Copyright | Copyright (C) 2006-2011 John Goerzen |
---|---|

License | BSD3 |

Maintainer | John Goerzen <jgoerzen@complete.org> |

Stability | provisional |

Portability | portable |

Safe Haskell | Safe |

Language | Haskell98 |

Tools for rendering sizes

Written by John Goerzen, jgoerzen@complete.org

## Synopsis

- renderNum :: (Ord a, Real a) => SizeOpts -> Int -> a -> String
- renderNums :: (Ord a, Real a) => SizeOpts -> Int -> [a] -> [String]
- parseNum :: (Read a, Fractional a) => SizeOpts -> Bool -> String -> Either String a
- parseNumInt :: (Read a, Integral a) => SizeOpts -> Bool -> String -> Either String a
- quantifyNum :: (Ord a, Real a, Floating b, Ord b) => SizeOpts -> a -> (b, Char)
- quantifyNums :: (Ord a, Real a, Floating b, Ord b) => SizeOpts -> [a] -> ([b], Char)
- data SizeOpts = SizeOpts {}
- binaryOpts :: SizeOpts
- siOpts :: SizeOpts

# Documentation

Render a number into a string, based on the given quantities. This is
useful for displaying quantities in terms of bytes or in SI units. Give this
function the `SizeOpts`

for the desired output, and a precision (number of
digits to the right of the decimal point), and you get a string output.

Here are some examples:

Data.Quantity> renderNum binaryOpts 0 1048576 "1M" Data.Quantity> renderNum binaryOpts 2 10485760 "10.00M" Data.Quantity> renderNum binaryOpts 3 1048576 "1.000M" Data.Quantity> renderNum binaryOpts 3 1500000 "1.431M" Data.Quantity> renderNum binaryOpts 2 (1500 ** 3) "3.14G"

Data.Quantity> renderNum siOpts 2 1024 "1.02k" Data.Quantity> renderNum siOpts 2 1048576 "1.05M" Data.Quantity> renderNum siOpts 2 0.001 "1.00m" Data.Quantity> renderNum siOpts 2 0.0001 "100.00u"

If you want more control over the output, see `quantifyNum`

.

:: (Ord a, Real a) | |

=> SizeOpts | |

-> Int | Prevision of the result |

-> [a] | The numbers to examine |

-> [String] | Result |

Like `renderNum`

, but operates on a list of numbers. The first number
in the list will be evaluated for the suffix. The same suffix and scale will
be used for the remaining items in the list. See `renderNum`

for more
examples.

Also, unlike `renderNum`

, the %f instead of %g printf format is used so that
"scientific" notation is avoided in the output.

Examples:

*Data.Quantity> renderNums binaryOpts 3 [1500000, 10240, 104857600] ["1.431M","0.010M","100.000M"] *Data.Quantity> renderNums binaryOpts 3 [1500, 10240, 104857600] ["1.465K","10.000K","102400.000K"]

:: (Read a, Fractional a) | |

=> SizeOpts | Information on how to parse this data |

-> Bool | Whether to perform a case-insensitive match |

-> String | The string to parse |

-> Either String a |

Parses a String, possibly generated by `renderNum`

. Parses the suffix
and applies it to the number, which is read via the Read class.

Returns Left "error message" on error, or Right number on successful parse.

If you want an Integral result, the convenience function `parseNumInt`

is for
you.

:: (Read a, Integral a) | |

=> SizeOpts | Information on how to parse this data |

-> Bool | Whether to perform a case-insensitive match |

-> String | The string to parse |

-> Either String a |

Parse a number as with `parseNum`

, but return the result as
an `Integral`

. Any type such as Integer, Int, etc. can be used for the
result type.

This function simply calls `round`

on the result of `parseNum`

. A
`Double`

is used internally for the parsing of the numeric component.

By using this function, a user can still say something like 1.5M and get an integral result.

quantifyNum :: (Ord a, Real a, Floating b, Ord b) => SizeOpts -> a -> (b, Char) Source #

Takes a number and returns a new (quantity, suffix) combination. The space character is used as the suffix for items around 0.

quantifyNums :: (Ord a, Real a, Floating b, Ord b) => SizeOpts -> [a] -> ([b], Char) Source #

Like `quantifyNum`

, but takes a list of numbers. The first number in
the list will be evaluated for the suffix. The same suffix and scale will
be used for the remaining items in the list. Please see `renderNums`

for
an example of how this works.

It is invalid to use this function on an empty list.

The options for `quantifyNum`

and `renderNum`

binaryOpts :: SizeOpts Source #

Predefined definitions for byte measurement in groups of 1024, from 0 to 2**80