tame/tamer/src/xir/flat.rs

486 lines
15 KiB
Rust
Raw Normal View History

// XIR flat (XIRF)
//
// Copyright (C) 2014-2022 Ryan Specialty Group, 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/>.
//! Lightly-parsed XIR as a flat stream (XIRF).
//!
//! XIRF lightly parses a raw XIR [`TokenStream`] into a stream of
//! [`XirfToken`]s that are,
//! like a [`TokenStream`],
//! flat in structure.
//! It provides the following features over raw XIR:
//!
//! 1. All closing tags must correspond to a matching opening tag at the
//! same depth;
//! 2. [`XirfToken`] exposes the [`Depth`] of each opening/closing tag;
//! 3. Attribute tokens are parsed into [`Attr`] objects;
//! 4. Documents must begin with an element and end with the closing of
//! that element;
//! 5. Parsing will fail if input ends before all elements have been
//! closed.
//!
//! XIRF lowering does not perform any dynamic memory allocation;
//! maximum element nesting depth is set statically depending on the needs
//! of the caller.
use super::{
attr::{Attr, AttrParseError, AttrParseState},
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
CloseSpan, OpenSpan, QName, Token as XirToken, TokenStream, Whitespace,
};
use crate::{
diagnose::{Annotate, AnnotatedSpan, Diagnostic},
parse::{
tamer: xir::parse: Attribute parser generator This is the first parser generator for the parsing framework. I've been waiting quite a while to do this because I wanted to be sure that I understood how I intended to write the attribute parsers manually. Now that I'm about to start parsing source XML files, it is necessary to have a parser generator. Typically one thinks of a parser generator as a separate program that generates code for some language, but that is not always the case---that represents a lack of expressiveness in the language itself (e.g. C). Here, I simply use Rust's macro system, which should be a concept familiar to someone coming from a language like Lisp. This also resolves where I stand on parser combinators with respect to this abstraction: they both accomplish the exact same thing (composition of smaller parsers), but this abstraction doesn't do so in the typical functional way. But the end result is the same. The parser generated by this abstraction will be optimized an inlined in the same manner as the hand-written parsers. Since they'll be tightly coupled with an element parser (which too will have a parser generator), I expect that most attribute parsers will simply be inlined; they exist as separate parsers conceptually, for the same reason that you'd use parser combinators. It's worth mentioning that this awkward reliance on dead state for a lookahead token to determine when aggregation is complete rubs me the wrong way, but resolving it would involve reintroducing the XIR AttrEnd that I had previously removed. I'll keep fighting with myself on this, but I want to get a bit further before I determine if it's worth the tradeoff of reintroducing (more complex IR but simplified parsing). DEV-7145
2022-06-13 11:17:21 -04:00
Context, Object, ParseState, ParsedResult, Token, Transition,
TransitionResult,
},
span::Span,
sym::SymbolId,
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
xir::EleSpan,
};
use arrayvec::ArrayVec;
use std::{error::Error, fmt::Display};
/// Tag nesting depth
/// (`0` represents the root).
#[derive(Debug, Clone, PartialEq, Eq)]
tamer: obj::xmlo::reader: Begin conversion to ParseState This begins to transition XmloReader into a ParseState. Unlike previous changes where ParseStates were composed into a single ParseState, this is instead a lowering operation that will take the output of one Parser and provide it to another. The mess in ld::poc (...which still needs to be refactored and removed) shows the concept, which will be abstracted away. This won't actually get to the ASG in order to test that that this works with the wip-xmlo-xir-reader flag on (development hasn't gotten that far yet), but since it type-checks, it should conceptually work. Wiring lowering operations together is something that I've been dreading for months, but my approach of only abstracting after-the-fact has helped to guide a sane approach for this. For some definition of "sane". It's also worth noting that AsgBuilder will too become a ParseState implemented as another lowering operation, so: XIR -> XIRF -> XMLO -> ASG These steps will all be streaming, with iteration happening only at the topmost level. For this reason, it's important that ASG not be responsible for doing that pull, and further we should propagate Parsed::Incomplete rather than filtering it out and looping an indeterminate number of times outside of the toplevel. One final note: the choice of 64 for the maximum depth is entirely arbitrary and should be more than generous; it'll be finalized at some point in the future once I actually evaluate what maximum depth is reasonable based on how the system is used, with some added growing room. DEV-10863
2022-03-22 13:56:43 -04:00
pub struct Depth(pub usize);
impl Display for Depth {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
Display::fmt(&self.0, f)
}
}
/// A lightly-parsed XIRF object.
///
/// Certain XIR [`Token`]s are formed into a single object,
/// such as an [`Attr`].
/// Other objects retain the same format as their underlying token,
/// but are still validated to ensure that they are well-formed and that
/// the XML is well-structured.
#[derive(Debug, Clone, PartialEq, Eq)]
pub enum XirfToken {
/// Opening tag of an element.
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
Open(QName, OpenSpan, Depth),
/// Closing tag of an element.
///
/// If the name is [`None`],
/// then the tag is self-closing.
/// If the name is [`Some`],
/// then the tag is guaranteed to be balanced
/// (matching the depth of its opening tag).
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
Close(Option<QName>, CloseSpan, Depth),
/// An attribute and its value.
///
/// The associated [`Span`]s can be found on the enclosed [`Attr`]
/// object.
Attr(Attr),
/// Comment node.
Comment(SymbolId, Span),
/// Character data as part of an element.
///
/// See also [`CData`](XirfToken::CData) variant.
Text(SymbolId, Span),
/// CData node (`<![CDATA[...]]>`).
///
/// _Warning: It is up to the caller to ensure that the string `]]>` is
/// not present in the text!_
/// This is intended for reading existing XML data where CData is
/// already present,
/// not for producing new CData safely!
CData(SymbolId, Span),
/// Similar to `Text`,
/// but intended for use where only whitespace is allowed,
/// such as alignment of attributes.
Whitespace(Whitespace, Span),
}
impl Token for XirfToken {
fn span(&self) -> Span {
use XirfToken::*;
match self {
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
Open(_, OpenSpan(span, _), _)
| Close(_, CloseSpan(span, _), _)
| Comment(_, span)
| Text(_, span)
| CData(_, span)
| Whitespace(_, span) => *span,
Attr(attr) => attr.span(),
}
}
}
tamer: xir::parse: Attribute parser generator This is the first parser generator for the parsing framework. I've been waiting quite a while to do this because I wanted to be sure that I understood how I intended to write the attribute parsers manually. Now that I'm about to start parsing source XML files, it is necessary to have a parser generator. Typically one thinks of a parser generator as a separate program that generates code for some language, but that is not always the case---that represents a lack of expressiveness in the language itself (e.g. C). Here, I simply use Rust's macro system, which should be a concept familiar to someone coming from a language like Lisp. This also resolves where I stand on parser combinators with respect to this abstraction: they both accomplish the exact same thing (composition of smaller parsers), but this abstraction doesn't do so in the typical functional way. But the end result is the same. The parser generated by this abstraction will be optimized an inlined in the same manner as the hand-written parsers. Since they'll be tightly coupled with an element parser (which too will have a parser generator), I expect that most attribute parsers will simply be inlined; they exist as separate parsers conceptually, for the same reason that you'd use parser combinators. It's worth mentioning that this awkward reliance on dead state for a lookahead token to determine when aggregation is complete rubs me the wrong way, but resolving it would involve reintroducing the XIR AttrEnd that I had previously removed. I'll keep fighting with myself on this, but I want to get a bit further before I determine if it's worth the tradeoff of reintroducing (more complex IR but simplified parsing). DEV-7145
2022-06-13 11:17:21 -04:00
impl Object for XirfToken {}
impl Display for XirfToken {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
use XirfToken::*;
match self {
Open(qname, span, _) => {
Display::fmt(&XirToken::Open(*qname, *span), f)
}
Close(oqname, span, _) => {
Display::fmt(&XirToken::Close(*oqname, *span), f)
}
Attr(attr) => Display::fmt(&attr, f),
Comment(sym, span) => {
Display::fmt(&XirToken::Comment(*sym, *span), f)
}
Text(sym, span) => Display::fmt(&XirToken::Text(*sym, *span), f),
CData(sym, span) => Display::fmt(&XirToken::CData(*sym, *span), f),
Whitespace(ws, span) => {
Display::fmt(&XirToken::Whitespace(*ws, *span), f)
}
}
}
}
impl From<Attr> for XirfToken {
fn from(attr: Attr) -> Self {
Self::Attr(attr)
}
}
/// XIRF-compatible attribute parser.
pub trait FlatAttrParseState<const MAX_DEPTH: usize> =
tamer: Replace ParseStatus::Dead with generic lookahead Oh what a tortured journey. I had originally tried to avoid formalizing lookahead for all parsers by pretending that it was only needed for dead state transitions (that is---states that have no transitions for a given input token), but then I needed to yield information for aggregation. So I added the ability to override the token for `Dead` to yield that, in addition to the token. But then I also needed to yield lookahead for error conditions. It was a mess that didn't make sense. This eliminates `ParseStatus::Dead` entirely and fully integrates the lookahead token in `Parser` that was previously implemented. Notably, the lookahead token is encapsulated in `TransitionResult` and unavailable to `ParseState` implementations, forcing them to rely on `Parser` for recursion. This not only prevents `ParseState` from recursing, but also simplifies delegation by removing the need to manually handle tokens of lookahead. The awkward case here is XIRT, which does not follow the streaming parsing convention, because it was conceived before the parsing framework. It needs to go away, but doing so right now would be a lot of work, so it has to stick around for a little bit longer until the new parser generators can be used instead. It is a persistent thorn in my side, going against the grain. `Parser` will immediately recurse if it sees a token of lookahead with an incomplete parse. This is because stitched parsers will frequently yield a dead state indication when they're done parsing, and there's no use in propagating an `Incomplete` status down the entire lowering pipeline. But, that does mean that the toplevel is not the only thing recursing. _But_, the behavior doesn't really change, in the sense that it would infinitely recurse down the entire lowering stack (though there'd be an opportunity to detect that). This should never happen with a correct parser, but it's not worth the effort right now to try to force such a thing with Rust's type system. Something like TLA+ is better suited here as an aid, but it shouldn't be necessary with clear implementations and proper test cases. Parser generators will also ensure such a thing cannot occur. I had hoped to remove ParseStatus entirely in favor of Parsed, but there's a lot of type inference that happens based on the fact that `ParseStatus` has a `ParseState` type parameter; `Parsed` has only `Object`. It is desirable for a public-facing `Parsed` to not be tied to `ParseState`, since consumers need not be concerned with such a heavy type; however, we _do_ want that heavy type internally, as it carries a lot of useful information that allows for significant and powerful type inference, which in turn creates expressive and convenient APIs. DEV-7145
2022-07-11 23:49:57 -04:00
ParseState<Token = XirToken, Object = Attr>
where
Self: Default,
<Self as ParseState>::Error: Into<XirToXirfError>,
StateContext<MAX_DEPTH>: AsMut<<Self as ParseState>::Context>;
/// Stack of element [`QName`] and [`Span`] pairs,
/// representing the current level of nesting.
///
/// This storage is statically allocated,
/// allowing XIRF's parser to avoid memory allocation entirely.
type ElementStack<const MAX_DEPTH: usize> = ArrayVec<(QName, Span), MAX_DEPTH>;
/// XIRF document parser state.
///
/// This parser is a pushdown automaton that parses a single XML document.
#[derive(Debug, Default, PartialEq, Eq)]
pub enum XirToXirf<const MAX_DEPTH: usize, SA = AttrParseState>
where
SA: FlatAttrParseState<MAX_DEPTH>,
{
/// Document parsing has not yet begun.
#[default]
PreRoot,
/// Parsing nodes.
NodeExpected,
/// Delegating to attribute parser.
AttrExpected(SA),
/// End of document has been reached.
Done,
}
pub type StateContext<const MAX_DEPTH: usize> =
Context<ElementStack<MAX_DEPTH>>;
impl<const MAX_DEPTH: usize, SA> ParseState for XirToXirf<MAX_DEPTH, SA>
where
SA: FlatAttrParseState<MAX_DEPTH>,
{
type Token = XirToken;
type Object = XirfToken;
type Error = XirToXirfError;
type Context = StateContext<MAX_DEPTH>;
fn parse_token(
self,
tok: Self::Token,
stack: &mut Self::Context,
) -> TransitionResult<Self> {
use XirToXirf::{AttrExpected, Done, NodeExpected, PreRoot};
match (self, tok) {
// Comments are permitted before and after the first root element.
(st @ (PreRoot | Done), XirToken::Comment(sym, span)) => {
Transition(st).ok(XirfToken::Comment(sym, span))
}
(PreRoot, tok @ XirToken::Open(..)) => Self::parse_node(tok, stack),
(PreRoot, tok) => {
Transition(PreRoot).err(XirToXirfError::RootOpenExpected(tok))
}
(NodeExpected, tok) => Self::parse_node(tok, stack),
tamer: Replace ParseStatus::Dead with generic lookahead Oh what a tortured journey. I had originally tried to avoid formalizing lookahead for all parsers by pretending that it was only needed for dead state transitions (that is---states that have no transitions for a given input token), but then I needed to yield information for aggregation. So I added the ability to override the token for `Dead` to yield that, in addition to the token. But then I also needed to yield lookahead for error conditions. It was a mess that didn't make sense. This eliminates `ParseStatus::Dead` entirely and fully integrates the lookahead token in `Parser` that was previously implemented. Notably, the lookahead token is encapsulated in `TransitionResult` and unavailable to `ParseState` implementations, forcing them to rely on `Parser` for recursion. This not only prevents `ParseState` from recursing, but also simplifies delegation by removing the need to manually handle tokens of lookahead. The awkward case here is XIRT, which does not follow the streaming parsing convention, because it was conceived before the parsing framework. It needs to go away, but doing so right now would be a lot of work, so it has to stick around for a little bit longer until the new parser generators can be used instead. It is a persistent thorn in my side, going against the grain. `Parser` will immediately recurse if it sees a token of lookahead with an incomplete parse. This is because stitched parsers will frequently yield a dead state indication when they're done parsing, and there's no use in propagating an `Incomplete` status down the entire lowering pipeline. But, that does mean that the toplevel is not the only thing recursing. _But_, the behavior doesn't really change, in the sense that it would infinitely recurse down the entire lowering stack (though there'd be an opportunity to detect that). This should never happen with a correct parser, but it's not worth the effort right now to try to force such a thing with Rust's type system. Something like TLA+ is better suited here as an aid, but it shouldn't be necessary with clear implementations and proper test cases. Parser generators will also ensure such a thing cannot occur. I had hoped to remove ParseStatus entirely in favor of Parsed, but there's a lot of type inference that happens based on the fact that `ParseStatus` has a `ParseState` type parameter; `Parsed` has only `Object`. It is desirable for a public-facing `Parsed` to not be tied to `ParseState`, since consumers need not be concerned with such a heavy type; however, we _do_ want that heavy type internally, as it carries a lot of useful information that allows for significant and powerful type inference, which in turn creates expressive and convenient APIs. DEV-7145
2022-07-11 23:49:57 -04:00
(AttrExpected(sa), tok) => sa.delegate(
tok,
stack,
|sa| Transition(AttrExpected(sa)),
|| Transition(NodeExpected),
),
(Done, tok) => Transition(Done).dead(tok),
}
}
/// Whether all elements have been closed.
///
/// Parsing will fail if there are any open elements.
/// Intuitively,
/// this means that the parser must have encountered the closing tag
/// for the root element.
fn is_accepting(&self) -> bool {
// TODO: It'd be nice if we could also return additional context to
// aid the user in diagnosing the problem,
// e.g. what element(s) still need closing.
*self == XirToXirf::Done
}
}
impl<const MAX_DEPTH: usize, SA> Display for XirToXirf<MAX_DEPTH, SA>
where
SA: FlatAttrParseState<MAX_DEPTH>,
{
fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
use XirToXirf::*;
match self {
PreRoot => write!(f, "expecting document root"),
NodeExpected => write!(f, "expecting a node"),
AttrExpected(sa) => Display::fmt(sa, f),
Done => write!(f, "done parsing"),
}
}
}
impl<const MAX_DEPTH: usize, SA> XirToXirf<MAX_DEPTH, SA>
where
SA: FlatAttrParseState<MAX_DEPTH>,
{
/// Parse a token while in a state expecting a node.
fn parse_node(
tok: <Self as ParseState>::Token,
stack: &mut ElementStack<MAX_DEPTH>,
) -> TransitionResult<Self> {
use XirToXirf::{AttrExpected, Done, NodeExpected};
use XirfToken::*;
match tok {
XirToken::Open(qname, span) if stack.len() == MAX_DEPTH => {
Transition(NodeExpected).err(XirToXirfError::MaxDepthExceeded {
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
open: (qname, span.tag_span()),
max: Depth(MAX_DEPTH),
})
}
XirToken::Open(qname, span) => {
let depth = stack.len();
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
stack.push((qname, span.tag_span()));
// Delegate to the attribute parser until it is complete.
Transition(AttrExpected(SA::default())).ok(Open(
qname,
span,
Depth(depth),
))
}
XirToken::Close(close_oqname, close_span) => {
match (close_oqname, stack.pop()) {
(_, None) => unreachable!("parser should be in Done state"),
(Some(qname), Some((open_qname, open_span)))
if qname != open_qname =>
{
Transition(NodeExpected).err(
XirToXirfError::UnbalancedTag {
open: (open_qname, open_span),
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
close: (qname, close_span.tag_span()),
},
)
}
// Final closing tag (for root node) completes the document.
(..) if stack.len() == 0 => Transition(Done).ok(Close(
close_oqname,
close_span,
Depth(0),
)),
(..) => {
let depth = stack.len();
Transition(NodeExpected).ok(Close(
close_oqname,
close_span,
Depth(depth),
))
}
}
}
XirToken::Comment(sym, span) => {
Transition(NodeExpected).ok(Comment(sym, span))
}
XirToken::Text(sym, span) => {
Transition(NodeExpected).ok(Text(sym, span))
}
XirToken::CData(sym, span) => {
Transition(NodeExpected).ok(CData(sym, span))
}
XirToken::Whitespace(ws, span) => {
Transition(NodeExpected).ok(Whitespace(ws, span))
}
// We should transition to `State::Attr` before encountering any
// of these tokens.
XirToken::AttrName(..)
| XirToken::AttrValue(..)
Revert "tamer: xir: Initial re-introduction of AttrEnd" This reverts commit b973d36862a4a2aaf53fb0b25fba01b57e5a7463. Alright, I'm getting sick of fighting with myself on this. But rather than just removing the last commit, I'm going to keep it around, so that my thoughts are clearly documented for my future quarrels with myself. Firstly: this added more overhead than I wanted it to. While it wasn't significant, it did add 100--150ms to one of our largest systems, up from ~2.8s, which seems a bit much for a token that's really just meant to make life easier for the parser. Further, it seems that all I've managed to do is push my original problem to a different layer---this started as a means to resolve having to emit both an object and an error simultaneously in the case where aggregate attribute parsing has completed, but we encounter an error on the next token (e.g. an unexpected element). But XIRF, if it's missing AttrEnd, should throw an error, but should also recover. Recovery is easy---just assume that it was present---_but then we don't emit a XIRF `AttrEnd` token_, which is necessary for downstream systems. So we'd need to either: (a) emit both a token and an error; or (b) panic. But if we're doing (a), then the need for `AttrEnd` goes away, because it solves the original problem (though the other concerns of the previous commit still stand). (b) is not ideal at all, even though the missing token does represent an internal system error; it's not something the user can correct. But, given that it's something that the user cannot correct, doesn't that imply that it's an awkward thing to include in the token stream? So back to `AttrEnd` being an awkward PITA to have. So, given (a), I'll just do that: errors will become more of a "hey, this error just occurred, but I'm trying to recover---here's an object that you should use if you choose to continue parsing, but it may or may not be what you're looking for; proceed with caution". That flips the original script: I imagined having external systems feed recovery tokens, but this encapsulates recovery within the parser, which really is more appropriate, though less flexible than having an omniscient external recovery system; such a monolith was always an awkward concept and would be difficult to implement cleanly. This can also potentially be implemented as a generalization of the Dead state change that allowed an object to be emitted alongside the lookahead/error. Anyway, back to where I was...I'm sure I'll look back on this in the future shaking my head, reflecting on how naive I was. DEV-7145
2022-06-29 11:02:18 -04:00
| XirToken::AttrValueFragment(..) => {
unreachable!("attribute token in NodeExpected state: {tok:?}")
}
}
}
}
/// Produce a streaming parser lowering a XIR [`TokenStream`] into a XIRF
/// stream.
pub fn parse<const MAX_DEPTH: usize>(
toks: impl TokenStream,
) -> impl Iterator<Item = ParsedResult<XirToXirf<MAX_DEPTH>>> {
XirToXirf::<MAX_DEPTH>::parse(toks)
}
/// Parsing error from [`XirToXirf`].
#[derive(Debug, Eq, PartialEq)]
pub enum XirToXirfError {
/// Opening root element tag was expected.
RootOpenExpected(XirToken),
/// Opening tag exceeds the maximum nesting depth for this parser.
MaxDepthExceeded { open: (QName, Span), max: Depth },
/// The closing tag does not match the opening tag at the same level of
/// nesting.
UnbalancedTag {
open: (QName, Span),
close: (QName, Span),
},
/// Error from the attribute parser.
AttrError(AttrParseError),
}
impl Display for XirToXirfError {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
use XirToXirfError::*;
match self {
RootOpenExpected(_tok) => {
write!(f, "missing opening root element",)
}
MaxDepthExceeded {
open: (_name, _),
max,
} => {
write!(
f,
"maximum XML element nesting depth of `{max}` exceeded"
)
}
UnbalancedTag {
open: (open_name, _),
close: (_close_name, _),
} => {
write!(f, "expected closing tag for `{open_name}`")
}
AttrError(e) => Display::fmt(e, f),
}
}
}
impl Error for XirToXirfError {
fn source(&self) -> Option<&(dyn Error + 'static)> {
match self {
Self::AttrError(e) => Some(e),
_ => None,
}
}
}
impl Diagnostic for XirToXirfError {
fn describe(&self) -> Vec<AnnotatedSpan> {
use XirToXirfError::*;
match self {
RootOpenExpected(tok) => {
// TODO: Should the span be the first byte,
// or should we delegate that question to an e.g. `SpanLike`?
tok.span()
.error("an opening root node was expected here")
.into()
}
MaxDepthExceeded {
open: (_, span),
max,
} => span
.error(format!(
"this opening tag increases the level of nesting \
past the limit of {max}"
))
.into(),
UnbalancedTag {
open: (open_name, open_span),
close: (_close_name, close_span),
} => {
// TODO: hint saying that the nesting could be wrong, etc;
// we can't just suggest a replacement,
// since that's not necessarily the problem
vec![
open_span
.note(format!("element `{open_name}` is opened here")),
// No need to state the close name since the source line
// will be highlighted by the diagnostic message.
close_span.error(format!("expected `</{open_name}>`")),
]
}
AttrError(e) => e.describe(),
}
}
}
impl From<AttrParseError> for XirToXirfError {
fn from(e: AttrParseError) -> Self {
Self::AttrError(e)
}
}
#[cfg(test)]
tamer: xir: Introduce {Ele,Open,Close}Span This isn't conceptally all that significant of a change, but there was a lot of modify to get it working. I would generally separate this into a commit for the implementation and another commit for the integration, but I decided to keep things together. This serves a role similar to AttrSpan---this allows deriving a span representing the element name from a span representing the entire XIR token. This will provide more useful context for errors---including the tag delimiter(s) means that we care about the fact that an element is in that position (as opposed to some other type of node) within the context of an error. However, if we are expecting an element but take issue with the element name itself, we want to place emphasis on that instead. This also starts to consider the issue of span contexts---a blob of detached data that is `Span` is useful for error context, but it's not useful for manipulation or deriving additional information. For that, we need to encode additional context, and this is an attempt at that. I am interested in the concept of providing Spans that are guaranteed to actually make sense---that are instantiated and manipulated with APIs that ensure consistency. But such a thing buys us very little, practically speaking, over what I have now for TAMER, and so I don't expect to actually implement that for this project; I'll leave that for a personal project. TAMER's already take a lot of my personal interests and it can cause me a lot of grief sometimes (with regards to letting my aspirations cause me more work). DEV-7145
2022-06-24 13:51:49 -04:00
pub mod test;