[lldb] Rewrite stripping and padding and add a unit test

Author: JDevlieghere

lldb/include/lldb/Core/Statusline.h

@@ -12,6 +12,7 @@
 #include "lldb/lldb-forward.h"
 #include "llvm/ADT/StringRef.h"
 #include <csignal>
+#include <cstdint>
 #include <string>
 
 namespace lldb_private {
@@ -33,6 +34,10 @@ class Statusline {
   /// Inform the statusline that the terminal dimensions have changed.
   void TerminalSizeChanged();
 
+protected:
+  /// Pad and trim the given string to fit to the given width.
+  static std::string TrimAndPad(std::string str, size_t width);
+
 private:
   /// Draw the statusline with the given text.
   void Draw(std::string msg);

lldb/source/Core/Statusline.cpp

@@ -18,7 +18,10 @@
 #include "lldb/Utility/LLDBLog.h"
 #include "lldb/Utility/Log.h"
 #include "lldb/Utility/StreamString.h"
+#include "llvm/ADT/StringRef.h"
 #include "llvm/Support/Locale.h"
+#include <algorithm>
+#include <cstdint>
 
 #define ESCAPE "\x1b"
 #define ANSI_NORMAL ESCAPE "[0m"
@@ -70,36 +73,58 @@ void Statusline::Disable() {
   SetScrollWindow(m_terminal_height);
 }
 
+std::string Statusline::TrimAndPad(std::string str, size_t max_width) {
+  size_t column_width = ColumnWidth(str);
+
+  // Trim the string.
+  if (column_width > max_width) {
+    size_t min_width_idx = max_width;
+    size_t min_width = column_width;
+
+    // Use a StringRef for more efficient slicing in the loop below.
+    llvm::StringRef str_ref = str;
+
+    // Keep extending the string to find the minimum column width to make sure
+    // we include as many ANSI escape characters or Unicode code units as
+    // possible. This is far from the most efficient way to do this, but it's
+    // means our stripping code doesn't need to be ANSI and Unicode aware and
+    // should be relatively cold code path.
+    for (size_t i = column_width; i < str.length(); ++i) {
+      size_t stripped_width = ColumnWidth(str_ref.take_front(i));
+      if (stripped_width <= column_width) {
+        min_width = stripped_width;
+        min_width_idx = i;
+      }
+    }
+
+    str = str.substr(0, min_width_idx);
+    column_width = min_width;
+  }
+
+  // Pad the string.
+  if (column_width < max_width)
+    str.append(max_width - column_width, ' ');
+
+  return str;
+}
+
 void Statusline::Draw(std::string str) {
   lldb::LockableStreamFileSP stream_sp = m_debugger.GetOutputStreamSP();
   if (!stream_sp)
     return;
 
-  static constexpr const size_t g_ellipsis_len = 3;
-
   UpdateTerminalProperties();
 
   m_last_str = str;
 
-  size_t column_width = ColumnWidth(str);
-
-  if (column_width + g_ellipsis_len >= m_terminal_width) {
-    // FIXME: If there are hidden characters (e.g. UTF-8, ANSI escape
-    // characters), this will strip the string more than necessary. Ideally we
-    // want to strip until column_width == m_terminal_width.
-    str = str.substr(0, m_terminal_width);
-    str.replace(m_terminal_width - g_ellipsis_len, g_ellipsis_len, "...");
-    column_width = ColumnWidth(str);
-  }
+  str = TrimAndPad(str, m_terminal_width);
 
   LockedStreamFile locked_stream = stream_sp->Lock();
   locked_stream << ANSI_SAVE_CURSOR;
   locked_stream.Printf(ANSI_TO_START_OF_ROW,
                        static_cast<unsigned>(m_terminal_height));
   locked_stream << ANSI_CLEAR_LINE;
   locked_stream << str;
-  if (column_width < m_terminal_width)
-    locked_stream << std::string(m_terminal_width - column_width, ' ');
   locked_stream << ANSI_NORMAL;
   locked_stream << ANSI_RESTORE_CURSOR;
 }

lldb/test/API/functionalities/statusline/TestStatusline.py

@@ -41,8 +41,8 @@ def test(self):
         )
 
         # Change the terminal dimensions and make sure it's reflected immediately.
-        self.child.setwinsize(terminal_height, 20)
-        self.child.expect(re.escape("a.out | main.c:2:..."))
+        self.child.setwinsize(terminal_height, 25)
+        self.child.expect(re.escape("a.out | main.c:2:11 | bre"))
         self.child.setwinsize(terminal_height, terminal_width)
 
         # Change the format.

lldb/unittests/Core/CMakeLists.txt

@@ -11,6 +11,7 @@ add_lldb_unittest(LLDBCoreTests
   RichManglingContextTest.cpp
   SourceLocationSpecTest.cpp
   SourceManagerTest.cpp
+  StatuslineTest.cpp
   TelemetryTest.cpp
   UniqueCStringMapTest.cpp
 

lldb/unittests/Core/StatuslineTest.cpp

@@ -0,0 +1,42 @@
+//===-- StatuslineTest.cpp ------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "lldb/Core/Statusline.h"
+#include "gtest/gtest.h"
+
+using namespace lldb_private;
+
+class TestStatusline : public Statusline {
+public:
+  using Statusline::TrimAndPad;
+};
+
+TEST(StatuslineTest, TestTrimAndPad) {
+  // Test basic ASCII.
+  EXPECT_EQ("     ", TestStatusline::TrimAndPad("", 5));
+  EXPECT_EQ("foo  ", TestStatusline::TrimAndPad("foo", 5));
+  EXPECT_EQ("fooba", TestStatusline::TrimAndPad("fooba", 5));
+  EXPECT_EQ("fooba", TestStatusline::TrimAndPad("foobar", 5));
+
+  // Simple test that ANSI escape codes don't contribute to the visible width.
+  EXPECT_EQ("\x1B[30m     ", TestStatusline::TrimAndPad("\x1B[30m", 5));
+  EXPECT_EQ("\x1B[30mfoo  ", TestStatusline::TrimAndPad("\x1B[30mfoo", 5));
+  EXPECT_EQ("\x1B[30mfooba", TestStatusline::TrimAndPad("\x1B[30mfooba", 5));
+  EXPECT_EQ("\x1B[30mfooba", TestStatusline::TrimAndPad("\x1B[30mfoobar", 5));
+
+  // Test that we include as many escape codes as we can.
+  EXPECT_EQ("fooba\x1B[30m", TestStatusline::TrimAndPad("fooba\x1B[30m", 5));
+  EXPECT_EQ("fooba\x1B[30m\x1B[34m",
+            TestStatusline::TrimAndPad("fooba\x1B[30m\x1B[34m", 5));
+  EXPECT_EQ("fooba\x1B[30m\x1B[34m",
+            TestStatusline::TrimAndPad("fooba\x1B[30m\x1B[34mr", 5));
+
+  // Test Unicode.
+  EXPECT_EQ("❤️    ", TestStatusline::TrimAndPad("❤️", 5));
+  EXPECT_EQ("    ❤️", TestStatusline::TrimAndPad("    ❤️", 5));
+}