clang-tools-extra/pseudo/tool/HTMLForest.cpp

   1 //===-- HTMLForest.cpp - browser-based parse forest explorer
   2 //---------------===//
   3 //
   4 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   5 // See https://llvm.org/LICENSE.txt for license information.
   6 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   7 //
   8 //===----------------------------------------------------------------------===//
   9 //
  10 // The plain text forest node dump (clang-pseudo -print-forest) is useful but
  11 // hard to reconcile with the code being examined, especially when it is large.
  12 //
  13 // HTMLForest produces a self-contained HTML file containing both the code and
  14 // the forest representation, linking them interactively with javascript.
  15 // At any given time, a single parse tree is shown (ambiguities resolved).
  16 // The user can switch between ambiguous alternatives.
  17 //
  18 // +-------+---------------+
  19 // |       |        +-----+|
  20 // | #tree |  #code |#info||
  21 // |       |        +-----+|
  22 // |       |               |
  23 // +-------+---------------+
  24 //
  25 // #tree is a hierarchical view of the nodes (nested <ul>s), like -print-forest.
  26 // (It is a simple tree, not a DAG, because ambiguities have been resolved).
  27 // Like -print-forest, trivial sequences are collapsed (expression~IDENTIFIER).
  28 //
  29 // #code is the source code, annotated with <span>s marking the node ranges.
  30 // These spans are usually invisible (exception: ambiguities are marked), but
  31 // they are used to show and change the selection.
  32 //
  33 // #info is a floating box that shows details of the currently selected node:
  34 //  - rule (for sequence nodes). Abbreviated rules are also shown.
  35 //  - alternatives (for ambiguous nodes). The user can choose an alternative.
  36 //  - ancestors. The parent nodes show how this node fits in translation-unit.
  37 //
  38 // There are two types of 'active' node:
  39 //  - *highlight* is what the cursor is over, and is colored blue.
  40 //    Near ancestors are shaded faintly (onion-skin) to show local structure.
  41 //  - *selection* is set by clicking.
  42 //    The #info box shows the selection, and selected nodes have a dashed ring.
  43 //
  44 //===----------------------------------------------------------------------===//
  45
  46 #include "clang-pseudo/Forest.h"
  47 #include "clang-pseudo/grammar/Grammar.h"
  48 #include "llvm/ADT/StringExtras.h"
  49 #include "llvm/Support/JSON.h"
  50 #include "llvm/Support/raw_ostream.h"
  51 namespace clang {
  52 namespace pseudo {
  53 namespace {
  54
  55 // Defines const char HTMLForest_css[] = "...contents of HTMLForest.css..."; etc
  56 #include "HTMLForestResources.inc"
  57
  58 struct Writer {
  59   llvm::raw_ostream &Out;
  60   const Grammar &G;
  61   const ForestNode &Root;
  62   const TokenStream &Stream;
  63
  64   void write() {
  65     Out << "<!doctype html>\n";
  66     tag("html", [&] {
  67       tag("head", [&] {
  68         tag("title", [&] { Out << "HTMLForest"; });
  69         tag("script", [&] { Out << HTMLForest_js; });
  70         tag("style", [&] { Out << HTMLForest_css; });
  71         tag("script", [&] {
  72           Out << "var forest=";
  73           writeForestJSON();
  74           Out << ";";
  75         });
  76         tag("pre id='hidden-code' hidden", [&] { writeCode(); });
  77       });
  78       tag("body", [&] { Out << HTMLForest_html; });
  79     });
  80   }
  81
  82   void writeCode();
  83   void writeForestJSON();
  84   void tag(llvm::StringRef Opener, llvm::function_ref<void()> Body) {
  85     Out << "<" << Opener << ">";
  86     Body();
  87     Out << "</" << Opener.split(' ').first << ">\n";
  88   }
  89 };
  90
  91 void Writer::writeCode() {
  92   // This loop (whitespace logic) is cribbed from TokenStream::Print.
  93   bool FirstToken = true;
  94   unsigned LastLine = -1;
  95   StringRef LastText;
  96   for (const auto &T : Stream.tokens()) {
  97     StringRef Text = T.text();
  98     if (FirstToken) {
  99       FirstToken = false;
 100     } else if (T.Line == LastLine) {
 101       if (LastText.data() + LastText.size() != Text.data())
 102         Out << ' ';
 103     } else {
 104       Out << " \n"; // Extra space aids selection.
 105       Out.indent(T.Indent);
 106     }
 107     Out << "<span class='token' id='t" << Stream.index(T) << "'>";
 108     llvm::printHTMLEscaped(Text, Out);
 109     Out << "</span>";
 110     LastLine = T.Line;
 111     LastText = Text;
 112   }
 113   if (!FirstToken)
 114     Out << '\n';
 115 }
 116
 117 // Writes a JSON array of forest nodes. Items are e.g.:
 118 //   {kind:'sequence', symbol:'compound-stmt', children:[5,8,33],
 119 //   rule:'compound-stmt := ...'} {kind:'terminal', symbol:'VOID', token:'t52'}
 120 //   {kind:'ambiguous', symbol:'type-specifier', children:[3,100] selected:3}
 121 //   {kind:'opaque', symbol:'statement-seq', firstToken:'t5', lastToken:'t6'}
 122 void Writer::writeForestJSON() {
 123   // This is the flat array of nodes: the index into this array is the node ID.
 124   std::vector<std::pair<const ForestNode *, /*End*/ Token::Index>> Sequence;
 125   llvm::DenseMap<const ForestNode *, unsigned> Index;
 126   auto AssignID = [&](const ForestNode *N, Token::Index End) -> unsigned {
 127     auto R = Index.try_emplace(N, Sequence.size());
 128     if (R.second)
 129       Sequence.push_back({N, End});
 130     return R.first->second;
 131   };
 132   AssignID(&Root, Stream.tokens().size());
 133   auto TokenID = [](Token::Index I) { return ("t" + llvm::Twine(I)).str(); };
 134
 135   llvm::json::OStream Out(this->Out, 2);
 136   Out.array([&] {
 137     for (unsigned I = 0; I < Sequence.size(); ++I) {
 138       const ForestNode *N = Sequence[I].first;
 139       Token::Index End = Sequence[I].second;
 140       Out.object([&] {
 141         Out.attribute("symbol", G.symbolName(N->symbol()));
 142         switch (N->kind()) {
 143         case ForestNode::Terminal:
 144           Out.attribute("kind", "terminal");
 145           Out.attribute("token", TokenID(N->startTokenIndex()));
 146           break;
 147         case ForestNode::Sequence:
 148           Out.attribute("kind", "sequence");
 149           Out.attribute("rule", G.dumpRule(N->rule()));
 150           break;
 151         case ForestNode::Ambiguous:
 152           Out.attribute("kind", "ambiguous");
 153           Out.attribute("selected", AssignID(N->children().front(), End));
 154           break;
 155         case ForestNode::Opaque:
 156           Out.attribute("kind", "opaque");
 157           Out.attribute("firstToken", TokenID(N->startTokenIndex()));
 158           // [firstToken, lastToken] is a closed range.
 159           // If empty, lastToken is omitted.
 160           if (N->startTokenIndex() != End)
 161             Out.attribute("lastToken", TokenID(End - 1));
 162           break;
 163         }
 164         auto Children = N->children();
 165         if (!Children.empty())
 166           Out.attributeArray("children", [&] {
 167             for (unsigned I = 0; I < Children.size(); ++I)
 168               Out.value(AssignID(Children[I],
 169                                  I + 1 == Children.size()
 170                                      ? End
 171                                      : Children[I + 1]->startTokenIndex()));
 172           });
 173       });
 174     }
 175   });
 176 }
 177
 178 } // namespace
 179
 180 // We only accept the derived stream here.
 181 // FIXME: allow the original stream instead?
 182 void writeHTMLForest(llvm::raw_ostream &OS, const Grammar &G,
 183                      const ForestNode &Root, const TokenStream &Stream) {
 184   Writer{OS, G, Root, Stream}.write();
 185 }
 186
 187 } // namespace pseudo
 188 } // namespace clang