singledocparser.cpp 11.8 KB
Newer Older
Jesse Beder's avatar
Jesse Beder committed
1
2
3
4
5
#include <algorithm>
#include <cstdio>
#include <sstream>

#include "collectionstack.h"  // IWYU pragma: keep
6
#include "scanner.h"
Jesse Beder's avatar
Jesse Beder committed
7
#include "singledocparser.h"
8
9
#include "tag.h"
#include "token.h"
Alan Griffiths's avatar
Alan Griffiths committed
10
#include "yaml-cpp/depthguard.h"
11
#include "yaml-cpp/emitterstyle.h"
Jesse Beder's avatar
Jesse Beder committed
12
13
14
#include "yaml-cpp/eventhandler.h"
#include "yaml-cpp/exceptions.h"  // IWYU pragma: keep
#include "yaml-cpp/mark.h"
15
#include "yaml-cpp/null.h"
16

Jesse Beder's avatar
Jesse Beder committed
17
18
19
20
21
namespace YAML {
SingleDocParser::SingleDocParser(Scanner& scanner, const Directives& directives)
    : m_scanner(scanner),
      m_directives(directives),
      m_pCollectionStack(new CollectionStack),
22
      m_anchors{},
Jesse Beder's avatar
Jesse Beder committed
23
24
      m_curAnchor(0) {}

25
SingleDocParser::~SingleDocParser() = default;
Jesse Beder's avatar
Jesse Beder committed
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

// HandleDocument
// . Handles the next document
// . Throws a ParserException on error.
void SingleDocParser::HandleDocument(EventHandler& eventHandler) {
  assert(!m_scanner.empty());  // guaranteed that there are tokens
  assert(!m_curAnchor);

  eventHandler.OnDocumentStart(m_scanner.peek().mark);

  // eat doc start
  if (m_scanner.peek().type == Token::DOC_START)
    m_scanner.pop();

  // recurse!
  HandleNode(eventHandler);

  eventHandler.OnDocumentEnd();

  // and finally eat any doc ends we see
  while (!m_scanner.empty() && m_scanner.peek().type == Token::DOC_END)
    m_scanner.pop();
}

void SingleDocParser::HandleNode(EventHandler& eventHandler) {
51
  DepthGuard<500> depthguard(depth, m_scanner.mark(), ErrorMsg::BAD_FILE);
Alan Griffiths's avatar
Alan Griffiths committed
52

Jesse Beder's avatar
Jesse Beder committed
53
54
55
56
57
58
59
60
61
62
63
  // an empty node *is* a possibility
  if (m_scanner.empty()) {
    eventHandler.OnNull(m_scanner.mark(), NullAnchor);
    return;
  }

  // save location
  Mark mark = m_scanner.peek().mark;

  // special case: a value node by itself must be a map, with no header
  if (m_scanner.peek().type == Token::VALUE) {
64
    eventHandler.OnMapStart(mark, "?", NullAnchor, EmitterStyle::Default);
Jesse Beder's avatar
Jesse Beder committed
65
66
67
68
69
70
71
72
73
74
75
76
77
    HandleMap(eventHandler);
    eventHandler.OnMapEnd();
    return;
  }

  // special case: an alias node
  if (m_scanner.peek().type == Token::ALIAS) {
    eventHandler.OnAlias(mark, LookupAnchor(mark, m_scanner.peek().value));
    m_scanner.pop();
    return;
  }

  std::string tag;
78
  std::string anchor_name;
Jesse Beder's avatar
Jesse Beder committed
79
  anchor_t anchor;
80
81
82
83
  ParseProperties(tag, anchor, anchor_name);

  if (!anchor_name.empty())
    eventHandler.OnAnchor(mark, anchor_name);
Jesse Beder's avatar
Jesse Beder committed
84

85
86
87
88
89
90
  // after parsing properties, an empty node is again a possibility
  if (m_scanner.empty()) {
    eventHandler.OnNull(mark, anchor);
    return;
  }

Jesse Beder's avatar
Jesse Beder committed
91
92
  const Token& token = m_scanner.peek();

Chen's avatar
Chen committed
93
94
95
96
97
98
  // add non-specific tags
  if (tag.empty())
    tag = (token.type == Token::NON_PLAIN_SCALAR ? "!" : "?");
  
  if (token.type == Token::PLAIN_SCALAR 
      && tag.compare("?") == 0 && IsNullString(token.value)) {
Jesse Beder's avatar
Jesse Beder committed
99
100
101
102
103
104
105
106
107
108
109
110
111
    eventHandler.OnNull(mark, anchor);
    m_scanner.pop();
    return;
  }

  // now split based on what kind of node we should be
  switch (token.type) {
    case Token::PLAIN_SCALAR:
    case Token::NON_PLAIN_SCALAR:
      eventHandler.OnScalar(mark, tag, anchor, token.value);
      m_scanner.pop();
      return;
    case Token::FLOW_SEQ_START:
112
113
114
115
      eventHandler.OnSequenceStart(mark, tag, anchor, EmitterStyle::Flow);
      HandleSequence(eventHandler);
      eventHandler.OnSequenceEnd();
      return;
Jesse Beder's avatar
Jesse Beder committed
116
    case Token::BLOCK_SEQ_START:
117
      eventHandler.OnSequenceStart(mark, tag, anchor, EmitterStyle::Block);
Jesse Beder's avatar
Jesse Beder committed
118
119
120
121
      HandleSequence(eventHandler);
      eventHandler.OnSequenceEnd();
      return;
    case Token::FLOW_MAP_START:
122
123
124
125
      eventHandler.OnMapStart(mark, tag, anchor, EmitterStyle::Flow);
      HandleMap(eventHandler);
      eventHandler.OnMapEnd();
      return;
Jesse Beder's avatar
Jesse Beder committed
126
    case Token::BLOCK_MAP_START:
127
      eventHandler.OnMapStart(mark, tag, anchor, EmitterStyle::Block);
Jesse Beder's avatar
Jesse Beder committed
128
129
130
131
132
133
134
      HandleMap(eventHandler);
      eventHandler.OnMapEnd();
      return;
    case Token::KEY:
      // compact maps can only go in a flow sequence
      if (m_pCollectionStack->GetCurCollectionType() ==
          CollectionType::FlowSeq) {
135
        eventHandler.OnMapStart(mark, tag, anchor, EmitterStyle::Flow);
Jesse Beder's avatar
Jesse Beder committed
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
        HandleMap(eventHandler);
        eventHandler.OnMapEnd();
        return;
      }
      break;
    default:
      break;
  }

  if (tag == "?")
    eventHandler.OnNull(mark, anchor);
  else
    eventHandler.OnScalar(mark, tag, anchor, "");
}

void SingleDocParser::HandleSequence(EventHandler& eventHandler) {
  // split based on start token
  switch (m_scanner.peek().type) {
    case Token::BLOCK_SEQ_START:
      HandleBlockSequence(eventHandler);
      break;
    case Token::FLOW_SEQ_START:
      HandleFlowSequence(eventHandler);
      break;
    default:
      break;
  }
}

void SingleDocParser::HandleBlockSequence(EventHandler& eventHandler) {
  // eat start token
  m_scanner.pop();
  m_pCollectionStack->PushCollectionType(CollectionType::BlockSeq);

170
  while (true) {
Jesse Beder's avatar
Jesse Beder committed
171
172
173
174
175
176
177
178
179
180
181
182
183
    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_SEQ);

    Token token = m_scanner.peek();
    if (token.type != Token::BLOCK_ENTRY && token.type != Token::BLOCK_SEQ_END)
      throw ParserException(token.mark, ErrorMsg::END_OF_SEQ);

    m_scanner.pop();
    if (token.type == Token::BLOCK_SEQ_END)
      break;

    // check for null
    if (!m_scanner.empty()) {
184
185
186
187
      const Token& nextToken = m_scanner.peek();
      if (nextToken.type == Token::BLOCK_ENTRY ||
          nextToken.type == Token::BLOCK_SEQ_END) {
        eventHandler.OnNull(nextToken.mark, NullAnchor);
Jesse Beder's avatar
Jesse Beder committed
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
        continue;
      }
    }

    HandleNode(eventHandler);
  }

  m_pCollectionStack->PopCollectionType(CollectionType::BlockSeq);
}

void SingleDocParser::HandleFlowSequence(EventHandler& eventHandler) {
  // eat start token
  m_scanner.pop();
  m_pCollectionStack->PushCollectionType(CollectionType::FlowSeq);

203
  while (true) {
Jesse Beder's avatar
Jesse Beder committed
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_SEQ_FLOW);

    // first check for end
    if (m_scanner.peek().type == Token::FLOW_SEQ_END) {
      m_scanner.pop();
      break;
    }

    // then read the node
    HandleNode(eventHandler);

    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_SEQ_FLOW);

    // now eat the separator (or could be a sequence end, which we ignore - but
    // if it's neither, then it's a bad node)
    Token& token = m_scanner.peek();
    if (token.type == Token::FLOW_ENTRY)
      m_scanner.pop();
    else if (token.type != Token::FLOW_SEQ_END)
      throw ParserException(token.mark, ErrorMsg::END_OF_SEQ_FLOW);
  }

  m_pCollectionStack->PopCollectionType(CollectionType::FlowSeq);
}

void SingleDocParser::HandleMap(EventHandler& eventHandler) {
  // split based on start token
  switch (m_scanner.peek().type) {
    case Token::BLOCK_MAP_START:
      HandleBlockMap(eventHandler);
      break;
    case Token::FLOW_MAP_START:
      HandleFlowMap(eventHandler);
      break;
    case Token::KEY:
      HandleCompactMap(eventHandler);
      break;
    case Token::VALUE:
      HandleCompactMapWithNoKey(eventHandler);
      break;
    default:
      break;
  }
}

void SingleDocParser::HandleBlockMap(EventHandler& eventHandler) {
  // eat start token
  m_scanner.pop();
  m_pCollectionStack->PushCollectionType(CollectionType::BlockMap);

256
  while (true) {
Jesse Beder's avatar
Jesse Beder committed
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_MAP);

    Token token = m_scanner.peek();
    if (token.type != Token::KEY && token.type != Token::VALUE &&
        token.type != Token::BLOCK_MAP_END)
      throw ParserException(token.mark, ErrorMsg::END_OF_MAP);

    if (token.type == Token::BLOCK_MAP_END) {
      m_scanner.pop();
      break;
    }

    // grab key (if non-null)
    if (token.type == Token::KEY) {
      m_scanner.pop();
      HandleNode(eventHandler);
    } else {
      eventHandler.OnNull(token.mark, NullAnchor);
    }

    // now grab value (optional)
    if (!m_scanner.empty() && m_scanner.peek().type == Token::VALUE) {
      m_scanner.pop();
      HandleNode(eventHandler);
    } else {
      eventHandler.OnNull(token.mark, NullAnchor);
    }
  }

  m_pCollectionStack->PopCollectionType(CollectionType::BlockMap);
}

void SingleDocParser::HandleFlowMap(EventHandler& eventHandler) {
  // eat start token
  m_scanner.pop();
  m_pCollectionStack->PushCollectionType(CollectionType::FlowMap);

295
  while (true) {
Jesse Beder's avatar
Jesse Beder committed
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_MAP_FLOW);

    Token& token = m_scanner.peek();
    const Mark mark = token.mark;
    // first check for end
    if (token.type == Token::FLOW_MAP_END) {
      m_scanner.pop();
      break;
    }

    // grab key (if non-null)
    if (token.type == Token::KEY) {
      m_scanner.pop();
      HandleNode(eventHandler);
    } else {
      eventHandler.OnNull(mark, NullAnchor);
    }

    // now grab value (optional)
    if (!m_scanner.empty() && m_scanner.peek().type == Token::VALUE) {
      m_scanner.pop();
      HandleNode(eventHandler);
    } else {
      eventHandler.OnNull(mark, NullAnchor);
    }

    if (m_scanner.empty())
      throw ParserException(m_scanner.mark(), ErrorMsg::END_OF_MAP_FLOW);

    // now eat the separator (or could be a map end, which we ignore - but if
    // it's neither, then it's a bad node)
    Token& nextToken = m_scanner.peek();
    if (nextToken.type == Token::FLOW_ENTRY)
      m_scanner.pop();
    else if (nextToken.type != Token::FLOW_MAP_END)
      throw ParserException(nextToken.mark, ErrorMsg::END_OF_MAP_FLOW);
  }

  m_pCollectionStack->PopCollectionType(CollectionType::FlowMap);
}

// . Single "key: value" pair in a flow sequence
void SingleDocParser::HandleCompactMap(EventHandler& eventHandler) {
  m_pCollectionStack->PushCollectionType(CollectionType::CompactMap);

  // grab key
  Mark mark = m_scanner.peek().mark;
  m_scanner.pop();
  HandleNode(eventHandler);

  // now grab value (optional)
  if (!m_scanner.empty() && m_scanner.peek().type == Token::VALUE) {
    m_scanner.pop();
    HandleNode(eventHandler);
  } else {
    eventHandler.OnNull(mark, NullAnchor);
  }

  m_pCollectionStack->PopCollectionType(CollectionType::CompactMap);
}

// . Single ": value" pair in a flow sequence
void SingleDocParser::HandleCompactMapWithNoKey(EventHandler& eventHandler) {
  m_pCollectionStack->PushCollectionType(CollectionType::CompactMap);

  // null key
  eventHandler.OnNull(m_scanner.peek().mark, NullAnchor);

  // grab value
  m_scanner.pop();
  HandleNode(eventHandler);

  m_pCollectionStack->PopCollectionType(CollectionType::CompactMap);
}

// ParseProperties
// . Grabs any tag or anchor tokens and deals with them.
374
375
void SingleDocParser::ParseProperties(std::string& tag, anchor_t& anchor,
                                      std::string& anchor_name) {
Jesse Beder's avatar
Jesse Beder committed
376
  tag.clear();
377
  anchor_name.clear();
Jesse Beder's avatar
Jesse Beder committed
378
379
  anchor = NullAnchor;

380
  while (true) {
Jesse Beder's avatar
Jesse Beder committed
381
382
383
384
385
386
387
388
    if (m_scanner.empty())
      return;

    switch (m_scanner.peek().type) {
      case Token::TAG:
        ParseTag(tag);
        break;
      case Token::ANCHOR:
389
        ParseAnchor(anchor, anchor_name);
Jesse Beder's avatar
Jesse Beder committed
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
        break;
      default:
        return;
    }
  }
}

void SingleDocParser::ParseTag(std::string& tag) {
  Token& token = m_scanner.peek();
  if (!tag.empty())
    throw ParserException(token.mark, ErrorMsg::MULTIPLE_TAGS);

  Tag tagInfo(token);
  tag = tagInfo.Translate(m_directives);
  m_scanner.pop();
}

407
void SingleDocParser::ParseAnchor(anchor_t& anchor, std::string& anchor_name) {
Jesse Beder's avatar
Jesse Beder committed
408
409
410
411
  Token& token = m_scanner.peek();
  if (anchor)
    throw ParserException(token.mark, ErrorMsg::MULTIPLE_ANCHORS);

412
  anchor_name = token.value;
Jesse Beder's avatar
Jesse Beder committed
413
414
415
416
417
418
419
420
421
422
423
424
425
  anchor = RegisterAnchor(token.value);
  m_scanner.pop();
}

anchor_t SingleDocParser::RegisterAnchor(const std::string& name) {
  if (name.empty())
    return NullAnchor;

  return m_anchors[name] = ++m_curAnchor;
}

anchor_t SingleDocParser::LookupAnchor(const Mark& mark,
                                       const std::string& name) const {
Rosen Penev's avatar
Rosen Penev committed
426
  auto it = m_anchors.find(name);
427
428
429
430
431
  if (it == m_anchors.end()) {
    std::stringstream ss;
    ss << ErrorMsg::UNKNOWN_ANCHOR << name;
    throw ParserException(mark, ss.str());
  }
Jesse Beder's avatar
Jesse Beder committed
432
433
434

  return it->second;
}
435
}  // namespace YAML