (e.g. 'f' -> '0f' ; 'hello' -> '68656c6c6f')
  • %z string/binary/buffer to base64
  • %Z string/binary/buffer to base64url
  • %O for objects (call string-kit's inspect() with ultra-minimal options)
  • %I call string-kit's inspect()
  • %Y call string-kit's inspect(), but do not inspect non-enumerable
  • %E call string-kit's inspectError()
  • %J to JSON (call JSON.stringify()
  • %D drop, the argument does not produce anything but is eaten anyway
  • %F filter function existing in the this context, e.g. %[filter:%a%a]F
  • %a argument for a filter function

Few examples:

var format = require( 'string-kit' ).format ;

console.log( format( 'This company regains %d%% of market share.' , 36 ) ) ;
// Output: 'This company regains 36% of market share.'

console.log( format( '11/8=%f' , 11/8 ) ) ;
// Output: '11/8=1.375'

console.log( format( 'Hexa %h %x' , 11 , 11 ) ) ;
// Output: 'Hexa b 0b'

We can insert a number between the % sign and the letter of the specifier, this way, rather than using the next argument, it uses the Nth argument, this is the absolute position:

console.log( format( '%2s%1s%3s' , 'A' , 'B' , 'C' ) ) ; // 'BAC'

Also, the internal pointer is moved anyway, so the Nth format specifier still use the Nth argument if it doesn't specify any position:

console.log( format( '%2s%s%s' , 'A' , 'B' , 'C' ) ) ; // 'BBC'

If the number is preceded by a plus or a minus sign, the relative position is used rather than the absolute position.

console.log( format( '%+1s%-1s%s' , 'A' , 'B' , 'C' ) ) ; // 'BAC'

Use case: language.

var hello = {
	en: 'Hello %s %s!' ,
	jp: 'Konnichiwa %2s %1s!'
} ;

console.log( format( hello[ lang ] , firstName , lastName ) ) ;
// Output the appropriate greeting in a language.
// In japanese the last name will come before the first name,
// but the argument list doesn't need to be changed.

Some specifiers accept parameters: there are between bracket, inserted before the letter, e.g.: %[L5]s. See the specifier parameters section.

The mysterious %[]F format specifier is used when we want custom formatter. Firstly we need to build an object containing one or many functions. Then, format() should be used with call(), to pass the functions collection as the this context.

The %[ is followed by the function's name, followed by a :, followed by a variable list of arguments using %a. It is still possible to use relative and absolute positionning. The whole format specifier is finished when a ]F is encountered.

Example:

var filters = {
	fxy: function( a , b ) { return '' + ( a * a + b ) ; }
} ;

console.log( format.call( filters , '%s%[fxy:%a%a]F' , 'f(x,y)=' , 5 , 3 ) ) ;
// Output: 'f(x,y)=28'

console.log( format.call( filters , '%s%[fxy:%+1a%-1a]F' , 'f(x,y)=' , 5 , 3 ) ) ;
// Output: 'f(x,y)=14'

Specifiers parameters

A parameter consists in a letter, then one character (letter or not letter), that could be followed by any number of non-letter characters. E.g. %[L5]s, the L parameter that produce left-padding to force a 5 characters-length.

A special parameter (specific for that specifier) consists in any number of non-letter characters and must be the first parameter. E.g.:

  • %[.2]f, the special parameter for the f specifier (float), it rounds the number to the second decimal place.
  • %[.2L5]f, mixing both the special and normal parameters, the special parameter comes first (round the the second decimal place), then comes the generic and normal parameter L (left-padding)

When a parameter needs a boolean, + denotes true, while - denotes false.

Generic parameters -- they almost always exists for any specifier and use an upper-case parameter letter:

  • L integer: it produces left-padding (using space) and force the length to the value
  • R integer: same than L but produce right-padding instead

Specifier's specific parameters :

  • %f %e %K %k %p %P:
    • integer: the number precision in digits, e.g. %[3]f (12.345 -> 12.3)
    • integer .: rounds to this integer place, e.g. %[3.]f (12345 -> 12000)
    • . integer: rounds to this decimal place, e.g. %[.2]f (1.2345 -> 1.23)
    • . integer !: rounds to this decimal place and force displaying 0 up to this decimal place, e.g. %[.2!]f (12.9 -> 12.90 ; 12 -> 12.00) (useful for prices)
    • . integer ?: rounds to this decimal place and force displaying 0 up to this decimal place if there is at least one non-zero in the decimal part, e.g. %[.2?]f (12.9 -> 12.90 ; 12 -> 12)
    • z integer: it produces zero padding for the integer part forcing at least integer digits, e.g. %[z5]f (12 -> 00012)
    • g char: set a group separator character for thousands, e.g. %[g,]f (123456 -> 12,345)
    • g: use the default group separator for thousands, e.g. %[g]f (123456 -> 12 345)
  • %O %I %Y %E:
    • integer: the depth of nested object inspection
    • c boolean: if true, can use ANSI color, if false, it will not use colors. E.g. %[c+]I.

Style markup reference

Since v0.3.x we can add styles (color, bold, italic, and so on...) using the ^ caret:

var format = require( 'string-kit' ).format ;
console.log( format( 'This is ^rred^ and ^bblue^:!' ) ) ;
// Output: 'This is red and blue!' with 'red' written in red and 'blue' written in blue.

Style markups:

  • ^^ write a single caret ^
  • ^b switch to blue
  • ^B switch to bright blue
  • ^c switch to cyan
  • ^C switch to bright cyan
  • ^g switch to green
  • ^G switch to bright green
  • ^k switch to black
  • ^K switch to bright black
  • ^m switch to magenta
  • ^M switch to bright magenta
  • ^r switch to red
  • ^R switch to bright red
  • ^w switch to white
  • ^W switch to bright white
  • ^y switch to yellow (i.e. brown or orange)
  • ^Y switch to bright yellow (i.e. yellow)
  • ^_ switch to underline
  • ^/ switch to italic
  • ^! switch to inverse (inverse background and foreground color)
  • ^+ switch to bold
  • ^- switch to dim
  • ^: reset the style
  • ^ (caret followed by a space) reset the style and write a single space
  • ^# background modifier: next color will be a background color instead of a foreground color, e.g.: 'Some ^#^r^bred background text' will write red background in blue over red.

Note: as soon as the format string has one style markup, a style reset will be added at the end of the string.

New: Complex Markup

The complex markup syntax starts with ^[ follows with the markup keyword or key:value, and ends with a ].

Complex markup in the key:value format

  • fg (or aliases: fgColor, color, c) set the foreground color, the value can be one of the ANSI color (red, brightRed, etc), it can also be any color declared in a Palette for methods of object supporting Palette, it can be a color-code (e.g.: #aa5577) if both the terminal and the method support true-color.
  • bg (or alias: bgColor) set the background color, the supported values are exactly the same than for foreground color.

E.g: This output “This is pink!” with the word pink written in pink color: console.log( format( 'This is: ^[fg:#f9b]pink^:!' ) )

Complex markup NOT in the key:value format

  • Any ANSI color (red, brightRed, etc) or color code (e.g.: #aa5577) will be considered as the value for the foreground color.
  • Named background color should be prefixed by bg and followed by camel-case (e.g. red becomes bgRed), for color code you have to use the key:value syntax (see above, e.g.: bg:#aa5577).
  • Other styles like bold, italic, underline, inverse, dim are also supported.

E.g: This output “This is pink!” with the word pink written in pink color: console.log( format( 'This is: ^[#f9b]pink^:!' ) )

.format.count( formatString )

  • formatString String a string containing some sprintf()-like formating

It just counts the number of format specifier in the formatString.

.inspect( [options] , variable )

  • options Object display options, the following key are possible:
    • style string this is the style to use, the value can be:
      • 'none': (default) normal output suitable for console.log() or writing into a file
      • 'inline': like 'none', but without newlines
      • 'color': colorful output suitable for terminal
      • 'html': html output
    • tab: string override the default tab string of the style
    • depth: number depth limit, default: 3
    • maxLength: number length limit for strings, default: 250
    • outputMaxLength: number length limit for the inspect output string, default: 5000
    • noFunc: boolean do not display functions
    • noDescriptor: boolean do not display descriptor information
    • noArrayProperty: boolean do not display array properties
    • noType: boolean do not display type and constructor
    • enumOnly: boolean only display enumerable properties
    • funcDetails: boolean display function's details
    • proto: boolean display object's prototype
    • sort: boolean sort the keys
    • minimal: boolean imply noFunc: true, noDescriptor: true, noType: true, enumOnly: true, proto: false and funcDetails: false. Display a minimal JSON-like output.
    • protoBlackList: Set of blacklisted object prototype (will not recurse inside it)
    • propertyBlackList: Set of blacklisted property names (will not even display it)
    • useInspect: boolean use .inspect() method when available on an object (default to false)
  • variable mixed anything we want to inspect/debug

It inspect a variable, and return a string ready to be displayed with console.log(), or even as HTML output.

It produces a slightly better output than node's util.inspect(), with more options to control what should be displayed.

Since options come first, it is possible to use bind() to create some custom variable inspector.

For example:

var colorInspect = require( 'string-kit' ).inspect.bind( undefined , { style: 'color' } ) ;

Escape functions collection

.escape.shellArg( str )

  • str String the string to filter

It escapes the string so that it will be suitable as a shell command's argument.

.escape.regExp( str ) , .escape.regExpPattern( str )

  • str String the string to filter

It escapes the string so that it will be suitable to inject it in a regular expression's pattern as a literal string.

Example of a search and replace from a user's input:

var result = data.replace(
	new RegExp( stringKit.escape.regExp( userInputSearch ) , 'g' ) ,
	stringKit.escape.regExpReplacement( userInputReplace )
) ;

.escape.regExpReplacement( str )

  • str String the string to filter

It escapes the string so that it will be suitable as a literal string for a regular expression's replacement.

.escape.html( str )

  • str String the string to filter

It escapes the string so that it will be suitable as HTML content.

Only < > & are replaced by HTML entities.

.escape.htmlAttr( str )

  • str String the string to filter

It escapes the string so that it will be suitable as an HTML tag attribute's value.

Only < > & " are replaced by HTML entities.

It assumes valid HTML: the attribute's value should be into double quote, not in single quote.

.escape.htmlSpecialChars( str )

  • str String the string to filter

It escapes all HTML special characters, < > & " ' are replaced by HTML entities.

.escape.control( str )

  • str String the string to filter

It escapes all ASCII control characters (code lesser than or equals to 0x1F, or backspace).

Carriage return, newline and tabulation are respectively replaced by \r, \n and \t. Other characters are replaced by the unicode notation, e.g. NUL is replaced by \x00.

Full BDD spec generated by Mocha: