averello/swift-mirror
1
0
Fork 0
mirror of https://github.com/apple/swift.git synced 2025-12-21 12:14:44 +01:00
Code Issues Packages Projects Releases Wiki Activity

[stdlib] Add tests for _isValidUTF8()

Browse Source 
Replaces the tests for the removed _numTrailingBytes()
This commit is contained in:
Patrick Pijnappel
2016-02-28 22:18:21 +11:00
parent 3543664088
commit e02d7e7350
1 changed files with 31 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
validation-test/stdlib/Unicode.swift
View File

@@ -787,32 +787,41 @@ UnicodeScalarTests.test("init") {
var UTF8Decoder = TestSuite("UTF8Decoder")
UTF8Decoder.test("Internal/_numTrailingBytes") {
for i in UInt8(0x00)...UInt8(0x7f) {
expectEqual(0, UTF8._numTrailingBytes(i), "i=\(i)")
UTF8Decoder.test("Internal/_isValidUTF8") {
// Ensure we accept all valid scalars
func ensureValid(scalar: UnicodeScalar) {
var data: UInt32 = 0
var i: UInt32 = 0
Swift.UTF8.encode(scalar) { cp in
data |= UInt32(cp) << (i*8)
i += 1
}
expectEqual(FastUTF8._isValidUTF8(data), true, "data=\(asHex(data))")
}
for i in UInt8(0x80)...UInt8(0xc1) {
expectEqual(4, UTF8._numTrailingBytes(i), "i=\(i)")
for i in 0..<0xd800 { ensureValid(UnicodeScalar(i)) }
for i in 0xe000...0x10ffff { ensureValid(UnicodeScalar(i)) }
// Ensure we have no false positives
var n = 0
func countValidSequences(head head: Range<UInt32>, tail: Range<UInt32>) {
for cu0 in head {
for rest in tail {
let data = rest << 8 | cu0
if FastUTF8._isValidUTF8(data) { n += 1 }
}
}
}
for i in UInt8(0xc2)...UInt8(0xdf) {
expectEqual(1, UTF8._numTrailingBytes(i), "i=\(i)")
}
for i in UInt8(0xe0)...UInt8(0xef) {
expectEqual(2, UTF8._numTrailingBytes(i), "i=\(i)")
}
for i in UInt8(0xf0)...UInt8(0xf4) {
expectEqual(3, UTF8._numTrailingBytes(i), "i=\(i)")
}
for i in UInt8(0xf5)...UInt8(0xfe) {
expectEqual(4, UTF8._numTrailingBytes(i), "i=\(i)")
}
// Separate test for 0xff because of:
// <rdar://problem/17376512> Range UInt8(0x00)...UInt8(0xff) invokes a
// runtime trap
var i = UInt8(0xff)
expectEqual(4, UTF8._numTrailingBytes(i), "i=\(i)")
countValidSequences(head: 0x00...0x7f, tail: 0...0)
countValidSequences(head: 0xc0...0xdf, tail: 0...0xff)
countValidSequences(head: 0xe0...0xef, tail: 0...0xffff)
countValidSequences(head: 0xf0...0xf7, tail: 0...0xffffff)
expectEqual(n, 0x10f800, "n=\(asHex(n))") // 0x10ffff minus surrogates
}
UTF8Decoder.test("Empty") {
expectTrue(checkDecodeUTF8([], [], []))
}