summaryrefslogtreecommitdiff
path: root/src/core/model
diff options
context:
space:
mode:
Diffstat (limited to 'src/core/model')
-rw-r--r--src/core/model/Ontology.cpp193
-rw-r--r--src/core/model/Ontology.hpp297
-rw-r--r--src/core/model/Syntax.cpp58
-rw-r--r--src/core/model/Syntax.hpp196
4 files changed, 700 insertions, 44 deletions
diff --git a/src/core/model/Ontology.cpp b/src/core/model/Ontology.cpp
index 8829139..3af727d 100644
--- a/src/core/model/Ontology.cpp
+++ b/src/core/model/Ontology.cpp
@@ -20,8 +20,9 @@
#include <queue>
#include <set>
-#include <core/common/RttiBuilder.hpp>
#include <core/common/Exceptions.hpp>
+#include <core/common/RttiBuilder.hpp>
+#include <core/common/Utils.hpp>
#include "Ontology.hpp"
@@ -169,52 +170,60 @@ static NodeVector<Node> pathTo(const Node *start, Logger &logger,
return shortest;
}
+struct CollectState {
+ Node *n;
+ size_t depth;
+
+ CollectState(Node *n, size_t depth) : n(n), depth(depth) {}
+};
+
template <typename F>
static NodeVector<Node> collect(const Node *start, F match)
{
// result
NodeVector<Node> res;
// queue for breadth-first search of graph.
- std::queue<Rooted<Node>> q;
+ std::queue<CollectState> q;
// put the initial node on the stack.
- q.push(const_cast<Node *>(start));
+ q.push(CollectState(const_cast<Node *>(start), 0));
// set of visited nodes.
std::unordered_set<const Node *> visited;
while (!q.empty()) {
- Rooted<Node> n = q.front();
+ CollectState state = q.front();
q.pop();
// do not proceed if this node was already visited.
- if (!visited.insert(n.get()).second) {
+ if (!visited.insert(state.n).second) {
continue;
}
- if (n->isa(&RttiTypes::StructuredClass)) {
- Rooted<StructuredClass> strct = n.cast<StructuredClass>();
+ if (state.n->isa(&RttiTypes::Descriptor)) {
+ Rooted<Descriptor> strct{static_cast<Descriptor *>(state.n)};
// look through all fields.
NodeVector<FieldDescriptor> fields = strct->getFieldDescriptors();
for (auto fd : fields) {
// note matches.
- if (match(fd)) {
+ if (match(fd, state.depth)) {
res.push_back(fd);
}
// only continue in the TREE field.
if (fd->getFieldType() == FieldDescriptor::FieldType::TREE) {
- q.push(fd);
+ q.push(CollectState(fd.get(), state.depth));
}
}
} else {
// otherwise this is a FieldDescriptor.
- Rooted<FieldDescriptor> field = n.cast<FieldDescriptor>();
+ Rooted<FieldDescriptor> field{
+ static_cast<FieldDescriptor *>(state.n)};
// and we proceed by visiting all permitted children.
for (auto c : field->getChildrenWithSubclasses()) {
// note matches.
- if (match(c)) {
+ if (match(c, state.depth)) {
res.push_back(c);
}
// We only continue our search via transparent children.
if (c->isTransparent()) {
- q.push(c);
+ q.push(CollectState(c.get(), state.depth + 1));
}
}
}
@@ -222,28 +231,59 @@ static NodeVector<Node> collect(const Node *start, F match)
return res;
}
+static std::vector<SyntaxDescriptor> collectPermittedTokens(
+ const Node *start, Handle<Domain> domain)
+{
+ // gather SyntaxDescriptors for structure children first.
+ std::vector<SyntaxDescriptor> res;
+ collect(start, [&res](Handle<Node> n, size_t depth) {
+ SyntaxDescriptor stx;
+ if (n->isa(&RttiTypes::FieldDescriptor)) {
+ stx = n.cast<FieldDescriptor>()->getSyntaxDescriptor(depth);
+ } else {
+ stx = n.cast<Descriptor>()->getSyntaxDescriptor(depth);
+ }
+ // do not add trivial SyntaxDescriptors.
+ if (!stx.isEmpty()) {
+ res.push_back(stx);
+ }
+ return false;
+ });
+ // gather SyntaxDescriptors for AnnotationClasses.
+ for (auto a : domain->getAnnotationClasses()) {
+ SyntaxDescriptor stx = a->getSyntaxDescriptor();
+ if (!stx.isEmpty()) {
+ res.push_back(stx);
+ }
+ }
+ return res;
+}
+
/* Class FieldDescriptor */
FieldDescriptor::FieldDescriptor(Manager &mgr, Handle<Type> primitiveType,
Handle<Descriptor> parent, FieldType fieldType,
- std::string name, bool optional)
+ std::string name, bool optional,
+ WhitespaceMode whitespaceMode)
: Node(mgr, std::move(name), parent),
children(this),
fieldType(fieldType),
primitiveType(acquire(primitiveType)),
optional(optional),
- primitive(true)
+ primitive(true),
+ whitespaceMode(whitespaceMode)
{
}
FieldDescriptor::FieldDescriptor(Manager &mgr, Handle<Descriptor> parent,
FieldType fieldType, std::string name,
- bool optional)
+ bool optional, WhitespaceMode whitespaceMode)
: Node(mgr, std::move(name), parent),
children(this),
fieldType(fieldType),
optional(optional),
- primitive(false)
+ primitive(false),
+ whitespaceMode(whitespaceMode)
{
}
@@ -272,6 +312,25 @@ bool FieldDescriptor::doValidate(Logger &logger) const
} else {
valid = valid & validateName(logger);
}
+ // check start and end token.
+ if (!startToken.special && !startToken.token.empty() &&
+ !Utils::isUserDefinedToken(startToken.token)) {
+ // TODO: Correct error message.
+ logger.error(std::string("Field \"") + getName() +
+ "\" has an invalid custom start token: " +
+ startToken.token,
+ *this);
+ valid = false;
+ }
+ if (!endToken.special && !endToken.token.empty() &&
+ !Utils::isUserDefinedToken(endToken.token)) {
+ // TODO: Correct error message.
+ logger.error(std::string("Field \"") + getName() +
+ "\" has an invalid custom end token: " +
+ endToken.token,
+ *this);
+ valid = false;
+ }
// check consistency of FieldType with the rest of the FieldDescriptor.
if (primitive) {
@@ -325,7 +384,7 @@ bool FieldDescriptor::doValidate(Logger &logger) const
}
static void gatherSubclasses(
- std::unordered_set<const StructuredClass *>& visited,
+ std::unordered_set<const StructuredClass *> &visited,
NodeVector<StructuredClass> &res, Handle<StructuredClass> strct)
{
// this check is to prevent cycles.
@@ -334,7 +393,7 @@ static void gatherSubclasses(
}
for (auto sub : strct->getSubclasses()) {
// this check is to prevent cycles.
- if(visited.count(sub.get())){
+ if (visited.count(sub.get())) {
continue;
}
res.push_back(sub);
@@ -381,7 +440,7 @@ NodeVector<Node> FieldDescriptor::pathTo(Handle<FieldDescriptor> field,
NodeVector<FieldDescriptor> FieldDescriptor::getDefaultFields() const
{
// TODO: In principle a cast would be nicer here, but for now we copy.
- NodeVector<Node> nodes = collect(this, [](Handle<Node> n) {
+ NodeVector<Node> nodes = collect(this, [](Handle<Node> n, size_t depth) {
if (!n->isa(&RttiTypes::FieldDescriptor)) {
return false;
}
@@ -396,6 +455,16 @@ NodeVector<FieldDescriptor> FieldDescriptor::getDefaultFields() const
return res;
}
+std::vector<SyntaxDescriptor> FieldDescriptor::getPermittedTokens() const
+{
+ if (getParent() == nullptr ||
+ getParent().cast<Descriptor>()->getParent() == nullptr) {
+ return std::vector<SyntaxDescriptor>();
+ }
+ return collectPermittedTokens(
+ this, getParent().cast<Descriptor>()->getParent().cast<Domain>());
+}
+
/* Class Descriptor */
void Descriptor::doResolve(ResolutionState &state)
@@ -443,6 +512,25 @@ bool Descriptor::doValidate(Logger &logger) const
}
valid = valid & attributesDescriptor->validate(logger);
}
+
+ // check start and end token.
+ if (!startToken.special && !startToken.token.empty() &&
+ !Utils::isUserDefinedToken(startToken.token)) {
+ logger.error(std::string("Descriptor \"") + getName() +
+ "\" has an invalid custom start token: " +
+ startToken.token,
+ *this);
+ valid = false;
+ }
+ if (!endToken.special && !endToken.token.empty() &&
+ !Utils::isUserDefinedToken(endToken.token)) {
+ logger.error(std::string("Descriptor \"") + getName() +
+ "\" has an invalid custom end token: " +
+ endToken.token,
+ *this);
+ valid = false;
+ }
+
// check that only one FieldDescriptor is of type TREE.
auto fds = Descriptor::getFieldDescriptors();
bool hasTREE = false;
@@ -483,7 +571,7 @@ std::pair<NodeVector<Node>, bool> Descriptor::pathTo(
NodeVector<FieldDescriptor> Descriptor::getDefaultFields() const
{
// TODO: In principle a cast would be nicer here, but for now we copy.
- NodeVector<Node> nodes = collect(this, [](Handle<Node> n) {
+ NodeVector<Node> nodes = collect(this, [](Handle<Node> n, size_t depth) {
if (!n->isa(&RttiTypes::FieldDescriptor)) {
return false;
}
@@ -501,7 +589,7 @@ NodeVector<FieldDescriptor> Descriptor::getDefaultFields() const
NodeVector<StructuredClass> Descriptor::getPermittedChildren() const
{
// TODO: In principle a cast would be nicer here, but for now we copy.
- NodeVector<Node> nodes = collect(this, [](Handle<Node> n) {
+ NodeVector<Node> nodes = collect(this, [](Handle<Node> n, size_t depth) {
return n->isa(&RttiTypes::StructuredClass);
});
NodeVector<StructuredClass> res;
@@ -669,6 +757,14 @@ std::pair<Rooted<FieldDescriptor>, bool> Descriptor::createFieldDescriptor(
return std::make_pair(fd, sorted);
}
+std::vector<SyntaxDescriptor> Descriptor::getPermittedTokens() const
+{
+ if (getParent() == nullptr) {
+ return std::vector<SyntaxDescriptor>();
+ }
+ return collectPermittedTokens(this, getParent().cast<Domain>());
+}
+
/* Class StructuredClass */
StructuredClass::StructuredClass(Manager &mgr, std::string name,
@@ -709,6 +805,16 @@ bool StructuredClass::doValidate(Logger &logger) const
logger.error(cardinality.toString() + " is not a cardinality!", *this);
valid = false;
}
+
+ // check short token.
+ if (!shortToken.special && !shortToken.token.empty() &&
+ !Utils::isUserDefinedToken(shortToken.token)) {
+ logger.error(std::string("Descriptor \"") + getName() +
+ "\" has an invalid custom short form token: " +
+ shortToken.token,
+ *this);
+ valid = false;
+ }
// check the validity of this superclass.
if (superclass != nullptr) {
valid = valid & superclass->validate(logger);
@@ -961,6 +1067,51 @@ Rooted<AnnotationClass> Ontology::createAnnotationClass(std::string name)
new AnnotationClass(getManager(), std::move(name), this)};
}
+static void gatherTokenDescriptors(
+ Handle<Descriptor> desc, std::vector<TokenDescriptor *> &res,
+ std::unordered_set<FieldDescriptor *> &visited)
+{
+ // add the TokenDescriptors for the Descriptor itself.
+ if (!desc->getStartToken().isEmpty()) {
+ res.push_back(desc->getStartTokenPointer());
+ }
+ if (!desc->getEndToken().isEmpty()) {
+ res.push_back(desc->getEndTokenPointer());
+ }
+ // add the TokenDescriptors for its FieldDescriptors.
+ for (auto fd : desc->getFieldDescriptors()) {
+ if (!visited.insert(fd.get()).second) {
+ continue;
+ }
+ if (!fd->getStartToken().isEmpty()) {
+ res.push_back(fd->getStartTokenPointer());
+ }
+ if (!fd->getEndToken().isEmpty()) {
+ res.push_back(fd->getEndTokenPointer());
+ }
+ }
+}
+
+std::vector<TokenDescriptor *> Domain::getAllTokenDescriptors() const
+{
+ std::vector<TokenDescriptor *> res;
+ // note all fields that are already visited because FieldReferences might
+ // lead to doubled fields.
+ std::unordered_set<FieldDescriptor *> visited;
+ // add the TokenDescriptors for the StructuredClasses (and their fields).
+ for (auto s : structuredClasses) {
+ if (!s->getShortToken().isEmpty()) {
+ res.push_back(s->getShortTokenPointer());
+ }
+ gatherTokenDescriptors(s, res, visited);
+ }
+ // add the TokenDescriptors for the AnnotationClasses (and their fields).
+ for (auto a : annotationClasses) {
+ gatherTokenDescriptors(a, res, visited);
+ }
+ return res;
+}
+
/* Type registrations */
namespace RttiTypes {
diff --git a/src/core/model/Ontology.hpp b/src/core/model/Ontology.hpp
index e1fbe96..d682bdf 100644
--- a/src/core/model/Ontology.hpp
+++ b/src/core/model/Ontology.hpp
@@ -168,11 +168,13 @@
#ifndef _OUSIA_MODEL_DOMAIN_HPP_
#define _OUSIA_MODEL_DOMAIN_HPP_
+#include <core/common/Whitespace.hpp>
#include <core/managed/ManagedContainer.hpp>
#include <core/RangeSet.hpp>
#include "Node.hpp"
#include "RootNode.hpp"
+#include "Syntax.hpp"
#include "Typesystem.hpp"
namespace ousia {
@@ -226,6 +228,9 @@ private:
Owned<Type> primitiveType;
bool optional;
bool primitive;
+ TokenDescriptor startToken;
+ TokenDescriptor endToken;
+ WhitespaceMode whitespaceMode;
protected:
bool doValidate(Logger &logger) const override;
@@ -234,39 +239,46 @@ public:
/**
* This is the constructor for primitive fields.
*
- * @param mgr is the global Manager instance.
- * @param parent is a handle of the Descriptor node that has this
- * FieldDescriptor.
- * @param primitiveType is a handle to some Type in some Typesystem of which
- * one instance is allowed to fill this field.
- * @param name is the name of this field.
- * @param optional should be set to 'false' is this field needs to be
- * filled in order for an instance of the parent
- * Descriptor to be valid.
+ * @param mgr is the global Manager instance.
+ * @param parent is a handle of the Descriptor node that has this
+ * FieldDescriptor.
+ * @param primitiveType is a handle to some Type in some Typesystem of
+ *which
+ * one instance is allowed to fill this field.
+ * @param name is the name of this field.
+ * @param optional should be set to 'false' is this field needs to be
+ * filled in order for an instance of the parent
+ * Descriptor to be valid.
+ * @param whitespaceMode the WhitespaceMode to be used when an instance of
+ * this FieldDescriptor is parsed.
*/
FieldDescriptor(Manager &mgr, Handle<Type> primitiveType,
Handle<Descriptor> parent,
FieldType fieldType = FieldType::TREE,
- std::string name = "", bool optional = false);
+ std::string name = "", bool optional = false,
+ WhitespaceMode whitespaceMode = WhitespaceMode::COLLAPSE);
/**
* This is the constructor for non-primitive fields. You have to provide
* children here later on.
*
- * @param mgr is the global Manager instance.
- * @param parent is a handle of the Descriptor node that has this
- * FieldDescriptor.
- * @param fieldType is the FieldType of this FieldDescriptor, either
- * TREE for the main or default structure or SUBTREE
- * for supporting structures.
- * @param name is the name of this field.
- * @param optional should be set to 'false' is this field needs to be
- * filled in order for an instance of the parent
- * Descriptor to be valid.
+ * @param mgr is the global Manager instance.
+ * @param parent is a handle of the Descriptor node that has this
+ * FieldDescriptor.
+ * @param fieldType is the FieldType of this FieldDescriptor, either
+ * TREE for the main or default structure or SUBTREE
+ * for supporting structures.
+ * @param name is the name of this field.
+ * @param optional should be set to 'false' is this field needs to be
+ * filled in order for an instance of the parent
+ * Descriptor to be valid.
+ * @param whitespaceMode the WhitespaceMode to be used when an instance of
+ * this FieldDescriptor is parsed.
*/
FieldDescriptor(Manager &mgr, Handle<Descriptor> parent = nullptr,
FieldType fieldType = FieldType::TREE,
- std::string name = "", bool optional = false);
+ std::string name = "", bool optional = false,
+ WhitespaceMode whitespaceMode = WhitespaceMode::COLLAPSE);
/**
* Returns a const reference to the NodeVector of StructuredClasses whose
@@ -455,6 +467,109 @@ public:
return std::move(name);
}
}
+
+ /**
+ * Returns a pointer to the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * Note that this does not invalidate the FieldDescriptor. So use with
+ * care.
+ *
+ * @return a pointer to the start TokenDescriptor.
+ */
+ TokenDescriptor *getStartTokenPointer() { return &startToken; }
+
+ /**
+ * Returns a copy of the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * @return a copy of the start TokenDescriptor.
+ */
+ TokenDescriptor getStartToken() const { return startToken; }
+
+ /**
+ * Sets the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * @param st the new start TokenDescriptor.
+ */
+ void setStartToken(TokenDescriptor st)
+ {
+ invalidate();
+ startToken = st;
+ }
+
+ /**
+ * Returns a pointer to the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @return a pointer to the end TokenDescriptor.
+ */
+ TokenDescriptor *getEndTokenPointer() { return &endToken; }
+
+ /**
+ * Returns a copy of the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @return a copy of the end TokenDescriptor.
+ */
+ TokenDescriptor getEndToken() const { return endToken; }
+
+ /**
+ * Sets the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @param e the new end TokenDescriptor.
+ */
+ void setEndToken(TokenDescriptor e)
+ {
+ invalidate();
+ endToken = e;
+ }
+
+ /**
+ * Returns the WhitespaceMode to be used when an instance of this
+ * FieldDescriptor is parsed.
+ *
+ * @return the WhitespaceMode to be used when an instance of this
+ * FieldDescriptor is parsed.
+ */
+ WhitespaceMode getWhitespaceMode() const { return whitespaceMode; }
+
+ /**
+ * Sets the WhitespaceMode to be used when an instance of this
+ * FieldDescriptor is parsed.
+ *
+ * @param wm the WhitespaceMode to be used when an instance of this
+ * FieldDescriptor is parsed.
+ */
+ WhitespaceMode setWhitespaceMode(WhitespaceMode wm)
+ {
+ return whitespaceMode = wm;
+ }
+
+ /**
+ * Returns the SyntaxDescriptor for this FieldDescriptor.
+ *
+ * @return the SyntaxDescriptor for this FieldDescriptor.
+ */
+ SyntaxDescriptor getSyntaxDescriptor(ssize_t depth = -1)
+ {
+ SyntaxDescriptor stx{startToken.id, endToken.id, Tokens::Empty,
+ const_cast<FieldDescriptor *>(this), depth};
+ return stx;
+ }
+
+ /**
+ * Returns a vector of SyntaxDescriptors, one for each Descriptor
+ * (StructuredClasses, AnnotationClasses or FieldDescriptors) that is
+ * permitted as child of this FieldDescriptor. This also makes use
+ * of transparency.
+ *
+ * @return a vector of SyntaxDescriptors, one for each Descriptor that is
+ * permitted as child of this FieldDescriptor
+ */
+ std::vector<SyntaxDescriptor> getPermittedTokens() const;
};
/**
@@ -478,7 +593,10 @@ public:
* </A>
* \endcode
*
- * key="value" inside the A-node would be an attribute, while <key>value</key>
+ * key="value" inside the A-node would be an attribute, while
+ * \code{.xml}
+ * <key>value</key>
+ * \endcode
* would be a primitive field. While equivalent in XML the semantics are
* different: An attribute describes indeed attributes, features of one single
* node whereas a primitive field describes the _content_ of a node.
@@ -490,6 +608,8 @@ class Descriptor : public Node {
private:
Owned<StructType> attributesDescriptor;
NodeVector<FieldDescriptor> fieldDescriptors;
+ TokenDescriptor startToken;
+ TokenDescriptor endToken;
bool addAndSortFieldDescriptor(Handle<FieldDescriptor> fd, Logger &logger);
@@ -738,6 +858,85 @@ public:
* of an instance of this Descriptor in the structure tree.
*/
NodeVector<StructuredClass> getPermittedChildren() const;
+
+ /**
+ * Returns a pointer to the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * @return a pointer to the start TokenDescriptor.
+ */
+ TokenDescriptor *getStartTokenPointer() { return &startToken; }
+
+ /**
+ * Returns a copy of the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * @return a copy of the start TokenDescriptor.
+ */
+ TokenDescriptor getStartToken() const { return startToken; }
+
+ /**
+ * Sets the start TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor starts.
+ *
+ * @param st the new start TokenDescriptor.
+ */
+ void setStartToken(TokenDescriptor st)
+ {
+ invalidate();
+ startToken = st;
+ }
+
+ /**
+ * Returns a pointer to the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @return a pointer to the end TokenDescriptor.
+ */
+ TokenDescriptor *getEndTokenPointer() { return &endToken; }
+
+ /**
+ * Returns a copy of the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @return a copy of the end TokenDescriptor.
+ */
+ TokenDescriptor getEndToken() const { return endToken; }
+
+ /**
+ * Sets the end TokenDescriptor. This Token is used as a
+ * signifier during parsing that an instance of this FieldDescriptor ends.
+ *
+ * @param e the new end TokenDescriptor.
+ */
+ void setEndToken(TokenDescriptor e)
+ {
+ invalidate();
+ endToken = e;
+ }
+
+ /**
+ * Returns the SyntaxDescriptor for this Descriptor.
+ *
+ * @return the SyntaxDescriptor for this Descriptor.
+ */
+ virtual SyntaxDescriptor getSyntaxDescriptor(ssize_t depth = -1)
+ {
+ SyntaxDescriptor stx{startToken.id, endToken.id, Tokens::Empty,
+ const_cast<Descriptor *>(this), depth};
+ return stx;
+ }
+
+ /**
+ * Returns a vector of SyntaxDescriptors, one for each Descriptor
+ * (StructuredClasses, AnnotationClasses or FieldDescriptors) that is
+ * permitted as child of this Descriptor. This also makes use
+ * of transparency.
+ *
+ * @return a vector of SyntaxDescriptors, one for each Descriptor that is
+ * permitted as child of this Descriptor.
+ */
+ std::vector<SyntaxDescriptor> getPermittedTokens() const;
};
/*
* TODO: We should discuss Cardinalities one more time. Is it smart to define
@@ -824,6 +1023,7 @@ private:
NodeVector<StructuredClass> subclasses;
bool transparent;
bool root;
+ TokenDescriptor shortToken;
/**
* Helper method for getFieldDescriptors.
@@ -981,6 +1181,50 @@ public:
invalidate();
root = std::move(r);
}
+
+ /**
+ * Returns a pointer to the short TokenDescriptor. During parsing an
+ * occurence of this token will be translated to an empty instance of this
+ * StructuredClass.
+ *
+ * @return a pointer to the short TokenDescriptor.
+ */
+ TokenDescriptor *getShortTokenPointer() { return &shortToken; }
+
+ /**
+ * Returns a copy of the short TokenDescriptor. During parsing an
+ * occurence of this token will be translated to an empty instance of this
+ * StructuredClass.
+ *
+ * @return a copy of the short TokenDescriptor.
+ */
+ TokenDescriptor getShortToken() const { return shortToken; }
+
+ /**
+ * Sets the short TokenDescriptor. During parsing an
+ * occurence of this token will be translated to an empty instance of this
+ * StructuredClass.
+ *
+ * @param s the new short TokenDescriptor.
+ */
+ void setShortToken(TokenDescriptor s)
+ {
+ invalidate();
+ shortToken = s;
+ }
+
+ /**
+ * Returns the SyntaxDescriptor for this StructuredClass.
+ *
+ * @return the SyntaxDescriptor for this StructuredClass.
+ */
+ SyntaxDescriptor getSyntaxDescriptor(ssize_t depth = -1) override
+ {
+ SyntaxDescriptor stx{getStartToken().id, getEndToken().id,
+ shortToken.id, const_cast<StructuredClass *>(this),
+ depth};
+ return stx;
+ }
};
/**
@@ -1207,6 +1451,13 @@ public:
{
ontologies.insert(ontologies.end(), ds.begin(), ds.end());
}
+
+ /**
+ * Returns all TokenDescriptors of classes and fields in this Ontology.
+ *
+ * @return all TokenDescriptors of classes and fields in this Ontology.
+ */
+ std::vector<TokenDescriptor *> getAllTokenDescriptors() const;
};
namespace RttiTypes {
@@ -1219,4 +1470,4 @@ extern const Rtti Ontology;
}
}
-#endif /* _OUSIA_MODEL_DOMAIN_HPP_ */ \ No newline at end of file
+#endif /* _OUSIA_MODEL_DOMAIN_HPP_ */
diff --git a/src/core/model/Syntax.cpp b/src/core/model/Syntax.cpp
new file mode 100644
index 0000000..9dbaccc
--- /dev/null
+++ b/src/core/model/Syntax.cpp
@@ -0,0 +1,58 @@
+/*
+ Ousía
+ Copyright (C) 2014, 2015 Benjamin Paaßen, Andreas Stöckel
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#include "Syntax.hpp"
+
+#include "Domain.hpp"
+
+namespace ousia {
+
+/* Class TokenSyntaxDescriptor */
+
+bool SyntaxDescriptor::isAnnotation() const
+{
+ return descriptor->isa(&RttiTypes::AnnotationClass);
+}
+bool SyntaxDescriptor::isFieldDescriptor() const
+{
+ return descriptor->isa(&RttiTypes::FieldDescriptor);
+}
+bool SyntaxDescriptor::isStruct() const
+{
+ return descriptor->isa(&RttiTypes::StructuredClass);
+}
+
+void SyntaxDescriptor::insertIntoTokenSet(TokenSet &set) const
+{
+ if (start != Tokens::Empty) {
+ set.insert(start);
+ }
+ if (end != Tokens::Empty) {
+ set.insert(end);
+ }
+ if (shortForm != Tokens::Empty) {
+ set.insert(shortForm);
+ }
+}
+
+bool SyntaxDescriptor::isEmpty() const
+{
+ return start == Tokens::Empty && end == Tokens::Empty &&
+ shortForm == Tokens::Empty;
+}
+} \ No newline at end of file
diff --git a/src/core/model/Syntax.hpp b/src/core/model/Syntax.hpp
new file mode 100644
index 0000000..4da3408
--- /dev/null
+++ b/src/core/model/Syntax.hpp
@@ -0,0 +1,196 @@
+/*
+ Ousía
+ Copyright (C) 2014, 2015 Benjamin Paaßen, Andreas Stöckel
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>.
+*/
+
+/**
+ * @file Syntax.hpp
+ *
+ * This header contains the Descriptor classes for user definable syntax for
+ * Document entities or fields. These classes are referenced in Ontology.hpp.
+ */
+
+#ifndef _OUSIA_MODEL_SYNTAX_HPP_
+#define _OUSIA_MODEL_SYNTAX_HPP_
+
+#include <core/common/Token.hpp>
+#include "Node.hpp"
+
+namespace ousia {
+
+/**
+ * Class to describe a single token that shall be used as user-defined syntax.
+ */
+struct TokenDescriptor {
+ /**
+ * The string content of this token, if it is not a special one.
+ */
+ std::string token;
+ /**
+ * A flag to be set true if this TokenDescriptor uses a special token.
+ */
+ bool special;
+ /**
+ * An id to uniquely identify this token.
+ */
+ TokenId id;
+
+ /**
+ * Constructor for non-special tokens. The special flag is set to false and
+ * the id to Tokens::Empty.
+ *
+ * @param token The string content of this token, if it is not a special
+ * one.
+ */
+ TokenDescriptor(std::string token = std::string())
+ : token(std::move(token)), special(false), id(Tokens::Empty)
+ {
+ }
+
+ /**
+ * Constructor for special tokens. The token is set to an empty string and
+ * the special flag to true.
+ *
+ * @param id the id of the special token.
+ */
+ TokenDescriptor(TokenId id) : special(true), id(id) {}
+
+ /**
+ * Returns true if and only if neither a string nor an ID is given.
+ *
+ * @return true if and only if neither a string nor an ID is given.
+ */
+ bool isEmpty() const { return token.empty() && id == Tokens::Empty; }
+};
+
+/**
+ * Class describing the user defined syntax for a StructuredClass,
+ * AnnotationClass or FieldDescriptor.
+ *
+ * This class is used during parsing of a Document. It is used to describe
+ * the tokens relevant for one Descriptor that could be created at this point
+ * during parsing.
+ */
+struct SyntaxDescriptor {
+ /**
+ * Possible start token or Tokens::Empty if no token is set.
+ */
+ TokenId start;
+
+ /**
+ * Possible end token or Tokens::Empty if no token is set.
+ */
+ TokenId end;
+
+ /**
+ * Possible representation token or Tokens::Empty if no token is set.
+ */
+ TokenId shortForm;
+
+ /*
+ * The Descriptor this SyntaxDescriptor belongs to. As this may be
+ * a FieldDescriptor as well as a class Descriptor (StructuredClass or
+ * AnnotationClass) we can only use the class Node as inner argument here.
+ */
+ Rooted<Node> descriptor;
+ /*
+ * Given the current leaf in the parsed document the depth of a
+ * SyntaxDescriptor is defined as the number of transparent elements that
+ * would be needed to construct an instance of the referenced descriptor.
+ */
+ ssize_t depth;
+
+ /**
+ * Default constructor, sets all token ids to Tokens::Empty and the
+ * descriptor handle to nullptr.
+ */
+ SyntaxDescriptor()
+ : start(Tokens::Empty),
+ end(Tokens::Empty),
+ shortForm(Tokens::Empty),
+ descriptor(nullptr),
+ depth(-1)
+ {
+ }
+
+ /**
+ * Member initializer constructor.
+ *
+ * @param start is a possible start token.
+ * @param end is a possible end token.
+ * @param shortForm is a possible short form token.
+ * @param descriptor The Descriptor this SyntaxDescriptor belongs to.
+ * @param depth Given the current leaf in the parsed document the depth of a
+ * SyntaxDescriptor is defined as the number of transparent elements that
+ * would be needed to construct an instance of the referenced descriptor.
+ */
+ SyntaxDescriptor(TokenId start, TokenId end, TokenId shortForm,
+ Handle<Node> descriptor, ssize_t depth)
+ : start(start),
+ end(end),
+ shortForm(shortForm),
+ descriptor(descriptor),
+ depth(depth)
+ {
+ }
+
+ /**
+ * Inserts all tokens referenced in this SyntaxDescriptor into the
+ * given TokenSet. Skips token ids set to Tokens::Empty.
+ *
+ * @param set is the TokenSet instance into which the Tokens should be
+ * inserted.
+ */
+ void insertIntoTokenSet(TokenSet &set) const;
+
+ /**
+ * Returns true if and only if this SyntaxDescriptor belongs to an
+ * AnnotationClass.
+ *
+ * @return true if and only if this SyntaxDescriptor belongs to an
+ * AnnotationClass.
+ */
+ bool isAnnotation() const;
+
+ /**
+ * Returns true if and only if this SyntaxDescriptor belongs to a
+ * StrcturedClass.
+ *
+ * @return true if and only if this SyntaxDescriptor belongs to a
+ * StrcturedClass.
+ */
+ bool isStruct() const;
+
+ /**
+ * Returns true if and only if this SyntaxDescriptor belongs to a
+ * FieldDescriptor.
+ *
+ * @return true if and only if this SyntaxDescriptor belongs to a
+ * FieldDescriptor.
+ */
+ bool isFieldDescriptor() const;
+
+ /**
+ * Returns true if and only if this SyntaxDescriptor has only empty
+ * entries in start, end and short.
+ *
+ * @return true if and only if this SyntaxDescriptor has only empty
+ * entries in start, end and short.
+ */
+ bool isEmpty() const;
+};
+}
+#endif \ No newline at end of file
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-30 17:58:58 +0000