Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
16 commits
Select commit Hold shift + click to select a range
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions Package.swift
Original file line number Diff line number Diff line change
Expand Up @@ -67,6 +67,11 @@ let package = Package(
.testTarget(
name: "ReerJSONTests",
dependencies: ["ReerJSON"],
exclude: [
"Info.plist",
"Models/instruments.json",
"Models/json_to_swift.rb",
],
resources: [
.copy("Models/apache_builds.json"),
.copy("Models/canada.json"),
Expand Down
35 changes: 28 additions & 7 deletions Sources/ReerJSON/JSON.swift
Original file line number Diff line number Diff line change
Expand Up @@ -78,10 +78,6 @@ extension JSON {
var string: String? {
guard let cString = yyjson_get_str(pointer) else { return nil }
let length = yyjson_get_len(pointer)
if memchr(cString, 0, length) != nil {
let rawBuffer = UnsafeRawBufferPointer(start: cString, count: length)
return String(bytes: rawBuffer, encoding: .utf8)
}
let buffer = UnsafeBufferPointer(
start: UnsafeRawPointer(cString).assumingMemoryBound(to: UInt8.self),
count: length
Expand Down Expand Up @@ -159,6 +155,9 @@ extension JSON {
@inline(__always)
func integer<T: FixedWidthInteger>() -> T? {
guard let cString = yyjson_get_raw(pointer) else { return nil }
if let value = Self.parsePlainInteger(cString, count: yyjson_get_len(pointer)) {
return T(exactly: value)
}
var convertedVal = yyjson_val()
var error = yyjson_read_err()
guard let _ = yyjson_read_number(cString, &convertedVal, YYJSON_READ_ALLOW_EXT_NUMBER, nil, &error) else {
Expand All @@ -169,14 +168,36 @@ extension JSON {
} else if yyjson_is_sint(&convertedVal) {
return T(exactly: yyjson_get_sint(&convertedVal))
} else if yyjson_is_real(&convertedVal) {
let real = unsafe_yyjson_get_real(&convertedVal)
if let value = T(exactly: real) { return value }
// try Int128/UInt128
// An integer literal only reads as real when it overflows 64 bits;
// its rounded double must not be accepted (e.g. -2^63 - 1 → Int64.min).
if strpbrk(cString, ".eE") != nil, let value = T(exactly: unsafe_yyjson_get_real(&convertedVal)) {
return value
}
// Exact parse for integer literals, e.g. Int128/UInt128.
return T(String(cString: cString))
}
return nil
}

/// Parses `-?[0-9]{1,18}` without going through `yyjson_read_number`.
/// Anything else (fractions, exponents, JSON5 forms, 19+ digits) returns
/// nil; 18 decimal digits cannot overflow `Int64`.
@inline(__always)
private static func parsePlainInteger(_ chars: UnsafePointer<CChar>, count: Int) -> Int64? {
guard count > 0, count <= 18 else { return nil }
let negative = chars[0] == 0x2D
var index = negative ? 1 : 0
guard index < count else { return nil }
var value: Int64 = 0
while index < count {
let digit = UInt8(bitPattern: chars[index]) &- 0x30
guard digit < 10 else { return nil }
value = value &* 10 &+ Int64(digit)
index &+= 1
}
return negative ? -value : value
}

var rawString: String? {
guard let cString = yyjson_get_raw(pointer) else { return nil }
return String(cString: cString)
Expand Down
185 changes: 139 additions & 46 deletions Sources/ReerJSON/JSONDecoderImpl.swift
Original file line number Diff line number Diff line change
Expand Up @@ -1016,7 +1016,7 @@ private final class PreTransformKeyedDecodingContainer<K: CodingKey>: KeyedDecod
while let keyPtr = yyjson_obj_iter_next(&iter) {
if let keyCString = yyjson_get_str(keyPtr), let valuePtr = yyjson_obj_iter_get_val(keyPtr) {
let jsonKey = String(cString: keyCString)
result[Self._convertFromSnakeCase(jsonKey)]._setIfNil(to: JSON(pointer: valuePtr))
result[SnakeCaseKeyConverter.convertFromSnakeCase(jsonKey)]._setIfNil(to: JSON(pointer: valuePtr))
}
}
case .custom(let converter):
Expand Down Expand Up @@ -1343,51 +1343,6 @@ private final class PreTransformKeyedDecodingContainer<K: CodingKey>: KeyedDecod
}
return int
}

private static func _convertFromSnakeCase(_ stringKey: String) -> String {
guard !stringKey.isEmpty else { return stringKey }

// Find the first non-underscore character
guard let firstNonUnderscore = stringKey.firstIndex(where: { $0 != "_" }) else {
// Reached the end without finding an _
return stringKey
}

// Find the last non-underscore character
var lastNonUnderscore = stringKey.index(before: stringKey.endIndex)
while lastNonUnderscore > firstNonUnderscore && stringKey[lastNonUnderscore] == "_" {
stringKey.formIndex(before: &lastNonUnderscore)
}

let keyRange = firstNonUnderscore...lastNonUnderscore
let leadingUnderscoreRange = stringKey.startIndex..<firstNonUnderscore
let trailingUnderscoreRange = stringKey.index(after: lastNonUnderscore)..<stringKey.endIndex

let components = stringKey[keyRange].split(separator: "_")
let joinedString: String
if components.count == 1 {
// No underscores in key, leave the word as is - maybe already camel cased
joinedString = String(stringKey[keyRange])
} else {
joinedString = ([components[0].lowercased()] + components[1...].map { $0.capitalized }).joined()
}

// Do a cheap isEmpty check before creating and appending potentially empty strings
let result: String
if (leadingUnderscoreRange.isEmpty && trailingUnderscoreRange.isEmpty) {
result = joinedString
} else if (!leadingUnderscoreRange.isEmpty && !trailingUnderscoreRange.isEmpty) {
// Both leading and trailing underscores
result = String(stringKey[leadingUnderscoreRange]) + joinedString + String(stringKey[trailingUnderscoreRange])
} else if (!leadingUnderscoreRange.isEmpty) {
// Just leading
result = String(stringKey[leadingUnderscoreRange]) + joinedString
} else {
// Just trailing
result = joinedString + String(stringKey[trailingUnderscoreRange])
}
return result
}
}

// MARK: - UnkeyedDecodingContainer
Expand Down Expand Up @@ -1610,3 +1565,141 @@ private struct JSONUnkeyedDecodingContainer: UnkeyedDecodingContainer {
return int
}
}

// MARK: - Snake case key conversion

enum SnakeCaseKeyConverter {
/// Same result as `convertFromSnakeCaseReference`, with a byte-level fast
/// path for the common ASCII keys. Components that are not purely ASCII
/// letters or purely digits use the reference, since `capitalized` has
/// word-boundary rules of its own.
static func convertFromSnakeCase(_ stringKey: String) -> String {
var key = stringKey
switch key.withUTF8(fastConvert) {
case .unchanged: return stringKey
case .converted(let result): return result
case .unsupported: return convertFromSnakeCaseReference(stringKey)
}
}

private enum FastResult {
case unchanged
case converted(String)
case unsupported
}

private static let underscore = UInt8(ascii: "_")

private static func fastConvert(_ bytes: UnsafeBufferPointer<UInt8>) -> FastResult {
guard let first = bytes.firstIndex(where: { $0 != underscore }) else { return .unchanged }
var last = bytes.count - 1
while bytes[last] == underscore { last -= 1 }

var componentCount = 0
var index = first
while index <= last {
if bytes[index] == underscore {
index += 1
continue
}
componentCount += 1
var hasLetter = false, hasDigit = false
while index <= last && bytes[index] != underscore {
let byte = bytes[index]
if byte >= 0x80 { return .unsupported }
if (byte | 0x20) >= UInt8(ascii: "a") && (byte | 0x20) <= UInt8(ascii: "z") {
hasLetter = true
} else if byte >= UInt8(ascii: "0") && byte <= UInt8(ascii: "9") {
hasDigit = true
} else {
return .unsupported
}
index += 1
}
if hasLetter && hasDigit { return .unsupported }
}
// A single component (no inner underscores) is left as is.
if componentCount <= 1 { return .unchanged }

let result = String(unsafeUninitializedCapacity: bytes.count) { out in
var count = 0
for i in 0..<first {
out[count] = bytes[i]; count += 1
}
var isFirstComponent = true
var index = first
while index <= last {
if bytes[index] == underscore {
index += 1
continue
}
var isComponentStart = true
while index <= last && bytes[index] != underscore {
let byte = bytes[index]
let isLetter = byte > UInt8(ascii: "9")
if isLetter {
let uppercase = isComponentStart && !isFirstComponent
out[count] = uppercase ? byte & ~0x20 : byte | 0x20
} else {
out[count] = byte
}
count += 1
isComponentStart = false
index += 1
}
isFirstComponent = false
}
for i in (last + 1)..<bytes.count {
out[count] = bytes[i]; count += 1
}
return count
}
return .converted(result)
}

/// Foundation's algorithm; the reference for `convertFromSnakeCase`.
static func convertFromSnakeCaseReference(_ stringKey: String) -> String {
guard !stringKey.isEmpty else { return stringKey }

// Find the first non-underscore character
guard let firstNonUnderscore = stringKey.firstIndex(where: { $0 != "_" }) else {
// Reached the end without finding an _
return stringKey
}

// Find the last non-underscore character
var lastNonUnderscore = stringKey.index(before: stringKey.endIndex)
while lastNonUnderscore > firstNonUnderscore && stringKey[lastNonUnderscore] == "_" {
stringKey.formIndex(before: &lastNonUnderscore)
}

let keyRange = firstNonUnderscore...lastNonUnderscore
let leadingUnderscoreRange = stringKey.startIndex..<firstNonUnderscore
let trailingUnderscoreRange = stringKey.index(after: lastNonUnderscore)..<stringKey.endIndex

let components = stringKey[keyRange].split(separator: "_")
let joinedString: String
if components.count == 1 {
// No underscores in key, leave the word as is - maybe already camel cased
joinedString = String(stringKey[keyRange])
} else {
joinedString = ([components[0].lowercased()] + components[1...].map { $0.capitalized }).joined()
}

// Do a cheap isEmpty check before creating and appending potentially empty strings
let result: String
if (leadingUnderscoreRange.isEmpty && trailingUnderscoreRange.isEmpty) {
result = joinedString
} else if (!leadingUnderscoreRange.isEmpty && !trailingUnderscoreRange.isEmpty) {
// Both leading and trailing underscores
result = String(stringKey[leadingUnderscoreRange]) + joinedString + String(stringKey[trailingUnderscoreRange])
} else if (!leadingUnderscoreRange.isEmpty) {
// Just leading
result = String(stringKey[leadingUnderscoreRange]) + joinedString
} else {
// Just trailing
result = joinedString + String(stringKey[trailingUnderscoreRange])
}
return result
}
}
Loading
Loading