Add utf module

Currently only has valid_utf8 and read_utf8

3 files changed, 69 insertions, 1 deletions

diff --git a/test/.gitignore b/test/.gitignore
index 8bf01f1..f1a5a11 100644
--- a/test/.gitignore
+++ b/test/.gitignore
@@ -10,3 +10,4 @@
 /test-observers
 /test-htmlattrtext
 /test-package
+/test-utf
diff --git a/test/Makefile.am b/test/Makefile.am
index abc1f9c..5248547 100644
--- a/test/Makefile.am
+++ b/test/Makefile.am
@@ -3,7 +3,7 @@ MAINTAINERCLEANFILES = Makefile.in
 AM_CXXFLAGS = @DEFINES@ -I$(top_srcdir)/src
 
 TESTS = test-url test-http test-args test-xdg test-paths test-strings \
-        test-observers test-htmlattrtext test-package
+        test-observers test-htmlattrtext test-package test-utf
 
 check_PROGRAMS = $(TESTS)
 
@@ -32,3 +32,6 @@ test_htmlattrtext_LDADD = $(top_builddir)/src/libattrstr.a
 
 test_package_SOURCES = test-package.cc test.hh
 test_package_LDADD = $(top_builddir)/src/libtp.a
+
+test_utf_SOURCES = test-utf.cc test.hh
+test_utf_LDADD = $(top_builddir)/src/libtp.a
diff --git a/test/test-utf.cc b/test/test-utf.cc
new file mode 100644
index 0000000..7096a47
--- /dev/null
+++ b/test/test-utf.cc
@@ -0,0 +1,64 @@
+// -*- mode: c++; c-basic-offset: 2; -*-
+
+#include "common.hh"
+#include "test.hh"
+
+#include <stdarg.h>
+
+#include "utf.hh"
+
+namespace {
+
+bool test_empty() {
+  ASSERT_EQ(true, valid_utf8(""));
+  ASSERT_EQ(true, valid_utf8("", 0));
+  ASSERT_EQ(true, valid_utf8("\xff", 0));
+  auto str = "";
+  ASSERT_EQ(str, read_utf8(str, 0, nullptr));
+  ASSERT_EQ(str, read_utf8(str, 10, nullptr));
+  return true;
+}
+
+bool test_good(char const* str, ...) {
+  ASSERT_EQ(true, valid_utf8(str));
+  va_list args;
+  va_start(args, str);
+  auto pos = str;
+  auto end = str + strlen(str) + 1;
+  uint32_t value;
+  while (true) {
+    uint32_t expected = va_arg(args, uint32_t);
+    auto next = read_utf8(pos, end - pos, &value);
+    if (expected == 0) {
+      ASSERT_EQ(pos, next);
+      ASSERT_EQ(expected, value);
+      break;
+    } else {
+      ASSERT_EQ(false, pos == next);
+      ASSERT_EQ(false, next == nullptr);
+      ASSERT_EQ(expected, value);
+    }
+    pos = next;
+  }
+  va_end(args);
+  return true;
+}
+
+bool test_bad(char const* str) {
+  ASSERT_EQ(false, valid_utf8(str));
+  return true;
+}
+
+}  // namespace
+
+int main(void) {
+  BEFORE;
+  RUN(test_empty());
+  RUN(test_good("$", 0x24, 0));
+  RUN(test_good("\xc2\xa2", 0xa2, 0));
+  RUN(test_good("\xe2\x82\xac", 0x20ac, 0));
+  RUN(test_good("\xf0\x90\x8d\x88", 0x10348, 0));
+  RUN(test_bad("\xf0\x82\x82\xac"));
+  AFTER;
+}
+