Support: Add PathV2 implementation.

llvm-svn: 120329

include/llvm/Support/PathV2.h

@@ -24,6 +24,9 @@
 //
 //===----------------------------------------------------------------------===//
 
+#ifndef LLVM_SYSTEM_PATHV2_H
+#define LLVM_SYSTEM_PATHV2_H
+
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/Twine.h"
 #include "llvm/Support/DataTypes.h"
@@ -64,13 +67,20 @@ namespace path {
 ///
 class const_iterator {
   StringRef Path;      //< The entire path.
-  StringRef Component; //< The current component.
+  StringRef Component; //< The current component. Not necessarily in Path.
+  size_t    Position;  //< The iterators current position within Path.
+
+  // An end iterator has Position = Path.size() + 1.
+  friend const_iterator begin(const StringRef &path);
+  friend const_iterator end(const StringRef &path);
 
 public:
   typedef const StringRef value_type;
+  typedef ptrdiff_t difference_type;
   typedef value_type &reference;
   typedef value_type *pointer;
   typedef std::bidirectional_iterator_tag iterator_category;
+
   reference operator*() const;
   pointer   operator->() const;
   const_iterator &operator++();    // preincrement
@@ -79,11 +89,18 @@ public:
   const_iterator &operator--(int); // postdecrement
   bool operator==(const const_iterator &RHS) const;
   bool operator!=(const const_iterator &RHS) const;
-
-  const_iterator(); //< Default construct end iterator.
-  const_iterator(const StringRef &path);
 };
 
+/// @brief Get begin iterator over \a path.
+/// @param path Input path.
+/// @returns Iterator initialized with the first component of \a path.
+const_iterator begin(const StringRef &path);
+
+/// @brief Get end iterator over \a path.
+/// @param path Input path.
+/// @returns Iterator initialized to the end of \a path.
+const_iterator end(const StringRef &path);
+
 /// @}
 /// @name Lexical Modifiers
 /// @{
@@ -136,7 +153,10 @@ error_code replace_extension(SmallVectorImpl<char> &path,
 /// @param component The component to be appended to \a path.
 /// @returns errc::success if \a component has been appended to \a path,
 ///          otherwise a platform specific error_code.
-error_code append(SmallVectorImpl<char> &path, const Twine &component);
+error_code append(SmallVectorImpl<char> &path, const Twine &a,
+                                               const Twine &b = "",
+                                               const Twine &c = "",
+                                               const Twine &d = "");
 
 /// @brief Append to path.
 ///
@@ -978,3 +998,5 @@ public:
 } // end namespace fs
 } // end namespace sys
 } // end namespace llvm
+
+#endif

lib/Support/CMakeLists.txt

@@ -64,6 +64,7 @@ add_llvm_library(LLVMSupport
   Memory.cpp
   Mutex.cpp
   Path.cpp
+  PathV2.cpp
   Process.cpp
   Program.cpp
   RWMutex.cpp
@@ -79,6 +80,7 @@ add_llvm_library(LLVMSupport
   Unix/Memory.inc
   Unix/Mutex.inc
   Unix/Path.inc
+  Unix/PathV2.inc
   Unix/Process.inc
   Unix/Program.inc
   Unix/RWMutex.inc
@@ -92,6 +94,7 @@ add_llvm_library(LLVMSupport
   Windows/Memory.inc
   Windows/Mutex.inc
   Windows/Path.inc
+  Windows/PathV2.inc
   Windows/Process.inc
   Windows/Program.inc
   Windows/RWMutex.inc

lib/Support/PathV2.cpp

@@ -0,0 +1,409 @@
+//===-- PathV2.cpp - Implement OS Path Concept ------------------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+//  This file implements the operating system PathV2 API.
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/Support/PathV2.h"
+#include "llvm/Support/ErrorHandling.h"
+#include <cctype>
+
+namespace {
+  using llvm::StringRef;
+
+  bool is_separator(const char value) {
+    switch(value) {
+#ifdef LLVM_ON_WIN32
+    case '\\': // fall through
+#endif
+    case '/': return true;
+    default: return false;
+    }
+  }
+
+#ifdef LLVM_ON_WIN32
+  const StringRef separators = "\\/";
+  const char       prefered_separator = '\\';
+#else
+  const StringRef separators = "/";
+  const char      prefered_separator = '/';
+#endif
+
+  StringRef find_first_component(const StringRef  &path) {
+    // Look for this first component in the following order.
+    // * empty (in this case we return an empty string)
+    // * either C: or {//,\\}net.
+    // * {/,\}
+    // * {.,..}
+    // * {file,directory}name
+
+    if (path.empty())
+      return path;
+
+    // C:
+    if (path.size() >= 2 && std::isalpha(path[0]) && path[1] == ':')
+      return StringRef(path.begin(), 2);
+
+    // //net
+    if ((path.size() > 2) &&
+        (path.startswith("\\\\") || path.startswith("//")) &&
+        (path[2] != '\\' && path[2] != '/')) {
+      // Find the next directory separator.
+      size_t end = path.find_first_of("\\/", 2);
+      if (end == StringRef::npos)
+        return path;
+      else
+        return StringRef(path.begin(), end);
+    }
+
+    // {/,\}
+    if (path[0] == '\\' || path[0] == '/')
+      return StringRef(path.begin(), 1);
+
+    if (path.startswith(".."))
+      return StringRef(path.begin(), 2);
+
+    if (path[0] == '.')
+      return StringRef(path.begin(), 1);
+
+    // * {file,directory}name
+    size_t end = path.find_first_of("\\/", 2);
+    if (end == StringRef::npos)
+      return path;
+    else
+      return StringRef(path.begin(), end);
+
+    return StringRef();
+  }
+}
+
+namespace llvm {
+namespace sys  {
+namespace path {
+
+const_iterator begin(const StringRef &path) {
+  const_iterator i;
+  i.Path      = path;
+  i.Component = find_first_component(path);
+  i.Position  = 0;
+  return i;
+}
+
+const_iterator end(const StringRef &path) {
+  const_iterator i;
+  i.Path      = path;
+  i.Position  = path.size();
+  return i;
+}
+
+const_iterator::reference const_iterator::operator*() const {
+  return Component;
+}
+
+const_iterator::pointer const_iterator::operator->() const {
+  return &Component;
+}
+
+const_iterator &const_iterator::operator++() {
+  assert(Position < Path.size() && "Tried to increment past end!");
+
+  // Increment Position to past the current component
+  Position += Component.size();
+
+  // Check for end.
+  if (Position == Path.size()) {
+    Component = StringRef();
+    return *this;
+  }
+
+  // Both POSIX and Windows treat paths that begin with exactly two separators
+  // specially.
+  bool was_net = Component.size() > 2 &&
+    is_separator(Component[0]) &&
+    Component[1] == Component[0] &&
+    !is_separator(Component[2]);
+
+  // Handle separators.
+  if (is_separator(Path[Position])) {
+    // Root dir.
+    if (was_net
+#ifdef LLVM_ON_WIN32
+        // c:/
+        || Component.endswith(":")
+#endif
+        ) {
+      Component = StringRef(Path.begin() + Position, 1);
+      return *this;
+    }
+
+    // Skip extra separators.
+    while (Position != Path.size() &&
+           is_separator(Path[Position])) {
+      ++Position;
+    }
+
+    // Treat trailing '/' as a '.'.
+    if (Position == Path.size()) {
+      --Position;
+      Component = ".";
+      return *this;
+    }
+  }
+
+  // Find next component.
+  size_t end_pos = Path.find_first_of(separators, Position);
+  if (end_pos == StringRef::npos)
+    end_pos = Path.size();
+  Component = StringRef(Path.begin() + Position, end_pos - Position);
+
+  return *this;
+}
+
+bool const_iterator::operator==(const const_iterator &RHS) const {
+  return Path.begin() == RHS.Path.begin() &&
+         Position == RHS.Position;
+}
+
+bool const_iterator::operator!=(const const_iterator &RHS) const {
+  return !(*this == RHS);
+}
+
+error_code root_path(const StringRef &path, StringRef &result) {
+  const_iterator b = begin(path),
+                 pos = b,
+                 e = end(path);
+  if (b != e) {
+    bool has_net = b->size() > 2 && is_separator((*b)[0]) && (*b)[1] == (*b)[0];
+    bool has_drive =
+#ifdef LLVM_ON_WIN32
+      b->endswith(":");
+#else
+      false;
+#endif
+
+    if (has_net || has_drive) {
+      if ((++pos != e) && is_separator((*pos)[0])) {
+        // {C:/,//net/}, so get the first two components.
+        result = StringRef(path.begin(), b->size() + pos->size());
+        return make_error_code(errc::success);
+      } else {
+        // just {C:,//net}, return the first component.
+        result = *b;
+        return make_error_code(errc::success);
+      }
+    }
+
+    // POSIX style root directory.
+    if (is_separator((*b)[0])) {
+      result = *b;
+      return make_error_code(errc::success);
+    }
+
+    // No root_path.
+    result = StringRef();
+    return make_error_code(errc::success);
+  }
+
+  // No path :(.
+  result = StringRef();
+  return make_error_code(errc::success);
+}
+
+error_code root_name(const StringRef &path, StringRef &result) {
+  const_iterator b = begin(path),
+                 e = end(path);
+  if (b != e) {
+    bool has_net = b->size() > 2 && is_separator((*b)[0]) && (*b)[1] == (*b)[0];
+    bool has_drive =
+#ifdef LLVM_ON_WIN32
+      b->endswith(":");
+#else
+      false;
+#endif
+
+    if (has_net || has_drive) {
+      // just {C:,//net}, return the first component.
+      result = *b;
+      return make_error_code(errc::success);
+    }
+  }
+
+  // No path or no name.
+  result = StringRef();
+  return make_error_code(errc::success);
+}
+
+error_code root_directory(const StringRef &path, StringRef &result) {
+  const_iterator b = begin(path),
+                 pos = b,
+                 e = end(path);
+  if (b != e) {
+    bool has_net = b->size() > 2 && is_separator((*b)[0]) && (*b)[1] == (*b)[0];
+    bool has_drive =
+#ifdef LLVM_ON_WIN32
+      b->endswith(":");
+#else
+      false;
+#endif
+
+    if ((has_net || has_drive) &&
+        // {C:,//net}, skip to the next component.
+        (++pos != e) && is_separator((*pos)[0])) {
+      result = *pos;
+      return make_error_code(errc::success);
+    }
+
+    // POSIX style root directory.
+    if (!has_net && is_separator((*b)[0])) {
+      result = *b;
+      return make_error_code(errc::success);
+    }
+  }
+
+  // No path or no root.
+  result = StringRef();
+  return make_error_code(errc::success);
+}
+
+error_code has_root_name(const Twine &path, bool &result) {
+  SmallString<128> storage;
+  StringRef p = path.toStringRef(storage);
+
+  if (error_code ec = root_name(p, p)) return ec;
+  result = !p.empty();
+  return make_error_code(errc::success);
+}
+
+error_code has_root_directory(const Twine &path, bool &result) {
+  SmallString<128> storage;
+  StringRef p = path.toStringRef(storage);
+
+  if (error_code ec = root_directory(p, p)) return ec;
+  result = !p.empty();
+  return make_error_code(errc::success);
+}
+
+error_code relative_path(const StringRef &path, StringRef &result) {
+  StringRef root;
+  if (error_code ec = root_path(path, root)) return ec;
+  result = StringRef(path.begin() + root.size(), path.size() - root.size());
+  return make_error_code(errc::success);
+}
+
+error_code append(SmallVectorImpl<char> &path, const Twine &a,
+                                               const Twine &b,
+                                               const Twine &c,
+                                               const Twine &d) {
+  SmallString<32> a_storage;
+  SmallString<32> b_storage;
+  SmallString<32> c_storage;
+  SmallString<32> d_storage;
+
+  SmallVector<StringRef, 4> components;
+  if (!a.isTriviallyEmpty()) components.push_back(a.toStringRef(a_storage));
+  if (!b.isTriviallyEmpty()) components.push_back(b.toStringRef(b_storage));
+  if (!c.isTriviallyEmpty()) components.push_back(c.toStringRef(c_storage));
+  if (!d.isTriviallyEmpty()) components.push_back(d.toStringRef(d_storage));
+
+  for (SmallVectorImpl<StringRef>::const_iterator i = components.begin(),
+                                                  e = components.end();
+                                                  i != e; ++i) {
+    bool path_has_sep = !path.empty() && is_separator(path[path.size() - 1]);
+    bool component_has_sep = !i->empty() && is_separator((*i)[0]);
+    bool is_root_name;
+    if (error_code ec = has_root_name(*i, is_root_name)) return ec;
+
+    if (path_has_sep) {
+      // Strip separators from beginning of component.
+      size_t loc = i->find_first_not_of(separators);
+      StringRef c = StringRef(i->begin() + loc, i->size() - loc);
+
+      // Append it.
+      path.append(c.begin(), c.end());
+      continue;
+    }
+
+    if (!component_has_sep && !(path.empty() && is_root_name)) {
+      // Add a separator.
+      path.push_back(prefered_separator);
+    }
+
+    path.append(i->begin(), i->end());
+  }
+
+  return make_error_code(errc::success);
+}
+
+error_code make_absolute(SmallVectorImpl<char> &path) {
+  StringRef p(path.data(), path.size());
+
+  bool rootName, rootDirectory;
+  if (error_code ec = has_root_name(p, rootName)) return ec;
+  if (error_code ec = has_root_directory(p, rootDirectory)) return ec;
+
+  // Already absolute.
+  if (rootName && rootDirectory)
+    return make_error_code(errc::success);
+
+  // All of the following conditions will need the current directory.
+  SmallString<128> current_dir;
+  if (error_code ec = current_path(current_dir)) return ec;
+
+  // Relative path. Prepend the current directory.
+  if (!rootName && !rootDirectory) {
+    // Append path to the current directory.
+    if (error_code ec = append(current_dir, p)) return ec;
+    // Set path to the result.
+    path.swap(current_dir);
+    return make_error_code(errc::success);
+  }
+
+  if (!rootName && rootDirectory) {
+    StringRef cdrn;
+    if (error_code ec = root_name(current_dir, cdrn)) return ec;
+    SmallString<128> curDirRootName(cdrn.begin(), cdrn.end());
+    if (error_code ec = append(curDirRootName, p)) return ec;
+    // Set path to the result.
+    path.swap(curDirRootName);
+    return make_error_code(errc::success);
+  }
+
+  if (rootName && !rootDirectory) {
+    StringRef pRootName;
+    StringRef bRootDirectory;
+    StringRef bRelativePath;
+    StringRef pRelativePath;
+    if (error_code ec = root_name(p, pRootName)) return ec;
+    if (error_code ec = root_directory(current_dir, bRootDirectory)) return ec;
+    if (error_code ec = relative_path(current_dir, bRelativePath)) return ec;
+    if (error_code ec = relative_path(p, pRelativePath)) return ec;
+
+    SmallString<128> res;
+    if (error_code ec = append(res, pRootName, bRootDirectory,
+                                    bRelativePath, pRelativePath)) return ec;
+    path.swap(res);
+    return make_error_code(errc::success);
+  }
+
+  llvm_unreachable("All rootName and rootDirectory combinations should have "
+                   "occurred above!");
+}
+
+}
+}
+}
+
+// Include the truly platform-specific parts.
+#if defined(LLVM_ON_UNIX)
+#include "Unix/PathV2.inc"
+#endif
+#if defined(LLVM_ON_WIN32)
+#include "Windows/PathV2.inc"
+#endif

lib/Support/Unix/PathV2.inc

@@ -0,0 +1,39 @@
+//===- llvm/Support/Unix/PathV2.cpp - Unix Path Implementation --*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file implements the Unix specific implementation of the PathV2 API.
+//
+//===----------------------------------------------------------------------===//
+
+//===----------------------------------------------------------------------===//
+//=== WARNING: Implementation here must contain only generic UNIX code that
+//===          is guaranteed to work on *all* UNIX variants.
+//===----------------------------------------------------------------------===//
+
+#include "Unix.h"
+
+namespace llvm {
+namespace sys  {
+namespace path {
+
+error_code current_path(SmallVectorImpl<char> &result) {
+  long size = ::pathconf(".", _PC_PATH_MAX);
+  result.reserve(size + 1);
+  result.set_size(size + 1);
+
+  if (::getcwd(result.data(), result.size()) == 0)
+    return error_code(errno, system_category());
+
+  result.set_size(strlen(result.data()));
+  return make_error_code(errc::success);
+}
+
+} // end namespace path
+} // end namespace sys
+} // end namespace llvm

lib/Support/Windows/PathV2.inc

@@ -0,0 +1,71 @@
+//===- llvm/Support/Win32/PathV2.cpp - Windows Path Impl --------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file implements the Windows specific implementation of the PathV2 API.
+//
+//===----------------------------------------------------------------------===//
+
+//===----------------------------------------------------------------------===//
+//=== WARNING: Implementation here must contain only generic Windows code that
+//===          is guaranteed to work on *all* Windows variants.
+//===----------------------------------------------------------------------===//
+
+#include "Windows.h"
+
+namespace llvm {
+namespace sys  {
+namespace path {
+
+error_code current_path(SmallVectorImpl<char> &result) {
+  SmallVector<wchar_t, 128> cur_path;
+  cur_path.reserve(128);
+retry_cur_dir:
+  DWORD len = ::GetCurrentDirectoryW(cur_path.capacity(), cur_path.data());
+
+  // A zero return value indicates a failure other than insufficient space.
+  if (len == 0)
+    return make_error_code(windows_error(::GetLastError()));
+
+  // If there's insufficient space, the len returned is larger than the len
+  // given.
+  if (len > cur_path.capacity()) {
+    cur_path.reserve(len);
+    goto retry_cur_dir;
+  }
+
+  cur_path.set_size(len);
+  // cur_path now holds the current directory in utf-16. Convert to utf-8.
+
+  // Find out how much space we need. Sadly, this function doesn't return the
+  // size needed unless you tell it the result size is 0, which means you
+  // _always_ have to call it twice.
+  len = ::WideCharToMultiByte(CP_UTF8, NULL,
+                              cur_path.data(), cur_path.size(),
+                              result.data(), 0,
+                              NULL, NULL);
+
+  if (len == 0)
+    return make_error_code(windows_error(::GetLastError()));
+
+  result.reserve(len);
+  result.set_size(len);
+  // Now do the actual conversion.
+  len = ::WideCharToMultiByte(CP_UTF8, NULL,
+                              cur_path.data(), cur_path.size(),
+                              result.data(), result.size(),
+                              NULL, NULL);
+  if (len == 0)
+    return make_error_code(windows_error(::GetLastError()));
+
+  return make_error_code(errc::success);
+}
+
+} // end namespace path
+} // end namespace sys
+} // end namespace llvm

unittests/System/Path.cpp

@@ -7,15 +7,85 @@
 //
 //===----------------------------------------------------------------------===//
 
-// For now, just test that the header file parses.
 #include "llvm/Support/PathV2.h"
 
 #include "gtest/gtest.h"
 
+using namespace llvm;
+
 namespace {
 
 TEST(System, Path) {
-  // TODO: Add tests!
+  SmallVector<StringRef, 40> paths;
+  paths.push_back("");
+  paths.push_back(".");
+  paths.push_back("..");
+  paths.push_back("foo");
+  paths.push_back("/");
+  paths.push_back("/foo");
+  paths.push_back("foo/");
+  paths.push_back("/foo/");
+  paths.push_back("foo/bar");
+  paths.push_back("/foo/bar");
+  paths.push_back("//net");
+  paths.push_back("//net/foo");
+  paths.push_back("///foo///");
+  paths.push_back("///foo///bar");
+  paths.push_back("/.");
+  paths.push_back("./");
+  paths.push_back("/..");
+  paths.push_back("../");
+  paths.push_back("foo/.");
+  paths.push_back("foo/..");
+  paths.push_back("foo/./");
+  paths.push_back("foo/./bar");
+  paths.push_back("foo/..");
+  paths.push_back("foo/../");
+  paths.push_back("foo/../bar");
+  paths.push_back("c:");
+  paths.push_back("c:/");
+  paths.push_back("c:foo");
+  paths.push_back("c:/foo");
+  paths.push_back("c:foo/");
+  paths.push_back("c:/foo/");
+  paths.push_back("c:/foo/bar");
+  paths.push_back("prn:");
+  paths.push_back("c:\\");
+  paths.push_back("c:foo");
+  paths.push_back("c:\\foo");
+  paths.push_back("c:foo\\");
+  paths.push_back("c:\\foo\\");
+  paths.push_back("c:\\foo/");
+  paths.push_back("c:/foo\\bar");
+
+  for (SmallVector<StringRef, 40>::const_iterator i = paths.begin(),
+                                                  e = paths.end();
+                                                  i != e;
+                                                  ++i) {
+    outs() << *i << " =>\n    Iteration: [";
+    for (sys::path::const_iterator ci = sys::path::begin(*i),
+                                   ce = sys::path::end(*i);
+                                   ci != ce;
+                                   ++ci) {
+      outs() << *ci << ',';
+    }
+    outs() << "]\n";
+
+    StringRef res;
+    SmallString<16> temp_store;
+    if (error_code ec = sys::path::root_path(*i, res)) ASSERT_FALSE(ec.message().c_str());
+    outs() << "    root_path: " << res << '\n';
+    if (error_code ec = sys::path::root_name(*i, res)) ASSERT_FALSE(ec.message().c_str());
+    outs() << "    root_name: " << res << '\n';
+    if (error_code ec = sys::path::root_directory(*i, res)) ASSERT_FALSE(ec.message().c_str());
+    outs() << "    root_directory: " << res << '\n';
+
+    temp_store = *i;
+    if (error_code ec = sys::path::make_absolute(temp_store)) ASSERT_FALSE(ec.message().c_str());
+    outs() << "    make_absolute: " << temp_store << '\n';
+
+    outs().flush();
+  }
 }
 
 } // anonymous namespace