Symbolizer refactoring: ExtractToken and friends

Reviewed at http://reviews.llvm.org/D7867



git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@231027 91177308-0d34-0410-b5e6-96231b3b80d8

5 files changed, 89 insertions, 39 deletions

diff --git a/lib/sanitizer_common/sanitizer_symbolizer_internal.h b/lib/sanitizer_common/sanitizer_symbolizer_internal.h
index f9853c316..30ec00208 100644
--- a/lib/sanitizer_common/sanitizer_symbolizer_internal.h
+++ b/lib/sanitizer_common/sanitizer_symbolizer_internal.h
@@ -18,6 +18,14 @@
 
 namespace __sanitizer {
 
+// Parsing helpers, 'str' is searched for delimiter(s) and a string or uptr
+// is extracted. When extracting a string, a newly allocated (using
+// InternalAlloc) and null-terminataed buffer is returned. They return a pointer
+// to the next characted after the found delimiter.
+const char *ExtractToken(const char *str, const char *delims, char **result);
+const char *ExtractInt(const char *str, const char *delims, int *result);
+const char *ExtractUptr(const char *str, const char *delims, uptr *result);
+
 class SymbolizerTool {
  public:
   // Can't declare pure virtual functions in sanitizer runtimes:
diff --git a/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc b/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc
index 7a091580d..082e271a1 100644
--- a/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc
+++ b/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc
@@ -11,11 +11,42 @@
 // run-time libraries.
 //===----------------------------------------------------------------------===//
 
+#include "sanitizer_allocator_internal.h"
 #include "sanitizer_internal_defs.h"
 #include "sanitizer_symbolizer_internal.h"
 
 namespace __sanitizer {
 
+const char *ExtractToken(const char *str, const char *delims, char **result) {
+  uptr prefix_len = internal_strcspn(str, delims);
+  *result = (char*)InternalAlloc(prefix_len + 1);
+  internal_memcpy(*result, str, prefix_len);
+  (*result)[prefix_len] = '\0';
+  const char *prefix_end = str + prefix_len;
+  if (*prefix_end != '\0') prefix_end++;
+  return prefix_end;
+}
+
+const char *ExtractInt(const char *str, const char *delims, int *result) {
+  char *buff;
+  const char *ret = ExtractToken(str, delims, &buff);
+  if (buff != 0) {
+    *result = (int)internal_atoll(buff);
+  }
+  InternalFree(buff);
+  return ret;
+}
+
+const char *ExtractUptr(const char *str, const char *delims, uptr *result) {
+  char *buff;
+  const char *ret = ExtractToken(str, delims, &buff);
+  if (buff != 0) {
+    *result = (uptr)internal_atoll(buff);
+  }
+  InternalFree(buff);
+  return ret;
+}
+
 Symbolizer *Symbolizer::GetOrInit() {
   SpinMutexLock l(&init_mu_);
   if (symbolizer_)
diff --git a/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc b/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc
index 33865690f..feff9654f 100644
--- a/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc
+++ b/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc
@@ -52,45 +52,6 @@ static const char *DemangleCXXABI(const char *name) {
   return name;
 }
 
-// Extracts the prefix of "str" that consists of any characters not
-// present in "delims" string, and copies this prefix to "result", allocating
-// space for it.
-// Returns a pointer to "str" after skipping extracted prefix and first
-// delimiter char.
-static const char *ExtractToken(const char *str, const char *delims,
-                                char **result) {
-  uptr prefix_len = internal_strcspn(str, delims);
-  *result = (char*)InternalAlloc(prefix_len + 1);
-  internal_memcpy(*result, str, prefix_len);
-  (*result)[prefix_len] = '\0';
-  const char *prefix_end = str + prefix_len;
-  if (*prefix_end != '\0') prefix_end++;
-  return prefix_end;
-}
-
-// Same as ExtractToken, but converts extracted token to integer.
-static const char *ExtractInt(const char *str, const char *delims,
-                              int *result) {
-  char *buff;
-  const char *ret = ExtractToken(str, delims, &buff);
-  if (buff != 0) {
-    *result = (int)internal_atoll(buff);
-  }
-  InternalFree(buff);
-  return ret;
-}
-
-static const char *ExtractUptr(const char *str, const char *delims,
-                               uptr *result) {
-  char *buff;
-  const char *ret = ExtractToken(str, delims, &buff);
-  if (buff != 0) {
-    *result = (uptr)internal_atoll(buff);
-  }
-  InternalFree(buff);
-  return ret;
-}
-
 // Parses one or more two-line strings in the following format:
 //   <function_name>
 //   <file_name>:<line_number>[:<column_number>]
diff --git a/lib/sanitizer_common/tests/CMakeLists.txt b/lib/sanitizer_common/tests/CMakeLists.txt
index 75008db3b..b062c5a02 100644
--- a/lib/sanitizer_common/tests/CMakeLists.txt
+++ b/lib/sanitizer_common/tests/CMakeLists.txt
@@ -28,6 +28,7 @@ set(SANITIZER_UNITTESTS
   sanitizer_stacktrace_test.cc
   sanitizer_stoptheworld_test.cc
   sanitizer_suppressions_test.cc
+  sanitizer_symbolizer_test.cc
   sanitizer_test_main.cc
   sanitizer_thread_registry_test.cc)
 
diff --git a/lib/sanitizer_common/tests/sanitizer_symbolizer_test.cc b/lib/sanitizer_common/tests/sanitizer_symbolizer_test.cc
new file mode 100644
index 000000000..aab3ef5b6
--- /dev/null
+++ b/lib/sanitizer_common/tests/sanitizer_symbolizer_test.cc
@@ -0,0 +1,49 @@
+//===-- sanitizer_symbolizer_test.cc --------------------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// Tests for sanitizer_symbolizer.h and sanitizer_symbolizer_internal.h
+//
+//===----------------------------------------------------------------------===//
+
+#include "sanitizer_common/sanitizer_allocator_internal.h"
+#include "sanitizer_common/sanitizer_symbolizer_internal.h"
+#include "gtest/gtest.h"
+
+namespace __sanitizer {
+
+TEST(Symbolizer, ExtractToken) {
+  char *token;
+  const char *rest;
+
+  rest = ExtractToken("a;b;c", ";", &token);
+  EXPECT_STREQ("a", token);
+  EXPECT_STREQ("b;c", rest);
+  InternalFree(token);
+
+  rest = ExtractToken("aaa-bbb.ccc", ";.-*", &token);
+  EXPECT_STREQ("aaa", token);
+  EXPECT_STREQ("bbb.ccc", rest);
+  InternalFree(token);
+}
+
+TEST(Symbolizer, ExtractInt) {
+  int token;
+  const char *rest = ExtractInt("123,456;789", ";,", &token);
+  EXPECT_EQ(123, token);
+  EXPECT_STREQ("456;789", rest);
+}
+
+TEST(Symbolizer, ExtractUptr) {
+  uptr token;
+  const char *rest = ExtractUptr("123,456;789", ";,", &token);
+  EXPECT_EQ(123U, token);
+  EXPECT_STREQ("456;789", rest);
+}
+
+}  // namespace __sanitizer