unittests/Support/UnicodeTest.cpp

   1 //===- unittests/Support/UnicodeTest.cpp - Unicode.h tests ----------------===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8
   9 #include "llvm/Support/Unicode.h"
  10 #include "gtest/gtest.h"
  11
  12 namespace llvm {
  13 namespace sys {
  14 namespace unicode {
  15 namespace {
  16
  17 TEST(Unicode, columnWidthUTF8) {
  18   EXPECT_EQ(0, columnWidthUTF8(""));
  19   EXPECT_EQ(1, columnWidthUTF8(" "));
  20   EXPECT_EQ(1, columnWidthUTF8("a"));
  21   EXPECT_EQ(1, columnWidthUTF8("~"));
  22
  23   EXPECT_EQ(6, columnWidthUTF8("abcdef"));
  24
  25   EXPECT_EQ(-1, columnWidthUTF8("\x01"));
  26   EXPECT_EQ(-1, columnWidthUTF8("aaaaaaaaaa\x01"));
  27   EXPECT_EQ(-1, columnWidthUTF8("\342\200\213")); // 200B ZERO WIDTH SPACE
  28
  29   // 00AD SOFT HYPHEN is displayed on most terminals as a space or a dash. Some
  30   // text editors display it only when a line is broken at it, some use it as a
  31   // line-break hint, but don't display. We choose terminal-oriented
  32   // interpretation.
  33   EXPECT_EQ(1, columnWidthUTF8("\302\255"));
  34
  35   EXPECT_EQ(0, columnWidthUTF8("\314\200"));     // 0300 COMBINING GRAVE ACCENT
  36   EXPECT_EQ(1, columnWidthUTF8("\340\270\201")); // 0E01 THAI CHARACTER KO KAI
  37   EXPECT_EQ(2, columnWidthUTF8("\344\270\200")); // CJK UNIFIED IDEOGRAPH-4E00
  38
  39   EXPECT_EQ(4, columnWidthUTF8("\344\270\200\344\270\200"));
  40   EXPECT_EQ(3, columnWidthUTF8("q\344\270\200"));
  41   EXPECT_EQ(3, columnWidthUTF8("\314\200\340\270\201\344\270\200"));
  42
  43   // Invalid UTF-8 strings, columnWidthUTF8 should error out.
  44   EXPECT_EQ(-2, columnWidthUTF8("\344"));
  45   EXPECT_EQ(-2, columnWidthUTF8("\344\270"));
  46   EXPECT_EQ(-2, columnWidthUTF8("\344\270\033"));
  47   EXPECT_EQ(-2, columnWidthUTF8("\344\270\300"));
  48   EXPECT_EQ(-2, columnWidthUTF8("\377\366\355"));
  49
  50   EXPECT_EQ(-2, columnWidthUTF8("qwer\344"));
  51   EXPECT_EQ(-2, columnWidthUTF8("qwer\344\270"));
  52   EXPECT_EQ(-2, columnWidthUTF8("qwer\344\270\033"));
  53   EXPECT_EQ(-2, columnWidthUTF8("qwer\344\270\300"));
  54   EXPECT_EQ(-2, columnWidthUTF8("qwer\377\366\355"));
  55
  56   // UTF-8 sequences longer than 4 bytes correspond to unallocated Unicode
  57   // characters.
  58   EXPECT_EQ(-2, columnWidthUTF8("\370\200\200\200\200"));     // U+200000
  59   EXPECT_EQ(-2, columnWidthUTF8("\374\200\200\200\200\200")); // U+4000000
  60 }
  61
  62 TEST(Unicode, isPrintable) {
  63   EXPECT_FALSE(isPrintable(0)); // <control-0000>-<control-001F>
  64   EXPECT_FALSE(isPrintable(0x01));
  65   EXPECT_FALSE(isPrintable(0x1F));
  66   EXPECT_TRUE(isPrintable(' '));
  67   EXPECT_TRUE(isPrintable('A'));
  68   EXPECT_TRUE(isPrintable('~'));
  69   EXPECT_FALSE(isPrintable(0x7F)); // <control-007F>..<control-009F>
  70   EXPECT_FALSE(isPrintable(0x90));
  71   EXPECT_FALSE(isPrintable(0x9F));
  72
  73   EXPECT_TRUE(isPrintable(0xAC));
  74   EXPECT_TRUE(isPrintable(0xAD)); // SOFT HYPHEN is displayed on most terminals
  75                                   // as either a space or a dash.
  76   EXPECT_TRUE(isPrintable(0xAE));
  77
  78   EXPECT_TRUE(isPrintable(0x0377));  // GREEK SMALL LETTER PAMPHYLIAN DIGAMMA
  79   EXPECT_FALSE(isPrintable(0x0378)); // <reserved-0378>..<reserved-0379>
  80
  81   EXPECT_FALSE(isPrintable(0x0600)); // ARABIC NUMBER SIGN
  82
  83   EXPECT_FALSE(isPrintable(0x1FFFF)); // <reserved-1F774>..<noncharacter-1FFFF>
  84   EXPECT_TRUE(isPrintable(0x20000));  // CJK UNIFIED IDEOGRAPH-20000
  85
  86   EXPECT_FALSE(isPrintable(0x10FFFF)); // noncharacter
  87 }
  88
  89 } // namespace
  90 } // namespace unicode
  91 } // namespace sys
  92 } // namespace llvm