1 //===-- MsgPackDocument.cpp - MsgPack Document --------------------------*-===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 ///
9 /// This file implements a class that exposes a simple in-memory representation
10 /// of a document of MsgPack objects, that can be read from MsgPack, written to
11 /// MsgPack, and inspected and modified in memory. This is intended to be a
12 /// lighter-weight (in terms of memory allocations) replacement for
13 /// MsgPackTypes.
14 ///
15 //===----------------------------------------------------------------------===//
16 
17 #include "llvm/BinaryFormat/MsgPackDocument.h"
18 #include "llvm/BinaryFormat/MsgPackWriter.h"
19 
20 using namespace llvm;
21 using namespace msgpack;
22 
23 // Convert this DocNode into an empty array.
24 void DocNode::convertToArray() { *this = getDocument()->getArrayNode(); }
25 
26 // Convert this DocNode into an empty map.
27 void DocNode::convertToMap() { *this = getDocument()->getMapNode(); }
28 
29 /// Find the key in the MapDocNode.
30 DocNode::MapTy::iterator MapDocNode::find(StringRef S) {
31   return find(getDocument()->getNode(S));
32 }
33 
34 /// Member access for MapDocNode. The string data must remain valid for the
35 /// lifetime of the Document.
36 DocNode &MapDocNode::operator[](StringRef S) {
37   return (*this)[getDocument()->getNode(S)];
38 }
39 
40 /// Member access for MapDocNode.
41 DocNode &MapDocNode::operator[](DocNode Key) {
42   assert(!Key.isEmpty());
43   DocNode &N = (*Map)[Key];
44   if (N.isEmpty()) {
45     // Ensure a new element has its KindAndDoc initialized.
46     N = getDocument()->getEmptyNode();
47   }
48   return N;
49 }
50 
51 /// Member access for MapDocNode for integer key.
52 DocNode &MapDocNode::operator[](int Key) {
53   return (*this)[getDocument()->getNode(Key)];
54 }
55 DocNode &MapDocNode::operator[](unsigned Key) {
56   return (*this)[getDocument()->getNode(Key)];
57 }
58 DocNode &MapDocNode::operator[](int64_t Key) {
59   return (*this)[getDocument()->getNode(Key)];
60 }
61 DocNode &MapDocNode::operator[](uint64_t Key) {
62   return (*this)[getDocument()->getNode(Key)];
63 }
64 
65 /// Array element access. This extends the array if necessary.
66 DocNode &ArrayDocNode::operator[](size_t Index) {
67   if (size() <= Index) {
68     // Ensure new elements have their KindAndDoc initialized.
69     Array->resize(Index + 1, getDocument()->getEmptyNode());
70   }
71   return (*Array)[Index];
72 }
73 
74 // Convenience assignment operators. This only works if the destination
75 // DocNode has an associated Document, i.e. it was not constructed using the
76 // default constructor. The string one does not copy, so the string must
77 // remain valid for the lifetime of the Document. Use fromString to avoid
78 // that restriction.
79 DocNode &DocNode::operator=(StringRef Val) {
80   *this = getDocument()->getNode(Val);
81   return *this;
82 }
83 DocNode &DocNode::operator=(MemoryBufferRef Val) {
84   *this = getDocument()->getNode(Val);
85   return *this;
86 }
87 DocNode &DocNode::operator=(bool Val) {
88   *this = getDocument()->getNode(Val);
89   return *this;
90 }
91 DocNode &DocNode::operator=(int Val) {
92   *this = getDocument()->getNode(Val);
93   return *this;
94 }
95 DocNode &DocNode::operator=(unsigned Val) {
96   *this = getDocument()->getNode(Val);
97   return *this;
98 }
99 DocNode &DocNode::operator=(int64_t Val) {
100   *this = getDocument()->getNode(Val);
101   return *this;
102 }
103 DocNode &DocNode::operator=(uint64_t Val) {
104   *this = getDocument()->getNode(Val);
105   return *this;
106 }
107 
108 // A level in the document reading stack.
109 struct StackLevel {
110   StackLevel(DocNode Node, size_t StartIndex, size_t Length,
111              DocNode *MapEntry = nullptr)
112       : Node(Node), Index(StartIndex), End(StartIndex + Length),
113         MapEntry(MapEntry) {}
114   DocNode Node;
115   size_t Index;
116   size_t End;
117   // Points to map entry when we have just processed a map key.
118   DocNode *MapEntry;
119   DocNode MapKey;
120 };
121 
122 // Read a document from a binary msgpack blob, merging into anything already in
123 // the Document.
124 // The blob data must remain valid for the lifetime of this Document (because a
125 // string object in the document contains a StringRef into the original blob).
126 // If Multi, then this sets root to an array and adds top-level objects to it.
127 // If !Multi, then it only reads a single top-level object, even if there are
128 // more, and sets root to that.
129 // Returns false if failed due to illegal format or merge error.
130 
131 bool Document::readFromBlob(
132     StringRef Blob, bool Multi,
133     function_ref<int(DocNode *DestNode, DocNode SrcNode, DocNode MapKey)>
134         Merger) {
135   msgpack::Reader MPReader(Blob);
136   SmallVector<StackLevel, 4> Stack;
137   if (Multi) {
138     // Create the array for multiple top-level objects.
139     Root = getArrayNode();
140     Stack.push_back(StackLevel(Root, 0, (size_t)-1));
141   }
142   do {
143     // On to next element (or key if doing a map key next).
144     // Read the value.
145     Object Obj;
146     if (!MPReader.read(Obj)) {
147       if (Multi && Stack.size() == 1) {
148         // OK to finish here as we've just done a top-level element with Multi
149         break;
150       }
151       return false; // Finished too early
152     }
153     // Convert it into a DocNode.
154     DocNode Node;
155     switch (Obj.Kind) {
156     case Type::Nil:
157       Node = getNode();
158       break;
159     case Type::Int:
160       Node = getNode(Obj.Int);
161       break;
162     case Type::UInt:
163       Node = getNode(Obj.UInt);
164       break;
165     case Type::Boolean:
166       Node = getNode(Obj.Bool);
167       break;
168     case Type::Float:
169       Node = getNode(Obj.Float);
170       break;
171     case Type::String:
172       Node = getNode(Obj.Raw);
173       break;
174     case Type::Binary:
175       Node = getNode(MemoryBufferRef(Obj.Raw, ""));
176       break;
177     case Type::Map:
178       Node = getMapNode();
179       break;
180     case Type::Array:
181       Node = getArrayNode();
182       break;
183     default:
184       return false; // Raw and Extension not supported
185     }
186 
187     // Store it.
188     DocNode *DestNode = nullptr;
189     if (Stack.empty())
190       DestNode = &Root;
191     else if (Stack.back().Node.getKind() == Type::Array) {
192       // Reading an array entry.
193       auto &Array = Stack.back().Node.getArray();
194       DestNode = &Array[Stack.back().Index++];
195     } else {
196       auto &Map = Stack.back().Node.getMap();
197       if (!Stack.back().MapEntry) {
198         // Reading a map key.
199         Stack.back().MapKey = Node;
200         Stack.back().MapEntry = &Map[Node];
201         continue;
202       }
203       // Reading the value for the map key read in the last iteration.
204       DestNode = Stack.back().MapEntry;
205       Stack.back().MapEntry = nullptr;
206       ++Stack.back().Index;
207     }
208     int MergeResult = 0;
209     if (!DestNode->isEmpty()) {
210       // In a merge, there is already a value at this position. Call the
211       // callback to attempt to resolve the conflict. The resolution must result
212       // in an array or map if Node is an array or map respectively.
213       DocNode MapKey = !Stack.empty() && !Stack.back().MapKey.isEmpty()
214                            ? Stack.back().MapKey
215                            : getNode();
216       MergeResult = Merger(DestNode, Node, MapKey);
217       if (MergeResult < 0)
218         return false; // Merge conflict resolution failed
219       assert(!((Node.isMap() && !DestNode->isMap()) ||
220                (Node.isArray() && !DestNode->isArray())));
221     } else
222       *DestNode = Node;
223 
224     // See if we're starting a new array or map.
225     switch (DestNode->getKind()) {
226     case msgpack::Type::Array:
227     case msgpack::Type::Map:
228       Stack.push_back(StackLevel(*DestNode, MergeResult, Obj.Length, nullptr));
229       break;
230     default:
231       break;
232     }
233 
234     // Pop finished stack levels.
235     while (!Stack.empty()) {
236       if (Stack.back().MapEntry)
237         break;
238       if (Stack.back().Index != Stack.back().End)
239         break;
240       Stack.pop_back();
241     }
242   } while (!Stack.empty());
243   return true;
244 }
245 
246 struct WriterStackLevel {
247   DocNode Node;
248   DocNode::MapTy::iterator MapIt;
249   DocNode::ArrayTy::iterator ArrayIt;
250   bool OnKey;
251 };
252 
253 /// Write a MsgPack document to a binary MsgPack blob.
254 void Document::writeToBlob(std::string &Blob) {
255   Blob.clear();
256   raw_string_ostream OS(Blob);
257   msgpack::Writer MPWriter(OS);
258   SmallVector<WriterStackLevel, 4> Stack;
259   DocNode Node = getRoot();
260   for (;;) {
261     switch (Node.getKind()) {
262     case Type::Array:
263       MPWriter.writeArraySize(Node.getArray().size());
264       Stack.push_back(
265           {Node, DocNode::MapTy::iterator(), Node.getArray().begin(), false});
266       break;
267     case Type::Map:
268       MPWriter.writeMapSize(Node.getMap().size());
269       Stack.push_back(
270           {Node, Node.getMap().begin(), DocNode::ArrayTy::iterator(), true});
271       break;
272     case Type::Nil:
273       MPWriter.writeNil();
274       break;
275     case Type::Boolean:
276       MPWriter.write(Node.getBool());
277       break;
278     case Type::Int:
279       MPWriter.write(Node.getInt());
280       break;
281     case Type::UInt:
282       MPWriter.write(Node.getUInt());
283       break;
284     case Type::String:
285       MPWriter.write(Node.getString());
286       break;
287     case Type::Binary:
288       MPWriter.write(Node.getBinary());
289       break;
290     case Type::Empty:
291       llvm_unreachable("unhandled empty msgpack node");
292     default:
293       llvm_unreachable("unhandled msgpack object kind");
294     }
295     // Pop finished stack levels.
296     while (!Stack.empty()) {
297       if (Stack.back().Node.getKind() == Type::Map) {
298         if (Stack.back().MapIt != Stack.back().Node.getMap().end())
299           break;
300       } else {
301         if (Stack.back().ArrayIt != Stack.back().Node.getArray().end())
302           break;
303       }
304       Stack.pop_back();
305     }
306     if (Stack.empty())
307       break;
308     // Get the next value.
309     if (Stack.back().Node.getKind() == Type::Map) {
310       if (Stack.back().OnKey) {
311         // Do the key of a key,value pair in a map.
312         Node = Stack.back().MapIt->first;
313         Stack.back().OnKey = false;
314       } else {
315         Node = Stack.back().MapIt->second;
316         ++Stack.back().MapIt;
317         Stack.back().OnKey = true;
318       }
319     } else {
320       Node = *Stack.back().ArrayIt;
321       ++Stack.back().ArrayIt;
322     }
323   }
324 }
325