README
¶
casefolded 
import "github.com/fvbommel/sortorder/casefolded"
Case-folded sort orders and comparison functions.
These sort characters as the lowest unicode value that is equivalent to that character, ignoring case.
Not all Unicode special cases are supported.
This is a separate sub-package because this needs to pull in the Unicode tables in the standard library, which can add significantly to the size of binaries.
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func NaturalLess ¶
NaturalLess compares two strings using natural ordering. This means that e.g. "abc2" < "abc12".
This is the simple case-folded version, which means that letters are considered equal if strings.SimpleFold says they are. For example, "abc2" < "ABC12" < "abc100" and 'k' == '\u212a' (the Kelvin symbol).
Non-digit sequences and numbers are compared separately. The former are compared rune-by-rune using the lowest equivalent runes, while digits are compared numerically (except that the number of leading zeros is used as a tie-breaker, so e.g. "2" < "02")
Limitations:
- only ASCII digits (0-9) are considered.
- comparisons are done on a rune-by-rune basis, so some special case equivalences like 'ß' == 'SS" are not supported.
- Special cases like Turkish 'i' == 'İ' (and not regular dotless 'I') are not supported either.
Types ¶
type Natural ¶
type Natural []string
Natural implements sort.Interface to sort strings in natural order. This means that e.g. "abc2" < "abc12".
This is the simple case-folded version, which means that letters are considered equal if strings.SimpleFold says they are. For example, "abc2" < "ABC12" < "abc100" and 'k' == '\u212a' (the Kelvin symbol).
Non-digit sequences and numbers are compared separately. The former are compared rune-by-rune using the lowest equivalent runes, while digits are compared numerically (except that the number of leading zeros is used as a tie-breaker, so e.g. "2" < "02")
Limitations:
- only ASCII digits (0-9) are considered.
- comparisons are done on a rune-by-rune basis, so some special case equivalences like 'ß' == 'SS" are not supported.
- Special cases like Turkish 'i' == 'İ' (and not regular dotless 'I') are not supported either.
Source Files