xo-reader2: #q supports literal dictionaries

This commit is contained in:
Roland Conybeare 2026-03-17 21:04:08 -04:00
commit 25fd378c78
19 changed files with 919 additions and 57 deletions

View file

@ -88,6 +88,10 @@ set(SELF_SRCS
ISyntaxStateMachine_DExpectQArraySsm.cpp
IPrintable_DExpectQArraySsm.cpp
DExpectQDictSsm.cpp
facet/ISyntaxStateMachine_DExpectQDictSsm.cpp
facet/IPrintable_DExpectQDictSsm.cpp
DProgressSsm.cpp
ISyntaxStateMachine_DProgressSsm.cpp
IPrintable_DProgressSsm.cpp
@ -107,8 +111,6 @@ xo_dependency(${SELF_LIB} xo_type)
xo_dependency(${SELF_LIB} xo_tokenizer2)
xo_dependency(${SELF_LIB} xo_expression2)
#xo_dependency(${SELF_LIB} reflect)
#xo_dependency(${SELF_LIB} xo_object2)
#xo_dependency(${SELF_LIB} xo_printable2)
#xo_dependency(${SELF_LIB} xo_flatstring)
xo_dependency(${SELF_LIB} subsys)
#xo_dependency(${SELF_LIB} indentlog)

View file

@ -0,0 +1,270 @@
/** @file DExpectQDictSsm.cpp
*
* @author Roland Conybeare, Mar 2026
**/
#include "ExpectQDictSsm.hpp"
#include "ExpectQLiteralSsm.hpp"
namespace xo {
using xo::print::APrintable;
namespace scm {
const char *
QDictXst::_descr(enum code x)
{
switch (x) {
case code::invalid: break;
case code::qdict_0: return "qdict_0";
case code::qdict_1a: return "qdict_1a";
case code::qdict_1b: return "qdict_1b";
case code::qdict_1c: return "qdict_1c";
case code::qdict_1d: return "qdict_1d";
case code::qdict_2: return "qdict_2";
case code::N: break;
}
return "?QDictXst";
}
DExpectQDictSsm::DExpectQDictSsm() : state_{QDictXst::code::qdict_0} {}
obj<ASyntaxStateMachine,DExpectQDictSsm>
DExpectQDictSsm::make(DArena & parser_mm)
{
return obj<ASyntaxStateMachine,DExpectQDictSsm>(_make(parser_mm));
}
DExpectQDictSsm *
DExpectQDictSsm::_make(DArena & parser_mm)
{
void * mem = parser_mm.alloc_for<DExpectQDictSsm>();
return new (mem) DExpectQDictSsm();
}
void
DExpectQDictSsm::start(ParserStateMachine * p_psm)
{
DArena::Checkpoint ckp = p_psm->parser_alloc().checkpoint();
p_psm->push_ssm(ckp, DExpectQDictSsm::make(p_psm->parser_alloc()));
}
syntaxstatetype
DExpectQDictSsm::ssm_type() const noexcept {
return syntaxstatetype::expect_qdict;
}
std::string_view
DExpectQDictSsm::get_expect_str() const {
switch (state_.code()) {
case QDictXst::code::qdict_0:
return "leftbrace";
case QDictXst::code::qdict_1a:
return "symbol|rightbrace";
case QDictXst::code::qdict_1b:
return "colon";
case QDictXst::code::qdict_1c:
return "literal";
case QDictXst::code::qdict_1d:
return "semicolon|rightbrace";
case QDictXst::code::qdict_2:
return "(done)";
case QDictXst::code::invalid:
case QDictXst::code::N:
break;
}
return "?DExpectQDictSsm";
}
void
DExpectQDictSsm::on_leftbrace_token(const Token & tk,
ParserStateMachine * p_psm)
{
if (state_.code() == QDictXst::code::qdict_0) {
constexpr DDictionary::size_type hint_cap = 8;
this->state_ = QDictXst(QDictXst::code::qdict_1a);
this->dict_ = DDictionary::empty(p_psm->expr_alloc(), hint_cap);
return;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_rightbrace_token(const Token & tk,
ParserStateMachine * p_psm)
{
if ((state_.code() == QDictXst::code::qdict_1a)
|| (state_.code() == QDictXst::code::qdict_1d))
{
this->state_ = QDictXst(QDictXst::code::qdict_2);
obj<AGCObject> lit = obj<AGCObject,DDictionary>(dict_);
p_psm->pop_ssm();
p_psm->on_quoted_literal(lit);
return;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_symbol_token(const Token & tk,
ParserStateMachine * p_psm)
{
if (state_.code() == QDictXst::code::qdict_1a) {
this->state_ = QDictXst(QDictXst::code::qdict_1b);
this->key_ = DString::from_view(p_psm->expr_alloc(), std::string_view(tk.text()));
return;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_colon_token(const Token & tk,
ParserStateMachine * p_psm)
{
if (state_.code() == QDictXst::code::qdict_1b) {
this->state_ = QDictXst(QDictXst::code::qdict_1c);
DExpectQLiteralSsm::start(p_psm,
false /*!cxl_on_rightparen*/);
return;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_semicolon_token(const Token & tk,
ParserStateMachine * p_psm)
{
if (state_.code() == QDictXst::code::qdict_1d) {
this->state_ = QDictXst(QDictXst::code::qdict_1a);
return;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_token(const Token & tk,
ParserStateMachine * p_psm)
{
switch(tk.tk_type())
{
case tokentype::tk_leftbrace:
this->on_leftbrace_token(tk, p_psm);
return;
case tokentype::tk_symbol:
this->on_symbol_token(tk, p_psm);
return;
case tokentype::tk_rightbrace:
this->on_rightbrace_token(tk, p_psm);
return;
case tokentype::tk_colon:
this->on_colon_token(tk, p_psm);
return;
case tokentype::tk_semicolon:
this->on_semicolon_token(tk, p_psm);
return;
case tokentype::tk_rightparen:
case tokentype::tk_comma:
case tokentype::tk_lambda:
case tokentype::tk_def:
case tokentype::tk_deftype:
case tokentype::tk_if:
case tokentype::tk_singleassign:
case tokentype::tk_string:
case tokentype::tk_f64:
case tokentype::tk_i64:
case tokentype::tk_bool:
case tokentype::tk_invalid:
case tokentype::tk_quote:
case tokentype::tk_leftparen:
case tokentype::tk_leftbracket:
case tokentype::tk_rightbracket:
case tokentype::tk_leftangle:
case tokentype::tk_rightangle:
case tokentype::tk_cmple:
case tokentype::tk_cmpge:
case tokentype::tk_dot:
case tokentype::tk_doublecolon:
case tokentype::tk_assign:
case tokentype::tk_yields:
case tokentype::tk_plus:
case tokentype::tk_minus:
case tokentype::tk_star:
case tokentype::tk_slash:
case tokentype::tk_cmpeq:
case tokentype::tk_cmpne:
case tokentype::tk_nil:
case tokentype::tk_type:
case tokentype::tk_then:
case tokentype::tk_else:
case tokentype::tk_let:
case tokentype::tk_in:
case tokentype::tk_end:
case tokentype::N:
break;
}
Super::illegal_token(tk, p_psm);
}
void
DExpectQDictSsm::on_quoted_literal(obj<AGCObject> lit,
ParserStateMachine * p_psm)
{
if (state_.code() == QDictXst::code::qdict_1c) {
// adjoin (key,value) pair into dictionary
this->state_ = QDictXst(QDictXst::code::qdict_1d);
assert(dict_);
bool ok = dict_->upsert(p_psm->expr_alloc(), DDictionary::pair_type(key_, lit));
this->key_ = nullptr;
assert(ok);
return;
}
Super::illegal_quoted_literal(lit, p_psm);
}
bool
DExpectQDictSsm::pretty(const ppindentinfo & ppii) const
{
obj<AGCObject,DDictionary> dict(dict_);
obj<APrintable,DDictionary> dict_pr(dict_);
return ppii.pps()->pretty_struct(ppii,
"DExpectQDictSsm",
refrtag("state", state_),
refrtag("expect", this->get_expect_str()),
refrtag("key", key_, key_),
refrtag("dict", dict_pr));
}
} /*namespace scm*/
} /*namespace xo*/
/* end DExpectQDictSsm.cpp */

View file

@ -132,6 +132,8 @@ namespace xo {
case tokentype::N:
break;
}
Super::illegal_token(tk, p_psm);
}
void

View file

@ -6,14 +6,10 @@
#include "ExpectQLiteralSsm.hpp"
#include "ExpectQListSsm.hpp"
#include "ExpectQArraySsm.hpp"
#include "ExpectQDictSsm.hpp"
#include <xo/object2/Float.hpp>
#include <xo/object2/Integer.hpp>
//#include "ssm/ISyntaxStateMachine_DExpectFormalArgSsm.hpp"
//#include <xo/expression2/DVariable.hpp>
//#include <xo/expression2/detail/IGCObject_DVariable.hpp>
//#include <xo/printable2/Printable.hpp>
//#include <xo/alloc2/arena/IAllocator_DArena.hpp>
//#include <xo/facet/FacetRegistry.hpp>
#include <xo/stringtable2/String.hpp>
#include <xo/indentlog/scope.hpp>
namespace xo {
@ -95,6 +91,10 @@ namespace xo {
this->on_i64_token(tk, p_psm);
return;
case tokentype::tk_string:
this->on_string_token(tk, p_psm);
return;
case tokentype::tk_leftparen:
this->on_leftparen_token(tk, p_psm);
return;
@ -111,6 +111,10 @@ namespace xo {
this->on_rightbracket_token(tk, p_psm);
return;
case tokentype::tk_leftbrace:
this->on_leftbrace_token(tk, p_psm);
return;
case tokentype::tk_comma:
case tokentype::tk_lambda:
case tokentype::tk_def:
@ -119,12 +123,10 @@ namespace xo {
case tokentype::tk_symbol:
case tokentype::tk_colon:
case tokentype::tk_singleassign:
case tokentype::tk_string:
case tokentype::tk_bool:
case tokentype::tk_semicolon:
case tokentype::tk_invalid:
case tokentype::tk_quote:
case tokentype::tk_leftbrace:
case tokentype::tk_rightbrace:
case tokentype::tk_leftangle:
case tokentype::tk_rightangle:
@ -176,49 +178,16 @@ namespace xo {
p_psm->on_quoted_literal(literal);
}
#ifdef NOT_YET
void
DExpectQLiteralSsm::_accept_formal(obj<AAllocator> expr_alloc,
DArena & parser_alloc,
const DUniqueString * param_name,
TypeDescr param_type)
DExpectQLiteralSsm::on_string_token(const Token & tk,
ParserStateMachine * p_psm)
{
/* note: param_type can be nullptr */
TypeRef typeref
= TypeRef::dwim(TypeRef::prefix_type::from_chars("formal"), param_type);
auto literal = obj<AGCObject,DString>(DString::from_view(p_psm->expr_alloc(),
std::string_view(tk.text())));
DVariable * var = DVariable::make(expr_alloc,
param_name,
typeref);
// need AGCObject facet to use DArray here.
// May want to have gc feature that allows it to use
// FacetRegistry on memory that stores obj<AExpression,..>
//
// In this case doesn't matter since DExpectQLiteralSsm not actually collected!
obj<AGCObject,DVariable> var_o(var);
if (argl_->size() == argl_->capacity()) {
// need to expand argl_ capacity.
// If DArena were to allow it (i.e. offer a realloc() feature,
// could do this in place since this SSM is at the top of the parser stack.
obj<AAllocator,DArena> mm(&parser_alloc);
DArray * argl_2x = DArray::empty(mm, 2 * argl_->capacity());
for (DArray::size_type i = 0, n = argl_->size(); i < n; ++i) {
// TODO: prefer non-bounds-checked access here
argl_2x->push_back(argl_->at(i));
}
// update in place
this->argl_ = argl_2x;
}
this->argl_->push_back(var_o);
p_psm->pop_ssm();
p_psm->on_quoted_literal(literal);
}
#endif
void
DExpectQLiteralSsm::on_leftparen_token(const Token & tk,
@ -268,6 +237,17 @@ namespace xo {
Super::illegal_token(tk, p_psm);
}
void
DExpectQLiteralSsm::on_leftbrace_token(const Token & tk,
ParserStateMachine * p_psm)
{
// replace self with specialized version for parsing a literal dict
p_psm->pop_ssm();
DExpectQDictSsm::start(p_psm);
p_psm->on_token(tk);
}
bool
DExpectQLiteralSsm::pretty(const ppindentinfo & ppii) const
{

View file

@ -25,6 +25,7 @@
#include "ExpectExprSsm.hpp"
#include "ExpectQLiteralSsm.hpp"
#include "ExpectQListSsm.hpp"
#include "ExpectQDictSsm.hpp"
#include "ExpectQArraySsm.hpp"
#include <xo/printable2/detail/APrintable.hpp>
@ -107,6 +108,9 @@ namespace xo {
FacetRegistry::register_impl<ASyntaxStateMachine, DExpectQListSsm>();
FacetRegistry::register_impl<APrintable, DExpectQListSsm>();
FacetRegistry::register_impl<ASyntaxStateMachine, DExpectQDictSsm>();
FacetRegistry::register_impl<APrintable, DExpectQDictSsm>();
FacetRegistry::register_impl<ASyntaxStateMachine, DExpectQArraySsm>();
FacetRegistry::register_impl<APrintable, DExpectQArraySsm>();
@ -128,6 +132,7 @@ namespace xo {
log && log(xtag("DExpectExprSsm.tseq", typeseq::id<DExpectExprSsm>()));
log && log(xtag("DExpectQLiteralSsm.tseq", typeseq::id<DExpectQLiteralSsm>()));
log && log(xtag("DExpectQListSsm.tseq", typeseq::id<DExpectQListSsm>()));
log && log(xtag("DExpectQDictSsm.tseq", typeseq::id<DExpectQDictSsm>()));
log && log(xtag("DExpectQArraySsm.tseq", typeseq::id<DExpectQArraySsm>()));
log && log(xtag("DProgressSsm.tseq", typeseq::id<DProgressSsm>()));

View file

@ -0,0 +1,28 @@
/** @file IPrintable_DExpectQDictSsm.cpp
*
* Generated automagically from ingredients:
* 1. code generator:
* [xo-facet/codegen/genfacet]
* arguments:
* --input [idl/IPrintable_DExpectQDictSsm.json5]
* 2. jinja2 template for abstract facet .hpp file:
* [iface_facet_any.hpp.j2]
* 3. idl for facet methods
* [idl/IPrintable_DExpectQDictSsm.json5]
**/
#include "expect_qdict/IPrintable_DExpectQDictSsm.hpp"
namespace xo {
namespace scm {
auto
IPrintable_DExpectQDictSsm::pretty(const DExpectQDictSsm & self, const ppindentinfo & ppii) -> bool
{
return self.pretty(ppii);
}
} /*namespace scm*/
} /*namespace xo*/
/* end IPrintable_DExpectQDictSsm.cpp */

View file

@ -0,0 +1,84 @@
/** @file ISyntaxStateMachine_DExpectQDictSsm.cpp
*
* Generated automagically from ingredients:
* 1. code generator:
* [xo-facet/codegen/genfacet]
* arguments:
* --input [idl/ISyntaxStateMachine_DExpectQDictSsm.json5]
* 2. jinja2 template for abstract facet .hpp file:
* [iface_facet_any.hpp.j2]
* 3. idl for facet methods
* [idl/ISyntaxStateMachine_DExpectQDictSsm.json5]
**/
#include "expect_qdict/ISyntaxStateMachine_DExpectQDictSsm.hpp"
namespace xo {
namespace scm {
auto
ISyntaxStateMachine_DExpectQDictSsm::ssm_type(const DExpectQDictSsm & self) noexcept -> syntaxstatetype
{
return self.ssm_type();
}
auto
ISyntaxStateMachine_DExpectQDictSsm::get_expect_str(const DExpectQDictSsm & self) noexcept -> std::string_view
{
return self.get_expect_str();
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_token(DExpectQDictSsm & self, const Token & tk, ParserStateMachine * p_psm) -> void
{
self.on_token(tk, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_symbol(DExpectQDictSsm & self, std::string_view sym, ParserStateMachine * p_psm) -> void
{
self.on_parsed_symbol(sym, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_typedescr(DExpectQDictSsm & self, TypeDescr td, ParserStateMachine * p_psm) -> void
{
self.on_parsed_typedescr(td, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_type(DExpectQDictSsm & self, obj<AType> type, ParserStateMachine * p_psm) -> void
{
self.on_parsed_type(type, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_formal(DExpectQDictSsm & self, const DUniqueString * param_name, TypeDescr param_type, ParserStateMachine * p_psm) -> void
{
self.on_parsed_formal(param_name, param_type, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_formal_with_token(DExpectQDictSsm & self, const DUniqueString * param_name, TypeDescr param_type, const Token & tk, ParserStateMachine * p_psm) -> void
{
self.on_parsed_formal_with_token(param_name, param_type, tk, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_formal_arglist(DExpectQDictSsm & self, DArray * arglist, ParserStateMachine * p_psm) -> void
{
self.on_parsed_formal_arglist(arglist, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_expression(DExpectQDictSsm & self, obj<AExpression> expr, ParserStateMachine * p_psm) -> void
{
self.on_parsed_expression(expr, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_parsed_expression_with_token(DExpectQDictSsm & self, obj<AExpression> expr, const Token & tk, ParserStateMachine * p_psm) -> void
{
self.on_parsed_expression_with_token(expr, tk, p_psm);
}
auto
ISyntaxStateMachine_DExpectQDictSsm::on_quoted_literal(DExpectQDictSsm & self, obj<AGCObject> lit, ParserStateMachine * p_psm) -> void
{
self.on_quoted_literal(lit, p_psm);
}
} /*namespace scm*/
} /*namespace xo*/
/* end ISyntaxStateMachine_DExpectQDictSsm.cpp */

View file

@ -51,6 +51,8 @@ namespace xo {
return "expect-qlist";
case syntaxstatetype::expect_qarray:
return "expect-qarray";
case syntaxstatetype::expect_qdict:
return "expect-qdict";
case syntaxstatetype::N:
break;
}