[YAML] Add support for non-printable characters

LLVM IR function names which disable mangling start with '\01'
(https://www.llvm.org/docs/LangRef.html#identifiers).

When an identifier like "\01@abc@" gets dumped to MIR, it is quoted, but
only with single quotes.

http://www.yaml.org/spec/1.2/spec.html#id2770814:

"The allowed character range explicitly excludes the C0 control block
allowed), the surrogate block #xD800-#xDFFF, #xFFFE, and #xFFFF."

http://www.yaml.org/spec/1.2/spec.html#id2776092:

"All non-printable characters must be escaped.
[...]
Note that escape sequences are only interpreted in double-quoted scalars."

This patch adds support for printing escaped non-printable characters
between double quotes if needed.

Should also fix PR31743.

Differential Revision: https://reviews.llvm.org/D41290

llvm-svn: 320996

7 files changed, 55 insertions, 22 deletions

diff --git a/llvm/lib/ObjectYAML/CodeViewYAMLDebugSections.cpp b/llvm/lib/ObjectYAML/CodeViewYAMLDebugSections.cpp
index 60b0ea28030..6debd8ab0c6 100644
--- a/llvm/lib/ObjectYAML/CodeViewYAMLDebugSections.cpp
+++ b/llvm/lib/ObjectYAML/CodeViewYAMLDebugSections.cpp
@@ -66,7 +66,7 @@ LLVM_YAML_IS_SEQUENCE_VECTOR(CrossModuleExport)
 LLVM_YAML_IS_SEQUENCE_VECTOR(YAMLCrossModuleImport)
 LLVM_YAML_IS_SEQUENCE_VECTOR(YAMLFrameData)
 
-LLVM_YAML_DECLARE_SCALAR_TRAITS(HexFormattedString, false)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(HexFormattedString, QuotingType::None)
 LLVM_YAML_DECLARE_ENUM_TRAITS(DebugSubsectionKind)
 LLVM_YAML_DECLARE_ENUM_TRAITS(FileChecksumKind)
 LLVM_YAML_DECLARE_BITSET_TRAITS(LineFlags)
diff --git a/llvm/lib/ObjectYAML/CodeViewYAMLSymbols.cpp b/llvm/lib/ObjectYAML/CodeViewYAMLSymbols.cpp
index dbe4e2a6d6f..199a65a2870 100644
--- a/llvm/lib/ObjectYAML/CodeViewYAMLSymbols.cpp
+++ b/llvm/lib/ObjectYAML/CodeViewYAMLSymbols.cpp
@@ -42,8 +42,8 @@ using namespace llvm::yaml;
 LLVM_YAML_IS_FLOW_SEQUENCE_VECTOR(TypeIndex)
 
 // We only need to declare these, the definitions are in CodeViewYAMLTypes.cpp
-LLVM_YAML_DECLARE_SCALAR_TRAITS(APSInt, false)
-LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeIndex, false)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(APSInt, QuotingType::None)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeIndex, QuotingType::None)
 
 LLVM_YAML_DECLARE_ENUM_TRAITS(SymbolKind)
 LLVM_YAML_DECLARE_ENUM_TRAITS(FrameCookieKind)
@@ -62,7 +62,7 @@ LLVM_YAML_DECLARE_ENUM_TRAITS(ThunkOrdinal)
 
 LLVM_YAML_STRONG_TYPEDEF(StringRef, TypeName)
 
-LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeName, true)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeName, QuotingType::Single)
 
 StringRef ScalarTraits<TypeName>::input(StringRef S, void *V, TypeName &T) {
   return ScalarTraits<StringRef>::input(S, V, T.value);
diff --git a/llvm/lib/ObjectYAML/CodeViewYAMLTypes.cpp b/llvm/lib/ObjectYAML/CodeViewYAMLTypes.cpp
index 887328cf508..ba4ad9382ce 100644
--- a/llvm/lib/ObjectYAML/CodeViewYAMLTypes.cpp
+++ b/llvm/lib/ObjectYAML/CodeViewYAMLTypes.cpp
@@ -48,8 +48,8 @@ LLVM_YAML_IS_SEQUENCE_VECTOR(OneMethodRecord)
 LLVM_YAML_IS_SEQUENCE_VECTOR(VFTableSlotKind)
 LLVM_YAML_IS_FLOW_SEQUENCE_VECTOR(TypeIndex)
 
-LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeIndex, false)
-LLVM_YAML_DECLARE_SCALAR_TRAITS(APSInt, false)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(TypeIndex, QuotingType::None)
+LLVM_YAML_DECLARE_SCALAR_TRAITS(APSInt, QuotingType::None)
 
 LLVM_YAML_DECLARE_ENUM_TRAITS(TypeLeafKind)
 LLVM_YAML_DECLARE_ENUM_TRAITS(PointerToMemberRepresentation)
diff --git a/llvm/lib/ObjectYAML/MachOYAML.cpp b/llvm/lib/ObjectYAML/MachOYAML.cpp
index 85079f2605f..e00a4ea9307 100644
--- a/llvm/lib/ObjectYAML/MachOYAML.cpp
+++ b/llvm/lib/ObjectYAML/MachOYAML.cpp
@@ -52,7 +52,9 @@ StringRef ScalarTraits<char_16>::input(StringRef Scalar, void *, char_16 &Val) {
   return StringRef();
 }
 
-bool ScalarTraits<char_16>::mustQuote(StringRef S) { return needsQuotes(S); }
+QuotingType ScalarTraits<char_16>::mustQuote(StringRef S) {
+  return needsQuotes(S);
+}
 
 void ScalarTraits<uuid_t>::output(const uuid_t &Val, void *, raw_ostream &Out) {
   Out.write_uuid(Val);
@@ -75,7 +77,9 @@ StringRef ScalarTraits<uuid_t>::input(StringRef Scalar, void *, uuid_t &Val) {
   return StringRef();
 }
 
-bool ScalarTraits<uuid_t>::mustQuote(StringRef S) { return needsQuotes(S); }
+QuotingType ScalarTraits<uuid_t>::mustQuote(StringRef S) {
+  return needsQuotes(S);
+}
 
 void MappingTraits<MachOYAML::FileHeader>::mapping(
     IO &IO, MachOYAML::FileHeader &FileHdr) {
diff --git a/llvm/lib/Support/Statistic.cpp b/llvm/lib/Support/Statistic.cpp
index 23718bb0e9c..544ae2d0983 100644
--- a/llvm/lib/Support/Statistic.cpp
+++ b/llvm/lib/Support/Statistic.cpp
@@ -168,9 +168,10 @@ void llvm::PrintStatisticsJSON(raw_ostream &OS) {
   const char *delim = "";
   for (const Statistic *Stat : Stats.Stats) {
     OS << delim;
-    assert(!yaml::needsQuotes(Stat->getDebugType()) &&
+    assert(yaml::needsQuotes(Stat->getDebugType()) == yaml::QuotingType::None &&
            "Statistic group/type name is simple.");
-    assert(!yaml::needsQuotes(Stat->getName()) && "Statistic name is simple");
+    assert(yaml::needsQuotes(Stat->getName()) == yaml::QuotingType::None &&
+           "Statistic name is simple");
     OS << "\t\"" << Stat->getDebugType() << '.' << Stat->getName() << "\": "
        << Stat->getValue();
     delim = ",\n";
diff --git a/llvm/lib/Support/Timer.cpp b/llvm/lib/Support/Timer.cpp
index 3386f2660f3..0c85faecca8 100644
--- a/llvm/lib/Support/Timer.cpp
+++ b/llvm/lib/Support/Timer.cpp
@@ -362,8 +362,10 @@ void TimerGroup::printAll(raw_ostream &OS) {
 
 void TimerGroup::printJSONValue(raw_ostream &OS, const PrintRecord &R,
                                 const char *suffix, double Value) {
-  assert(!yaml::needsQuotes(Name) && "TimerGroup name needs no quotes");
-  assert(!yaml::needsQuotes(R.Name) && "Timer name needs no quotes");
+  assert(yaml::needsQuotes(Name) == yaml::QuotingType::None &&
+         "TimerGroup name needs no quotes");
+  assert(yaml::needsQuotes(R.Name) == yaml::QuotingType::None &&
+         "Timer name needs no quotes");
   OS << "\t\"time." << Name << '.' << R.Name << suffix << "\": " << Value;
 }
 
diff --git a/llvm/lib/Support/YAMLTraits.cpp b/llvm/lib/Support/YAMLTraits.cpp
index a80adfda830..05ca40f0301 100644
--- a/llvm/lib/Support/YAMLTraits.cpp
+++ b/llvm/lib/Support/YAMLTraits.cpp
@@ -19,6 +19,7 @@
 #include "llvm/Support/Format.h"
 #include "llvm/Support/LineIterator.h"
 #include "llvm/Support/MemoryBuffer.h"
+#include "llvm/Support/Unicode.h"
 #include "llvm/Support/YAMLParser.h"
 #include "llvm/Support/raw_ostream.h"
 #include <algorithm>
@@ -330,7 +331,7 @@ void Input::endBitSetScalar() {
   }
 }
 
-void Input::scalarString(StringRef &S, bool) {
+void Input::scalarString(StringRef &S, QuotingType) {
   if (ScalarHNode *SN = dyn_cast<ScalarHNode>(CurrentNode)) {
     S = SN->value();
   } else {
@@ -338,7 +339,7 @@ void Input::scalarString(StringRef &S, bool) {
   }
 }
 
-void Input::blockScalarString(StringRef &S) { scalarString(S, false); }
+void Input::blockScalarString(StringRef &S) { scalarString(S, QuotingType::None); }
 
 void Input::setError(HNode *hnode, const Twine &message) {
   assert(hnode && "HNode must not be NULL");
@@ -617,7 +618,7 @@ void Output::endBitSetScalar() {
   this->outputUpToEndOfLine(" ]");
 }
 
-void Output::scalarString(StringRef &S, bool MustQuote) {
+void Output::scalarString(StringRef &S, QuotingType MustQuote) {
   this->newLineCheck();
   if (S.empty()) {
     // Print '' for the empty string because leaving the field empty is not
@@ -625,27 +626,52 @@ void Output::scalarString(StringRef &S, bool MustQuote) {
     this->outputUpToEndOfLine("''");
     return;
   }
-  if (!MustQuote) {
+  if (MustQuote == QuotingType::None) {
     // Only quote if we must.
     this->outputUpToEndOfLine(S);
     return;
   }
+
   unsigned i = 0;
   unsigned j = 0;
   unsigned End = S.size();
-  output("'"); // Starting single quote.
   const char *Base = S.data();
+
+  const char *const Quote = MustQuote == QuotingType::Single ? "'" : "\"";
+  const char QuoteChar = MustQuote == QuotingType::Single ? '\'' : '"';
+
+  output(Quote); // Starting quote.
+
+  // When using single-quoted strings, any single quote ' must be doubled to be
+  // escaped.
+  // When using double-quoted strings, print \x + hex for non-printable ASCII
+  // characters, and escape double quotes.
   while (j < End) {
-    // Escape a single quote by doubling it.
-    if (S[j] == '\'') {
-      output(StringRef(&Base[i], j - i + 1));
-      output("'");
+    if (S[j] == QuoteChar) {                  // Escape quotes.
+      output(StringRef(&Base[i], j - i));     // "flush".
+      if (MustQuote == QuotingType::Double) { // Print it as \"
+        output(StringLiteral("\\"));
+        output(StringRef(Quote, 1));
+      } else {                       // Single
+        output(StringLiteral("''")); // Print it as ''
+      }
+      i = j + 1;
+    } else if (MustQuote == QuotingType::Double &&
+               !sys::unicode::isPrintable(S[j])) {
+      output(StringRef(&Base[i], j - i)); // "flush"
+      output(StringLiteral("\\x"));
+
+      // Output the byte 0x0F as \x0f.
+      auto FormattedHex = format_hex_no_prefix(S[j], 2);
+      Out << FormattedHex;
+      Column += 4; // one for the '\', one for the 'x', and two for the hex
+
       i = j + 1;
     }
     ++j;
   }
   output(StringRef(&Base[i], j - i));
-  this->outputUpToEndOfLine("'"); // Ending single quote.
+  this->outputUpToEndOfLine(Quote); // Ending quote.
 }
 
 void Output::blockScalarString(StringRef &S) {