sow-tools/internal/topdata/wiki_slug.go

package topdata

import (
	"html"
	"strings"
	"unicode"

	"golang.org/x/text/unicode/norm"
)

var reservedCanonicalWikiFirstSegments = map[string]struct{}{
	"admin":     {},
	"api":       {},
	"category":  {},
	"compose":   {},
	"edit":      {},
	"namespace": {},
	"search":    {},
}

func canonicalWikiSegment(value string) string {
	value = html.UnescapeString(strings.TrimSpace(value))
	var b strings.Builder
	lastSeparator := false
	for _, r := range norm.NFKD.String(value) {
		transliterated := transliterateWikiSlugRune(unicode.ToLower(r))
		switch {
		case (r >= 'a' && r <= 'z') || (r >= 'A' && r <= 'Z') || (r >= '0' && r <= '9'):
			b.WriteRune(r)
			lastSeparator = false
		case r == '_':
			if b.Len() > 0 && !lastSeparator {
				b.WriteRune(r)
				lastSeparator = true
			}
		case unicode.Is(unicode.Mn, r):
			continue
		case isWikiSlugJoinerPunctuation(r):
			continue
		case transliterated != "":
			b.WriteString(matchWikiTransliterationCase(r, transliterated))
			lastSeparator = false
		default:
			if b.Len() > 0 && !lastSeparator {
				b.WriteByte('_')
				lastSeparator = true
			}
		}
	}
	return strings.Trim(b.String(), "_")
}

func canonicalWikiSegmentFoldedKey(value string) string {
	return strings.Join(strings.Fields(strings.ToLower(strings.ReplaceAll(canonicalWikiSegment(value), "_", " "))), " ")
}

func canonicalWikiSlashPath(value string) string {
	segments := []string{}
	for _, part := range strings.Split(strings.Trim(strings.TrimSpace(value), "/"), "/") {
		canonical := canonicalWikiSegment(part)
		if canonical != "" {
			segments = append(segments, canonical)
		}
	}
	return strings.Join(segments, "/")
}

func canonicalWikiSlashPathFoldedKey(value string) string {
	segments := []string{}
	for _, part := range strings.Split(strings.Trim(strings.TrimSpace(value), "/"), "/") {
		folded := canonicalWikiSegmentFoldedKey(part)
		if folded != "" {
			segments = append(segments, folded)
		}
	}
	return strings.Join(segments, "/")
}

func canonicalWikiSlashPathSegments(value string) []string {
	segments := []string{}
	for _, part := range strings.Split(strings.Trim(strings.TrimSpace(value), "/"), "/") {
		canonical := canonicalWikiSegment(part)
		if canonical != "" {
			segments = append(segments, canonical)
		}
	}
	return segments
}

func isReservedCanonicalWikiFirstSegment(path string) bool {
	segments := canonicalWikiSlashPathSegments(path)
	if len(segments) == 0 {
		return false
	}
	first := strings.ToLower(segments[0])
	if isCanonicalWikiNumericSegment(first) {
		return true
	}
	_, reserved := reservedCanonicalWikiFirstSegments[first]
	return reserved
}

func isCanonicalWikiNumericSegment(segment string) bool {
	if segment == "" {
		return false
	}
	for _, r := range segment {
		if r < '0' || r > '9' {
			return false
		}
	}
	return true
}

func isNestedCanonicalWikiPath(path string) bool {
	return len(canonicalWikiSlashPathSegments(path)) > 2
}

func matchWikiTransliterationCase(r rune, value string) string {
	if unicode.IsUpper(r) {
		if len(value) == 1 {
			return strings.ToUpper(value)
		}
		return strings.ToUpper(value[:1]) + value[1:]
	}
	return value
}

func canonicalWikiPath(segments ...string) string {
	out := []string{}
	for _, segment := range segments {
		canonical := canonicalWikiSegment(segment)
		if canonical != "" {
			out = append(out, canonical)
		}
	}
	return strings.Join(out, "/")
}

func slugifyWikiText(value string, fallback string) string {
	value = html.UnescapeString(strings.TrimSpace(value))
	var b strings.Builder
	lastDash := false
	for _, r := range norm.NFKD.String(strings.ToLower(value)) {
		transliterated := transliterateWikiSlugRune(r)
		switch {
		case r >= 'a' && r <= 'z':
			b.WriteRune(r)
			lastDash = false
		case r >= '0' && r <= '9':
			b.WriteRune(r)
			lastDash = false
		case unicode.Is(unicode.Mn, r):
			continue
		case isWikiSlugJoinerPunctuation(r):
			continue
		case transliterated != "":
			b.WriteString(transliterated)
			lastDash = false
		default:
			if b.Len() > 0 && !lastDash {
				b.WriteByte('-')
				lastDash = true
			}
		}
	}
	slug := strings.Trim(b.String(), "-")
	if slug == "" {
		return fallback
	}
	return slug
}

func isWikiSlugJoinerPunctuation(r rune) bool {
	switch r {
	case '\'', '"', '`', '´', '‘', '’', '‚', '‛', '“', '”', '„', '‟', '‹', '›', '«', '»', '′', '″', '＇', '＂':
		return true
	default:
		return false
	}
}

func transliterateWikiSlugRune(r rune) string {
	switch r {
	case 'æ':
		return "ae"
	case 'œ':
		return "oe"
	case 'ø':
		return "o"
	case 'ß':
		return "ss"
	case 'þ':
		return "th"
	case 'ð', 'đ':
		return "d"
	case 'ł':
		return "l"
	case 'ħ':
		return "h"
	case 'ı':
		return "i"
	case 'ŋ':
		return "n"
	case 'ŧ':
		return "t"
	default:
		return ""
	}
}