================
@@ -0,0 +1,234 @@
+//===----------------------------------------------------------------------===//
+//
+// 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 "Markdown.h"
+#include "llvm/ADT/SmallString.h"
+#include "llvm/ADT/SmallVector.h"
+
+namespace clang::doc::markdown {
+
+//===----------------------------------------------------------------------===//
+// Inline node print/dump
+//===----------------------------------------------------------------------===//
+
+LLVM_DUMP_METHOD void InlineNode::dump() const { print(llvm::errs()); }
+
+void TextNode::print(llvm::raw_ostream &OS) const {
+ OS << "TextNode: " << getText() << "\n";
+}
+
+void InlineCodeNode::print(llvm::raw_ostream &OS) const {
+ OS << "InlineCodeNode: " << getCode() << "\n";
+}
+
+void EmphasisNode::print(llvm::raw_ostream &OS) const {
+ OS << "EmphasisNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+void StrongNode::print(llvm::raw_ostream &OS) const {
+ OS << "StrongNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+//===----------------------------------------------------------------------===//
+// Block node print/dump
+//===----------------------------------------------------------------------===//
+
+LLVM_DUMP_METHOD void BlockNode::dump() const { print(llvm::errs()); }
+
+void ParagraphNode::print(llvm::raw_ostream &OS) const {
+ OS << "ParagraphNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+void HeadingNode::print(llvm::raw_ostream &OS) const {
+ OS << "HeadingNode: level=" << getLevel() << "\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+void FencedCodeNode::print(llvm::raw_ostream &OS) const {
+ OS << "FencedCodeNode: lang=" << getLang() << "\n" << getCode() << "\n";
+}
+
+void ListItemNode::print(llvm::raw_ostream &OS) const {
+ OS << "ListItemNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+LLVM_DUMP_METHOD void ListItemNode::dump() const { print(llvm::errs()); }
+
+void UnorderedListNode::print(llvm::raw_ostream &OS) const {
+ OS << "UnorderedListNode\n";
+ for (const auto &Item : Items)
+ Item.print(OS);
+}
+
+void OrderedListNode::print(llvm::raw_ostream &OS) const {
+ OS << "OrderedListNode: start=" << getStart() << "\n";
+ for (const auto &Item : Items)
+ Item.print(OS);
+}
+
+void BlockQuoteNode::print(llvm::raw_ostream &OS) const {
+ OS << "BlockQuoteNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+void ThematicBreakNode::print(llvm::raw_ostream &OS) const {
+ OS << "ThematicBreakNode\n";
+}
+
+void DocumentNode::print(llvm::raw_ostream &OS) const {
+ OS << "DocumentNode\n";
+ for (const auto &Child : Children)
+ Child.print(OS);
+}
+
+//===----------------------------------------------------------------------===//
+// Parser
+//===----------------------------------------------------------------------===//
+
+static bool isListMarker(llvm::StringRef Line) {
+ return Line.starts_with("- ") || Line.starts_with("* ") ||
+ Line.starts_with("+ ");
+}
+
+static bool isThematicBreak(llvm::StringRef Line) {
+ if (Line.empty())
+ return false;
+ char Marker = Line[0];
+ if (Marker != '-' && Marker != '*' && Marker != '_')
+ return false;
+ llvm::SmallString<8> Allowed;
+ Allowed += Marker;
+ Allowed += ' ';
+ if (Line.find_first_not_of(llvm::StringRef(Allowed)) !=
llvm::StringRef::npos)
+ return false;
+ return Line.count(Marker) >= 3;
+}
+
+DocumentNode *parseMarkdown(llvm::StringRef Text, ASTContext &Ctx) {
+ auto *Doc = Ctx.allocate<DocumentNode>();
+ Ctx.setRoot(Doc);
+
+ llvm::SmallVector<llvm::StringRef> Lines;
+ Text.split(Lines, '\n');
+
+ size_t I = 0;
+ while (I < Lines.size()) {
+ llvm::StringRef Line = Lines[I].trim();
+
+ if (Line.empty()) {
+ ++I;
+ continue;
+ }
+
+ // Thematic break must come before list and fenced code checks since
+ // "---" and "- - -" would otherwise match those patterns first.
+ if (isThematicBreak(Line)) {
+ auto *Node = Ctx.allocate<ThematicBreakNode>();
+ Doc->Children.push_back(*Node);
+ ++I;
+ continue;
+ }
+
+ // Fenced code block
+ if (Line.starts_with("```") || Line.starts_with("~~~")) {
+ char Fence = Line[0];
+ llvm::StringRef Lang = Line.drop_front(3).trim();
+ ++I;
+ llvm::SmallString<256> Code;
+ while (I < Lines.size()) {
+ llvm::StringRef Trimmed = Lines[I].trim();
+ if (Trimmed.size() >= 3 && Trimmed[0] == Fence && Trimmed[1] == Fence
&&
+ Trimmed[2] == Fence) {
+ ++I;
+ break;
+ }
+ if (!Code.empty())
+ Code += '\n';
+ Code += Lines[I];
+ ++I;
+ }
+ auto *Node = Ctx.allocate<FencedCodeNode>(Lang, Ctx.internString(Code));
----------------
ilovepi wrote:
I understand why you're allocating the Node, but why is it necessary to intern
the String? Based on earlier discussions, you only need a read only reference
to strings that outlive the AST context. You have that, from what I can tell,
though it may need to be explained in the documentation. Generally, we don't
want to be reallocating and copying substrings all over the place in
`clang-doc` or `clang`.
```suggestion
auto *Node = Ctx.allocate<FencedCodeNode>(Lang, Code);
```
https://github.com/llvm/llvm-project/pull/205609
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
