add antlr

This commit is contained in:
2020-09-29 23:16:42 +05:30
parent 34754a7130
commit 2b36e74ff5
181 changed files with 14689 additions and 7 deletions

View File

@@ -0,0 +1,112 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "RuleContext.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ATN {
public:
static const size_t INVALID_ALT_NUMBER = 0;
/// Used for runtime deserialization of ATNs from strings.
ATN();
ATN(ATN &&other);
ATN(ATNType grammarType, size_t maxTokenType);
virtual ~ATN();
std::vector<ATNState *> states;
/// Each subrule/rule is a decision point and we must track them so we
/// can go back later and build DFA predictors for them. This includes
/// all the rules, subrules, optional blocks, ()+, ()* etc...
std::vector<DecisionState *> decisionToState;
/// Maps from rule index to starting state number.
std::vector<RuleStartState *> ruleToStartState;
/// Maps from rule index to stop state number.
std::vector<RuleStopState *> ruleToStopState;
/// The type of the ATN.
ATNType grammarType;
/// The maximum value for any symbol recognized by a transition in the ATN.
size_t maxTokenType;
/// <summary>
/// For lexer ATNs, this maps the rule index to the resulting token type.
/// For parser ATNs, this maps the rule index to the generated bypass token
/// type if the
/// <seealso cref="ATNDeserializationOptions#isGenerateRuleBypassTransitions"/>
/// deserialization option was specified; otherwise, this is {@code null}.
/// </summary>
std::vector<size_t> ruleToTokenType;
/// For lexer ATNs, this is an array of {@link LexerAction} objects which may
/// be referenced by action transitions in the ATN.
std::vector<Ref<LexerAction>> lexerActions;
std::vector<TokensStartState *> modeToStartState;
ATN& operator = (ATN &other) NOEXCEPT;
ATN& operator = (ATN &&other) NOEXCEPT;
/// <summary>
/// Compute the set of valid tokens that can occur starting in state {@code s}.
/// If {@code ctx} is null, the set of tokens will not include what can follow
/// the rule surrounding {@code s}. In other words, the set will be
/// restricted to tokens reachable staying within {@code s}'s rule.
/// </summary>
virtual misc::IntervalSet nextTokens(ATNState *s, RuleContext *ctx) const;
/// <summary>
/// Compute the set of valid tokens that can occur starting in {@code s} and
/// staying in same rule. <seealso cref="Token#EPSILON"/> is in set if we reach end of
/// rule.
/// </summary>
virtual misc::IntervalSet const& nextTokens(ATNState *s) const;
virtual void addState(ATNState *state);
virtual void removeState(ATNState *state);
virtual int defineDecisionState(DecisionState *s);
virtual DecisionState *getDecisionState(size_t decision) const;
virtual size_t getNumberOfDecisions() const;
/// <summary>
/// Computes the set of input symbols which could follow ATN state number
/// {@code stateNumber} in the specified full {@code context}. This method
/// considers the complete parser context, but does not evaluate semantic
/// predicates (i.e. all predicates encountered during the calculation are
/// assumed true). If a path in the ATN exists from the starting state to the
/// <seealso cref="RuleStopState"/> of the outermost context without matching any
/// symbols, <seealso cref="Token#EOF"/> is added to the returned set.
/// <p/>
/// If {@code context} is {@code null}, it is treated as
/// <seealso cref="ParserRuleContext#EMPTY"/>.
/// </summary>
/// <param name="stateNumber"> the ATN state number </param>
/// <param name="context"> the full parse context </param>
/// <returns> The set of potentially valid input symbols which could follow the
/// specified state in the specified context. </returns>
/// <exception cref="IllegalArgumentException"> if the ATN does not contain a state with
/// number {@code stateNumber} </exception>
virtual misc::IntervalSet getExpectedTokens(size_t stateNumber, RuleContext *context) const;
std::string toString() const;
private:
mutable std::mutex _mutex;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,148 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
namespace antlr4 {
namespace atn {
/// <summary>
/// A tuple: (ATN state, predicted alt, syntactic, semantic context).
/// The syntactic context is a graph-structured stack node whose
/// path(s) to the root is the rule invocation(s)
/// chain used to arrive at the state. The semantic context is
/// the tree of semantic predicates encountered before reaching
/// an ATN state.
/// </summary>
class ANTLR4CPP_PUBLIC ATNConfig {
public:
struct Hasher
{
size_t operator()(ATNConfig const& k) const {
return k.hashCode();
}
};
struct Comparer {
bool operator()(ATNConfig const& lhs, ATNConfig const& rhs) const {
return (&lhs == &rhs) || (lhs == rhs);
}
};
using Set = std::unordered_set<Ref<ATNConfig>, Hasher, Comparer>;
/// The ATN state associated with this configuration.
ATNState * state;
/// What alt (or lexer rule) is predicted by this configuration.
const size_t alt;
/// The stack of invoking states leading to the rule/states associated
/// with this config. We track only those contexts pushed during
/// execution of the ATN simulator.
///
/// Can be shared between multiple ANTConfig instances.
Ref<PredictionContext> context;
/**
* We cannot execute predicates dependent upon local context unless
* we know for sure we are in the correct context. Because there is
* no way to do this efficiently, we simply cannot evaluate
* dependent predicates unless we are in the rule that initially
* invokes the ATN simulator.
*
* <p>
* closure() tracks the depth of how far we dip into the outer context:
* depth > 0. Note that it may not be totally accurate depth since I
* don't ever decrement. TODO: make it a boolean then</p>
*
* <p>
* For memory efficiency, the {@link #isPrecedenceFilterSuppressed} method
* is also backed by this field. Since the field is publicly accessible, the
* highest bit which would not cause the value to become negative is used to
* store this field. This choice minimizes the risk that code which only
* compares this value to 0 would be affected by the new purpose of the
* flag. It also ensures the performance of the existing {@link ATNConfig}
* constructors as well as certain operations like
* {@link ATNConfigSet#add(ATNConfig, DoubleKeyMap)} method are
* <em>completely</em> unaffected by the change.</p>
*/
size_t reachesIntoOuterContext;
/// Can be shared between multiple ATNConfig instances.
Ref<SemanticContext> semanticContext;
ATNConfig(ATNState *state, size_t alt, Ref<PredictionContext> const& context);
ATNConfig(ATNState *state, size_t alt, Ref<PredictionContext> const& context, Ref<SemanticContext> const& semanticContext);
ATNConfig(Ref<ATNConfig> const& c); // dup
ATNConfig(Ref<ATNConfig> const& c, ATNState *state);
ATNConfig(Ref<ATNConfig> const& c, ATNState *state, Ref<SemanticContext> const& semanticContext);
ATNConfig(Ref<ATNConfig> const& c, Ref<SemanticContext> const& semanticContext);
ATNConfig(Ref<ATNConfig> const& c, ATNState *state, Ref<PredictionContext> const& context);
ATNConfig(Ref<ATNConfig> const& c, ATNState *state, Ref<PredictionContext> const& context, Ref<SemanticContext> const& semanticContext);
ATNConfig(ATNConfig const&) = default;
virtual ~ATNConfig();
virtual size_t hashCode() const;
/**
* This method gets the value of the {@link #reachesIntoOuterContext} field
* as it existed prior to the introduction of the
* {@link #isPrecedenceFilterSuppressed} method.
*/
size_t getOuterContextDepth() const ;
bool isPrecedenceFilterSuppressed() const;
void setPrecedenceFilterSuppressed(bool value);
/// An ATN configuration is equal to another if both have
/// the same state, they predict the same alternative, and
/// syntactic/semantic contexts are the same.
bool operator == (const ATNConfig &other) const;
bool operator != (const ATNConfig &other) const;
virtual std::string toString();
std::string toString(bool showAlt);
private:
/**
* This field stores the bit mask for implementing the
* {@link #isPrecedenceFilterSuppressed} property as a bit within the
* existing {@link #reachesIntoOuterContext} field.
*/
static const size_t SUPPRESS_PRECEDENCE_FILTER;
};
} // namespace atn
} // namespace antlr4
// Hash function for ATNConfig.
namespace std {
using antlr4::atn::ATNConfig;
template <> struct hash<ATNConfig>
{
size_t operator() (const ATNConfig &x) const
{
return x.hashCode();
}
};
template <> struct hash<std::vector<Ref<ATNConfig>>>
{
size_t operator() (const std::vector<Ref<ATNConfig>> &vector) const
{
std::size_t seed = 0;
for (auto &config : vector) {
seed ^= config->hashCode() + 0x9e3779b9 + (seed << 6) + (seed >> 2);
}
return seed;
}
};
}

View File

@@ -0,0 +1,110 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "support/BitSet.h"
#include "atn/PredictionContext.h"
namespace antlr4 {
namespace atn {
/// Specialized set that can track info about the set, with support for combining similar configurations using a
/// graph-structured stack.
class ANTLR4CPP_PUBLIC ATNConfigSet {
public:
/// Track the elements as they are added to the set; supports get(i)
std::vector<Ref<ATNConfig>> configs;
// TODO: these fields make me pretty uncomfortable but nice to pack up info together, saves recomputation
// TODO: can we track conflicts as they are added to save scanning configs later?
size_t uniqueAlt;
/** Currently this is only used when we detect SLL conflict; this does
* not necessarily represent the ambiguous alternatives. In fact,
* I should also point out that this seems to include predicated alternatives
* that have predicates that evaluate to false. Computed in computeTargetState().
*/
antlrcpp::BitSet conflictingAlts;
// Used in parser and lexer. In lexer, it indicates we hit a pred
// while computing a closure operation. Don't make a DFA state from this.
bool hasSemanticContext;
bool dipsIntoOuterContext;
/// Indicates that this configuration set is part of a full context
/// LL prediction. It will be used to determine how to merge $. With SLL
/// it's a wildcard whereas it is not for LL context merge.
const bool fullCtx;
ATNConfigSet(bool fullCtx = true);
ATNConfigSet(const Ref<ATNConfigSet> &old);
virtual ~ATNConfigSet();
virtual bool add(const Ref<ATNConfig> &config);
/// <summary>
/// Adding a new config means merging contexts with existing configs for
/// {@code (s, i, pi, _)}, where {@code s} is the
/// <seealso cref="ATNConfig#state"/>, {@code i} is the <seealso cref="ATNConfig#alt"/>, and
/// {@code pi} is the <seealso cref="ATNConfig#semanticContext"/>. We use
/// {@code (s,i,pi)} as key.
/// <p/>
/// This method updates <seealso cref="#dipsIntoOuterContext"/> and
/// <seealso cref="#hasSemanticContext"/> when necessary.
/// </summary>
virtual bool add(const Ref<ATNConfig> &config, PredictionContextMergeCache *mergeCache);
virtual std::vector<ATNState *> getStates();
/**
* Gets the complete set of represented alternatives for the configuration
* set.
*
* @return the set of represented alternatives in this configuration set
*
* @since 4.3
*/
antlrcpp::BitSet getAlts();
virtual std::vector<Ref<SemanticContext>> getPredicates();
virtual Ref<ATNConfig> get(size_t i) const;
virtual void optimizeConfigs(ATNSimulator *interpreter);
bool addAll(const Ref<ATNConfigSet> &other);
bool operator == (const ATNConfigSet &other);
virtual size_t hashCode();
virtual size_t size();
virtual bool isEmpty();
virtual void clear();
virtual bool isReadonly();
virtual void setReadonly(bool readonly);
virtual std::string toString();
protected:
/// Indicates that the set of configurations is read-only. Do not
/// allow any code to manipulate the set; DFA states will point at
/// the sets and they must not change. This does not protect the other
/// fields; in particular, conflictingAlts is set after
/// we've made this readonly.
bool _readonly;
virtual size_t getHash(ATNConfig *c); // Hash differs depending on set type.
private:
size_t _cachedHashCode;
/// All configs but hashed by (s, i, _, pi) not including context. Wiped out
/// when we go readonly as this set becomes a DFA state.
std::unordered_map<size_t, ATNConfig *> _configLookup;
void InitializeInstanceFields();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,50 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "antlr4-common.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ATNDeserializationOptions {
private:
static ATNDeserializationOptions defaultOptions;
bool readOnly;
bool verifyATN;
bool generateRuleBypassTransitions;
public:
ATNDeserializationOptions();
ATNDeserializationOptions(ATNDeserializationOptions *options);
ATNDeserializationOptions(ATNDeserializationOptions const&) = default;
virtual ~ATNDeserializationOptions();
ATNDeserializationOptions& operator=(ATNDeserializationOptions const&) = default;
static const ATNDeserializationOptions& getDefaultOptions();
bool isReadOnly();
void makeReadOnly();
bool isVerifyATN();
void setVerifyATN(bool verify);
bool isGenerateRuleBypassTransitions();
void setGenerateRuleBypassTransitions(bool generate);
protected:
virtual void throwIfReadOnly();
private:
void InitializeInstanceFields();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,85 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/ATNDeserializationOptions.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ATNDeserializer {
public:
static const size_t SERIALIZED_VERSION;
/// This is the current serialized UUID.
// ml: defined as function to avoid the “static initialization order fiasco”.
static Guid SERIALIZED_UUID();
ATNDeserializer();
ATNDeserializer(const ATNDeserializationOptions& dso);
virtual ~ATNDeserializer();
static Guid toUUID(const unsigned short *data, size_t offset);
virtual ATN deserialize(const std::vector<uint16_t> &input);
virtual void verifyATN(const ATN &atn);
static void checkCondition(bool condition);
static void checkCondition(bool condition, const std::string &message);
static Transition *edgeFactory(const ATN &atn, size_t type, size_t src, size_t trg, size_t arg1, size_t arg2,
size_t arg3, const std::vector<misc::IntervalSet> &sets);
static ATNState *stateFactory(size_t type, size_t ruleIndex);
protected:
/// Determines if a particular serialized representation of an ATN supports
/// a particular feature, identified by the <seealso cref="UUID"/> used for serializing
/// the ATN at the time the feature was first introduced.
///
/// <param name="feature"> The <seealso cref="UUID"/> marking the first time the feature was
/// supported in the serialized ATN. </param>
/// <param name="actualUuid"> The <seealso cref="UUID"/> of the actual serialized ATN which is
/// currently being deserialized. </param>
/// <returns> {@code true} if the {@code actualUuid} value represents a
/// serialized ATN at or after the feature identified by {@code feature} was
/// introduced; otherwise, {@code false}. </returns>
virtual bool isFeatureSupported(const Guid &feature, const Guid &actualUuid);
void markPrecedenceDecisions(const ATN &atn);
Ref<LexerAction> lexerActionFactory(LexerActionType type, int data1, int data2);
private:
/// This is the earliest supported serialized UUID.
static Guid BASE_SERIALIZED_UUID();
/// This UUID indicates an extension of <seealso cref="BASE_SERIALIZED_UUID"/> for the
/// addition of precedence predicates.
static Guid ADDED_PRECEDENCE_TRANSITIONS();
/**
* This UUID indicates an extension of ADDED_PRECEDENCE_TRANSITIONS
* for the addition of lexer actions encoded as a sequence of
* LexerAction instances.
*/
static Guid ADDED_LEXER_ACTIONS();
/**
* This UUID indicates the serialized ATN contains two sets of
* IntervalSets, where the second set's values are encoded as
* 32-bit integers to support the full Unicode SMP range up to U+10FFFF.
*/
static Guid ADDED_UNICODE_SMP();
/// This list contains all of the currently supported UUIDs, ordered by when
/// the feature first appeared in this branch.
static std::vector<Guid>& SUPPORTED_UUIDS();
ATNDeserializationOptions deserializationOptions;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,61 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ATNSerializer {
public:
ATN *atn;
ATNSerializer(ATN *atn);
ATNSerializer(ATN *atn, const std::vector<std::string> &tokenNames);
virtual ~ATNSerializer();
/// <summary>
/// Serialize state descriptors, edge descriptors, and decision->state map
/// into list of ints:
///
/// grammar-type, (ANTLRParser.LEXER, ...)
/// max token type,
/// num states,
/// state-0-type ruleIndex, state-1-type ruleIndex, ... state-i-type
/// ruleIndex optional-arg ...
/// num rules,
/// rule-1-start-state rule-1-args, rule-2-start-state rule-2-args, ...
/// (args are token type,actionIndex in lexer else 0,0)
/// num modes,
/// mode-0-start-state, mode-1-start-state, ... (parser has 0 modes)
/// num sets
/// set-0-interval-count intervals, set-1-interval-count intervals, ...
/// num total edges,
/// src, trg, edge-type, edge arg1, optional edge arg2 (present always),
/// ...
/// num decisions,
/// decision-0-start-state, decision-1-start-state, ...
///
/// Convenient to pack into unsigned shorts to make as Java string.
/// </summary>
virtual std::vector<size_t> serialize();
virtual std::string decode(const std::wstring& data);
virtual std::string getTokenName(size_t t);
/// Used by Java target to encode short/int array as chars in string.
static std::wstring getSerializedAsString(ATN *atn);
static std::vector<size_t> getSerialized(ATN *atn);
static std::string getDecoded(ATN *atn, std::vector<std::string> &tokenNames);
private:
std::vector<std::string> _tokenNames;
void serializeUUID(std::vector<size_t> &data, Guid uuid);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,87 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATN.h"
#include "misc/IntervalSet.h"
#include "support/CPPUtils.h"
#include "atn/PredictionContext.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ATNSimulator {
public:
/// Must distinguish between missing edge and edge we know leads nowhere.
static const Ref<dfa::DFAState> ERROR;
const ATN &atn;
ATNSimulator(const ATN &atn, PredictionContextCache &sharedContextCache);
virtual ~ATNSimulator();
virtual void reset() = 0;
/**
* Clear the DFA cache used by the current instance. Since the DFA cache may
* be shared by multiple ATN simulators, this method may affect the
* performance (but not accuracy) of other parsers which are being used
* concurrently.
*
* @throws UnsupportedOperationException if the current instance does not
* support clearing the DFA.
*
* @since 4.3
*/
virtual void clearDFA();
virtual PredictionContextCache& getSharedContextCache();
virtual Ref<PredictionContext> getCachedContext(Ref<PredictionContext> const& context);
/// @deprecated Use <seealso cref="ATNDeserializer#deserialize"/> instead.
static ATN deserialize(const std::vector<uint16_t> &data);
/// @deprecated Use <seealso cref="ATNDeserializer#checkCondition(boolean)"/> instead.
static void checkCondition(bool condition);
/// @deprecated Use <seealso cref="ATNDeserializer#checkCondition(boolean, String)"/> instead.
static void checkCondition(bool condition, const std::string &message);
/// @deprecated Use <seealso cref="ATNDeserializer#edgeFactory"/> instead.
static Transition *edgeFactory(const ATN &atn, int type, int src, int trg, int arg1, int arg2, int arg3,
const std::vector<misc::IntervalSet> &sets);
/// @deprecated Use <seealso cref="ATNDeserializer#stateFactory"/> instead.
static ATNState *stateFactory(int type, int ruleIndex);
protected:
static antlrcpp::SingleWriteMultipleReadLock _stateLock; // Lock for DFA states.
static antlrcpp::SingleWriteMultipleReadLock _edgeLock; // Lock for the sparse edge map in DFA states.
/// <summary>
/// The context cache maps all PredictionContext objects that are equals()
/// to a single cached copy. This cache is shared across all contexts
/// in all ATNConfigs in all DFA states. We rebuild each ATNConfigSet
/// to use only cached nodes/graphs in addDFAState(). We don't want to
/// fill this during closure() since there are lots of contexts that
/// pop up but are not used ever again. It also greatly slows down closure().
/// <p/>
/// This cache makes a huge difference in memory and a little bit in speed.
/// For the Java grammar on java.*, it dropped the memory requirements
/// at the end from 25M to 16M. We don't store any of the full context
/// graphs in the DFA because they are limited to local context only,
/// but apparently there's a lot of repetition there as well. We optimize
/// the config contexts before storing the config set in the DFA states
/// by literally rebuilding them with cached subgraphs only.
/// <p/>
/// I tried a cache for use during closure operations, that was
/// whacked after each adaptivePredict(). It cost a little bit
/// more time I think and doesn't save on the overall footprint
/// so it's not worth the complexity.
/// </summary>
PredictionContextCache &_sharedContextCache;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,133 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "misc/IntervalSet.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// The following images show the relation of states and
/// <seealso cref="ATNState#transitions"/> for various grammar constructs.
///
/// <ul>
///
/// <li>Solid edges marked with an &#0949; indicate a required
/// <seealso cref="EpsilonTransition"/>.</li>
///
/// <li>Dashed edges indicate locations where any transition derived from
/// <seealso cref="Transition"/> might appear.</li>
///
/// <li>Dashed nodes are place holders for either a sequence of linked
/// <seealso cref="BasicState"/> states or the inclusion of a block representing a nested
/// construct in one of the forms below.</li>
///
/// <li>Nodes showing multiple outgoing alternatives with a {@code ...} support
/// any number of alternatives (one or more). Nodes without the {@code ...} only
/// support the exact number of alternatives shown in the diagram.</li>
///
/// </ul>
///
/// <h2>Basic Blocks</h2>
///
/// <h3>Rule</h3>
///
/// <embed src="images/Rule.svg" type="image/svg+xml"/>
///
/// <h3>Block of 1 or more alternatives</h3>
///
/// <embed src="images/Block.svg" type="image/svg+xml"/>
///
/// <h2>Greedy Loops</h2>
///
/// <h3>Greedy Closure: {@code (...)*}</h3>
///
/// <embed src="images/ClosureGreedy.svg" type="image/svg+xml"/>
///
/// <h3>Greedy Positive Closure: {@code (...)+}</h3>
///
/// <embed src="images/PositiveClosureGreedy.svg" type="image/svg+xml"/>
///
/// <h3>Greedy Optional: {@code (...)?}</h3>
///
/// <embed src="images/OptionalGreedy.svg" type="image/svg+xml"/>
///
/// <h2>Non-Greedy Loops</h2>
///
/// <h3>Non-Greedy Closure: {@code (...)*?}</h3>
///
/// <embed src="images/ClosureNonGreedy.svg" type="image/svg+xml"/>
///
/// <h3>Non-Greedy Positive Closure: {@code (...)+?}</h3>
///
/// <embed src="images/PositiveClosureNonGreedy.svg" type="image/svg+xml"/>
///
/// <h3>Non-Greedy Optional: {@code (...)??}</h3>
///
/// <embed src="images/OptionalNonGreedy.svg" type="image/svg+xml"/>
/// </summary>
class ANTLR4CPP_PUBLIC ATN;
class ANTLR4CPP_PUBLIC ATNState {
public:
ATNState();
ATNState(ATNState const&) = delete;
virtual ~ATNState();
ATNState& operator=(ATNState const&) = delete;
static const size_t INITIAL_NUM_TRANSITIONS = 4;
static const size_t INVALID_STATE_NUMBER = static_cast<size_t>(-1); // std::numeric_limits<size_t>::max();
enum {
ATN_INVALID_TYPE = 0,
BASIC = 1,
RULE_START = 2,
BLOCK_START = 3,
PLUS_BLOCK_START = 4,
STAR_BLOCK_START = 5,
TOKEN_START = 6,
RULE_STOP = 7,
BLOCK_END = 8,
STAR_LOOP_BACK = 9,
STAR_LOOP_ENTRY = 10,
PLUS_LOOP_BACK = 11,
LOOP_END = 12
};
static const std::vector<std::string> serializationNames;
size_t stateNumber = INVALID_STATE_NUMBER;
size_t ruleIndex = 0; // at runtime, we don't have Rule objects
bool epsilonOnlyTransitions = false;
public:
virtual size_t hashCode();
bool operator == (const ATNState &other);
/// Track the transitions emanating from this ATN state.
std::vector<Transition*> transitions;
virtual bool isNonGreedyExitState();
virtual std::string toString() const;
virtual void addTransition(Transition *e);
virtual void addTransition(size_t index, Transition *e);
virtual Transition* removeTransition(size_t index);
virtual size_t getStateType() = 0;
private:
/// Used to cache lookahead during parsing, not used during construction.
misc::IntervalSet _nextTokenWithinRule;
std::atomic<bool> _nextTokenUpdated { false };
friend class ATN;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,20 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "antlr4-common.h"
namespace antlr4 {
namespace atn {
/// Represents the type of recognizer an ATN applies to.
enum class ATNType {
LEXER = 0,
PARSER = 1,
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,24 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTState;
class ANTLR4CPP_PUBLIC AbstractPredicateTransition : public Transition {
public:
AbstractPredicateTransition(ATNState *target);
~AbstractPredicateTransition();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,33 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ActionTransition final : public Transition {
public:
const size_t ruleIndex;
const size_t actionIndex;
const bool isCtxDependent; // e.g., $i ref in action
ActionTransition(ATNState *target, size_t ruleIndex);
ActionTransition(ATNState *target, size_t ruleIndex, size_t actionIndex, bool isCtxDependent);
virtual SerializationType getSerializationType() const override;
virtual bool isEpsilon() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,68 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionEventInfo.h"
#include "support/BitSet.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This class represents profiling event information for an ambiguity.
/// Ambiguities are decisions where a particular input resulted in an SLL
/// conflict, followed by LL prediction also reaching a conflict state
/// (indicating a true ambiguity in the grammar).
///
/// <para>
/// This event may be reported during SLL prediction in cases where the
/// conflicting SLL configuration set provides sufficient information to
/// determine that the SLL conflict is truly an ambiguity. For example, if none
/// of the ATN configurations in the conflicting SLL configuration set have
/// traversed a global follow transition (i.e.
/// <seealso cref="ATNConfig#reachesIntoOuterContext"/> is 0 for all configurations), then
/// the result of SLL prediction for that input is known to be equivalent to the
/// result of LL prediction for that input.</para>
///
/// <para>
/// In some cases, the minimum represented alternative in the conflicting LL
/// configuration set is not equal to the minimum represented alternative in the
/// conflicting SLL configuration set. Grammars and inputs which result in this
/// scenario are unable to use <seealso cref="PredictionMode#SLL"/>, which in turn means
/// they cannot use the two-stage parsing strategy to improve parsing performance
/// for that input.</para>
/// </summary>
/// <seealso cref= ParserATNSimulator#reportAmbiguity </seealso>
/// <seealso cref= ANTLRErrorListener#reportAmbiguity
///
/// @since 4.3 </seealso>
class ANTLR4CPP_PUBLIC AmbiguityInfo : public DecisionEventInfo {
public:
/// The set of alternative numbers for this decision event that lead to a valid parse.
antlrcpp::BitSet ambigAlts;
/// <summary>
/// Constructs a new instance of the <seealso cref="AmbiguityInfo"/> class with the
/// specified detailed ambiguity information.
/// </summary>
/// <param name="decision"> The decision number </param>
/// <param name="configs"> The final configuration set identifying the ambiguous
/// alternatives for the current input </param>
/// <param name="ambigAlts"> The set of alternatives in the decision that lead to a valid parse.
/// The predicted alt is the min(ambigAlts) </param>
/// <param name="input"> The input token stream </param>
/// <param name="startIndex"> The start index for the current prediction </param>
/// <param name="stopIndex"> The index at which the ambiguity was identified during
/// prediction </param>
/// <param name="fullCtx"> {@code true} if the ambiguity was identified during LL
/// prediction; otherwise, {@code false} if the ambiguity was identified
/// during SLL prediction </param>
AmbiguityInfo(size_t decision, ATNConfigSet *configs, const antlrcpp::BitSet &ambigAlts, TokenStream *input,
size_t startIndex, size_t stopIndex, bool fullCtx);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,43 @@

/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/PredictionContext.h"
namespace antlr4 {
namespace atn {
class SingletonPredictionContext;
class ANTLR4CPP_PUBLIC ArrayPredictionContext : public PredictionContext {
public:
/// Parent can be empty only if full ctx mode and we make an array
/// from EMPTY and non-empty. We merge EMPTY by using null parent and
/// returnState == EMPTY_RETURN_STATE.
// Also here: we use a strong reference to our parents to avoid having them freed prematurely.
// See also SinglePredictionContext.
const std::vector<Ref<PredictionContext>> parents;
/// Sorted for merge, no duplicates; if present, EMPTY_RETURN_STATE is always last.
const std::vector<size_t> returnStates;
ArrayPredictionContext(Ref<SingletonPredictionContext> const& a);
ArrayPredictionContext(std::vector<Ref<PredictionContext>> const& parents_, std::vector<size_t> const& returnStates);
virtual ~ArrayPredictionContext();
virtual bool isEmpty() const override;
virtual size_t size() const override;
virtual Ref<PredictionContext> getParent(size_t index) const override;
virtual size_t getReturnState(size_t index) const override;
bool operator == (const PredictionContext &o) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,30 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
/// TODO: make all transitions sets? no, should remove set edges.
class ANTLR4CPP_PUBLIC AtomTransition final : public Transition {
public:
/// The token type or character value; or, signifies special label.
const size_t _label;
AtomTransition(ATNState *target, size_t label);
virtual SerializationType getSerializationType() const override;
virtual misc::IntervalSet label() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,22 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "antlr4-common.h"
#include "atn/BlockStartState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC BasicBlockStartState final : public BlockStartState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,21 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC BasicState final : public ATNState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,24 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
/// Terminal node of a simple {@code (a|b|c)} block.
class ANTLR4CPP_PUBLIC BlockEndState final : public ATNState {
public:
BlockStartState *startState = nullptr;
BlockEndState();
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,21 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionState.h"
namespace antlr4 {
namespace atn {
/// The start of a regular {@code (...)} block.
class ANTLR4CPP_PUBLIC BlockStartState : public DecisionState {
public:
~BlockStartState();
BlockEndState *endState = nullptr;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,47 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionEventInfo.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This class represents profiling event information for a context sensitivity.
/// Context sensitivities are decisions where a particular input resulted in an
/// SLL conflict, but LL prediction produced a single unique alternative.
///
/// <para>
/// In some cases, the unique alternative identified by LL prediction is not
/// equal to the minimum represented alternative in the conflicting SLL
/// configuration set. Grammars and inputs which result in this scenario are
/// unable to use <seealso cref="PredictionMode#SLL"/>, which in turn means they cannot use
/// the two-stage parsing strategy to improve parsing performance for that
/// input.</para>
/// </summary>
/// <seealso cref= ParserATNSimulator#reportContextSensitivity </seealso>
/// <seealso cref= ANTLRErrorListener#reportContextSensitivity
///
/// @since 4.3 </seealso>
class ANTLR4CPP_PUBLIC ContextSensitivityInfo : public DecisionEventInfo {
public:
/// <summary>
/// Constructs a new instance of the <seealso cref="ContextSensitivityInfo"/> class
/// with the specified detailed context sensitivity information.
/// </summary>
/// <param name="decision"> The decision number </param>
/// <param name="configs"> The final configuration set containing the unique
/// alternative identified by full-context prediction </param>
/// <param name="input"> The input token stream </param>
/// <param name="startIndex"> The start index for the current prediction </param>
/// <param name="stopIndex"> The index at which the context sensitivity was
/// identified during full-context prediction </param>
ContextSensitivityInfo(size_t decision, ATNConfigSet *configs, TokenStream *input, size_t startIndex, size_t stopIndex);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,70 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "antlr4-common.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This is the base class for gathering detailed information about prediction
/// events which occur during parsing.
///
/// Note that we could record the parser call stack at the time this event
/// occurred but in the presence of left recursive rules, the stack is kind of
/// meaningless. It's better to look at the individual configurations for their
/// individual stacks. Of course that is a <seealso cref="PredictionContext"/> object
/// not a parse tree node and so it does not have information about the extent
/// (start...stop) of the various subtrees. Examining the stack tops of all
/// configurations provide the return states for the rule invocations.
/// From there you can get the enclosing rule.
///
/// @since 4.3
/// </summary>
class ANTLR4CPP_PUBLIC DecisionEventInfo {
public:
/// <summary>
/// The invoked decision number which this event is related to.
/// </summary>
/// <seealso cref= ATN#decisionToState </seealso>
const size_t decision;
/// <summary>
/// The configuration set containing additional information relevant to the
/// prediction state when the current event occurred, or {@code null} if no
/// additional information is relevant or available.
/// </summary>
const ATNConfigSet *configs;
/// <summary>
/// The input token stream which is being parsed.
/// </summary>
const TokenStream *input;
/// <summary>
/// The token index in the input stream at which the current prediction was
/// originally invoked.
/// </summary>
const size_t startIndex;
/// <summary>
/// The token index in the input stream at which the current event occurred.
/// </summary>
const size_t stopIndex;
/// <summary>
/// {@code true} if the current event occurred during LL prediction;
/// otherwise, {@code false} if the input occurred during SLL prediction.
/// </summary>
const bool fullCtx;
DecisionEventInfo(size_t decision, ATNConfigSet *configs, TokenStream *input, size_t startIndex,
size_t stopIndex, bool fullCtx);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,227 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ContextSensitivityInfo.h"
#include "atn/AmbiguityInfo.h"
#include "atn/PredicateEvalInfo.h"
#include "atn/ErrorInfo.h"
namespace antlr4 {
namespace atn {
class LookaheadEventInfo;
/// <summary>
/// This class contains profiling gathered for a particular decision.
///
/// <para>
/// Parsing performance in ANTLR 4 is heavily influenced by both static factors
/// (e.g. the form of the rules in the grammar) and dynamic factors (e.g. the
/// choice of input and the state of the DFA cache at the time profiling
/// operations are started). For best results, gather and use aggregate
/// statistics from a large sample of inputs representing the inputs expected in
/// production before using the results to make changes in the grammar.</para>
///
/// @since 4.3
/// </summary>
class ANTLR4CPP_PUBLIC DecisionInfo {
public:
/// <summary>
/// The decision number, which is an index into <seealso cref="ATN#decisionToState"/>.
/// </summary>
const size_t decision;
/// <summary>
/// The total number of times <seealso cref="ParserATNSimulator#adaptivePredict"/> was
/// invoked for this decision.
/// </summary>
long long invocations = 0;
/// <summary>
/// The total time spent in <seealso cref="ParserATNSimulator#adaptivePredict"/> for
/// this decision, in nanoseconds.
///
/// <para>
/// The value of this field contains the sum of differential results obtained
/// by <seealso cref="System#nanoTime()"/>, and is not adjusted to compensate for JIT
/// and/or garbage collection overhead. For best accuracy, use a modern JVM
/// implementation that provides precise results from
/// <seealso cref="System#nanoTime()"/>, and perform profiling in a separate process
/// which is warmed up by parsing the input prior to profiling. If desired,
/// call <seealso cref="ATNSimulator#clearDFA"/> to reset the DFA cache to its initial
/// state before starting the profiling measurement pass.</para>
/// </summary>
long long timeInPrediction = 0;
/// <summary>
/// The sum of the lookahead required for SLL prediction for this decision.
/// Note that SLL prediction is used before LL prediction for performance
/// reasons even when <seealso cref="PredictionMode#LL"/> or
/// <seealso cref="PredictionMode#LL_EXACT_AMBIG_DETECTION"/> is used.
/// </summary>
long long SLL_TotalLook = 0;
/// <summary>
/// Gets the minimum lookahead required for any single SLL prediction to
/// complete for this decision, by reaching a unique prediction, reaching an
/// SLL conflict state, or encountering a syntax error.
/// </summary>
long long SLL_MinLook = 0;
/// <summary>
/// Gets the maximum lookahead required for any single SLL prediction to
/// complete for this decision, by reaching a unique prediction, reaching an
/// SLL conflict state, or encountering a syntax error.
/// </summary>
long long SLL_MaxLook = 0;
/// Gets the <seealso cref="LookaheadEventInfo"/> associated with the event where the
/// <seealso cref="#SLL_MaxLook"/> value was set.
Ref<LookaheadEventInfo> SLL_MaxLookEvent;
/// <summary>
/// The sum of the lookahead required for LL prediction for this decision.
/// Note that LL prediction is only used when SLL prediction reaches a
/// conflict state.
/// </summary>
long long LL_TotalLook = 0;
/// <summary>
/// Gets the minimum lookahead required for any single LL prediction to
/// complete for this decision. An LL prediction completes when the algorithm
/// reaches a unique prediction, a conflict state (for
/// <seealso cref="PredictionMode#LL"/>, an ambiguity state (for
/// <seealso cref="PredictionMode#LL_EXACT_AMBIG_DETECTION"/>, or a syntax error.
/// </summary>
long long LL_MinLook = 0;
/// <summary>
/// Gets the maximum lookahead required for any single LL prediction to
/// complete for this decision. An LL prediction completes when the algorithm
/// reaches a unique prediction, a conflict state (for
/// <seealso cref="PredictionMode#LL"/>, an ambiguity state (for
/// <seealso cref="PredictionMode#LL_EXACT_AMBIG_DETECTION"/>, or a syntax error.
/// </summary>
long long LL_MaxLook = 0;
/// <summary>
/// Gets the <seealso cref="LookaheadEventInfo"/> associated with the event where the
/// <seealso cref="#LL_MaxLook"/> value was set.
/// </summary>
Ref<LookaheadEventInfo> LL_MaxLookEvent;
/// <summary>
/// A collection of <seealso cref="ContextSensitivityInfo"/> instances describing the
/// context sensitivities encountered during LL prediction for this decision.
/// </summary>
/// <seealso cref= ContextSensitivityInfo </seealso>
std::vector<ContextSensitivityInfo> contextSensitivities;
/// <summary>
/// A collection of <seealso cref="ErrorInfo"/> instances describing the parse errors
/// identified during calls to <seealso cref="ParserATNSimulator#adaptivePredict"/> for
/// this decision.
/// </summary>
/// <seealso cref= ErrorInfo </seealso>
std::vector<ErrorInfo> errors;
/// <summary>
/// A collection of <seealso cref="AmbiguityInfo"/> instances describing the
/// ambiguities encountered during LL prediction for this decision.
/// </summary>
/// <seealso cref= AmbiguityInfo </seealso>
std::vector<AmbiguityInfo> ambiguities;
/// <summary>
/// A collection of <seealso cref="PredicateEvalInfo"/> instances describing the
/// results of evaluating individual predicates during prediction for this
/// decision.
/// </summary>
/// <seealso cref= PredicateEvalInfo </seealso>
std::vector<PredicateEvalInfo> predicateEvals;
/// <summary>
/// The total number of ATN transitions required during SLL prediction for
/// this decision. An ATN transition is determined by the number of times the
/// DFA does not contain an edge that is required for prediction, resulting
/// in on-the-fly computation of that edge.
///
/// <para>
/// If DFA caching of SLL transitions is employed by the implementation, ATN
/// computation may cache the computed edge for efficient lookup during
/// future parsing of this decision. Otherwise, the SLL parsing algorithm
/// will use ATN transitions exclusively.</para>
/// </summary>
/// <seealso cref= #SLL_ATNTransitions </seealso>
/// <seealso cref= ParserATNSimulator#computeTargetState </seealso>
/// <seealso cref= LexerATNSimulator#computeTargetState </seealso>
long long SLL_ATNTransitions = 0;
/// <summary>
/// The total number of DFA transitions required during SLL prediction for
/// this decision.
///
/// <para>If the ATN simulator implementation does not use DFA caching for SLL
/// transitions, this value will be 0.</para>
/// </summary>
/// <seealso cref= ParserATNSimulator#getExistingTargetState </seealso>
/// <seealso cref= LexerATNSimulator#getExistingTargetState </seealso>
long long SLL_DFATransitions = 0;
/// <summary>
/// Gets the total number of times SLL prediction completed in a conflict
/// state, resulting in fallback to LL prediction.
///
/// <para>Note that this value is not related to whether or not
/// <seealso cref="PredictionMode#SLL"/> may be used successfully with a particular
/// grammar. If the ambiguity resolution algorithm applied to the SLL
/// conflicts for this decision produce the same result as LL prediction for
/// this decision, <seealso cref="PredictionMode#SLL"/> would produce the same overall
/// parsing result as <seealso cref="PredictionMode#LL"/>.</para>
/// </summary>
long long LL_Fallback = 0;
/// <summary>
/// The total number of ATN transitions required during LL prediction for
/// this decision. An ATN transition is determined by the number of times the
/// DFA does not contain an edge that is required for prediction, resulting
/// in on-the-fly computation of that edge.
///
/// <para>
/// If DFA caching of LL transitions is employed by the implementation, ATN
/// computation may cache the computed edge for efficient lookup during
/// future parsing of this decision. Otherwise, the LL parsing algorithm will
/// use ATN transitions exclusively.</para>
/// </summary>
/// <seealso cref= #LL_DFATransitions </seealso>
/// <seealso cref= ParserATNSimulator#computeTargetState </seealso>
/// <seealso cref= LexerATNSimulator#computeTargetState </seealso>
long long LL_ATNTransitions = 0;
/// <summary>
/// The total number of DFA transitions required during LL prediction for
/// this decision.
///
/// <para>If the ATN simulator implementation does not use DFA caching for LL
/// transitions, this value will be 0.</para>
/// </summary>
/// <seealso cref= ParserATNSimulator#getExistingTargetState </seealso>
/// <seealso cref= LexerATNSimulator#getExistingTargetState </seealso>
long long LL_DFATransitions = 0;
/// <summary>
/// Constructs a new instance of the <seealso cref="DecisionInfo"/> class to contain
/// statistics for a particular decision.
/// </summary>
/// <param name="decision"> The decision number </param>
DecisionInfo(size_t decision);
std::string toString() const;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,30 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC DecisionState : public ATNState {
public:
int decision;
bool nonGreedy;
private:
void InitializeInstanceFields();
public:
DecisionState() {
InitializeInstanceFields();
}
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,27 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/SingletonPredictionContext.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC EmptyPredictionContext : public SingletonPredictionContext {
public:
EmptyPredictionContext();
virtual bool isEmpty() const override;
virtual size_t size() const override;
virtual Ref<PredictionContext> getParent(size_t index) const override;
virtual size_t getReturnState(size_t index) const override;
virtual std::string toString() const override;
virtual bool operator == (const PredictionContext &o) const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,39 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC EpsilonTransition final : public Transition {
public:
EpsilonTransition(ATNState *target);
EpsilonTransition(ATNState *target, size_t outermostPrecedenceReturn);
/**
* @return the rule index of a precedence rule for which this transition is
* returning from, where the precedence value is 0; otherwise, INVALID_INDEX.
*
* @see ATNConfig#isPrecedenceFilterSuppressed()
* @see ParserATNSimulator#applyPrecedenceFilter(ATNConfigSet)
* @since 4.4.1
*/
size_t outermostPrecedenceReturn();
virtual SerializationType getSerializationType() const override;
virtual bool isEpsilon() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
private:
const size_t _outermostPrecedenceReturn; // A rule index.
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,43 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionEventInfo.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This class represents profiling event information for a syntax error
/// identified during prediction. Syntax errors occur when the prediction
/// algorithm is unable to identify an alternative which would lead to a
/// successful parse.
/// </summary>
/// <seealso cref= Parser#notifyErrorListeners(Token, String, RecognitionException) </seealso>
/// <seealso cref= ANTLRErrorListener#syntaxError
///
/// @since 4.3 </seealso>
class ANTLR4CPP_PUBLIC ErrorInfo : public DecisionEventInfo {
public:
/// <summary>
/// Constructs a new instance of the <seealso cref="ErrorInfo"/> class with the
/// specified detailed syntax error information.
/// </summary>
/// <param name="decision"> The decision number </param>
/// <param name="configs"> The final configuration set reached during prediction
/// prior to reaching the <seealso cref="ATNSimulator#ERROR"/> state </param>
/// <param name="input"> The input token stream </param>
/// <param name="startIndex"> The start index for the current prediction </param>
/// <param name="stopIndex"> The index at which the syntax error was identified </param>
/// <param name="fullCtx"> {@code true} if the syntax error was identified during LL
/// prediction; otherwise, {@code false} if the syntax error was identified
/// during SLL prediction </param>
ErrorInfo(size_t decision, ATNConfigSet *configs, TokenStream *input, size_t startIndex, size_t stopIndex,
bool fullCtx);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,109 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "Token.h"
#include "support/BitSet.h"
#include "atn/PredictionContext.h"
#include "atn/ATNConfig.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC LL1Analyzer {
public:
/// Special value added to the lookahead sets to indicate that we hit
/// a predicate during analysis if {@code seeThruPreds==false}.
static const size_t HIT_PRED = Token::INVALID_TYPE;
const atn::ATN &_atn;
LL1Analyzer(const atn::ATN &atn);
virtual ~LL1Analyzer();
/// <summary>
/// Calculates the SLL(1) expected lookahead set for each outgoing transition
/// of an <seealso cref="ATNState"/>. The returned array has one element for each
/// outgoing transition in {@code s}. If the closure from transition
/// <em>i</em> leads to a semantic predicate before matching a symbol, the
/// element at index <em>i</em> of the result will be {@code null}.
/// </summary>
/// <param name="s"> the ATN state </param>
/// <returns> the expected symbols for each outgoing transition of {@code s}. </returns>
virtual std::vector<misc::IntervalSet> getDecisionLookahead(ATNState *s) const;
/// <summary>
/// Compute set of tokens that can follow {@code s} in the ATN in the
/// specified {@code ctx}.
/// <p/>
/// If {@code ctx} is {@code null} and the end of the rule containing
/// {@code s} is reached, <seealso cref="Token#EPSILON"/> is added to the result set.
/// If {@code ctx} is not {@code null} and the end of the outermost rule is
/// reached, <seealso cref="Token#EOF"/> is added to the result set.
/// </summary>
/// <param name="s"> the ATN state </param>
/// <param name="ctx"> the complete parser context, or {@code null} if the context
/// should be ignored
/// </param>
/// <returns> The set of tokens that can follow {@code s} in the ATN in the
/// specified {@code ctx}. </returns>
virtual misc::IntervalSet LOOK(ATNState *s, RuleContext *ctx) const;
/// <summary>
/// Compute set of tokens that can follow {@code s} in the ATN in the
/// specified {@code ctx}.
/// <p/>
/// If {@code ctx} is {@code null} and the end of the rule containing
/// {@code s} is reached, <seealso cref="Token#EPSILON"/> is added to the result set.
/// If {@code ctx} is not {@code null} and the end of the outermost rule is
/// reached, <seealso cref="Token#EOF"/> is added to the result set.
/// </summary>
/// <param name="s"> the ATN state </param>
/// <param name="stopState"> the ATN state to stop at. This can be a
/// <seealso cref="BlockEndState"/> to detect epsilon paths through a closure. </param>
/// <param name="ctx"> the complete parser context, or {@code null} if the context
/// should be ignored
/// </param>
/// <returns> The set of tokens that can follow {@code s} in the ATN in the
/// specified {@code ctx}. </returns>
virtual misc::IntervalSet LOOK(ATNState *s, ATNState *stopState, RuleContext *ctx) const;
/// <summary>
/// Compute set of tokens that can follow {@code s} in the ATN in the
/// specified {@code ctx}.
/// <p/>
/// If {@code ctx} is {@code null} and {@code stopState} or the end of the
/// rule containing {@code s} is reached, <seealso cref="Token#EPSILON"/> is added to
/// the result set. If {@code ctx} is not {@code null} and {@code addEOF} is
/// {@code true} and {@code stopState} or the end of the outermost rule is
/// reached, <seealso cref="Token#EOF"/> is added to the result set.
/// </summary>
/// <param name="s"> the ATN state. </param>
/// <param name="stopState"> the ATN state to stop at. This can be a
/// <seealso cref="BlockEndState"/> to detect epsilon paths through a closure. </param>
/// <param name="ctx"> The outer context, or {@code null} if the outer context should
/// not be used. </param>
/// <param name="look"> The result lookahead set. </param>
/// <param name="lookBusy"> A set used for preventing epsilon closures in the ATN
/// from causing a stack overflow. Outside code should pass
/// {@code new HashSet<ATNConfig>} for this argument. </param>
/// <param name="calledRuleStack"> A set used for preventing left recursion in the
/// ATN from causing a stack overflow. Outside code should pass
/// {@code new BitSet()} for this argument. </param>
/// <param name="seeThruPreds"> {@code true} to true semantic predicates as
/// implicitly {@code true} and "see through them", otherwise {@code false}
/// to treat semantic predicates as opaque and add <seealso cref="#HIT_PRED"/> to the
/// result if one is encountered. </param>
/// <param name="addEOF"> Add <seealso cref="Token#EOF"/> to the result if the end of the
/// outermost context is reached. This parameter has no effect if {@code ctx}
/// is {@code null}. </param>
protected:
virtual void _LOOK(ATNState *s, ATNState *stopState, Ref<PredictionContext> const& ctx, misc::IntervalSet &look,
ATNConfig::Set &lookBusy, antlrcpp::BitSet &calledRuleStack, bool seeThruPreds, bool addEOF) const;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,44 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNConfig.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC LexerATNConfig : public ATNConfig {
public:
LexerATNConfig(ATNState *state, int alt, Ref<PredictionContext> const& context);
LexerATNConfig(ATNState *state, int alt, Ref<PredictionContext> const& context, Ref<LexerActionExecutor> const& lexerActionExecutor);
LexerATNConfig(Ref<LexerATNConfig> const& c, ATNState *state);
LexerATNConfig(Ref<LexerATNConfig> const& c, ATNState *state, Ref<LexerActionExecutor> const& lexerActionExecutor);
LexerATNConfig(Ref<LexerATNConfig> const& c, ATNState *state, Ref<PredictionContext> const& context);
/**
* Gets the {@link LexerActionExecutor} capable of executing the embedded
* action(s) for the current configuration.
*/
Ref<LexerActionExecutor> getLexerActionExecutor() const;
bool hasPassedThroughNonGreedyDecision();
virtual size_t hashCode() const override;
bool operator == (const LexerATNConfig& other) const;
private:
/**
* This is the backing field for {@link #getLexerActionExecutor}.
*/
const Ref<LexerActionExecutor> _lexerActionExecutor;
const bool _passedThroughNonGreedyDecision;
static bool checkNonGreedyDecision(Ref<LexerATNConfig> const& source, ATNState *target);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,210 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNSimulator.h"
#include "atn/LexerATNConfig.h"
#include "atn/ATNConfigSet.h"
namespace antlr4 {
namespace atn {
/// "dup" of ParserInterpreter
class ANTLR4CPP_PUBLIC LexerATNSimulator : public ATNSimulator {
protected:
class SimState {
public:
virtual ~SimState();
protected:
size_t index;
size_t line;
size_t charPos;
dfa::DFAState *dfaState;
virtual void reset();
friend class LexerATNSimulator;
private:
void InitializeInstanceFields();
public:
SimState() {
InitializeInstanceFields();
}
};
public:
static const size_t MIN_DFA_EDGE = 0;
static const size_t MAX_DFA_EDGE = 127; // forces unicode to stay in ATN
protected:
/// <summary>
/// When we hit an accept state in either the DFA or the ATN, we
/// have to notify the character stream to start buffering characters
/// via <seealso cref="IntStream#mark"/> and record the current state. The current sim state
/// includes the current index into the input, the current line,
/// and current character position in that line. Note that the Lexer is
/// tracking the starting line and characterization of the token. These
/// variables track the "state" of the simulator when it hits an accept state.
/// <p/>
/// We track these variables separately for the DFA and ATN simulation
/// because the DFA simulation often has to fail over to the ATN
/// simulation. If the ATN simulation fails, we need the DFA to fall
/// back to its previously accepted state, if any. If the ATN succeeds,
/// then the ATN does the accept and the DFA simulator that invoked it
/// can simply return the predicted token type.
/// </summary>
Lexer *const _recog;
/// The current token's starting index into the character stream.
/// Shared across DFA to ATN simulation in case the ATN fails and the
/// DFA did not have a previous accept state. In this case, we use the
/// ATN-generated exception object.
size_t _startIndex;
/// line number 1..n within the input.
size_t _line;
/// The index of the character relative to the beginning of the line 0..n-1.
size_t _charPositionInLine;
public:
std::vector<dfa::DFA> &_decisionToDFA;
protected:
size_t _mode;
/// Used during DFA/ATN exec to record the most recent accept configuration info.
SimState _prevAccept;
public:
static int match_calls;
LexerATNSimulator(const ATN &atn, std::vector<dfa::DFA> &decisionToDFA, PredictionContextCache &sharedContextCache);
LexerATNSimulator(Lexer *recog, const ATN &atn, std::vector<dfa::DFA> &decisionToDFA, PredictionContextCache &sharedContextCache);
virtual ~LexerATNSimulator () {}
virtual void copyState(LexerATNSimulator *simulator);
virtual size_t match(CharStream *input, size_t mode);
virtual void reset() override;
virtual void clearDFA() override;
protected:
virtual size_t matchATN(CharStream *input);
virtual size_t execATN(CharStream *input, dfa::DFAState *ds0);
/// <summary>
/// Get an existing target state for an edge in the DFA. If the target state
/// for the edge has not yet been computed or is otherwise not available,
/// this method returns {@code null}.
/// </summary>
/// <param name="s"> The current DFA state </param>
/// <param name="t"> The next input symbol </param>
/// <returns> The existing target DFA state for the given input symbol
/// {@code t}, or {@code null} if the target state for this edge is not
/// already cached </returns>
virtual dfa::DFAState *getExistingTargetState(dfa::DFAState *s, size_t t);
/// <summary>
/// Compute a target state for an edge in the DFA, and attempt to add the
/// computed state and corresponding edge to the DFA.
/// </summary>
/// <param name="input"> The input stream </param>
/// <param name="s"> The current DFA state </param>
/// <param name="t"> The next input symbol
/// </param>
/// <returns> The computed target DFA state for the given input symbol
/// {@code t}. If {@code t} does not lead to a valid DFA state, this method
/// returns <seealso cref="#ERROR"/>. </returns>
virtual dfa::DFAState *computeTargetState(CharStream *input, dfa::DFAState *s, size_t t);
virtual size_t failOrAccept(CharStream *input, ATNConfigSet *reach, size_t t);
/// <summary>
/// Given a starting configuration set, figure out all ATN configurations
/// we can reach upon input {@code t}. Parameter {@code reach} is a return
/// parameter.
/// </summary>
void getReachableConfigSet(CharStream *input, ATNConfigSet *closure_, // closure_ as we have a closure() already
ATNConfigSet *reach, size_t t);
virtual void accept(CharStream *input, const Ref<LexerActionExecutor> &lexerActionExecutor, size_t startIndex, size_t index,
size_t line, size_t charPos);
virtual ATNState *getReachableTarget(Transition *trans, size_t t);
virtual std::unique_ptr<ATNConfigSet> computeStartState(CharStream *input, ATNState *p);
/// <summary>
/// Since the alternatives within any lexer decision are ordered by
/// preference, this method stops pursuing the closure as soon as an accept
/// state is reached. After the first accept state is reached by depth-first
/// search from {@code config}, all other (potentially reachable) states for
/// this rule would have a lower priority.
/// </summary>
/// <returns> {@code true} if an accept state is reached, otherwise
/// {@code false}. </returns>
virtual bool closure(CharStream *input, const Ref<LexerATNConfig> &config, ATNConfigSet *configs,
bool currentAltReachedAcceptState, bool speculative, bool treatEofAsEpsilon);
// side-effect: can alter configs.hasSemanticContext
virtual Ref<LexerATNConfig> getEpsilonTarget(CharStream *input, const Ref<LexerATNConfig> &config, Transition *t,
ATNConfigSet *configs, bool speculative, bool treatEofAsEpsilon);
/// <summary>
/// Evaluate a predicate specified in the lexer.
/// <p/>
/// If {@code speculative} is {@code true}, this method was called before
/// <seealso cref="#consume"/> for the matched character. This method should call
/// <seealso cref="#consume"/> before evaluating the predicate to ensure position
/// sensitive values, including <seealso cref="Lexer#getText"/>, <seealso cref="Lexer#getLine"/>,
/// and <seealso cref="Lexer#getCharPositionInLine"/>, properly reflect the current
/// lexer state. This method should restore {@code input} and the simulator
/// to the original state before returning (i.e. undo the actions made by the
/// call to <seealso cref="#consume"/>.
/// </summary>
/// <param name="input"> The input stream. </param>
/// <param name="ruleIndex"> The rule containing the predicate. </param>
/// <param name="predIndex"> The index of the predicate within the rule. </param>
/// <param name="speculative"> {@code true} if the current index in {@code input} is
/// one character before the predicate's location.
/// </param>
/// <returns> {@code true} if the specified predicate evaluates to
/// {@code true}. </returns>
virtual bool evaluatePredicate(CharStream *input, size_t ruleIndex, size_t predIndex, bool speculative);
virtual void captureSimState(CharStream *input, dfa::DFAState *dfaState);
virtual dfa::DFAState* addDFAEdge(dfa::DFAState *from, size_t t, ATNConfigSet *q);
virtual void addDFAEdge(dfa::DFAState *p, size_t t, dfa::DFAState *q);
/// <summary>
/// Add a new DFA state if there isn't one with this set of
/// configurations already. This method also detects the first
/// configuration containing an ATN rule stop state. Later, when
/// traversing the DFA, we will know which rule to accept.
/// </summary>
virtual dfa::DFAState *addDFAState(ATNConfigSet *configs);
public:
dfa::DFA& getDFA(size_t mode);
/// Get the text matched so far for the current token.
virtual std::string getText(CharStream *input);
virtual size_t getLine() const;
virtual void setLine(size_t line);
virtual size_t getCharPositionInLine();
virtual void setCharPositionInLine(size_t charPositionInLine);
virtual void consume(CharStream *input);
virtual std::string getTokenName(size_t t);
private:
void InitializeInstanceFields();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,66 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerActionType.h"
#include "antlr4-common.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Represents a single action which can be executed following the successful
/// match of a lexer rule. Lexer actions are used for both embedded action syntax
/// and ANTLR 4's new lexer command syntax.
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerAction {
public:
virtual ~LexerAction();
/// <summary>
/// Gets the serialization type of the lexer action.
/// </summary>
/// <returns> The serialization type of the lexer action. </returns>
virtual LexerActionType getActionType() const = 0;
/// <summary>
/// Gets whether the lexer action is position-dependent. Position-dependent
/// actions may have different semantics depending on the <seealso cref="CharStream"/>
/// index at the time the action is executed.
///
/// <para>Many lexer commands, including {@code type}, {@code skip}, and
/// {@code more}, do not check the input index during their execution.
/// Actions like this are position-independent, and may be stored more
/// efficiently as part of the <seealso cref="LexerATNConfig#lexerActionExecutor"/>.</para>
/// </summary>
/// <returns> {@code true} if the lexer action semantics can be affected by the
/// position of the input <seealso cref="CharStream"/> at the time it is executed;
/// otherwise, {@code false}. </returns>
virtual bool isPositionDependent() const = 0;
/// <summary>
/// Execute the lexer action in the context of the specified <seealso cref="Lexer"/>.
///
/// <para>For position-dependent actions, the input stream must already be
/// positioned correctly prior to calling this method.</para>
/// </summary>
/// <param name="lexer"> The lexer instance. </param>
virtual void execute(Lexer *lexer) = 0;
virtual size_t hashCode() const = 0;
virtual bool operator == (const LexerAction &obj) const = 0;
virtual bool operator != (const LexerAction &obj) const {
return !(*this == obj);
}
virtual std::string toString() const = 0;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,115 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "CharStream.h"
#include "atn/LexerAction.h"
namespace antlr4 {
namespace atn {
/// Represents an executor for a sequence of lexer actions which traversed during
/// the matching operation of a lexer rule (token).
///
/// <para>The executor tracks position information for position-dependent lexer actions
/// efficiently, ensuring that actions appearing only at the end of the rule do
/// not cause bloating of the <seealso cref="DFA"/> created for the lexer.</para>
class ANTLR4CPP_PUBLIC LexerActionExecutor : public std::enable_shared_from_this<LexerActionExecutor> {
public:
/// <summary>
/// Constructs an executor for a sequence of <seealso cref="LexerAction"/> actions. </summary>
/// <param name="lexerActions"> The lexer actions to execute. </param>
LexerActionExecutor(const std::vector<Ref<LexerAction>> &lexerActions);
virtual ~LexerActionExecutor();
/// <summary>
/// Creates a <seealso cref="LexerActionExecutor"/> which executes the actions for
/// the input {@code lexerActionExecutor} followed by a specified
/// {@code lexerAction}.
/// </summary>
/// <param name="lexerActionExecutor"> The executor for actions already traversed by
/// the lexer while matching a token within a particular
/// <seealso cref="LexerATNConfig"/>. If this is {@code null}, the method behaves as
/// though it were an empty executor. </param>
/// <param name="lexerAction"> The lexer action to execute after the actions
/// specified in {@code lexerActionExecutor}.
/// </param>
/// <returns> A <seealso cref="LexerActionExecutor"/> for executing the combine actions
/// of {@code lexerActionExecutor} and {@code lexerAction}. </returns>
static Ref<LexerActionExecutor> append(Ref<LexerActionExecutor> const& lexerActionExecutor,
Ref<LexerAction> const& lexerAction);
/// <summary>
/// Creates a <seealso cref="LexerActionExecutor"/> which encodes the current offset
/// for position-dependent lexer actions.
///
/// <para>Normally, when the executor encounters lexer actions where
/// <seealso cref="LexerAction#isPositionDependent"/> returns {@code true}, it calls
/// <seealso cref="IntStream#seek"/> on the input <seealso cref="CharStream"/> to set the input
/// position to the <em>end</em> of the current token. This behavior provides
/// for efficient DFA representation of lexer actions which appear at the end
/// of a lexer rule, even when the lexer rule matches a variable number of
/// characters.</para>
///
/// <para>Prior to traversing a match transition in the ATN, the current offset
/// from the token start index is assigned to all position-dependent lexer
/// actions which have not already been assigned a fixed offset. By storing
/// the offsets relative to the token start index, the DFA representation of
/// lexer actions which appear in the middle of tokens remains efficient due
/// to sharing among tokens of the same length, regardless of their absolute
/// position in the input stream.</para>
///
/// <para>If the current executor already has offsets assigned to all
/// position-dependent lexer actions, the method returns {@code this}.</para>
/// </summary>
/// <param name="offset"> The current offset to assign to all position-dependent
/// lexer actions which do not already have offsets assigned.
/// </param>
/// <returns> A <seealso cref="LexerActionExecutor"/> which stores input stream offsets
/// for all position-dependent lexer actions. </returns>
virtual Ref<LexerActionExecutor> fixOffsetBeforeMatch(int offset);
/// <summary>
/// Gets the lexer actions to be executed by this executor. </summary>
/// <returns> The lexer actions to be executed by this executor. </returns>
virtual std::vector<Ref<LexerAction>> getLexerActions() const;
/// <summary>
/// Execute the actions encapsulated by this executor within the context of a
/// particular <seealso cref="Lexer"/>.
///
/// <para>This method calls <seealso cref="IntStream#seek"/> to set the position of the
/// {@code input} <seealso cref="CharStream"/> prior to calling
/// <seealso cref="LexerAction#execute"/> on a position-dependent action. Before the
/// method returns, the input position will be restored to the same position
/// it was in when the method was invoked.</para>
/// </summary>
/// <param name="lexer"> The lexer instance. </param>
/// <param name="input"> The input stream which is the source for the current token.
/// When this method is called, the current <seealso cref="IntStream#index"/> for
/// {@code input} should be the start of the following token, i.e. 1
/// character past the end of the current token. </param>
/// <param name="startIndex"> The token start index. This value may be passed to
/// <seealso cref="IntStream#seek"/> to set the {@code input} position to the beginning
/// of the token. </param>
virtual void execute(Lexer *lexer, CharStream *input, size_t startIndex);
virtual size_t hashCode() const;
virtual bool operator == (const LexerActionExecutor &obj) const;
virtual bool operator != (const LexerActionExecutor &obj) const;
private:
const std::vector<Ref<LexerAction>> _lexerActions;
/// Caches the result of <seealso cref="#hashCode"/> since the hash code is an element
/// of the performance-critical <seealso cref="LexerATNConfig#hashCode"/> operation.
const size_t _hashCode;
size_t generateHashCode() const;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,55 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "antlr4-common.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Represents the serialization type of a <seealso cref="LexerAction"/>.
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
enum class LexerActionType : size_t {
/// <summary>
/// The type of a <seealso cref="LexerChannelAction"/> action.
/// </summary>
CHANNEL,
/// <summary>
/// The type of a <seealso cref="LexerCustomAction"/> action.
/// </summary>
CUSTOM,
/// <summary>
/// The type of a <seealso cref="LexerModeAction"/> action.
/// </summary>
MODE,
/// <summary>
/// The type of a <seealso cref="LexerMoreAction"/> action.
/// </summary>
MORE,
/// <summary>
/// The type of a <seealso cref="LexerPopModeAction"/> action.
/// </summary>
POP_MODE,
/// <summary>
/// The type of a <seealso cref="LexerPushModeAction"/> action.
/// </summary>
PUSH_MODE,
/// <summary>
/// The type of a <seealso cref="LexerSkipAction"/> action.
/// </summary>
SKIP,
/// <summary>
/// The type of a <seealso cref="LexerTypeAction"/> action.
/// </summary>
TYPE,
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,63 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
using antlr4::Lexer;
/// <summary>
/// Implements the {@code channel} lexer action by calling
/// <seealso cref="Lexer#setChannel"/> with the assigned channel.
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerChannelAction final : public LexerAction {
public:
/// <summary>
/// Constructs a new {@code channel} action with the specified channel value. </summary>
/// <param name="channel"> The channel value to pass to <seealso cref="Lexer#setChannel"/>. </param>
LexerChannelAction(int channel);
/// <summary>
/// Gets the channel to use for the <seealso cref="Token"/> created by the lexer.
/// </summary>
/// <returns> The channel to use for the <seealso cref="Token"/> created by the lexer. </returns>
int getChannel() const;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#CHANNEL"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#setChannel"/> with the
/// value provided by <seealso cref="#getChannel"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const int _channel;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,87 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Executes a custom lexer action by calling <seealso cref="Recognizer#action"/> with the
/// rule and action indexes assigned to the custom action. The implementation of
/// a custom action is added to the generated code for the lexer in an override
/// of <seealso cref="Recognizer#action"/> when the grammar is compiled.
///
/// <para>This class may represent embedded actions created with the <code>{...}</code>
/// syntax in ANTLR 4, as well as actions created for lexer commands where the
/// command argument could not be evaluated when the grammar was compiled.</para>
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerCustomAction final : public LexerAction {
public:
/// <summary>
/// Constructs a custom lexer action with the specified rule and action
/// indexes.
/// </summary>
/// <param name="ruleIndex"> The rule index to use for calls to
/// <seealso cref="Recognizer#action"/>. </param>
/// <param name="actionIndex"> The action index to use for calls to
/// <seealso cref="Recognizer#action"/>. </param>
LexerCustomAction(size_t ruleIndex, size_t actionIndex);
/// <summary>
/// Gets the rule index to use for calls to <seealso cref="Recognizer#action"/>.
/// </summary>
/// <returns> The rule index for the custom action. </returns>
size_t getRuleIndex() const;
/// <summary>
/// Gets the action index to use for calls to <seealso cref="Recognizer#action"/>.
/// </summary>
/// <returns> The action index for the custom action. </returns>
size_t getActionIndex() const;
/// <summary>
/// {@inheritDoc}
/// </summary>
/// <returns> This method returns <seealso cref="LexerActionType#CUSTOM"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// Gets whether the lexer action is position-dependent. Position-dependent
/// actions may have different semantics depending on the <seealso cref="CharStream"/>
/// index at the time the action is executed.
///
/// <para>Custom actions are position-dependent since they may represent a
/// user-defined embedded action which makes calls to methods like
/// <seealso cref="Lexer#getText"/>.</para>
/// </summary>
/// <returns> This method returns {@code true}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>Custom actions are implemented by calling <seealso cref="Lexer#action"/> with the
/// appropriate rule and action indexes.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const size_t _ruleIndex;
const size_t _actionIndex;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,82 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "RuleContext.h"
#include "atn/LexerAction.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This implementation of <seealso cref="LexerAction"/> is used for tracking input offsets
/// for position-dependent actions within a <seealso cref="LexerActionExecutor"/>.
///
/// <para>This action is not serialized as part of the ATN, and is only required for
/// position-dependent lexer actions which appear at a location other than the
/// end of a rule. For more information about DFA optimizations employed for
/// lexer actions, see <seealso cref="LexerActionExecutor#append"/> and
/// <seealso cref="LexerActionExecutor#fixOffsetBeforeMatch"/>.</para>
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerIndexedCustomAction final : public LexerAction {
public:
/// <summary>
/// Constructs a new indexed custom action by associating a character offset
/// with a <seealso cref="LexerAction"/>.
///
/// <para>Note: This class is only required for lexer actions for which
/// <seealso cref="LexerAction#isPositionDependent"/> returns {@code true}.</para>
/// </summary>
/// <param name="offset"> The offset into the input <seealso cref="CharStream"/>, relative to
/// the token start index, at which the specified lexer action should be
/// executed. </param>
/// <param name="action"> The lexer action to execute at a particular offset in the
/// input <seealso cref="CharStream"/>. </param>
LexerIndexedCustomAction(int offset, Ref<LexerAction> const& action);
/// <summary>
/// Gets the location in the input <seealso cref="CharStream"/> at which the lexer
/// action should be executed. The value is interpreted as an offset relative
/// to the token start index.
/// </summary>
/// <returns> The location in the input <seealso cref="CharStream"/> at which the lexer
/// action should be executed. </returns>
int getOffset() const;
/// <summary>
/// Gets the lexer action to execute.
/// </summary>
/// <returns> A <seealso cref="LexerAction"/> object which executes the lexer action. </returns>
Ref<LexerAction> getAction() const;
/// <summary>
/// {@inheritDoc}
/// </summary>
/// <returns> This method returns the result of calling <seealso cref="#getActionType"/>
/// on the <seealso cref="LexerAction"/> returned by <seealso cref="#getAction"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code true}. </returns>
virtual bool isPositionDependent() const override;
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const int _offset;
const Ref<LexerAction> _action;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,61 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Implements the {@code mode} lexer action by calling <seealso cref="Lexer#mode"/> with
/// the assigned mode.
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerModeAction final : public LexerAction {
public:
/// <summary>
/// Constructs a new {@code mode} action with the specified mode value. </summary>
/// <param name="mode"> The mode value to pass to <seealso cref="Lexer#mode"/>. </param>
LexerModeAction(int mode);
/// <summary>
/// Get the lexer mode this action should transition the lexer to.
/// </summary>
/// <returns> The lexer mode for this {@code mode} command. </returns>
int getMode();
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#MODE"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#mode"/> with the
/// value provided by <seealso cref="#getMode"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const int _mode;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,57 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Implements the {@code more} lexer action by calling <seealso cref="Lexer#more"/>.
///
/// <para>The {@code more} command does not have any parameters, so this action is
/// implemented as a singleton instance exposed by <seealso cref="#INSTANCE"/>.</para>
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerMoreAction final : public LexerAction {
public:
/// <summary>
/// Provides a singleton instance of this parameterless lexer action.
/// </summary>
static const Ref<LexerMoreAction> getInstance();
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#MORE"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#more"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
/// Constructs the singleton instance of the lexer {@code more} command.
LexerMoreAction();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,57 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Implements the {@code popMode} lexer action by calling <seealso cref="Lexer#popMode"/>.
///
/// <para>The {@code popMode} command does not have any parameters, so this action is
/// implemented as a singleton instance exposed by <seealso cref="#INSTANCE"/>.</para>
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerPopModeAction final : public LexerAction {
public:
/// <summary>
/// Provides a singleton instance of this parameterless lexer action.
/// </summary>
static const Ref<LexerPopModeAction> getInstance();
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#POP_MODE"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#popMode"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
/// Constructs the singleton instance of the lexer {@code popMode} command.
LexerPopModeAction();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,61 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Implements the {@code pushMode} lexer action by calling
/// <seealso cref="Lexer#pushMode"/> with the assigned mode.
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerPushModeAction final : public LexerAction {
public:
/// <summary>
/// Constructs a new {@code pushMode} action with the specified mode value. </summary>
/// <param name="mode"> The mode value to pass to <seealso cref="Lexer#pushMode"/>. </param>
LexerPushModeAction(int mode);
/// <summary>
/// Get the lexer mode this action should transition the lexer to.
/// </summary>
/// <returns> The lexer mode for this {@code pushMode} command. </returns>
int getMode() const;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#PUSH_MODE"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#pushMode"/> with the
/// value provided by <seealso cref="#getMode"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const int _mode;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,55 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerAction.h"
#include "atn/LexerActionType.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// Implements the {@code skip} lexer action by calling <seealso cref="Lexer#skip"/>.
///
/// <para>The {@code skip} command does not have any parameters, so this action is
/// implemented as a singleton instance exposed by <seealso cref="#INSTANCE"/>.</para>
///
/// @author Sam Harwell
/// @since 4.2
/// </summary>
class ANTLR4CPP_PUBLIC LexerSkipAction final : public LexerAction {
public:
/// Provides a singleton instance of this parameterless lexer action.
static const Ref<LexerSkipAction> getInstance();
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#SKIP"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#skip"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
/// Constructs the singleton instance of the lexer {@code skip} command.
LexerSkipAction();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,55 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/LexerActionType.h"
#include "atn/LexerAction.h"
namespace antlr4 {
namespace atn {
/// Implements the {@code type} lexer action by calling <seealso cref="Lexer#setType"/>
/// with the assigned type.
class ANTLR4CPP_PUBLIC LexerTypeAction : public LexerAction {
public:
/// <summary>
/// Constructs a new {@code type} action with the specified token type value. </summary>
/// <param name="type"> The type to assign to the token using <seealso cref="Lexer#setType"/>. </param>
LexerTypeAction(int type);
/// <summary>
/// Gets the type to assign to a token created by the lexer. </summary>
/// <returns> The type to assign to a token created by the lexer. </returns>
virtual int getType() const;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns <seealso cref="LexerActionType#TYPE"/>. </returns>
virtual LexerActionType getActionType() const override;
/// <summary>
/// {@inheritDoc} </summary>
/// <returns> This method returns {@code false}. </returns>
virtual bool isPositionDependent() const override;
/// <summary>
/// {@inheritDoc}
///
/// <para>This action is implemented by calling <seealso cref="Lexer#setType"/> with the
/// value provided by <seealso cref="#getType"/>.</para>
/// </summary>
virtual void execute(Lexer *lexer) override;
virtual size_t hashCode() const override;
virtual bool operator == (const LexerAction &obj) const override;
virtual std::string toString() const override;
private:
const int _type;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,42 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionEventInfo.h"
namespace antlr4 {
namespace atn {
/// This class represents profiling event information for tracking the lookahead
/// depth required in order to make a prediction.
class ANTLR4CPP_PUBLIC LookaheadEventInfo : public DecisionEventInfo {
public:
/// The alternative chosen by adaptivePredict(), not necessarily
/// the outermost alt shown for a rule; left-recursive rules have
/// user-level alts that differ from the rewritten rule with a (...) block
/// and a (..)* loop.
size_t predictedAlt = 0;
/// <summary>
/// Constructs a new instance of the <seealso cref="LookaheadEventInfo"/> class with
/// the specified detailed lookahead information.
/// </summary>
/// <param name="decision"> The decision number </param>
/// <param name="configs"> The final configuration set containing the necessary
/// information to determine the result of a prediction, or {@code null} if
/// the final configuration set is not available </param>
/// <param name="input"> The input token stream </param>
/// <param name="startIndex"> The start index for the current prediction </param>
/// <param name="stopIndex"> The index at which the prediction was finally made </param>
/// <param name="fullCtx"> {@code true} if the current lookahead is part of an LL
/// prediction; otherwise, {@code false} if the current lookahead is part of
/// an SLL prediction </param>
LookaheadEventInfo(size_t decision, ATNConfigSet *configs, size_t predictedAlt, TokenStream *input, size_t startIndex,
size_t stopIndex, bool fullCtx);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,22 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
/// Mark the end of a * or + loop.
class ANTLR4CPP_PUBLIC LoopEndState final : public ATNState {
public:
ATNState *loopBackState = nullptr;
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,25 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/SetTransition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC NotSetTransition final : public SetTransition {
public:
NotSetTransition(ATNState *target, const misc::IntervalSet &set);
virtual SerializationType getSerializationType() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,20 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNConfigSet.h"
#include "atn/ATNConfig.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC OrderedATNConfigSet : public ATNConfigSet {
protected:
virtual size_t getHash(ATNConfig *c) override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,102 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionInfo.h"
namespace antlr4 {
namespace atn {
class ProfilingATNSimulator;
/// This class provides access to specific and aggregate statistics gathered
/// during profiling of a parser.
class ANTLR4CPP_PUBLIC ParseInfo {
public:
ParseInfo(ProfilingATNSimulator *atnSimulator);
ParseInfo(ParseInfo const&) = default;
virtual ~ParseInfo();
ParseInfo& operator=(ParseInfo const&) = default;
/// <summary>
/// Gets an array of <seealso cref="DecisionInfo"/> instances containing the profiling
/// information gathered for each decision in the ATN.
/// </summary>
/// <returns> An array of <seealso cref="DecisionInfo"/> instances, indexed by decision
/// number. </returns>
virtual std::vector<DecisionInfo> getDecisionInfo();
/// <summary>
/// Gets the decision numbers for decisions that required one or more
/// full-context predictions during parsing. These are decisions for which
/// <seealso cref="DecisionInfo#LL_Fallback"/> is non-zero.
/// </summary>
/// <returns> A list of decision numbers which required one or more
/// full-context predictions during parsing. </returns>
virtual std::vector<size_t> getLLDecisions();
/// <summary>
/// Gets the total time spent during prediction across all decisions made
/// during parsing. This value is the sum of
/// <seealso cref="DecisionInfo#timeInPrediction"/> for all decisions.
/// </summary>
virtual long long getTotalTimeInPrediction();
/// <summary>
/// Gets the total number of SLL lookahead operations across all decisions
/// made during parsing. This value is the sum of
/// <seealso cref="DecisionInfo#SLL_TotalLook"/> for all decisions.
/// </summary>
virtual long long getTotalSLLLookaheadOps();
/// <summary>
/// Gets the total number of LL lookahead operations across all decisions
/// made during parsing. This value is the sum of
/// <seealso cref="DecisionInfo#LL_TotalLook"/> for all decisions.
/// </summary>
virtual long long getTotalLLLookaheadOps();
/// <summary>
/// Gets the total number of ATN lookahead operations for SLL prediction
/// across all decisions made during parsing.
/// </summary>
virtual long long getTotalSLLATNLookaheadOps();
/// <summary>
/// Gets the total number of ATN lookahead operations for LL prediction
/// across all decisions made during parsing.
/// </summary>
virtual long long getTotalLLATNLookaheadOps();
/// <summary>
/// Gets the total number of ATN lookahead operations for SLL and LL
/// prediction across all decisions made during parsing.
///
/// <para>
/// This value is the sum of <seealso cref="#getTotalSLLATNLookaheadOps"/> and
/// <seealso cref="#getTotalLLATNLookaheadOps"/>.</para>
/// </summary>
virtual long long getTotalATNLookaheadOps();
/// <summary>
/// Gets the total number of DFA states stored in the DFA cache for all
/// decisions in the ATN.
/// </summary>
virtual size_t getDFASize();
/// <summary>
/// Gets the total number of DFA states stored in the DFA cache for a
/// particular decision.
/// </summary>
virtual size_t getDFASize(size_t decision);
protected:
const ProfilingATNSimulator *_atnSimulator; // non-owning, we are created by this simulator.
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,904 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "PredictionMode.h"
#include "dfa/DFAState.h"
#include "atn/ATNSimulator.h"
#include "atn/PredictionContext.h"
#include "SemanticContext.h"
#include "atn/ATNConfig.h"
namespace antlr4 {
namespace atn {
/**
* The embodiment of the adaptive LL(*), ALL(*), parsing strategy.
*
* <p>
* The basic complexity of the adaptive strategy makes it harder to understand.
* We begin with ATN simulation to build paths in a DFA. Subsequent prediction
* requests go through the DFA first. If they reach a state without an edge for
* the current symbol, the algorithm fails over to the ATN simulation to
* complete the DFA path for the current input (until it finds a conflict state
* or uniquely predicting state).</p>
*
* <p>
* All of that is done without using the outer context because we want to create
* a DFA that is not dependent upon the rule invocation stack when we do a
* prediction. One DFA works in all contexts. We avoid using context not
* necessarily because it's slower, although it can be, but because of the DFA
* caching problem. The closure routine only considers the rule invocation stack
* created during prediction beginning in the decision rule. For example, if
* prediction occurs without invoking another rule's ATN, there are no context
* stacks in the configurations. When lack of context leads to a conflict, we
* don't know if it's an ambiguity or a weakness in the strong LL(*) parsing
* strategy (versus full LL(*)).</p>
*
* <p>
* When SLL yields a configuration set with conflict, we rewind the input and
* retry the ATN simulation, this time using full outer context without adding
* to the DFA. Configuration context stacks will be the full invocation stacks
* from the start rule. If we get a conflict using full context, then we can
* definitively say we have a true ambiguity for that input sequence. If we
* don't get a conflict, it implies that the decision is sensitive to the outer
* context. (It is not context-sensitive in the sense of context-sensitive
* grammars.)</p>
*
* <p>
* The next time we reach this DFA state with an SLL conflict, through DFA
* simulation, we will again retry the ATN simulation using full context mode.
* This is slow because we can't save the results and have to "interpret" the
* ATN each time we get that input.</p>
*
* <p>
* <strong>CACHING FULL CONTEXT PREDICTIONS</strong></p>
*
* <p>
* We could cache results from full context to predicted alternative easily and
* that saves a lot of time but doesn't work in presence of predicates. The set
* of visible predicates from the ATN start state changes depending on the
* context, because closure can fall off the end of a rule. I tried to cache
* tuples (stack context, semantic context, predicted alt) but it was slower
* than interpreting and much more complicated. Also required a huge amount of
* memory. The goal is not to create the world's fastest parser anyway. I'd like
* to keep this algorithm simple. By launching multiple threads, we can improve
* the speed of parsing across a large number of files.</p>
*
* <p>
* There is no strict ordering between the amount of input used by SLL vs LL,
* which makes it really hard to build a cache for full context. Let's say that
* we have input A B C that leads to an SLL conflict with full context X. That
* implies that using X we might only use A B but we could also use A B C D to
* resolve conflict. Input A B C D could predict alternative 1 in one position
* in the input and A B C E could predict alternative 2 in another position in
* input. The conflicting SLL configurations could still be non-unique in the
* full context prediction, which would lead us to requiring more input than the
* original A B C. To make a prediction cache work, we have to track the exact
* input used during the previous prediction. That amounts to a cache that maps
* X to a specific DFA for that context.</p>
*
* <p>
* Something should be done for left-recursive expression predictions. They are
* likely LL(1) + pred eval. Easier to do the whole SLL unless error and retry
* with full LL thing Sam does.</p>
*
* <p>
* <strong>AVOIDING FULL CONTEXT PREDICTION</strong></p>
*
* <p>
* We avoid doing full context retry when the outer context is empty, we did not
* dip into the outer context by falling off the end of the decision state rule,
* or when we force SLL mode.</p>
*
* <p>
* As an example of the not dip into outer context case, consider as super
* constructor calls versus function calls. One grammar might look like
* this:</p>
*
* <pre>
* ctorBody
* : '{' superCall? stat* '}'
* ;
* </pre>
*
* <p>
* Or, you might see something like</p>
*
* <pre>
* stat
* : superCall ';'
* | expression ';'
* | ...
* ;
* </pre>
*
* <p>
* In both cases I believe that no closure operations will dip into the outer
* context. In the first case ctorBody in the worst case will stop at the '}'.
* In the 2nd case it should stop at the ';'. Both cases should stay within the
* entry rule and not dip into the outer context.</p>
*
* <p>
* <strong>PREDICATES</strong></p>
*
* <p>
* Predicates are always evaluated if present in either SLL or LL both. SLL and
* LL simulation deals with predicates differently. SLL collects predicates as
* it performs closure operations like ANTLR v3 did. It delays predicate
* evaluation until it reaches and accept state. This allows us to cache the SLL
* ATN simulation whereas, if we had evaluated predicates on-the-fly during
* closure, the DFA state configuration sets would be different and we couldn't
* build up a suitable DFA.</p>
*
* <p>
* When building a DFA accept state during ATN simulation, we evaluate any
* predicates and return the sole semantically valid alternative. If there is
* more than 1 alternative, we report an ambiguity. If there are 0 alternatives,
* we throw an exception. Alternatives without predicates act like they have
* true predicates. The simple way to think about it is to strip away all
* alternatives with false predicates and choose the minimum alternative that
* remains.</p>
*
* <p>
* When we start in the DFA and reach an accept state that's predicated, we test
* those and return the minimum semantically viable alternative. If no
* alternatives are viable, we throw an exception.</p>
*
* <p>
* During full LL ATN simulation, closure always evaluates predicates and
* on-the-fly. This is crucial to reducing the configuration set size during
* closure. It hits a landmine when parsing with the Java grammar, for example,
* without this on-the-fly evaluation.</p>
*
* <p>
* <strong>SHARING DFA</strong></p>
*
* <p>
* All instances of the same parser share the same decision DFAs through a
* static field. Each instance gets its own ATN simulator but they share the
* same {@link #decisionToDFA} field. They also share a
* {@link PredictionContextCache} object that makes sure that all
* {@link PredictionContext} objects are shared among the DFA states. This makes
* a big size difference.</p>
*
* <p>
* <strong>THREAD SAFETY</strong></p>
*
* <p>
* The {@link ParserATNSimulator} locks on the {@link #decisionToDFA} field when
* it adds a new DFA object to that array. {@link #addDFAEdge}
* locks on the DFA for the current decision when setting the
* {@link DFAState#edges} field. {@link #addDFAState} locks on
* the DFA for the current decision when looking up a DFA state to see if it
* already exists. We must make sure that all requests to add DFA states that
* are equivalent result in the same shared DFA object. This is because lots of
* threads will be trying to update the DFA at once. The
* {@link #addDFAState} method also locks inside the DFA lock
* but this time on the shared context cache when it rebuilds the
* configurations' {@link PredictionContext} objects using cached
* subgraphs/nodes. No other locking occurs, even during DFA simulation. This is
* safe as long as we can guarantee that all threads referencing
* {@code s.edge[t]} get the same physical target {@link DFAState}, or
* {@code null}. Once into the DFA, the DFA simulation does not reference the
* {@link DFA#states} map. It follows the {@link DFAState#edges} field to new
* targets. The DFA simulator will either find {@link DFAState#edges} to be
* {@code null}, to be non-{@code null} and {@code dfa.edges[t]} null, or
* {@code dfa.edges[t]} to be non-null. The
* {@link #addDFAEdge} method could be racing to set the field
* but in either case the DFA simulator works; if {@code null}, and requests ATN
* simulation. It could also race trying to get {@code dfa.edges[t]}, but either
* way it will work because it's not doing a test and set operation.</p>
*
* <p>
* <strong>Starting with SLL then failing to combined SLL/LL (Two-Stage
* Parsing)</strong></p>
*
* <p>
* Sam pointed out that if SLL does not give a syntax error, then there is no
* point in doing full LL, which is slower. We only have to try LL if we get a
* syntax error. For maximum speed, Sam starts the parser set to pure SLL
* mode with the {@link BailErrorStrategy}:</p>
*
* <pre>
* parser.{@link Parser#getInterpreter() getInterpreter()}.{@link #setPredictionMode setPredictionMode}{@code (}{@link PredictionMode#SLL}{@code )};
* parser.{@link Parser#setErrorHandler setErrorHandler}(new {@link BailErrorStrategy}());
* </pre>
*
* <p>
* If it does not get a syntax error, then we're done. If it does get a syntax
* error, we need to retry with the combined SLL/LL strategy.</p>
*
* <p>
* The reason this works is as follows. If there are no SLL conflicts, then the
* grammar is SLL (at least for that input set). If there is an SLL conflict,
* the full LL analysis must yield a set of viable alternatives which is a
* subset of the alternatives reported by SLL. If the LL set is a singleton,
* then the grammar is LL but not SLL. If the LL set is the same size as the SLL
* set, the decision is SLL. If the LL set has size &gt; 1, then that decision
* is truly ambiguous on the current input. If the LL set is smaller, then the
* SLL conflict resolution might choose an alternative that the full LL would
* rule out as a possibility based upon better context information. If that's
* the case, then the SLL parse will definitely get an error because the full LL
* analysis says it's not viable. If SLL conflict resolution chooses an
* alternative within the LL set, them both SLL and LL would choose the same
* alternative because they both choose the minimum of multiple conflicting
* alternatives.</p>
*
* <p>
* Let's say we have a set of SLL conflicting alternatives {@code {1, 2, 3}} and
* a smaller LL set called <em>s</em>. If <em>s</em> is {@code {2, 3}}, then SLL
* parsing will get an error because SLL will pursue alternative 1. If
* <em>s</em> is {@code {1, 2}} or {@code {1, 3}} then both SLL and LL will
* choose the same alternative because alternative one is the minimum of either
* set. If <em>s</em> is {@code {2}} or {@code {3}} then SLL will get a syntax
* error. If <em>s</em> is {@code {1}} then SLL will succeed.</p>
*
* <p>
* Of course, if the input is invalid, then we will get an error for sure in
* both SLL and LL parsing. Erroneous input will therefore require 2 passes over
* the input.</p>
*/
class ANTLR4CPP_PUBLIC ParserATNSimulator : public ATNSimulator {
public:
/// Testing only!
ParserATNSimulator(const ATN &atn, std::vector<dfa::DFA> &decisionToDFA,
PredictionContextCache &sharedContextCache);
ParserATNSimulator(Parser *parser, const ATN &atn, std::vector<dfa::DFA> &decisionToDFA,
PredictionContextCache &sharedContextCache);
virtual void reset() override;
virtual void clearDFA() override;
virtual size_t adaptivePredict(TokenStream *input, size_t decision, ParserRuleContext *outerContext);
static const bool TURN_OFF_LR_LOOP_ENTRY_BRANCH_OPT;
std::vector<dfa::DFA> &decisionToDFA;
/** Implements first-edge (loop entry) elimination as an optimization
* during closure operations. See antlr/antlr4#1398.
*
* The optimization is to avoid adding the loop entry config when
* the exit path can only lead back to the same
* StarLoopEntryState after popping context at the rule end state
* (traversing only epsilon edges, so we're still in closure, in
* this same rule).
*
* We need to detect any state that can reach loop entry on
* epsilon w/o exiting rule. We don't have to look at FOLLOW
* links, just ensure that all stack tops for config refer to key
* states in LR rule.
*
* To verify we are in the right situation we must first check
* closure is at a StarLoopEntryState generated during LR removal.
* Then we check that each stack top of context is a return state
* from one of these cases:
*
* 1. 'not' expr, '(' type ')' expr. The return state points at loop entry state
* 2. expr op expr. The return state is the block end of internal block of (...)*
* 3. 'between' expr 'and' expr. The return state of 2nd expr reference.
* That state points at block end of internal block of (...)*.
* 4. expr '?' expr ':' expr. The return state points at block end,
* which points at loop entry state.
*
* If any is true for each stack top, then closure does not add a
* config to the current config set for edge[0], the loop entry branch.
*
* Conditions fail if any context for the current config is:
*
* a. empty (we'd fall out of expr to do a global FOLLOW which could
* even be to some weird spot in expr) or,
* b. lies outside of expr or,
* c. lies within expr but at a state not the BlockEndState
* generated during LR removal
*
* Do we need to evaluate predicates ever in closure for this case?
*
* No. Predicates, including precedence predicates, are only
* evaluated when computing a DFA start state. I.e., only before
* the lookahead (but not parser) consumes a token.
*
* There are no epsilon edges allowed in LR rule alt blocks or in
* the "primary" part (ID here). If closure is in
* StarLoopEntryState any lookahead operation will have consumed a
* token as there are no epsilon-paths that lead to
* StarLoopEntryState. We do not have to evaluate predicates
* therefore if we are in the generated StarLoopEntryState of a LR
* rule. Note that when making a prediction starting at that
* decision point, decision d=2, compute-start-state performs
* closure starting at edges[0], edges[1] emanating from
* StarLoopEntryState. That means it is not performing closure on
* StarLoopEntryState during compute-start-state.
*
* How do we know this always gives same prediction answer?
*
* Without predicates, loop entry and exit paths are ambiguous
* upon remaining input +b (in, say, a+b). Either paths lead to
* valid parses. Closure can lead to consuming + immediately or by
* falling out of this call to expr back into expr and loop back
* again to StarLoopEntryState to match +b. In this special case,
* we choose the more efficient path, which is to take the bypass
* path.
*
* The lookahead language has not changed because closure chooses
* one path over the other. Both paths lead to consuming the same
* remaining input during a lookahead operation. If the next token
* is an operator, lookahead will enter the choice block with
* operators. If it is not, lookahead will exit expr. Same as if
* closure had chosen to enter the choice block immediately.
*
* Closure is examining one config (some loopentrystate, some alt,
* context) which means it is considering exactly one alt. Closure
* always copies the same alt to any derived configs.
*
* How do we know this optimization doesn't mess up precedence in
* our parse trees?
*
* Looking through expr from left edge of stat only has to confirm
* that an input, say, a+b+c; begins with any valid interpretation
* of an expression. The precedence actually doesn't matter when
* making a decision in stat seeing through expr. It is only when
* parsing rule expr that we must use the precedence to get the
* right interpretation and, hence, parse tree.
*/
bool canDropLoopEntryEdgeInLeftRecursiveRule(ATNConfig *config) const;
virtual std::string getRuleName(size_t index);
virtual Ref<ATNConfig> precedenceTransition(Ref<ATNConfig> const& config, PrecedencePredicateTransition *pt,
bool collectPredicates, bool inContext, bool fullCtx);
void setPredictionMode(PredictionMode newMode);
PredictionMode getPredictionMode();
Parser* getParser();
virtual std::string getTokenName(size_t t);
virtual std::string getLookaheadName(TokenStream *input);
/// <summary>
/// Used for debugging in adaptivePredict around execATN but I cut
/// it out for clarity now that alg. works well. We can leave this
/// "dead" code for a bit.
/// </summary>
virtual void dumpDeadEndConfigs(NoViableAltException &nvae);
protected:
Parser *const parser;
/// <summary>
/// Each prediction operation uses a cache for merge of prediction contexts.
/// Don't keep around as it wastes huge amounts of memory. The merge cache
/// isn't synchronized but we're ok since two threads shouldn't reuse same
/// parser/atnsim object because it can only handle one input at a time.
/// This maps graphs a and b to merged result c. (a,b)->c. We can avoid
/// the merge if we ever see a and b again. Note that (b,a)->c should
/// also be examined during cache lookup.
/// </summary>
PredictionContextMergeCache mergeCache;
// LAME globals to avoid parameters!!!!! I need these down deep in predTransition
TokenStream *_input;
size_t _startIndex;
ParserRuleContext *_outerContext;
dfa::DFA *_dfa; // Reference into the decisionToDFA vector.
/// <summary>
/// Performs ATN simulation to compute a predicted alternative based
/// upon the remaining input, but also updates the DFA cache to avoid
/// having to traverse the ATN again for the same input sequence.
///
/// There are some key conditions we're looking for after computing a new
/// set of ATN configs (proposed DFA state):
/// if the set is empty, there is no viable alternative for current symbol
/// does the state uniquely predict an alternative?
/// does the state have a conflict that would prevent us from
/// putting it on the work list?
///
/// We also have some key operations to do:
/// add an edge from previous DFA state to potentially new DFA state, D,
/// upon current symbol but only if adding to work list, which means in all
/// cases except no viable alternative (and possibly non-greedy decisions?)
/// collecting predicates and adding semantic context to DFA accept states
/// adding rule context to context-sensitive DFA accept states
/// consuming an input symbol
/// reporting a conflict
/// reporting an ambiguity
/// reporting a context sensitivity
/// reporting insufficient predicates
///
/// cover these cases:
/// dead end
/// single alt
/// single alt + preds
/// conflict
/// conflict + preds
/// </summary>
virtual size_t execATN(dfa::DFA &dfa, dfa::DFAState *s0, TokenStream *input, size_t startIndex,
ParserRuleContext *outerContext);
/// <summary>
/// Get an existing target state for an edge in the DFA. If the target state
/// for the edge has not yet been computed or is otherwise not available,
/// this method returns {@code null}.
/// </summary>
/// <param name="previousD"> The current DFA state </param>
/// <param name="t"> The next input symbol </param>
/// <returns> The existing target DFA state for the given input symbol
/// {@code t}, or {@code null} if the target state for this edge is not
/// already cached </returns>
virtual dfa::DFAState* getExistingTargetState(dfa::DFAState *previousD, size_t t);
/// <summary>
/// Compute a target state for an edge in the DFA, and attempt to add the
/// computed state and corresponding edge to the DFA.
/// </summary>
/// <param name="dfa"> The DFA </param>
/// <param name="previousD"> The current DFA state </param>
/// <param name="t"> The next input symbol
/// </param>
/// <returns> The computed target DFA state for the given input symbol
/// {@code t}. If {@code t} does not lead to a valid DFA state, this method
/// returns <seealso cref="#ERROR"/>. </returns>
virtual dfa::DFAState *computeTargetState(dfa::DFA &dfa, dfa::DFAState *previousD, size_t t);
virtual void predicateDFAState(dfa::DFAState *dfaState, DecisionState *decisionState);
// comes back with reach.uniqueAlt set to a valid alt
virtual size_t execATNWithFullContext(dfa::DFA &dfa, dfa::DFAState *D, ATNConfigSet *s0,
TokenStream *input, size_t startIndex, ParserRuleContext *outerContext); // how far we got before failing over
virtual std::unique_ptr<ATNConfigSet> computeReachSet(ATNConfigSet *closure, size_t t, bool fullCtx);
/// <summary>
/// Return a configuration set containing only the configurations from
/// {@code configs} which are in a <seealso cref="RuleStopState"/>. If all
/// configurations in {@code configs} are already in a rule stop state, this
/// method simply returns {@code configs}.
/// <p/>
/// When {@code lookToEndOfRule} is true, this method uses
/// <seealso cref="ATN#nextTokens"/> for each configuration in {@code configs} which is
/// not already in a rule stop state to see if a rule stop state is reachable
/// from the configuration via epsilon-only transitions.
/// </summary>
/// <param name="configs"> the configuration set to update </param>
/// <param name="lookToEndOfRule"> when true, this method checks for rule stop states
/// reachable by epsilon-only transitions from each configuration in
/// {@code configs}.
/// </param>
/// <returns> {@code configs} if all configurations in {@code configs} are in a
/// rule stop state, otherwise return a new configuration set containing only
/// the configurations from {@code configs} which are in a rule stop state </returns>
virtual ATNConfigSet* removeAllConfigsNotInRuleStopState(ATNConfigSet *configs, bool lookToEndOfRule);
virtual std::unique_ptr<ATNConfigSet> computeStartState(ATNState *p, RuleContext *ctx, bool fullCtx);
/* parrt internal source braindump that doesn't mess up
* external API spec.
applyPrecedenceFilter is an optimization to avoid highly
nonlinear prediction of expressions and other left recursive
rules. The precedence predicates such as {3>=prec}? Are highly
context-sensitive in that they can only be properly evaluated
in the context of the proper prec argument. Without pruning,
these predicates are normal predicates evaluated when we reach
conflict state (or unique prediction). As we cannot evaluate
these predicates out of context, the resulting conflict leads
to full LL evaluation and nonlinear prediction which shows up
very clearly with fairly large expressions.
Example grammar:
e : e '*' e
| e '+' e
| INT
;
We convert that to the following:
e[int prec]
: INT
( {3>=prec}? '*' e[4]
| {2>=prec}? '+' e[3]
)*
;
The (..)* loop has a decision for the inner block as well as
an enter or exit decision, which is what concerns us here. At
the 1st + of input 1+2+3, the loop entry sees both predicates
and the loop exit also sees both predicates by falling off the
edge of e. This is because we have no stack information with
SLL and find the follow of e, which will hit the return states
inside the loop after e[4] and e[3], which brings it back to
the enter or exit decision. In this case, we know that we
cannot evaluate those predicates because we have fallen off
the edge of the stack and will in general not know which prec
parameter is the right one to use in the predicate.
Because we have special information, that these are precedence
predicates, we can resolve them without failing over to full
LL despite their context sensitive nature. We make an
assumption that prec[-1] <= prec[0], meaning that the current
precedence level is greater than or equal to the precedence
level of recursive invocations above us in the stack. For
example, if predicate {3>=prec}? is true of the current prec,
then one option is to enter the loop to match it now. The
other option is to exit the loop and the left recursive rule
to match the current operator in rule invocation further up
the stack. But, we know that all of those prec are lower or
the same value and so we can decide to enter the loop instead
of matching it later. That means we can strip out the other
configuration for the exit branch.
So imagine we have (14,1,$,{2>=prec}?) and then
(14,2,$-dipsIntoOuterContext,{2>=prec}?). The optimization
allows us to collapse these two configurations. We know that
if {2>=prec}? is true for the current prec parameter, it will
also be true for any prec from an invoking e call, indicated
by dipsIntoOuterContext. As the predicates are both true, we
have the option to evaluate them early in the decision start
state. We do this by stripping both predicates and choosing to
enter the loop as it is consistent with the notion of operator
precedence. It's also how the full LL conflict resolution
would work.
The solution requires a different DFA start state for each
precedence level.
The basic filter mechanism is to remove configurations of the
form (p, 2, pi) if (p, 1, pi) exists for the same p and pi. In
other words, for the same ATN state and predicate context,
remove any configuration associated with an exit branch if
there is a configuration associated with the enter branch.
It's also the case that the filter evaluates precedence
predicates and resolves conflicts according to precedence
levels. For example, for input 1+2+3 at the first +, we see
prediction filtering
[(11,1,[$],{3>=prec}?), (14,1,[$],{2>=prec}?), (5,2,[$],up=1),
(11,2,[$],up=1), (14,2,[$],up=1)],hasSemanticContext=true,dipsIntoOuterContext
to
[(11,1,[$]), (14,1,[$]), (5,2,[$],up=1)],dipsIntoOuterContext
This filters because {3>=prec}? evals to true and collapses
(11,1,[$],{3>=prec}?) and (11,2,[$],up=1) since early conflict
resolution based upon rules of operator precedence fits with
our usual match first alt upon conflict.
We noticed a problem where a recursive call resets precedence
to 0. Sam's fix: each config has flag indicating if it has
returned from an expr[0] call. then just don't filter any
config with that flag set. flag is carried along in
closure(). so to avoid adding field, set bit just under sign
bit of dipsIntoOuterContext (SUPPRESS_PRECEDENCE_FILTER).
With the change you filter "unless (p, 2, pi) was reached
after leaving the rule stop state of the LR rule containing
state p, corresponding to a rule invocation with precedence
level 0"
*/
/**
* This method transforms the start state computed by
* {@link #computeStartState} to the special start state used by a
* precedence DFA for a particular precedence value. The transformation
* process applies the following changes to the start state's configuration
* set.
*
* <ol>
* <li>Evaluate the precedence predicates for each configuration using
* {@link SemanticContext#evalPrecedence}.</li>
* <li>When {@link ATNConfig#isPrecedenceFilterSuppressed} is {@code false},
* remove all configurations which predict an alternative greater than 1,
* for which another configuration that predicts alternative 1 is in the
* same ATN state with the same prediction context. This transformation is
* valid for the following reasons:
* <ul>
* <li>The closure block cannot contain any epsilon transitions which bypass
* the body of the closure, so all states reachable via alternative 1 are
* part of the precedence alternatives of the transformed left-recursive
* rule.</li>
* <li>The "primary" portion of a left recursive rule cannot contain an
* epsilon transition, so the only way an alternative other than 1 can exist
* in a state that is also reachable via alternative 1 is by nesting calls
* to the left-recursive rule, with the outer calls not being at the
* preferred precedence level. The
* {@link ATNConfig#isPrecedenceFilterSuppressed} property marks ATN
* configurations which do not meet this condition, and therefore are not
* eligible for elimination during the filtering process.</li>
* </ul>
* </li>
* </ol>
*
* <p>
* The prediction context must be considered by this filter to address
* situations like the following.
* </p>
* <code>
* <pre>
* grammar TA;
* prog: statement* EOF;
* statement: letterA | statement letterA 'b' ;
* letterA: 'a';
* </pre>
* </code>
* <p>
* If the above grammar, the ATN state immediately before the token
* reference {@code 'a'} in {@code letterA} is reachable from the left edge
* of both the primary and closure blocks of the left-recursive rule
* {@code statement}. The prediction context associated with each of these
* configurations distinguishes between them, and prevents the alternative
* which stepped out to {@code prog} (and then back in to {@code statement}
* from being eliminated by the filter.
* </p>
*
* @param configs The configuration set computed by
* {@link #computeStartState} as the start state for the DFA.
* @return The transformed configuration set representing the start state
* for a precedence DFA at a particular precedence level (determined by
* calling {@link Parser#getPrecedence}).
*/
std::unique_ptr<ATNConfigSet> applyPrecedenceFilter(ATNConfigSet *configs);
virtual ATNState *getReachableTarget(Transition *trans, size_t ttype);
virtual std::vector<Ref<SemanticContext>> getPredsForAmbigAlts(const antlrcpp::BitSet &ambigAlts,
ATNConfigSet *configs, size_t nalts);
virtual std::vector<dfa::DFAState::PredPrediction*> getPredicatePredictions(const antlrcpp::BitSet &ambigAlts,
std::vector<Ref<SemanticContext>> const& altToPred);
/**
* This method is used to improve the localization of error messages by
* choosing an alternative rather than throwing a
* {@link NoViableAltException} in particular prediction scenarios where the
* {@link #ERROR} state was reached during ATN simulation.
*
* <p>
* The default implementation of this method uses the following
* algorithm to identify an ATN configuration which successfully parsed the
* decision entry rule. Choosing such an alternative ensures that the
* {@link ParserRuleContext} returned by the calling rule will be complete
* and valid, and the syntax error will be reported later at a more
* localized location.</p>
*
* <ul>
* <li>If a syntactically valid path or paths reach the end of the decision rule and
* they are semantically valid if predicated, return the min associated alt.</li>
* <li>Else, if a semantically invalid but syntactically valid path exist
* or paths exist, return the minimum associated alt.
* </li>
* <li>Otherwise, return {@link ATN#INVALID_ALT_NUMBER}.</li>
* </ul>
*
* <p>
* In some scenarios, the algorithm described above could predict an
* alternative which will result in a {@link FailedPredicateException} in
* the parser. Specifically, this could occur if the <em>only</em> configuration
* capable of successfully parsing to the end of the decision rule is
* blocked by a semantic predicate. By choosing this alternative within
* {@link #adaptivePredict} instead of throwing a
* {@link NoViableAltException}, the resulting
* {@link FailedPredicateException} in the parser will identify the specific
* predicate which is preventing the parser from successfully parsing the
* decision rule, which helps developers identify and correct logic errors
* in semantic predicates.
* </p>
*
* @param configs The ATN configurations which were valid immediately before
* the {@link #ERROR} state was reached
* @param outerContext The is the \gamma_0 initial parser context from the paper
* or the parser stack at the instant before prediction commences.
*
* @return The value to return from {@link #adaptivePredict}, or
* {@link ATN#INVALID_ALT_NUMBER} if a suitable alternative was not
* identified and {@link #adaptivePredict} should report an error instead.
*/
size_t getSynValidOrSemInvalidAltThatFinishedDecisionEntryRule(ATNConfigSet *configs,
ParserRuleContext *outerContext);
virtual size_t getAltThatFinishedDecisionEntryRule(ATNConfigSet *configs);
/** Walk the list of configurations and split them according to
* those that have preds evaluating to true/false. If no pred, assume
* true pred and include in succeeded set. Returns Pair of sets.
*
* Create a new set so as not to alter the incoming parameter.
*
* Assumption: the input stream has been restored to the starting point
* prediction, which is where predicates need to evaluate.
*/
std::pair<ATNConfigSet *, ATNConfigSet *> splitAccordingToSemanticValidity(ATNConfigSet *configs,
ParserRuleContext *outerContext);
/// <summary>
/// Look through a list of predicate/alt pairs, returning alts for the
/// pairs that win. A {@code NONE} predicate indicates an alt containing an
/// unpredicated config which behaves as "always true." If !complete
/// then we stop at the first predicate that evaluates to true. This
/// includes pairs with null predicates.
/// </summary>
virtual antlrcpp::BitSet evalSemanticContext(std::vector<dfa::DFAState::PredPrediction*> predPredictions,
ParserRuleContext *outerContext, bool complete);
/**
* Evaluate a semantic context within a specific parser context.
*
* <p>
* This method might not be called for every semantic context evaluated
* during the prediction process. In particular, we currently do not
* evaluate the following but it may change in the future:</p>
*
* <ul>
* <li>Precedence predicates (represented by
* {@link SemanticContext.PrecedencePredicate}) are not currently evaluated
* through this method.</li>
* <li>Operator predicates (represented by {@link SemanticContext.AND} and
* {@link SemanticContext.OR}) are evaluated as a single semantic
* context, rather than evaluating the operands individually.
* Implementations which require evaluation results from individual
* predicates should override this method to explicitly handle evaluation of
* the operands within operator predicates.</li>
* </ul>
*
* @param pred The semantic context to evaluate
* @param parserCallStack The parser context in which to evaluate the
* semantic context
* @param alt The alternative which is guarded by {@code pred}
* @param fullCtx {@code true} if the evaluation is occurring during LL
* prediction; otherwise, {@code false} if the evaluation is occurring
* during SLL prediction
*
* @since 4.3
*/
virtual bool evalSemanticContext(Ref<SemanticContext> const& pred, ParserRuleContext *parserCallStack,
size_t alt, bool fullCtx);
/* TODO: If we are doing predicates, there is no point in pursuing
closure operations if we reach a DFA state that uniquely predicts
alternative. We will not be caching that DFA state and it is a
waste to pursue the closure. Might have to advance when we do
ambig detection thought :(
*/
virtual void closure(Ref<ATNConfig> const& config, ATNConfigSet *configs, ATNConfig::Set &closureBusy,
bool collectPredicates, bool fullCtx, bool treatEofAsEpsilon);
virtual void closureCheckingStopState(Ref<ATNConfig> const& config, ATNConfigSet *configs, ATNConfig::Set &closureBusy,
bool collectPredicates, bool fullCtx, int depth, bool treatEofAsEpsilon);
/// Do the actual work of walking epsilon edges.
virtual void closure_(Ref<ATNConfig> const& config, ATNConfigSet *configs, ATNConfig::Set &closureBusy,
bool collectPredicates, bool fullCtx, int depth, bool treatEofAsEpsilon);
virtual Ref<ATNConfig> getEpsilonTarget(Ref<ATNConfig> const& config, Transition *t, bool collectPredicates,
bool inContext, bool fullCtx, bool treatEofAsEpsilon);
virtual Ref<ATNConfig> actionTransition(Ref<ATNConfig> const& config, ActionTransition *t);
virtual Ref<ATNConfig> predTransition(Ref<ATNConfig> const& config, PredicateTransition *pt, bool collectPredicates,
bool inContext, bool fullCtx);
virtual Ref<ATNConfig> ruleTransition(Ref<ATNConfig> const& config, RuleTransition *t);
/**
* Gets a {@link BitSet} containing the alternatives in {@code configs}
* which are part of one or more conflicting alternative subsets.
*
* @param configs The {@link ATNConfigSet} to analyze.
* @return The alternatives in {@code configs} which are part of one or more
* conflicting alternative subsets. If {@code configs} does not contain any
* conflicting subsets, this method returns an empty {@link BitSet}.
*/
virtual antlrcpp::BitSet getConflictingAlts(ATNConfigSet *configs);
/// <summary>
/// Sam pointed out a problem with the previous definition, v3, of
/// ambiguous states. If we have another state associated with conflicting
/// alternatives, we should keep going. For example, the following grammar
///
/// s : (ID | ID ID?) ';' ;
///
/// When the ATN simulation reaches the state before ';', it has a DFA
/// state that looks like: [12|1|[], 6|2|[], 12|2|[]]. Naturally
/// 12|1|[] and 12|2|[] conflict, but we cannot stop processing this node
/// because alternative to has another way to continue, via [6|2|[]].
/// The key is that we have a single state that has config's only associated
/// with a single alternative, 2, and crucially the state transitions
/// among the configurations are all non-epsilon transitions. That means
/// we don't consider any conflicts that include alternative 2. So, we
/// ignore the conflict between alts 1 and 2. We ignore a set of
/// conflicting alts when there is an intersection with an alternative
/// associated with a single alt state in the state->config-list map.
///
/// It's also the case that we might have two conflicting configurations but
/// also a 3rd nonconflicting configuration for a different alternative:
/// [1|1|[], 1|2|[], 8|3|[]]. This can come about from grammar:
///
/// a : A | A | A B ;
///
/// After matching input A, we reach the stop state for rule A, state 1.
/// State 8 is the state right before B. Clearly alternatives 1 and 2
/// conflict and no amount of further lookahead will separate the two.
/// However, alternative 3 will be able to continue and so we do not
/// stop working on this state. In the previous example, we're concerned
/// with states associated with the conflicting alternatives. Here alt
/// 3 is not associated with the conflicting configs, but since we can continue
/// looking for input reasonably, I don't declare the state done. We
/// ignore a set of conflicting alts when we have an alternative
/// that we still need to pursue.
/// </summary>
virtual antlrcpp::BitSet getConflictingAltsOrUniqueAlt(ATNConfigSet *configs);
virtual NoViableAltException noViableAlt(TokenStream *input, ParserRuleContext *outerContext,
ATNConfigSet *configs, size_t startIndex, bool deleteConfigs);
static size_t getUniqueAlt(ATNConfigSet *configs);
/// <summary>
/// Add an edge to the DFA, if possible. This method calls
/// <seealso cref="#addDFAState"/> to ensure the {@code to} state is present in the
/// DFA. If {@code from} is {@code null}, or if {@code t} is outside the
/// range of edges that can be represented in the DFA tables, this method
/// returns without adding the edge to the DFA.
/// <p/>
/// If {@code to} is {@code null}, this method returns {@code null}.
/// Otherwise, this method returns the <seealso cref="DFAState"/> returned by calling
/// <seealso cref="#addDFAState"/> for the {@code to} state.
/// </summary>
/// <param name="dfa"> The DFA </param>
/// <param name="from"> The source state for the edge </param>
/// <param name="t"> The input symbol </param>
/// <param name="to"> The target state for the edge
/// </param>
/// <returns> If {@code to} is {@code null}, this method returns {@code null};
/// otherwise this method returns the result of calling <seealso cref="#addDFAState"/>
/// on {@code to} </returns>
virtual dfa::DFAState *addDFAEdge(dfa::DFA &dfa, dfa::DFAState *from, ssize_t t, dfa::DFAState *to);
/// <summary>
/// Add state {@code D} to the DFA if it is not already present, and return
/// the actual instance stored in the DFA. If a state equivalent to {@code D}
/// is already in the DFA, the existing state is returned. Otherwise this
/// method returns {@code D} after adding it to the DFA.
/// <p/>
/// If {@code D} is <seealso cref="#ERROR"/>, this method returns <seealso cref="#ERROR"/> and
/// does not change the DFA.
/// </summary>
/// <param name="dfa"> The dfa </param>
/// <param name="D"> The DFA state to add </param>
/// <returns> The state stored in the DFA. This will be either the existing
/// state if {@code D} is already in the DFA, or {@code D} itself if the
/// state was not already present. </returns>
virtual dfa::DFAState *addDFAState(dfa::DFA &dfa, dfa::DFAState *D);
virtual void reportAttemptingFullContext(dfa::DFA &dfa, const antlrcpp::BitSet &conflictingAlts,
ATNConfigSet *configs, size_t startIndex, size_t stopIndex);
virtual void reportContextSensitivity(dfa::DFA &dfa, size_t prediction, ATNConfigSet *configs,
size_t startIndex, size_t stopIndex);
/// If context sensitive parsing, we know it's ambiguity not conflict.
virtual void reportAmbiguity(dfa::DFA &dfa,
dfa::DFAState *D, // the DFA state from execATN() that had SLL conflicts
size_t startIndex, size_t stopIndex,
bool exact,
const antlrcpp::BitSet &ambigAlts,
ATNConfigSet *configs); // configs that LL not SLL considered conflicting
private:
// SLL, LL, or LL + exact ambig detection?
PredictionMode _mode;
static bool getLrLoopSetting();
void InitializeInstanceFields();
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,25 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/BlockStartState.h"
namespace antlr4 {
namespace atn {
/// Start of {@code (A|B|...)+} loop. Technically a decision state, but
/// we don't use for code generation; somebody might need it, so I'm defining
/// it for completeness. In reality, the <seealso cref="PlusLoopbackState"/> node is the
/// real decision-making note for {@code A+}.
class ANTLR4CPP_PUBLIC PlusBlockStartState final : public BlockStartState {
public:
PlusLoopbackState *loopBackState = nullptr;
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,22 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionState.h"
namespace antlr4 {
namespace atn {
/// Decision state for {@code A+} and {@code (A|B)+}. It has two transitions:
/// one to the loop back to start of the block and one to exit.
class ANTLR4CPP_PUBLIC PlusLoopbackState final : public DecisionState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,29 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/AbstractPredicateTransition.h"
#include "SemanticContext.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC PrecedencePredicateTransition final : public AbstractPredicateTransition {
public:
const int precedence;
PrecedencePredicateTransition(ATNState *target, int precedence);
virtual SerializationType getSerializationType() const override;
virtual bool isEpsilon() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
Ref<SemanticContext::PrecedencePredicate> getPredicate() const;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,62 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionEventInfo.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// This class represents profiling event information for semantic predicate
/// evaluations which occur during prediction.
/// </summary>
/// <seealso cref= ParserATNSimulator#evalSemanticContext
///
/// @since 4.3 </seealso>
class ANTLR4CPP_PUBLIC PredicateEvalInfo : public DecisionEventInfo {
public:
/// The semantic context which was evaluated.
const Ref<SemanticContext> semctx;
/// <summary>
/// The alternative number for the decision which is guarded by the semantic
/// context <seealso cref="#semctx"/>. Note that other ATN
/// configurations may predict the same alternative which are guarded by
/// other semantic contexts and/or <seealso cref="SemanticContext#NONE"/>.
/// </summary>
const size_t predictedAlt;
/// The result of evaluating the semantic context <seealso cref="#semctx"/>.
const bool evalResult;
/// <summary>
/// Constructs a new instance of the <seealso cref="PredicateEvalInfo"/> class with the
/// specified detailed predicate evaluation information.
/// </summary>
/// <param name="decision"> The decision number </param>
/// <param name="input"> The input token stream </param>
/// <param name="startIndex"> The start index for the current prediction </param>
/// <param name="stopIndex"> The index at which the predicate evaluation was
/// triggered. Note that the input stream may be reset to other positions for
/// the actual evaluation of individual predicates. </param>
/// <param name="semctx"> The semantic context which was evaluated </param>
/// <param name="evalResult"> The results of evaluating the semantic context </param>
/// <param name="predictedAlt"> The alternative number for the decision which is
/// guarded by the semantic context {@code semctx}. See <seealso cref="#predictedAlt"/>
/// for more information. </param>
/// <param name="fullCtx"> {@code true} if the semantic context was
/// evaluated during LL prediction; otherwise, {@code false} if the semantic
/// context was evaluated during SLL prediction
/// </param>
/// <seealso cref= ParserATNSimulator#evalSemanticContext(SemanticContext, ParserRuleContext, int, boolean) </seealso>
/// <seealso cref= SemanticContext#eval(Recognizer, RuleContext) </seealso>
PredicateEvalInfo(size_t decision, TokenStream *input, size_t startIndex, size_t stopIndex,
Ref<SemanticContext> const& semctx, bool evalResult, size_t predictedAlt, bool fullCtx);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,39 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/AbstractPredicateTransition.h"
#include "SemanticContext.h"
namespace antlr4 {
namespace atn {
/// TODO: this is old comment:
/// A tree of semantic predicates from the grammar AST if label==SEMPRED.
/// In the ATN, labels will always be exactly one predicate, but the DFA
/// may have to combine a bunch of them as it collects predicates from
/// multiple ATN configurations into a single DFA state.
class ANTLR4CPP_PUBLIC PredicateTransition final : public AbstractPredicateTransition {
public:
const size_t ruleIndex;
const size_t predIndex;
const bool isCtxDependent; // e.g., $i ref in pred
PredicateTransition(ATNState *target, size_t ruleIndex, size_t predIndex, bool isCtxDependent);
virtual SerializationType getSerializationType() const override;
virtual bool isEpsilon() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
Ref<SemanticContext::Predicate> getPredicate() const;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,254 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "Recognizer.h"
#include "atn/ATN.h"
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
struct PredictionContextHasher;
struct PredictionContextComparer;
class PredictionContextMergeCache;
typedef std::unordered_set<Ref<PredictionContext>, PredictionContextHasher, PredictionContextComparer> PredictionContextCache;
class ANTLR4CPP_PUBLIC PredictionContext {
public:
/// Represents $ in local context prediction, which means wildcard.
/// *+x = *.
static const Ref<PredictionContext> EMPTY;
/// Represents $ in an array in full context mode, when $
/// doesn't mean wildcard: $ + x = [$,x]. Here,
/// $ = EMPTY_RETURN_STATE.
// ml: originally Integer.MAX_VALUE, which would be -1 for us, but this is already used in places where
// -1 is converted to unsigned, so we use a different value here. Any value does the job provided it doesn't
// conflict with real return states.
static const size_t EMPTY_RETURN_STATE = static_cast<size_t>(-10); // std::numeric_limits<size_t>::max() - 9;
private:
static const size_t INITIAL_HASH = 1;
public:
static size_t globalNodeCount;
const size_t id;
/// <summary>
/// Stores the computed hash code of this <seealso cref="PredictionContext"/>. The hash
/// code is computed in parts to match the following reference algorithm.
///
/// <pre>
/// private int referenceHashCode() {
/// int hash = <seealso cref="MurmurHash#initialize"/>(<seealso cref="#INITIAL_HASH"/>);
///
/// for (int i = 0; i < <seealso cref="#size()"/>; i++) {
/// hash = <seealso cref="MurmurHash#update"/>(hash, <seealso cref="#getParent"/>(i));
/// }
///
/// for (int i = 0; i < <seealso cref="#size()"/>; i++) {
/// hash = <seealso cref="MurmurHash#update"/>(hash, <seealso cref="#getReturnState"/>(i));
/// }
///
/// hash = <seealso cref="MurmurHash#finish"/>(hash, 2 * <seealso cref="#size()"/>);
/// return hash;
/// }
/// </pre>
/// </summary>
const size_t cachedHashCode;
protected:
PredictionContext(size_t cachedHashCode);
~PredictionContext();
public:
/// Convert a RuleContext tree to a PredictionContext graph.
/// Return EMPTY if outerContext is empty.
static Ref<PredictionContext> fromRuleContext(const ATN &atn, RuleContext *outerContext);
virtual size_t size() const = 0;
virtual Ref<PredictionContext> getParent(size_t index) const = 0;
virtual size_t getReturnState(size_t index) const = 0;
virtual bool operator == (const PredictionContext &o) const = 0;
/// This means only the EMPTY (wildcard? not sure) context is in set.
virtual bool isEmpty() const;
virtual bool hasEmptyPath() const;
virtual size_t hashCode() const;
protected:
static size_t calculateEmptyHashCode();
static size_t calculateHashCode(Ref<PredictionContext> parent, size_t returnState);
static size_t calculateHashCode(const std::vector<Ref<PredictionContext>> &parents,
const std::vector<size_t> &returnStates);
public:
// dispatch
static Ref<PredictionContext> merge(const Ref<PredictionContext> &a, const Ref<PredictionContext> &b,
bool rootIsWildcard, PredictionContextMergeCache *mergeCache);
/// <summary>
/// Merge two <seealso cref="SingletonPredictionContext"/> instances.
///
/// <p/>
///
/// Stack tops equal, parents merge is same; return left graph.<br/>
/// <embed src="images/SingletonMerge_SameRootSamePar.svg" type="image/svg+xml"/>
///
/// <p/>
///
/// Same stack top, parents differ; merge parents giving array node, then
/// remainders of those graphs. A new root node is created to point to the
/// merged parents.<br/>
/// <embed src="images/SingletonMerge_SameRootDiffPar.svg" type="image/svg+xml"/>
///
/// <p/>
///
/// Different stack tops pointing to same parent. Make array node for the
/// root where both element in the root point to the same (original)
/// parent.<br/>
/// <embed src="images/SingletonMerge_DiffRootSamePar.svg" type="image/svg+xml"/>
///
/// <p/>
///
/// Different stack tops pointing to different parents. Make array node for
/// the root where each element points to the corresponding original
/// parent.<br/>
/// <embed src="images/SingletonMerge_DiffRootDiffPar.svg" type="image/svg+xml"/>
/// </summary>
/// <param name="a"> the first <seealso cref="SingletonPredictionContext"/> </param>
/// <param name="b"> the second <seealso cref="SingletonPredictionContext"/> </param>
/// <param name="rootIsWildcard"> {@code true} if this is a local-context merge,
/// otherwise false to indicate a full-context merge </param>
/// <param name="mergeCache"> </param>
static Ref<PredictionContext> mergeSingletons(const Ref<SingletonPredictionContext> &a,
const Ref<SingletonPredictionContext> &b, bool rootIsWildcard, PredictionContextMergeCache *mergeCache);
/**
* Handle case where at least one of {@code a} or {@code b} is
* {@link #EMPTY}. In the following diagrams, the symbol {@code $} is used
* to represent {@link #EMPTY}.
*
* <h2>Local-Context Merges</h2>
*
* <p>These local-context merge operations are used when {@code rootIsWildcard}
* is true.</p>
*
* <p>{@link #EMPTY} is superset of any graph; return {@link #EMPTY}.<br>
* <embed src="images/LocalMerge_EmptyRoot.svg" type="image/svg+xml"/></p>
*
* <p>{@link #EMPTY} and anything is {@code #EMPTY}, so merged parent is
* {@code #EMPTY}; return left graph.<br>
* <embed src="images/LocalMerge_EmptyParent.svg" type="image/svg+xml"/></p>
*
* <p>Special case of last merge if local context.<br>
* <embed src="images/LocalMerge_DiffRoots.svg" type="image/svg+xml"/></p>
*
* <h2>Full-Context Merges</h2>
*
* <p>These full-context merge operations are used when {@code rootIsWildcard}
* is false.</p>
*
* <p><embed src="images/FullMerge_EmptyRoots.svg" type="image/svg+xml"/></p>
*
* <p>Must keep all contexts; {@link #EMPTY} in array is a special value (and
* null parent).<br>
* <embed src="images/FullMerge_EmptyRoot.svg" type="image/svg+xml"/></p>
*
* <p><embed src="images/FullMerge_SameRoot.svg" type="image/svg+xml"/></p>
*
* @param a the first {@link SingletonPredictionContext}
* @param b the second {@link SingletonPredictionContext}
* @param rootIsWildcard {@code true} if this is a local-context merge,
* otherwise false to indicate a full-context merge
*/
static Ref<PredictionContext> mergeRoot(const Ref<SingletonPredictionContext> &a,
const Ref<SingletonPredictionContext> &b, bool rootIsWildcard);
/**
* Merge two {@link ArrayPredictionContext} instances.
*
* <p>Different tops, different parents.<br>
* <embed src="images/ArrayMerge_DiffTopDiffPar.svg" type="image/svg+xml"/></p>
*
* <p>Shared top, same parents.<br>
* <embed src="images/ArrayMerge_ShareTopSamePar.svg" type="image/svg+xml"/></p>
*
* <p>Shared top, different parents.<br>
* <embed src="images/ArrayMerge_ShareTopDiffPar.svg" type="image/svg+xml"/></p>
*
* <p>Shared top, all shared parents.<br>
* <embed src="images/ArrayMerge_ShareTopSharePar.svg" type="image/svg+xml"/></p>
*
* <p>Equal tops, merge parents and reduce top to
* {@link SingletonPredictionContext}.<br>
* <embed src="images/ArrayMerge_EqualTop.svg" type="image/svg+xml"/></p>
*/
static Ref<PredictionContext> mergeArrays(const Ref<ArrayPredictionContext> &a,
const Ref<ArrayPredictionContext> &b, bool rootIsWildcard, PredictionContextMergeCache *mergeCache);
protected:
/// Make pass over all M parents; merge any equal() ones.
/// @returns true if the list has been changed (i.e. duplicates where found).
static bool combineCommonParents(std::vector<Ref<PredictionContext>> &parents);
public:
static std::string toDOTString(const Ref<PredictionContext> &context);
static Ref<PredictionContext> getCachedContext(const Ref<PredictionContext> &context,
PredictionContextCache &contextCache,
std::map<Ref<PredictionContext>, Ref<PredictionContext>> &visited);
// ter's recursive version of Sam's getAllNodes()
static std::vector<Ref<PredictionContext>> getAllContextNodes(const Ref<PredictionContext> &context);
static void getAllContextNodes_(const Ref<PredictionContext> &context,
std::vector<Ref<PredictionContext>> &nodes, std::set<PredictionContext *> &visited);
virtual std::string toString() const;
virtual std::string toString(Recognizer *recog) const;
std::vector<std::string> toStrings(Recognizer *recognizer, int currentState);
std::vector<std::string> toStrings(Recognizer *recognizer, const Ref<PredictionContext> &stop, int currentState);
};
struct PredictionContextHasher {
size_t operator () (const Ref<PredictionContext> &k) const {
return k->hashCode();
}
};
struct PredictionContextComparer {
bool operator () (const Ref<PredictionContext> &lhs, const Ref<PredictionContext> &rhs) const
{
if (lhs == rhs) // Object identity.
return true;
return (lhs->hashCode() == rhs->hashCode()) && (*lhs == *rhs);
}
};
class PredictionContextMergeCache {
public:
Ref<PredictionContext> put(Ref<PredictionContext> const& key1, Ref<PredictionContext> const& key2,
Ref<PredictionContext> const& value);
Ref<PredictionContext> get(Ref<PredictionContext> const& key1, Ref<PredictionContext> const& key2);
void clear();
std::string toString() const;
size_t count() const;
private:
std::unordered_map<Ref<PredictionContext>,
std::unordered_map<Ref<PredictionContext>, Ref<PredictionContext>, PredictionContextHasher, PredictionContextComparer>,
PredictionContextHasher, PredictionContextComparer> _data;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,436 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "support/BitSet.h"
namespace antlr4 {
namespace atn {
/**
* This enumeration defines the prediction modes available in ANTLR 4 along with
* utility methods for analyzing configuration sets for conflicts and/or
* ambiguities.
*/
enum class PredictionMode {
/**
* The SLL(*) prediction mode. This prediction mode ignores the current
* parser context when making predictions. This is the fastest prediction
* mode, and provides correct results for many grammars. This prediction
* mode is more powerful than the prediction mode provided by ANTLR 3, but
* may result in syntax errors for grammar and input combinations which are
* not SLL.
*
* <p>
* When using this prediction mode, the parser will either return a correct
* parse tree (i.e. the same parse tree that would be returned with the
* {@link #LL} prediction mode), or it will report a syntax error. If a
* syntax error is encountered when using the {@link #SLL} prediction mode,
* it may be due to either an actual syntax error in the input or indicate
* that the particular combination of grammar and input requires the more
* powerful {@link #LL} prediction abilities to complete successfully.</p>
*
* <p>
* This prediction mode does not provide any guarantees for prediction
* behavior for syntactically-incorrect inputs.</p>
*/
SLL,
/**
* The LL(*) prediction mode. This prediction mode allows the current parser
* context to be used for resolving SLL conflicts that occur during
* prediction. This is the fastest prediction mode that guarantees correct
* parse results for all combinations of grammars with syntactically correct
* inputs.
*
* <p>
* When using this prediction mode, the parser will make correct decisions
* for all syntactically-correct grammar and input combinations. However, in
* cases where the grammar is truly ambiguous this prediction mode might not
* report a precise answer for <em>exactly which</em> alternatives are
* ambiguous.</p>
*
* <p>
* This prediction mode does not provide any guarantees for prediction
* behavior for syntactically-incorrect inputs.</p>
*/
LL,
/**
* The LL(*) prediction mode with exact ambiguity detection. In addition to
* the correctness guarantees provided by the {@link #LL} prediction mode,
* this prediction mode instructs the prediction algorithm to determine the
* complete and exact set of ambiguous alternatives for every ambiguous
* decision encountered while parsing.
*
* <p>
* This prediction mode may be used for diagnosing ambiguities during
* grammar development. Due to the performance overhead of calculating sets
* of ambiguous alternatives, this prediction mode should be avoided when
* the exact results are not necessary.</p>
*
* <p>
* This prediction mode does not provide any guarantees for prediction
* behavior for syntactically-incorrect inputs.</p>
*/
LL_EXACT_AMBIG_DETECTION
};
class ANTLR4CPP_PUBLIC PredictionModeClass {
public:
/**
* Computes the SLL prediction termination condition.
*
* <p>
* This method computes the SLL prediction termination condition for both of
* the following cases.</p>
*
* <ul>
* <li>The usual SLL+LL fallback upon SLL conflict</li>
* <li>Pure SLL without LL fallback</li>
* </ul>
*
* <p><strong>COMBINED SLL+LL PARSING</strong></p>
*
* <p>When LL-fallback is enabled upon SLL conflict, correct predictions are
* ensured regardless of how the termination condition is computed by this
* method. Due to the substantially higher cost of LL prediction, the
* prediction should only fall back to LL when the additional lookahead
* cannot lead to a unique SLL prediction.</p>
*
* <p>Assuming combined SLL+LL parsing, an SLL configuration set with only
* conflicting subsets should fall back to full LL, even if the
* configuration sets don't resolve to the same alternative (e.g.
* {@code {1,2}} and {@code {3,4}}. If there is at least one non-conflicting
* configuration, SLL could continue with the hopes that more lookahead will
* resolve via one of those non-conflicting configurations.</p>
*
* <p>Here's the prediction termination rule them: SLL (for SLL+LL parsing)
* stops when it sees only conflicting configuration subsets. In contrast,
* full LL keeps going when there is uncertainty.</p>
*
* <p><strong>HEURISTIC</strong></p>
*
* <p>As a heuristic, we stop prediction when we see any conflicting subset
* unless we see a state that only has one alternative associated with it.
* The single-alt-state thing lets prediction continue upon rules like
* (otherwise, it would admit defeat too soon):</p>
*
* <p>{@code [12|1|[], 6|2|[], 12|2|[]]. s : (ID | ID ID?) ';' ;}</p>
*
* <p>When the ATN simulation reaches the state before {@code ';'}, it has a
* DFA state that looks like: {@code [12|1|[], 6|2|[], 12|2|[]]}. Naturally
* {@code 12|1|[]} and {@code 12|2|[]} conflict, but we cannot stop
* processing this node because alternative to has another way to continue,
* via {@code [6|2|[]]}.</p>
*
* <p>It also let's us continue for this rule:</p>
*
* <p>{@code [1|1|[], 1|2|[], 8|3|[]] a : A | A | A B ;}</p>
*
* <p>After matching input A, we reach the stop state for rule A, state 1.
* State 8 is the state right before B. Clearly alternatives 1 and 2
* conflict and no amount of further lookahead will separate the two.
* However, alternative 3 will be able to continue and so we do not stop
* working on this state. In the previous example, we're concerned with
* states associated with the conflicting alternatives. Here alt 3 is not
* associated with the conflicting configs, but since we can continue
* looking for input reasonably, don't declare the state done.</p>
*
* <p><strong>PURE SLL PARSING</strong></p>
*
* <p>To handle pure SLL parsing, all we have to do is make sure that we
* combine stack contexts for configurations that differ only by semantic
* predicate. From there, we can do the usual SLL termination heuristic.</p>
*
* <p><strong>PREDICATES IN SLL+LL PARSING</strong></p>
*
* <p>SLL decisions don't evaluate predicates until after they reach DFA stop
* states because they need to create the DFA cache that works in all
* semantic situations. In contrast, full LL evaluates predicates collected
* during start state computation so it can ignore predicates thereafter.
* This means that SLL termination detection can totally ignore semantic
* predicates.</p>
*
* <p>Implementation-wise, {@link ATNConfigSet} combines stack contexts but not
* semantic predicate contexts so we might see two configurations like the
* following.</p>
*
* <p>{@code (s, 1, x, {}), (s, 1, x', {p})}</p>
*
* <p>Before testing these configurations against others, we have to merge
* {@code x} and {@code x'} (without modifying the existing configurations).
* For example, we test {@code (x+x')==x''} when looking for conflicts in
* the following configurations.</p>
*
* <p>{@code (s, 1, x, {}), (s, 1, x', {p}), (s, 2, x'', {})}</p>
*
* <p>If the configuration set has predicates (as indicated by
* {@link ATNConfigSet#hasSemanticContext}), this algorithm makes a copy of
* the configurations to strip out all of the predicates so that a standard
* {@link ATNConfigSet} will merge everything ignoring predicates.</p>
*/
static bool hasSLLConflictTerminatingPrediction(PredictionMode mode, ATNConfigSet *configs);
/// <summary>
/// Checks if any configuration in {@code configs} is in a
/// <seealso cref="RuleStopState"/>. Configurations meeting this condition have
/// reached
/// the end of the decision rule (local context) or end of start rule (full
/// context).
/// </summary>
/// <param name="configs"> the configuration set to test </param>
/// <returns> {@code true} if any configuration in {@code configs} is in a
/// <seealso cref="RuleStopState"/>, otherwise {@code false} </returns>
static bool hasConfigInRuleStopState(ATNConfigSet *configs);
/// <summary>
/// Checks if all configurations in {@code configs} are in a
/// <seealso cref="RuleStopState"/>. Configurations meeting this condition have
/// reached
/// the end of the decision rule (local context) or end of start rule (full
/// context).
/// </summary>
/// <param name="configs"> the configuration set to test </param>
/// <returns> {@code true} if all configurations in {@code configs} are in a
/// <seealso cref="RuleStopState"/>, otherwise {@code false} </returns>
static bool allConfigsInRuleStopStates(ATNConfigSet *configs);
/**
* Full LL prediction termination.
*
* <p>Can we stop looking ahead during ATN simulation or is there some
* uncertainty as to which alternative we will ultimately pick, after
* consuming more input? Even if there are partial conflicts, we might know
* that everything is going to resolve to the same minimum alternative. That
* means we can stop since no more lookahead will change that fact. On the
* other hand, there might be multiple conflicts that resolve to different
* minimums. That means we need more look ahead to decide which of those
* alternatives we should predict.</p>
*
* <p>The basic idea is to split the set of configurations {@code C}, into
* conflicting subsets {@code (s, _, ctx, _)} and singleton subsets with
* non-conflicting configurations. Two configurations conflict if they have
* identical {@link ATNConfig#state} and {@link ATNConfig#context} values
* but different {@link ATNConfig#alt} value, e.g. {@code (s, i, ctx, _)}
* and {@code (s, j, ctx, _)} for {@code i!=j}.</p>
*
* <p>Reduce these configuration subsets to the set of possible alternatives.
* You can compute the alternative subsets in one pass as follows:</p>
*
* <p>{@code A_s,ctx = {i | (s, i, ctx, _)}} for each configuration in
* {@code C} holding {@code s} and {@code ctx} fixed.</p>
*
* <p>Or in pseudo-code, for each configuration {@code c} in {@code C}:</p>
*
* <pre>
* map[c] U= c.{@link ATNConfig#alt alt} # map hash/equals uses s and x, not
* alt and not pred
* </pre>
*
* <p>The values in {@code map} are the set of {@code A_s,ctx} sets.</p>
*
* <p>If {@code |A_s,ctx|=1} then there is no conflict associated with
* {@code s} and {@code ctx}.</p>
*
* <p>Reduce the subsets to singletons by choosing a minimum of each subset. If
* the union of these alternative subsets is a singleton, then no amount of
* more lookahead will help us. We will always pick that alternative. If,
* however, there is more than one alternative, then we are uncertain which
* alternative to predict and must continue looking for resolution. We may
* or may not discover an ambiguity in the future, even if there are no
* conflicting subsets this round.</p>
*
* <p>The biggest sin is to terminate early because it means we've made a
* decision but were uncertain as to the eventual outcome. We haven't used
* enough lookahead. On the other hand, announcing a conflict too late is no
* big deal; you will still have the conflict. It's just inefficient. It
* might even look until the end of file.</p>
*
* <p>No special consideration for semantic predicates is required because
* predicates are evaluated on-the-fly for full LL prediction, ensuring that
* no configuration contains a semantic context during the termination
* check.</p>
*
* <p><strong>CONFLICTING CONFIGS</strong></p>
*
* <p>Two configurations {@code (s, i, x)} and {@code (s, j, x')}, conflict
* when {@code i!=j} but {@code x=x'}. Because we merge all
* {@code (s, i, _)} configurations together, that means that there are at
* most {@code n} configurations associated with state {@code s} for
* {@code n} possible alternatives in the decision. The merged stacks
* complicate the comparison of configuration contexts {@code x} and
* {@code x'}. Sam checks to see if one is a subset of the other by calling
* merge and checking to see if the merged result is either {@code x} or
* {@code x'}. If the {@code x} associated with lowest alternative {@code i}
* is the superset, then {@code i} is the only possible prediction since the
* others resolve to {@code min(i)} as well. However, if {@code x} is
* associated with {@code j>i} then at least one stack configuration for
* {@code j} is not in conflict with alternative {@code i}. The algorithm
* should keep going, looking for more lookahead due to the uncertainty.</p>
*
* <p>For simplicity, I'm doing a equality check between {@code x} and
* {@code x'} that lets the algorithm continue to consume lookahead longer
* than necessary. The reason I like the equality is of course the
* simplicity but also because that is the test you need to detect the
* alternatives that are actually in conflict.</p>
*
* <p><strong>CONTINUE/STOP RULE</strong></p>
*
* <p>Continue if union of resolved alternative sets from non-conflicting and
* conflicting alternative subsets has more than one alternative. We are
* uncertain about which alternative to predict.</p>
*
* <p>The complete set of alternatives, {@code [i for (_,i,_)]}, tells us which
* alternatives are still in the running for the amount of input we've
* consumed at this point. The conflicting sets let us to strip away
* configurations that won't lead to more states because we resolve
* conflicts to the configuration with a minimum alternate for the
* conflicting set.</p>
*
* <p><strong>CASES</strong></p>
*
* <ul>
*
* <li>no conflicts and more than 1 alternative in set =&gt; continue</li>
*
* <li> {@code (s, 1, x)}, {@code (s, 2, x)}, {@code (s, 3, z)},
* {@code (s', 1, y)}, {@code (s', 2, y)} yields non-conflicting set
* {@code {3}} U conflicting sets {@code min({1,2})} U {@code min({1,2})} =
* {@code {1,3}} =&gt; continue
* </li>
*
* <li>{@code (s, 1, x)}, {@code (s, 2, x)}, {@code (s', 1, y)},
* {@code (s', 2, y)}, {@code (s'', 1, z)} yields non-conflicting set
* {@code {1}} U conflicting sets {@code min({1,2})} U {@code min({1,2})} =
* {@code {1}} =&gt; stop and predict 1</li>
*
* <li>{@code (s, 1, x)}, {@code (s, 2, x)}, {@code (s', 1, y)},
* {@code (s', 2, y)} yields conflicting, reduced sets {@code {1}} U
* {@code {1}} = {@code {1}} =&gt; stop and predict 1, can announce
* ambiguity {@code {1,2}}</li>
*
* <li>{@code (s, 1, x)}, {@code (s, 2, x)}, {@code (s', 2, y)},
* {@code (s', 3, y)} yields conflicting, reduced sets {@code {1}} U
* {@code {2}} = {@code {1,2}} =&gt; continue</li>
*
* <li>{@code (s, 1, x)}, {@code (s, 2, x)}, {@code (s', 3, y)},
* {@code (s', 4, y)} yields conflicting, reduced sets {@code {1}} U
* {@code {3}} = {@code {1,3}} =&gt; continue</li>
*
* </ul>
*
* <p><strong>EXACT AMBIGUITY DETECTION</strong></p>
*
* <p>If all states report the same conflicting set of alternatives, then we
* know we have the exact ambiguity set.</p>
*
* <p><code>|A_<em>i</em>|&gt;1</code> and
* <code>A_<em>i</em> = A_<em>j</em></code> for all <em>i</em>, <em>j</em>.</p>
*
* <p>In other words, we continue examining lookahead until all {@code A_i}
* have more than one alternative and all {@code A_i} are the same. If
* {@code A={{1,2}, {1,3}}}, then regular LL prediction would terminate
* because the resolved set is {@code {1}}. To determine what the real
* ambiguity is, we have to know whether the ambiguity is between one and
* two or one and three so we keep going. We can only stop prediction when
* we need exact ambiguity detection when the sets look like
* {@code A={{1,2}}} or {@code {{1,2},{1,2}}}, etc...</p>
*/
static size_t resolvesToJustOneViableAlt(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Determines if every alternative subset in {@code altsets} contains more
/// than one alternative.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
/// <returns> {@code true} if every <seealso cref="BitSet"/> in {@code altsets}
/// has
/// <seealso cref="BitSet#cardinality cardinality"/> &gt; 1, otherwise {@code
/// false} </returns>
static bool allSubsetsConflict(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Determines if any single alternative subset in {@code altsets} contains
/// exactly one alternative.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
/// <returns> {@code true} if {@code altsets} contains a <seealso
/// cref="BitSet"/> with
/// <seealso cref="BitSet#cardinality cardinality"/> 1, otherwise {@code false}
/// </returns>
static bool hasNonConflictingAltSet(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Determines if any single alternative subset in {@code altsets} contains
/// more than one alternative.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
/// <returns> {@code true} if {@code altsets} contains a <seealso
/// cref="BitSet"/> with
/// <seealso cref="BitSet#cardinality cardinality"/> &gt; 1, otherwise {@code
/// false} </returns>
static bool hasConflictingAltSet(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Determines if every alternative subset in {@code altsets} is equivalent.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
/// <returns> {@code true} if every member of {@code altsets} is equal to the
/// others, otherwise {@code false} </returns>
static bool allSubsetsEqual(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Returns the unique alternative predicted by all alternative subsets in
/// {@code altsets}. If no such alternative exists, this method returns
/// <seealso cref="ATN#INVALID_ALT_NUMBER"/>.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
static size_t getUniqueAlt(const std::vector<antlrcpp::BitSet> &altsets);
/// <summary>
/// Gets the complete set of represented alternatives for a collection of
/// alternative subsets. This method returns the union of each <seealso
/// cref="BitSet"/>
/// in {@code altsets}.
/// </summary>
/// <param name="altsets"> a collection of alternative subsets </param>
/// <returns> the set of represented alternatives in {@code altsets} </returns>
static antlrcpp::BitSet getAlts(const std::vector<antlrcpp::BitSet> &altsets);
/** Get union of all alts from configs. @since 4.5.1 */
static antlrcpp::BitSet getAlts(ATNConfigSet *configs);
/// <summary>
/// This function gets the conflicting alt subsets from a configuration set.
/// For each configuration {@code c} in {@code configs}:
///
/// <pre>
/// map[c] U= c.<seealso cref="ATNConfig#alt alt"/> # map hash/equals uses s and
/// x, not
/// alt and not pred
/// </pre>
/// </summary>
static std::vector<antlrcpp::BitSet> getConflictingAltSubsets(ATNConfigSet *configs);
/// <summary>
/// Get a map from state to alt subset from a configuration set. For each
/// configuration {@code c} in {@code configs}:
///
/// <pre>
/// map[c.<seealso cref="ATNConfig#state state"/>] U= c.<seealso
/// cref="ATNConfig#alt alt"/>
/// </pre>
/// </summary>
static std::map<ATNState*, antlrcpp::BitSet> getStateToAltMap(ATNConfigSet *configs);
static bool hasStateAssociatedWithOneAlt(ATNConfigSet *configs);
static size_t getSingleViableAlt(const std::vector<antlrcpp::BitSet> &altsets);
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,60 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ParserATNSimulator.h"
#include "atn/DecisionInfo.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC ProfilingATNSimulator : public ParserATNSimulator {
public:
ProfilingATNSimulator(Parser *parser);
virtual size_t adaptivePredict(TokenStream *input, size_t decision, ParserRuleContext *outerContext) override;
virtual std::vector<DecisionInfo> getDecisionInfo() const;
virtual dfa::DFAState* getCurrentState() const;
protected:
std::vector<DecisionInfo> _decisions;
int _sllStopIndex = 0;
int _llStopIndex = 0;
size_t _currentDecision = 0;
dfa::DFAState *_currentState;
/// <summary>
/// At the point of LL failover, we record how SLL would resolve the conflict so that
/// we can determine whether or not a decision / input pair is context-sensitive.
/// If LL gives a different result than SLL's predicted alternative, we have a
/// context sensitivity for sure. The converse is not necessarily true, however.
/// It's possible that after conflict resolution chooses minimum alternatives,
/// SLL could get the same answer as LL. Regardless of whether or not the result indicates
/// an ambiguity, it is not treated as a context sensitivity because LL prediction
/// was not required in order to produce a correct prediction for this decision and input sequence.
/// It may in fact still be a context sensitivity but we don't know by looking at the
/// minimum alternatives for the current input.
/// </summary>
size_t conflictingAltResolvedBySLL = 0;
virtual dfa::DFAState* getExistingTargetState(dfa::DFAState *previousD, size_t t) override;
virtual dfa::DFAState* computeTargetState(dfa::DFA &dfa, dfa::DFAState *previousD, size_t t) override;
virtual std::unique_ptr<ATNConfigSet> computeReachSet(ATNConfigSet *closure, size_t t, bool fullCtx) override;
virtual bool evalSemanticContext(Ref<SemanticContext> const& pred, ParserRuleContext *parserCallStack,
size_t alt, bool fullCtx) override;
virtual void reportAttemptingFullContext(dfa::DFA &dfa, const antlrcpp::BitSet &conflictingAlts, ATNConfigSet *configs,
size_t startIndex, size_t stopIndex) override;
virtual void reportContextSensitivity(dfa::DFA &dfa, size_t prediction, ATNConfigSet *configs,
size_t startIndex, size_t stopIndex) override;
virtual void reportAmbiguity(dfa::DFA &dfa, dfa::DFAState *D, size_t startIndex, size_t stopIndex, bool exact,
const antlrcpp::BitSet &ambigAlts, ATNConfigSet *configs) override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,29 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC RangeTransition final : public Transition {
public:
const size_t from;
const size_t to;
RangeTransition(ATNState *target, size_t from, size_t to);
virtual SerializationType getSerializationType() const override;
virtual misc::IntervalSet label() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,25 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC RuleStartState final : public ATNState {
public:
RuleStartState();
RuleStopState *stopState = nullptr;
bool isLeftRecursiveRule = false;
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,25 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
/// The last node in the ATN for a rule, unless that rule is the start symbol.
/// In that case, there is one transition to EOF. Later, we might encode
/// references to all calls to this rule to compute FOLLOW sets for
/// error handling.
class ANTLR4CPP_PUBLIC RuleStopState final : public ATNState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,40 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC RuleTransition : public Transition {
public:
/// Ptr to the rule definition object for this rule ref.
const size_t ruleIndex; // no Rule object at runtime
const int precedence;
/// What node to begin computations following ref to rule.
ATNState *followState;
/// @deprecated Use
/// <seealso cref="#RuleTransition(RuleStartState, size_t, int, ATNState)"/> instead.
RuleTransition(RuleStartState *ruleStart, size_t ruleIndex, ATNState *followState);
RuleTransition(RuleStartState *ruleStart, size_t ruleIndex, int precedence, ATNState *followState);
RuleTransition(RuleTransition const&) = delete;
RuleTransition& operator=(RuleTransition const&) = delete;
virtual SerializationType getSerializationType() const override;
virtual bool isEpsilon() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,222 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "Recognizer.h"
#include "support/CPPUtils.h"
namespace antlr4 {
namespace atn {
/// A tree structure used to record the semantic context in which
/// an ATN configuration is valid. It's either a single predicate,
/// a conjunction "p1 && p2", or a sum of products "p1||p2".
///
/// I have scoped the AND, OR, and Predicate subclasses of
/// SemanticContext within the scope of this outer class.
class ANTLR4CPP_PUBLIC SemanticContext : public std::enable_shared_from_this<SemanticContext> {
public:
struct Hasher
{
size_t operator()(Ref<SemanticContext> const& k) const {
return k->hashCode();
}
};
struct Comparer {
bool operator()(Ref<SemanticContext> const& lhs, Ref<SemanticContext> const& rhs) const {
if (lhs == rhs)
return true;
return (lhs->hashCode() == rhs->hashCode()) && (*lhs == *rhs);
}
};
using Set = std::unordered_set<Ref<SemanticContext>, Hasher, Comparer>;
/**
* The default {@link SemanticContext}, which is semantically equivalent to
* a predicate of the form {@code {true}?}.
*/
static const Ref<SemanticContext> NONE;
virtual ~SemanticContext();
virtual size_t hashCode() const = 0;
virtual std::string toString() const = 0;
virtual bool operator == (const SemanticContext &other) const = 0;
virtual bool operator != (const SemanticContext &other) const;
/// <summary>
/// For context independent predicates, we evaluate them without a local
/// context (i.e., null context). That way, we can evaluate them without
/// having to create proper rule-specific context during prediction (as
/// opposed to the parser, which creates them naturally). In a practical
/// sense, this avoids a cast exception from RuleContext to myruleContext.
/// <p/>
/// For context dependent predicates, we must pass in a local context so that
/// references such as $arg evaluate properly as _localctx.arg. We only
/// capture context dependent predicates in the context in which we begin
/// prediction, so we passed in the outer context here in case of context
/// dependent predicate evaluation.
/// </summary>
virtual bool eval(Recognizer *parser, RuleContext *parserCallStack) = 0;
/**
* Evaluate the precedence predicates for the context and reduce the result.
*
* @param parser The parser instance.
* @param parserCallStack
* @return The simplified semantic context after precedence predicates are
* evaluated, which will be one of the following values.
* <ul>
* <li>{@link #NONE}: if the predicate simplifies to {@code true} after
* precedence predicates are evaluated.</li>
* <li>{@code null}: if the predicate simplifies to {@code false} after
* precedence predicates are evaluated.</li>
* <li>{@code this}: if the semantic context is not changed as a result of
* precedence predicate evaluation.</li>
* <li>A non-{@code null} {@link SemanticContext}: the new simplified
* semantic context after precedence predicates are evaluated.</li>
* </ul>
*/
virtual Ref<SemanticContext> evalPrecedence(Recognizer *parser, RuleContext *parserCallStack);
static Ref<SemanticContext> And(Ref<SemanticContext> const& a, Ref<SemanticContext> const& b);
/// See also: ParserATNSimulator::getPredsForAmbigAlts.
static Ref<SemanticContext> Or(Ref<SemanticContext> const& a, Ref<SemanticContext> const& b);
class Predicate;
class PrecedencePredicate;
class Operator;
class AND;
class OR;
private:
static std::vector<Ref<PrecedencePredicate>> filterPrecedencePredicates(const Set &collection);
};
class ANTLR4CPP_PUBLIC SemanticContext::Predicate : public SemanticContext {
public:
const size_t ruleIndex;
const size_t predIndex;
const bool isCtxDependent; // e.g., $i ref in pred
protected:
Predicate();
public:
Predicate(size_t ruleIndex, size_t predIndex, bool isCtxDependent);
virtual bool eval(Recognizer *parser, RuleContext *parserCallStack) override;
virtual size_t hashCode() const override;
virtual bool operator == (const SemanticContext &other) const override;
virtual std::string toString() const override;
};
class ANTLR4CPP_PUBLIC SemanticContext::PrecedencePredicate : public SemanticContext {
public:
const int precedence;
protected:
PrecedencePredicate();
public:
PrecedencePredicate(int precedence);
virtual bool eval(Recognizer *parser, RuleContext *parserCallStack) override;
virtual Ref<SemanticContext> evalPrecedence(Recognizer *parser, RuleContext *parserCallStack) override;
virtual int compareTo(PrecedencePredicate *o);
virtual size_t hashCode() const override;
virtual bool operator == (const SemanticContext &other) const override;
virtual std::string toString() const override;
};
/**
* This is the base class for semantic context "operators", which operate on
* a collection of semantic context "operands".
*
* @since 4.3
*/
class ANTLR4CPP_PUBLIC SemanticContext::Operator : public SemanticContext {
public:
virtual ~Operator() override;
/**
* Gets the operands for the semantic context operator.
*
* @return a collection of {@link SemanticContext} operands for the
* operator.
*
* @since 4.3
*/
virtual std::vector<Ref<SemanticContext>> getOperands() const = 0;
};
/**
* A semantic context which is true whenever none of the contained contexts
* is false.
*/
class ANTLR4CPP_PUBLIC SemanticContext::AND : public SemanticContext::Operator {
public:
std::vector<Ref<SemanticContext>> opnds;
AND(Ref<SemanticContext> const& a, Ref<SemanticContext> const& b) ;
virtual std::vector<Ref<SemanticContext>> getOperands() const override;
virtual bool operator == (const SemanticContext &other) const override;
virtual size_t hashCode() const override;
/**
* The evaluation of predicates by this context is short-circuiting, but
* unordered.</p>
*/
virtual bool eval(Recognizer *parser, RuleContext *parserCallStack) override;
virtual Ref<SemanticContext> evalPrecedence(Recognizer *parser, RuleContext *parserCallStack) override;
virtual std::string toString() const override;
};
/**
* A semantic context which is true whenever at least one of the contained
* contexts is true.
*/
class ANTLR4CPP_PUBLIC SemanticContext::OR : public SemanticContext::Operator {
public:
std::vector<Ref<SemanticContext>> opnds;
OR(Ref<SemanticContext> const& a, Ref<SemanticContext> const& b);
virtual std::vector<Ref<SemanticContext>> getOperands() const override;
virtual bool operator == (const SemanticContext &other) const override;
virtual size_t hashCode() const override;
/**
* The evaluation of predicates by this context is short-circuiting, but
* unordered.
*/
virtual bool eval(Recognizer *parser, RuleContext *parserCallStack) override;
virtual Ref<SemanticContext> evalPrecedence(Recognizer *parser, RuleContext *parserCallStack) override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4
// Hash function for SemanticContext, used in the MurmurHash::update function
namespace std {
using antlr4::atn::SemanticContext;
template <> struct hash<SemanticContext>
{
size_t operator () (SemanticContext &x) const
{
return x.hashCode();
}
};
}

View File

@@ -0,0 +1,30 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// A transition containing a set of values. </summary>
class ANTLR4CPP_PUBLIC SetTransition : public Transition {
public:
const misc::IntervalSet set;
SetTransition(ATNState *target, const misc::IntervalSet &set);
virtual SerializationType getSerializationType() const override;
virtual misc::IntervalSet label() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,36 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/PredictionContext.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC SingletonPredictionContext : public PredictionContext {
public:
// Usually a parent is linked via a weak ptr. Not so here as we have kinda reverse reference chain.
// There are no child contexts stored here and often the parent context is left dangling when it's
// owning ATNState is released. In order to avoid having this context released as well (leaving all other contexts
// which got this one as parent with a null reference) we use a shared_ptr here instead, to keep those left alone
// parent contexts alive.
const Ref<PredictionContext> parent;
const size_t returnState;
SingletonPredictionContext(Ref<PredictionContext> const& parent, size_t returnState);
virtual ~SingletonPredictionContext();
static Ref<SingletonPredictionContext> create(Ref<PredictionContext> const& parent, size_t returnState);
virtual size_t size() const override;
virtual Ref<PredictionContext> getParent(size_t index) const override;
virtual size_t getReturnState(size_t index) const override;
virtual bool operator == (const PredictionContext &o) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,21 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/BlockStartState.h"
namespace antlr4 {
namespace atn {
/// The block that begins a closure loop.
class ANTLR4CPP_PUBLIC StarBlockStartState final : public BlockStartState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,35 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC StarLoopEntryState final : public DecisionState {
public:
StarLoopEntryState();
/**
* Indicates whether this state can benefit from a precedence DFA during SLL
* decision making.
*
* <p>This is a computed property that is calculated during ATN deserialization
* and stored for use in {@link ParserATNSimulator} and
* {@link ParserInterpreter}.</p>
*
* @see DFA#isPrecedenceDfa()
*/
bool isPrecedenceDecision = false;
StarLoopbackState *loopBackState = nullptr;
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,21 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/ATNState.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC StarLoopbackState final : public ATNState {
public:
StarLoopEntryState *getLoopEntryState();
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,21 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/DecisionState.h"
namespace antlr4 {
namespace atn {
/// The Tokens rule start state linking to each lexer rule start state.
class ANTLR4CPP_PUBLIC TokensStartState final : public DecisionState {
public:
virtual size_t getStateType() override;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,76 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "misc/IntervalSet.h"
namespace antlr4 {
namespace atn {
/// <summary>
/// An ATN transition between any two ATN states. Subclasses define
/// atom, set, epsilon, action, predicate, rule transitions.
/// <p/>
/// This is a one way link. It emanates from a state (usually via a list of
/// transitions) and has a target state.
/// <p/>
/// Since we never have to change the ATN transitions once we construct it,
/// we can fix these transitions as specific classes. The DFA transitions
/// on the other hand need to update the labels as it adds transitions to
/// the states. We'll use the term Edge for the DFA to distinguish them from
/// ATN transitions.
/// </summary>
class ANTLR4CPP_PUBLIC Transition {
public:
// constants for serialization
enum SerializationType {
EPSILON = 1,
RANGE = 2,
RULE = 3,
PREDICATE = 4, // e.g., {isType(input.LT(1))}?
ATOM = 5,
ACTION = 6,
SET = 7, // ~(A|B) or ~atom, wildcard, which convert to next 2
NOT_SET = 8,
WILDCARD = 9,
PRECEDENCE = 10,
};
static const std::vector<std::string> serializationNames;
/// The target of this transition.
// ml: this is a reference into the ATN.
ATNState *target;
virtual ~Transition();
protected:
Transition(ATNState *target);
public:
virtual SerializationType getSerializationType() const = 0;
/**
* Determines if the transition is an "epsilon" transition.
*
* <p>The default implementation returns {@code false}.</p>
*
* @return {@code true} if traversing this transition in the ATN does not
* consume an input symbol; otherwise, {@code false} if traversing this
* transition consumes (matches) an input symbol.
*/
virtual bool isEpsilon() const;
virtual misc::IntervalSet label() const;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const = 0;
virtual std::string toString() const;
Transition(Transition const&) = delete;
Transition& operator=(Transition const&) = delete;
};
} // namespace atn
} // namespace antlr4

View File

@@ -0,0 +1,25 @@
/* Copyright (c) 2012-2017 The ANTLR Project. All rights reserved.
* Use of this file is governed by the BSD 3-clause license that
* can be found in the LICENSE.txt file in the project root.
*/
#pragma once
#include "atn/Transition.h"
namespace antlr4 {
namespace atn {
class ANTLR4CPP_PUBLIC WildcardTransition final : public Transition {
public:
WildcardTransition(ATNState *target);
virtual SerializationType getSerializationType() const override;
virtual bool matches(size_t symbol, size_t minVocabSymbol, size_t maxVocabSymbol) const override;
virtual std::string toString() const override;
};
} // namespace atn
} // namespace antlr4