ReadViewSDK/Sources/RDEpubReaderView/EPUBCore/CFI/RDEPUBCFIParser.swift

import Foundation

public enum RDEPUBCFIParser {

    public static func parse(_ rawValue: String?) throws -> RDEPUBCFI {

        guard let rawValue = rawValue?.trimmingCharacters(in: .whitespacesAndNewlines),
              !rawValue.isEmpty else {
            throw RDEPUBCFIError.empty
        }

        let body = try unwrap(rawValue)

        let parts = body.split(separator: "!", maxSplits: 1, omittingEmptySubsequences: false)

        let packagePath = try parsePath(String(parts.first ?? ""))

        let contentBody = parts.count > 1 ? String(parts[1]) : ""

        let parsedContent = try parsePathWithOffset(contentBody)

        return RDEPUBCFI(
            packagePath: packagePath,
            contentPath: parsedContent.path,
            characterOffset: parsedContent.characterOffset,
            sideBias: parsedContent.sideBias,
            textAssertion: parsedContent.textAssertion
        )
    }

    public static func parseRange(_ rawValue: String?) throws -> RDEPUBCFIRange {

        guard let rawValue = rawValue?.trimmingCharacters(in: .whitespacesAndNewlines),
              !rawValue.isEmpty else {
            throw RDEPUBCFIError.empty
        }

        let body = try unwrap(rawValue)

        let rangeParts = body.split(separator: ",", omittingEmptySubsequences: false)
        guard rangeParts.count == 3 else {
            throw RDEPUBCFIError.unsupportedRange(rawValue)
        }

        let parentRaw = "epubcfi(\(rangeParts[0]))"

        let startRaw = "epubcfi(\(rangeParts[0])\(rangeParts[1]))"

        let endRaw = "epubcfi(\(rangeParts[0])\(rangeParts[2]))"
        return RDEPUBCFIRange(
            parent: try parse(parentRaw),
            start: try parse(startRaw),
            end: try parse(endRaw)
        )
    }

    private static func unwrap(_ rawValue: String) throws -> String {
        guard rawValue.hasPrefix("epubcfi("), rawValue.hasSuffix(")") else {
            throw RDEPUBCFIError.invalidWrapper(rawValue)
        }
        return String(rawValue.dropFirst("epubcfi(".count).dropLast())
    }

    private static func parsePathWithOffset(_ body: String) throws -> (
        path: RDEPUBCFIPath,
        characterOffset: Int?,
        sideBias: RDEPUBCFISideBias?,
        textAssertion: RDEPUBCFITextAssertion?
    ) {

        guard let offsetSeparator = firstIndexOutsideBrackets(of: ":", in: body) else {

            return (try parsePath(body), nil, nil, nil)
        }

        let pathPart = String(body[..<offsetSeparator])
        let offsetAndQualifiers = String(body[body.index(after: offsetSeparator)...])
        let path = try parsePath(pathPart)

        let offsetDigits = offsetAndQualifiers.prefix { $0.isNumber }
        let offset = Int(offsetDigits)

        let qualifierStart = offsetAndQualifiers.index(
            offsetAndQualifiers.startIndex,
            offsetBy: offsetDigits.count
        )
        let qualifiers = String(offsetAndQualifiers[qualifierStart...])
        return (
            path,
            offset,
            parseSideBias(from: qualifiers),
            parseTextAssertion(from: qualifiers)
        )
    }

    private static func parsePath(_ path: String) throws -> RDEPUBCFIPath {

        guard path.isEmpty || path.hasPrefix("/") else {
            throw RDEPUBCFIError.invalidPath(path)
        }

        let steps = path
            .split(separator: "/", omittingEmptySubsequences: true)
            .compactMap { parseStep(String($0)) }
        return RDEPUBCFIPath(steps: steps)
    }

    private static func parseStep(_ rawStep: String) -> RDEPUBCFIStep? {

        let indexPart = rawStep.split(separator: "[", maxSplits: 1, omittingEmptySubsequences: false).first.map(String.init) ?? rawStep
        guard let index = Int(indexPart) else { return nil }

        let idAssertion: String?
        if let open = rawStep.firstIndex(of: "["),
           let close = rawStep.lastIndex(of: "]"),
           open < close {
            idAssertion = String(rawStep[rawStep.index(after: open)..<close]).rd_nilIfEmpty
        } else {
            idAssertion = nil
        }
        return RDEPUBCFIStep(index: index, idAssertion: idAssertion)
    }

    private static func parseSideBias(from body: String) -> RDEPUBCFISideBias? {

        guard let markerRange = body.range(of: ";s=") else { return nil }
        let value = body[markerRange.upperBound...].prefix(1)
        return RDEPUBCFISideBias(rawValue: String(value))
    }

    private static func parseTextAssertion(from body: String) -> RDEPUBCFITextAssertion? {

        // Scan for the opening bracket, respecting backslash escapes
        guard let open = firstUnescapedIndex(of: "[", in: body) else { return nil }

        // Find the matching closing bracket, respecting backslash escapes and nesting
        var depth = 1
        var current = body.index(after: open)
        while current < body.endIndex && depth > 0 {
            if body[current] == "\\" {
                // Skip escaped character
                current = body.index(after: current)
                if current < body.endIndex {
                    current = body.index(after: current)
                }
                continue
            }
            if body[current] == "[" {
                depth += 1
            } else if body[current] == "]" {
                depth -= 1
            }
            if depth > 0 {
                current = body.index(after: current)
            }
        }

        guard depth == 0, current <= body.endIndex else { return nil }
        let close = current

        // Extract content, processing escape sequences left-to-right
        let raw = String(body[body.index(after: open)..<close])
        let unescaped = unescapeCFIText(raw)

        // Split by comma, but only commas that are not inside nested brackets
        let components = splitAssertionComponents(unescaped)
        guard components.count > 1 else {
            return RDEPUBCFITextAssertion(exact: unescaped)
        }
        return RDEPUBCFITextAssertion(
            prefix: components.first,
            exact: components.count > 1 ? components[1] : nil,
            suffix: components.count > 2 ? components[2] : nil
        )
    }

    /// Find the first occurrence of a character not preceded by a backslash
    private static func firstUnescapedIndex(of character: Character, in string: String) -> String.Index? {
        var index = string.startIndex
        while index < string.endIndex {
            if string[index] == "\\" {
                // Skip escaped character
                index = string.index(after: index)
                if index < string.endIndex {
                    index = string.index(after: index)
                }
                continue
            }
            if string[index] == character {
                return index
            }
            index = string.index(after: index)
        }
        return nil
    }

    /// Split assertion text by commas, but only those at the top level (not inside brackets)
    private static func splitAssertionComponents(_ text: String) -> [String] {
        var components: [String] = []
        var current = ""
        var depth = 0
        var index = text.startIndex
        while index < text.endIndex {
            let char = text[index]
            if char == "\\" {
                current.append(char)
                index = text.index(after: index)
                if index < text.endIndex {
                    current.append(text[index])
                    index = text.index(after: index)
                }
                continue
            }
            if char == "[" {
                depth += 1
            } else if char == "]" {
                depth = max(depth - 1, 0)
            } else if char == "," && depth == 0 {
                components.append(current)
                current = ""
                index = text.index(after: index)
                continue
            }
            current.append(char)
            index = text.index(after: index)
        }
        components.append(current)
        return components
    }

    /// Unescape CFI text assertion content: \[ → [, \] → ], \\ → \
    private static func unescapeCFIText(_ text: String) -> String {
        var result = ""
        var index = text.startIndex
        while index < text.endIndex {
            if text[index] == "\\" {
                let nextIndex = text.index(after: index)
                if nextIndex < text.endIndex {
                    let nextChar = text[nextIndex]
                    if nextChar == "[" || nextChar == "]" || nextChar == "\\" {
                        result.append(nextChar)
                        index = text.index(after: nextIndex)
                        continue
                    }
                }
                // Not a recognized escape — keep the backslash
                result.append(text[index])
            } else {
                result.append(text[index])
            }
            index = text.index(after: index)
        }
        return result
    }

    private static func firstIndexOutsideBrackets(of character: Character, in body: String) -> String.Index? {

        var bracketDepth = 0
        for index in body.indices {
            let current = body[index]
            if current == "[" {
                bracketDepth += 1
            } else if current == "]" {
                bracketDepth = max(bracketDepth - 1, 0)
            } else if current == character, bracketDepth == 0 {
                return index
            }
        }
        return nil
    }
}