tame/tamer/src/asg/graph/object/meta.rs

163 lines
4.8 KiB
Rust
Raw Normal View History

tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
// Metasyntactic variables represented on the ASG
//
// Copyright (C) 2014-2023 Ryan Specialty, LLC.
//
// This file is part of TAME.
//
// 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/>.
//! Metasyntactic variables on the ASG.
//!
//! Metasyntactic variables
//! (sometimes called "metavariables" herein for short)
//! have historically been a feature of the template system.
//! The canonical metavariable is the template parameter.
use super::{prelude::*, Ident, Tpl};
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
use crate::{
diagnose::Annotate,
diagnostic_todo,
f::Functor,
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
fmt::{DisplayWrapper, TtQuote},
parse::util::SPair,
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
span::Span,
};
use std::fmt::Display;
/// Metasyntactic variable (metavariable).
///
/// A metavariable is a lexical construct.
/// Its value is a lexeme that represents an [`Ident`],
/// whose meaning depends on the context in which the metavariable is
/// referenced.
/// Its lexeme may be composed of multiple [`Self::Lexeme`]s,
/// and may even be constructed dynamically based on the values of other
/// [`Meta`]s.
///
/// Metavariables are identified by being bound by an [`Ident`];
/// the symbol representing that identifier then acts as a metavariable.
#[derive(Debug, PartialEq, Eq)]
pub enum Meta {
Required(Span),
ConcatList(Span),
Lexeme(Span, SPair),
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
}
impl Meta {
/// Create a new metavariable without a value.
///
/// Metavariables with no value cannot be used in an expansion context.
/// Intuitively,
/// they act as required parameters.
pub fn new_required(span: Span) -> Self {
Self::Required(span)
}
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
pub fn span(&self) -> Span {
match self {
Self::Required(span)
| Self::ConcatList(span)
| Self::Lexeme(span, _) => *span,
}
}
/// Assign a lexeme to a metavariable.
///
/// In a template definition context,
/// this acts as a default value for this metavariable.
/// In an application context,
/// this has the effect of binding a value to this metavariable.
pub fn assign_lexeme(self, lexeme: SPair) -> Self {
match self {
Self::Required(span) => Self::Lexeme(span, lexeme),
Self::ConcatList(_) => diagnostic_todo!(
vec![lexeme.note("while parsing this lexeme")],
"append to ConcatList",
),
Self::Lexeme(_, _) => diagnostic_todo!(
vec![lexeme.note("while parsing this lexeme")],
"Lexeme => ConcatList",
),
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
}
}
}
impl From<&Meta> for Span {
fn from(meta: &Meta) -> Self {
meta.span()
}
}
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
impl Display for Meta {
fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
match self {
Self::Required(_) => {
write!(f, "metasyntactic parameter with required value")
}
Self::ConcatList(_) => {
write!(f, "metasyntactic concatenation list")
}
Self::Lexeme(_, spair) => {
write!(f, "lexeme {}", TtQuote::wrap(spair))
}
}
}
}
impl Functor<Span> for Meta {
fn map(self, f: impl FnOnce(Span) -> Span) -> Self::Target {
match self {
Self::Required(span) => Self::Required(f(span)),
Self::ConcatList(span) => Self::ConcatList(f(span)),
Self::Lexeme(span, spair) => Self::Lexeme(f(span), spair),
tamer: src::asg: Scaffolding for metasyntactic variables Also known as metavariables or template parameters. This is a bit of a tortured excursion, trying to figure out how I want to best represent this. I have a number of pages of hand-written notes that I'd like to distill over time, but the rendered graph ontology (via `asg-ontviz`) demonstrates the broad idea. `AirTpl::TplApply` highlights some remaining questions. What I had _wanted_ to do is to separate the concepts of application and expansion, and support partial application and such. But it's going to be too much work for now, when it isn't needed---partial application can be worked around by simply creating new templates and duplicating params, as we do today, although that sucks and is a maintenance issue. But I'd rather address that head-on in the future. So it's looking like Option B is going to be the approach for now, with templates being closed (as in, no free metavariables) and expanded at the same time. This simplifies the parser and error conditions significantly and makes it easier to utilize anonymous templates, since it'll still be the active context. My intent is to get at least the graph construction sorted out---not the actual expansion and binding yet---enough that I can use templates to represent parts of NIR that do not have proper graph representations or desugaring yet, so that I can spit them back out again in the `xmli` file and incrementally handle them. That was an option I had considered some months ago, but didn't want to entertain it at the time because I wasn't sure what doing so would look like; while it was an attractive approach since it pushes existing primitives into the template system (something I've wanted to do for years), I didn't want to potentially tank performance or compromise the design for it after I had spent so much effort on all of this so far. But my efforts have yielded a system that significantly exceeds my initial performance expectations, with a decent abstractions, and so this seems viable. DEV-13708
2023-03-15 11:49:13 -04:00
}
}
}
object_rel! {
/// Templates may expand into nearly any context,
/// and must therefore be able to contain just about anything.
Meta -> {
tree Meta,
cross Ident,
}
}
impl ObjectIndex<Meta> {
pub fn identify_as_tpl_param(
&self,
asg: &mut Asg,
oi_tpl: ObjectIndex<Tpl>,
name: SPair,
) -> ObjectIndex<Ident> {
oi_tpl
.declare_local(asg, name)
.add_edge_to(asg, *self, None)
}
pub fn assign_lexeme(self, asg: &mut Asg, lexeme: SPair) -> Self {
self.map_obj(asg, |meta| meta.assign_lexeme(lexeme))
}
pub fn close(self, asg: &mut Asg, close_span: Span) -> Self {
self.map_obj(asg, |meta| {
meta.map(|open_span| {
open_span.merge(close_span).unwrap_or(open_span)
})
})
}
}