Main methods to handle ANTLR ParseTree nodesΒΆ

ANTLR generates a visitor method for each rule in the grammar. The visitor method looks like:

std::any visitRULE(AslParser::RULEContext *ctx)

where RULE is the name of the rule.

The parameter of each visitor method is a pointer ctx to an object of class RULEContext, which is derived from ANTLR class ParseRuleContext.

The pointer ctx allows access to information in the node currently being visited, and to its children.

For instance, if our grammar contains the rule:

expr : expr (MUL|DIV) expr

then our visitor will have a method:

std::any visitExpr(AslParser::ExprContext *ctx)

Inside this method we can, for instance:

  • Check whether the node has a MUL child:

    if (ctx->MUL()) ...

  • Get the string that matched the MUL regex:

    ctx->MUL()->getText()

    (not very useful for MUL, but very useful e.g. for ID or INTVAL)

  • Check whether the rule has some expr child:

    if (ctx->expr())...

  • Get how many expr children the rule has:

    ctx->expr().size()

    Note: Only when more than one expr are expected in the rule/subrule, ctx->expr() is a std::vector<ParseRuleContext>. Otherwise, ctx->expr() is the only expr child.

  • Get the i-th expr child (counting from 0):

    ctx->expr(i)

  • Get the total number of children of any kind:

    ctx->children.size()

  • Get the i-th child of any kind:

    ctx->children[i]

  • Both ctx->expr() and ctx->children can be iterated if needed:

    for (int i=0; i<ctx->expr().size(); ++i) visit(ctx->expr(i));
for (auto ex : ctx->expr()) visit(ex);

for (int i=0; i<ctx->children.size(); ++i) visit(ctx->children[i]);
for (auto ch : ctx->children) visit(ch);

    The last two iterations are equivalent to a simple call to visitChildren(ctx).