[lldb] Support custom LLVM formatting for variables (#81196)

Adds support for applying LLVM formatting to variables.

The reason for this is to support cases such as the following.

Let's say you have two separate bytes that you want to print as a
combined hex value. Consider the following summary string:

```
${var.byte1%x}${var.byte2%x}
```

The output of this will be: `0x120x34`. That is, a `0x` prefix is
unconditionally applied to each byte. This is unlike printf formatting
where you must include the `0x` yourself.

Currently, there's no way to do this with summary strings, instead
you'll need a summary provider in python or c++.

This change introduces formatting support using LLVM's formatter system.
This allows users to achieve the desired custom formatting using:

```
${var.byte1:x-}${var.byte2:x-}
```

Here, each variable is suffixed with `:x-`. This is passed to the LLVM
formatter as `{0:x-}`. For integer values, `x` declares the output as
hex, and `-` declares that no `0x` prefix is to be used. Further, one
could write:

```
${var.byte1:x-2}${var.byte2:x-2}
```

Where the added `2` results in these bytes being written with a minimum
of 2 digits.

An alternative considered was to add a new format specifier that would
print hex values without the `0x` prefix. The reason that approach was
not taken is because in addition to forcing a `0x` prefix, hex values
are also forced to use leading zeros. This approach lets the user have
full control over formatting.

lldb/docs/use/variable.rst

@@ -460,6 +460,15 @@ summary strings, regardless of the format they have applied to their types. To
 do that, you can use %format inside an expression path, as in ${var.x->x%u},
 which would display the value of x as an unsigned integer.
 
+Additionally, custom output can be achieved by using an LLVM format string,
+commencing with the ``:`` marker. To illustrate, compare ``${var.byte%x}`` and
+``${var.byte:x-}``. The former uses lldb's builtin hex formatting (``x``),
+which unconditionally inserts a ``0x`` prefix, and also zero pads the value to
+match the size of the type. The latter uses ``llvm::formatv`` formatting
+(``:x-``), and will print only the hex value, with no ``0x`` prefix, and no
+padding. This raw control is useful when composing multiple pieces into a
+larger whole.
+
 You can also use some other special format markers, not available for formats
 themselves, but which carry a special meaning when used in this context:
 

lldb/source/Core/FormatEntity.cpp

@@ -57,6 +57,7 @@
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/StringRef.h"
 #include "llvm/Support/Compiler.h"
+#include "llvm/Support/Regex.h"
 #include "llvm/TargetParser/Triple.h"
 
 #include <cctype>
@@ -658,6 +659,38 @@ static char ConvertValueObjectStyleToChar(
   return '\0';
 }
 
+static llvm::Regex LLVMFormatPattern{"x[-+]?\\d*|n|d", llvm::Regex::IgnoreCase};
+
+static bool DumpValueWithLLVMFormat(Stream &s, llvm::StringRef options,
+                                    ValueObject &valobj) {
+  std::string formatted;
+  std::string llvm_format = ("{0:" + options + "}").str();
+
+  // Options supported by format_provider<T> for integral arithmetic types.
+  // See table in FormatProviders.h.
+
+  auto type_info = valobj.GetTypeInfo();
+  if (type_info & eTypeIsInteger && LLVMFormatPattern.match(options)) {
+    if (type_info & eTypeIsSigned) {
+      bool success = false;
+      int64_t integer = valobj.GetValueAsSigned(0, &success);
+      if (success)
+        formatted = llvm::formatv(llvm_format.data(), integer);
+    } else {
+      bool success = false;
+      uint64_t integer = valobj.GetValueAsUnsigned(0, &success);
+      if (success)
+        formatted = llvm::formatv(llvm_format.data(), integer);
+    }
+  }
+
+  if (formatted.empty())
+    return false;
+
+  s.Write(formatted.data(), formatted.size());
+  return true;
+}
+
 static bool DumpValue(Stream &s, const SymbolContext *sc,
                       const ExecutionContext *exe_ctx,
                       const FormatEntity::Entry &entry, ValueObject *valobj) {
@@ -728,9 +761,12 @@ static bool DumpValue(Stream &s, const SymbolContext *sc,
     return RunScriptFormatKeyword(s, sc, exe_ctx, valobj, entry.string.c_str());
   }
 
-  llvm::StringRef subpath(entry.string);
+  auto split = llvm::StringRef(entry.string).split(':');
+  auto subpath = split.first;
+  auto llvm_format = split.second;
+
   // simplest case ${var}, just print valobj's value
-  if (entry.string.empty()) {
+  if (subpath.empty()) {
     if (entry.printf_format.empty() && entry.fmt == eFormatDefault &&
         entry.number == ValueObject::eValueObjectRepresentationStyleValue)
       was_plain_var = true;
@@ -739,22 +775,19 @@ static bool DumpValue(Stream &s, const SymbolContext *sc,
     target = valobj;
   } else // this is ${var.something} or multiple .something nested
   {
-    if (entry.string[0] == '[')
+    if (subpath[0] == '[')
       was_var_indexed = true;
     ScanBracketedRange(subpath, close_bracket_index,
                        var_name_final_if_array_range, index_lower,
                        index_higher);
 
     Status error;
 
-    const std::string &expr_path = entry.string;
-
-    LLDB_LOGF(log, "[Debugger::FormatPrompt] symbol to expand: %s",
-              expr_path.c_str());
+    LLDB_LOG(log, "[Debugger::FormatPrompt] symbol to expand: {0}", subpath);
 
     target =
         valobj
-            ->GetValueForExpressionPath(expr_path.c_str(), &reason_to_stop,
+            ->GetValueForExpressionPath(subpath, &reason_to_stop,
                                         &final_value_type, options, &what_next)
             .get();
 
@@ -883,8 +916,18 @@ static bool DumpValue(Stream &s, const SymbolContext *sc,
   }
 
   if (!is_array_range) {
-    LLDB_LOGF(log,
-              "[Debugger::FormatPrompt] dumping ordinary printable output");
+    if (!llvm_format.empty()) {
+      if (DumpValueWithLLVMFormat(s, llvm_format, *target)) {
+        LLDB_LOGF(log, "dumping using llvm format");
+        return true;
+      } else {
+        LLDB_LOG(
+            log,
+            "empty output using llvm format '{0}' - with type info flags {1}",
+            entry.printf_format, target->GetTypeInfo());
+      }
+    }
+    LLDB_LOGF(log, "dumping ordinary printable output");
     return target->DumpPrintableRepresentation(s, val_obj_display,
                                                custom_format);
   } else {
@@ -2227,6 +2270,13 @@ static Status ParseInternal(llvm::StringRef &format, Entry &parent_entry,
           if (error.Fail())
             return error;
 
+          auto [_, llvm_format] = llvm::StringRef(entry.string).split(':');
+          if (!LLVMFormatPattern.match(llvm_format)) {
+            error.SetErrorStringWithFormat("invalid llvm format: '%s'",
+                                           llvm_format.data());
+            return error;
+          }
+
           if (verify_is_thread_id) {
             if (entry.type != Entry::Type::ThreadID &&
                 entry.type != Entry::Type::ThreadProtocolID) {

lldb/test/API/functionalities/data-formatter/custom-printf-summary/Makefile

@@ -0,0 +1,2 @@
+C_SOURCES := main.c
+include Makefile.rules

lldb/test/API/functionalities/data-formatter/custom-printf-summary/TestCustomSummaryLLVMFormat.py

@@ -0,0 +1,20 @@
+import lldb
+from lldbsuite.test.lldbtest import *
+import lldbsuite.test.lldbutil as lldbutil
+
+
+class TestCase(TestBase):
+    def test_raw_bytes(self):
+        self.build()
+        lldbutil.run_to_source_breakpoint(self, "break here", lldb.SBFileSpec("main.c"))
+        self.runCmd("type summary add -s '${var.ubyte:x-2}${var.sbyte:x-2}!' Bytes")
+        self.expect("v bytes", substrs=[" = 3001!"])
+
+    def test_bad_format(self):
+        self.build()
+        lldbutil.run_to_source_breakpoint(self, "break here", lldb.SBFileSpec("main.c"))
+        self.expect(
+            "type summary add -s '${var.ubyte:y}!' Bytes",
+            error=True,
+            substrs=["invalid llvm format"],
+        )

lldb/test/API/functionalities/data-formatter/custom-printf-summary/main.c

@@ -0,0 +1,13 @@
+#include <stdint.h>
+#include <stdio.h>
+
+struct Bytes {
+  uint8_t ubyte;
+  int8_t sbyte;
+};
+
+int main() {
+  struct Bytes bytes = {0x30, 0x01};
+  (void)bytes;
+  printf("break here\n");
+}