spsays-bff/pkg/shorturl/parse.go

package shorturl

import (
	"net/url"
	"strings"
)

// Parsed contains analytics-friendly components extracted from a long URL.
// Unrecognized paths produce ContentType = "unknown" with the other fields
// left empty.
type Parsed struct {
	Path           string
	ContentType    string
	BookCode       string
	Division       string
	Chapter        string
	Verse          string
	Slug           string
	HasHighlight   bool
	HighlightParam string
}

// ParsePath decomposes a long URL (path + query) into analytics columns.
// It expects a same-site relative path; callers should run ValidatePath first.
func ParsePath(longURL string) Parsed {
	u, err := url.Parse("http://x" + longURL)
	if err != nil {
		return Parsed{Path: longURL, ContentType: "unknown"}
	}

	hl := u.Query().Get("hl")
	out := Parsed{
		Path:           u.Path,
		HasHighlight:   hl != "",
		HighlightParam: truncate(hl, 1024),
		ContentType:    "unknown",
	}

	parts := splitPath(u.Path)
	if len(parts) == 0 {
		return out
	}

	switch parts[0] {
	case "books":
		out = parseBooks(parts, out)
	case "letters":
		if len(parts) >= 2 {
			out.ContentType = "letter"
			out.Slug = parts[1]
		}
	case "transcriptions":
		if len(parts) >= 2 {
			out.ContentType = "transcription"
			out.Slug = parts[1]
		}
	case "essays":
		if len(parts) >= 2 {
			out.ContentType = "essay"
			out.Slug = parts[1]
		}
	case "journals":
		if len(parts) >= 2 {
			out.ContentType = "journal"
			out.Slug = parts[1]
		}
		if len(parts) >= 3 {
			out.Chapter = parts[2]
		}
	case "magazines":
		if len(parts) >= 2 {
			out.ContentType = "magazine"
			out.Slug = parts[1]
		}
		if len(parts) >= 3 {
			out.Chapter = parts[2]
		}
	case "other-documents":
		if len(parts) >= 2 {
			out.ContentType = "other_document"
			out.Slug = parts[1]
		}
		if len(parts) >= 3 {
			out.Chapter = parts[2]
		}
	case "glossary":
		out.ContentType = "glossary"
		if len(parts) >= 2 {
			out.Slug = parts[1]
		}
	}
	return out
}

// verseBookCodes are the books whose URL pattern has a per-verse leaf.
var verseBookCodes = map[string]bool{
	"bg":     true,
	"sb1962": true,
	"iso":    true,
	"noi":    true,
	"bs":     true,
	"mm":     true,
	"nbs":    true,
}

// divisionVerseBookCodes have an extra division segment (canto for SB, līlā for CC).
var divisionVerseBookCodes = map[string]bool{
	"sb": true,
	"cc": true,
}

func parseBooks(parts []string, out Parsed) Parsed {
	if len(parts) < 2 {
		out.ContentType = "book"
		return out
	}
	code := parts[1]
	out.BookCode = code

	if divisionVerseBookCodes[code] {
		// /books/sb/{canto}/{chapter}/{verse}
		// /books/cc/{lila}/{chapter}/{verse}
		if len(parts) >= 5 {
			out.ContentType = "verse"
			out.Division = parts[2]
			out.Chapter = parts[3]
			out.Verse = parts[4]
		} else if len(parts) >= 4 {
			out.ContentType = "verse_index"
			out.Division = parts[2]
			out.Chapter = parts[3]
		} else {
			out.ContentType = "book"
		}
		return out
	}

	if verseBookCodes[code] {
		// /books/{code}/{chapter}/{verse}  (BG, SB1962, BS, MM, NBS)
		// /books/{code}/{slug}             (ISO, NOI — single-segment leaf)
		if len(parts) >= 4 {
			out.ContentType = "verse"
			out.Chapter = parts[2]
			out.Verse = parts[3]
		} else if len(parts) >= 3 {
			out.ContentType = "verse"
			out.Slug = parts[2]
		} else {
			out.ContentType = "book"
		}
		return out
	}

	// All other book codes are prose. KB has /kb/{volume}/{slug}; others /code/{slug}.
	if code == "kb" && len(parts) >= 4 {
		out.ContentType = "prose"
		out.Division = parts[2] // volume number
		out.Slug = parts[3]
		return out
	}
	if len(parts) >= 3 {
		out.ContentType = "prose"
		out.Slug = parts[2]
		return out
	}
	out.ContentType = "book"
	return out
}

func splitPath(p string) []string {
	p = strings.Trim(p, "/")
	if p == "" {
		return nil
	}
	parts := strings.Split(p, "/")
	for i, s := range parts {
		if decoded, err := url.PathUnescape(s); err == nil {
			parts[i] = decoded
		}
	}
	return parts
}

func truncate(s string, n int) string {
	if len(s) <= n {
		return s
	}
	return s[:n]
}