Add DependencyExtractor skeleton.

Summary:
Add the DependencyExtractor class, which takes an AST and extracts
the dependencies using RecursiveVisitor.

It currently does nothing.

Create a CLI for it to be able to test it.

Reviewed By: tmikov

Differential Revision: D21091835

fbshipit-source-id: a7ac6caa021ef610000acd486976ad6252b852b7

CMakeLists.txt

@@ -580,6 +580,7 @@ list(APPEND HERMES_TEST_DEPS
   hbc-attribute
   hbc-deltaprep
   hbc-diff
+  dependency-extractor
   )
 
 set(HERMES_LIT_TEST_PARAMS
@@ -590,6 +591,7 @@ set(HERMES_LIT_TEST_PARAMS
   jit_enabled=${HERMESVM_JIT}
   jit_disassembler_enabled=${HERMESVM_JIT_DISASSEMBLER}
   hbc_deltaprep=${HERMES_TOOLS_OUTPUT_DIR}/hbc-deltaprep
+  dependency_extractor=${HERMES_TOOLS_OUTPUT_DIR}/dependency-extractor
   FileCheck=${HERMES_TOOLS_OUTPUT_DIR}//FileCheck
   hermes=${HERMES_TOOLS_OUTPUT_DIR}/hermes
   hermesc=${HERMES_TOOLS_OUTPUT_DIR}/hermesc

include/hermes/DependencyExtractor/DependencyExtractor.h

@@ -0,0 +1,74 @@
+/*
+ * Copyright (c) Facebook, Inc. and its affiliates.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+#ifndef HERMES_AST_DEPENDENCYEXTRACTOR_H
+#define HERMES_AST_DEPENDENCYEXTRACTOR_H
+
+#include "hermes/AST/Context.h"
+#include "hermes/AST/ESTree.h"
+
+namespace hermes {
+
+/// The kind of dependency, used to distinguish between the different ways to
+/// pull an external module into the file.
+enum class DependencyKind {
+  // import {x} from 'source';
+  ESM,
+  // import type {x} from 'source';
+  Type,
+  // require('source');
+  Require,
+  // import('source');
+  Async,
+  // JSResource('source');
+  Resource,
+  // PrefetchedJSResource('source');
+  PrefetchedResource,
+  // graphql`fragment source` and others;
+  GraphQL,
+};
+
+inline const char *dependencyKindStr(DependencyKind kind) {
+  switch (kind) {
+    case DependencyKind::ESM:
+      return "ESM";
+    case DependencyKind::Type:
+      return "Type";
+    case DependencyKind::Require:
+      return "Require";
+    case DependencyKind::Async:
+      return "Async";
+    case DependencyKind::Resource:
+      return "Resource";
+    case DependencyKind::PrefetchedResource:
+      return "PrefetchedResource";
+    case DependencyKind::GraphQL:
+      return "GraphQL";
+  }
+}
+
+/// Represents a single dependency from a source file.
+struct Dependency {
+  /// The target file name of the dependency.
+  /// For example, `require('name')` is used to indicate a Require dependency on
+  /// 'name'.
+  std::string name;
+
+  /// The kind of the dependency.
+  DependencyKind kind;
+};
+
+/// Extract any dependencies in \p node and its children.
+/// \return a list of Dependency indicating what the target is, as well as the
+/// kind of dependency.
+std::vector<Dependency> extractDependencies(
+    Context &astContext,
+    ESTree::Node *node);
+
+} // namespace hermes
+
+#endif

lib/CMakeLists.txt

@@ -53,6 +53,7 @@ add_subdirectory(VM)
 add_subdirectory(Inst)
 add_subdirectory(FrontEndDefs)
 add_subdirectory(CompilerDriver)
+add_subdirectory(DependencyExtractor)
 add_subdirectory(ADT)
 add_subdirectory(AST)
 add_subdirectory(FlowParser)

lib/DependencyExtractor/CMakeLists.txt

@@ -0,0 +1,15 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+set(source_files
+  DependencyExtractor.cpp
+)
+
+add_hermes_library(hermesDependencyExtractor STATIC ${source_files}
+  LINK_LIBS
+  hermesAST
+  hermesParser
+  hermesSupport
+)

lib/DependencyExtractor/DependencyExtractor.cpp

@@ -0,0 +1,52 @@
+/*
+ * Copyright (c) Facebook, Inc. and its affiliates.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+#include "hermes/DependencyExtractor/DependencyExtractor.h"
+
+#include "hermes/AST/RecursiveVisitor.h"
+
+namespace hermes {
+
+using namespace hermes::ESTree;
+
+/// Visitor for extracting dependencies from a given node.
+/// Reports errors on encountering nodes which should introduce dependencies
+/// but are malformed.
+class DependencyExtractor {
+  /// Storage for the extracted dependencies.
+  std::vector<Dependency> deps_{};
+
+  /// SourceErrorManager for reporting errors, e.g. for invalid require() calls.
+  SourceErrorManager &sm_;
+
+ public:
+  DependencyExtractor(Context &astContext)
+      : sm_(astContext.getSourceErrorManager()) {}
+
+  std::vector<Dependency> &getDeps() {
+    return deps_;
+  }
+
+  /// Perform the extraction on whole AST.
+  void doIt(Node *rootNode) {
+    visitESTreeNode(*this, rootNode);
+  }
+
+  /// Stub which catches all nodes we don't need to directly extract
+  /// dependencies from.
+  void visit(Node *node) {
+    visitESTreeChildren(*this, node);
+  }
+};
+
+std::vector<Dependency> extractDependencies(Context &astContext, Node *node) {
+  DependencyExtractor extract{astContext};
+  extract.doIt(node);
+  return std::move(extract.getDeps());
+}
+
+} // namespace hermes

test/lit.cfg

@@ -78,3 +78,4 @@ config.substitutions.append(("%hdb", lit_config.params["hdb"].replace('\\', '/')
 config.substitutions.append(("%hbcdump", lit_config.params["hbcdump"].replace('\\', '/')))
 config.substitutions.append(("%hbc-deltaprep", lit_config.params["hbc_deltaprep"].replace('\\', '/')))
 config.substitutions.append(("%hbc-diff", lit_config.params["hbc_diff"].replace('\\', '/')))
+config.substitutions.append(("%dependency-extractor", lit_config.params["dependency_extractor"].replace('\\', '/')))

tools/CMakeLists.txt

@@ -15,6 +15,7 @@ add_subdirectory(hbc-attribute)
 add_subdirectory(jsi)
 add_subdirectory(emhermesc)
 add_subdirectory(fuzzers)
+add_subdirectory(dependency-extractor)
 
 if (HERMESVM_API_TRACE)
   add_subdirectory(synth)

tools/dependency-extractor/CMakeLists.txt

@@ -0,0 +1,17 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+set(HERMES_LINK_COMPONENTS LLVHSupport)
+
+add_hermes_tool(dependency-extractor
+  dependency-extractor.cpp
+  ${ALL_HEADER_FILES}
+  )
+
+target_link_libraries(dependency-extractor
+  hermesAST
+  hermesDependencyExtractor
+  hermesParser
+)

tools/dependency-extractor/dependency-extractor.cpp

@@ -0,0 +1,70 @@
+/*
+ * Copyright (c) Facebook, Inc. and its affiliates.
+ *
+ * This source code is licensed under the MIT license found in the
+ * LICENSE file in the root directory of this source tree.
+ */
+
+#include "hermes/DependencyExtractor/DependencyExtractor.h"
+#include "hermes/Parser/JSParser.h"
+
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/FileSystem.h"
+#include "llvm/Support/InitLLVM.h"
+#include "llvm/Support/MemoryBuffer.h"
+#include "llvm/Support/PrettyStackTrace.h"
+#include "llvm/Support/Signals.h"
+#include "llvm/Support/raw_ostream.h"
+
+static llvm::cl::opt<std::string> InputFilename(
+    llvm::cl::desc("input file"),
+    llvm::cl::Positional);
+
+using namespace hermes;
+
+int main(int argc, char **argv) {
+  // Normalize the arg vector.
+  llvm::InitLLVM initLLVM(argc, argv);
+  llvm::sys::PrintStackTraceOnErrorSignal("dependency-extractor");
+  llvm::PrettyStackTraceProgram X(argc, argv);
+  llvm::llvm_shutdown_obj Y;
+  llvm::cl::ParseCommandLineOptions(
+      argc, argv, "Hermes JS dependency extractor\n");
+
+  llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> fileBufOrErr =
+      llvm::MemoryBuffer::getFile(InputFilename);
+
+  if (!fileBufOrErr) {
+    llvm::errs() << "Error: fail to open file: " << InputFilename << ": "
+                 << fileBufOrErr.getError().message() << "\n";
+    return -1;
+  }
+
+  auto fileBuf = std::move(fileBufOrErr.get());
+
+  auto context = std::make_shared<Context>();
+#if HERMES_PARSE_JSX
+  context->setParseJSX(true);
+#endif
+#if HERMES_PARSE_FLOW
+  context->setParseFlow(true);
+#endif
+
+  int fileBufId =
+      context->getSourceErrorManager().addNewSourceBuffer(std::move(fileBuf));
+  auto mode = parser::FullParse;
+
+  parser::JSParser jsParser(*context, fileBufId, mode);
+  llvm::Optional<ESTree::ProgramNode *> parsedJs = jsParser.parse();
+
+  if (!parsedJs)
+    return -1;
+
+  auto deps = extractDependencies(*context, *parsedJs);
+  for (const auto &dep : deps) {
+    llvm::outs() << dependencyKindStr(dep.kind) << " | " << dep.name << '\n';
+  }
+
+  return 0;
+}