Compare commits

...

10 Commits

Author SHA1 Message Date
Tom Alexander
288350daef
Iterate over the bytes instead of characters when counting brackets.
All checks were successful
rust-test Build rust-test has succeeded
rust-build Build rust-build has succeeded
2023-08-28 03:52:21 -04:00
Tom Alexander
c683516620
Switch inline source blocks to using bracket depth from OrgSource instead of from the context.
This is for the same reasons as footnote references.
2023-08-28 03:04:33 -04:00
Tom Alexander
e731e8ff6b
Switch inline babel call to using bracket depth from OrgSource instead of from the context.
This is for the same reasons as footnote references.
2023-08-28 03:04:33 -04:00
Tom Alexander
4c2037ec44
Switch subscript and superscript to using bracket depth from OrgSource instead of from the context.
This is for the same reasons as footnote references.
2023-08-28 03:04:33 -04:00
Tom Alexander
a46b358549
Switch citations to using bracket depth from OrgSource instead of from the context.
This is for the same reasons as footnote references.
2023-08-28 03:04:32 -04:00
Tom Alexander
ec813e3b3f
Switch to using bracket depth from OrgSource instead of from the context for footnote references.
It is currently unknown if this will produce a performance increase, but unless it has a significant performance penalty we are going to go forward with this change because it makes it more explicit which values need to be read deeply from other elements (therefore needing to be in the context) vs values that can be bound to the exit matcher since they are only used within the confines of the current element.

I suspect we will get a performance boost since it will be reducing the nodes that need to be walked in the context but maintaining bracket depth count over the entire document instead of only inside elements that need balanced brackets could cost us.
2023-08-28 03:04:32 -04:00
Tom Alexander
f11f7bcc73
Keep track of bracket, brace, and parenthesis depth when iterating over the OrgSource. 2023-08-28 01:18:46 -04:00
Tom Alexander
9e0e5f6f0a
Remove line number limit for LaTeX fragments. 2023-08-28 01:18:46 -04:00
Tom Alexander
16e788c36c
Add tests for LaTeX fragments and text markup that span more than three lines.
The documentation currently states that the body for these cannot span more than three lines but that is not the behavior I am seeing from emacs in practice. Waiting on a mailing list response to tell me if this is a documentation error or a parser error.
2023-08-28 01:18:46 -04:00
Tom Alexander
1952d175c0
Record line number in OrgSource.
This will be used for elements who have limits on the number of lines inside of them. This includes LaTeX fragment bodies and text markup.
2023-08-28 01:18:45 -04:00
13 changed files with 393 additions and 500 deletions

View File

@ -0,0 +1,22 @@
# Extra open
[cite/a/b-_/foo:unbalancedglobal[prefix;keyprefix @foo keysuffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;unbalancedkey[prefix @foo keysuffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;keyprefix @foo unbalancedkey[suffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;keyprefix @foo keysuffix;unbalancedglobal[suffix]
# Extra close
[cite/a/b-_/foo:unbalancedglobal]prefix;keyprefix @foo keysuffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;unbalancedkey]prefix @foo keysuffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;keyprefix @foo unbalancedkey]suffix;globalsuffix]
[cite/a/b-_/foo:globalprefix;keyprefix @foo keysuffix;unbalancedglobal]suffix]
# balanced:
[cite/a/b-_/foo:gl[obalpref]ix;ke[ypref]ix @foo ke[ysuff]ix;gl[obalsuff]ix]

View File

@ -0,0 +1,2 @@
[fn:2:This footnote [ has balanced ] brackets inside it]
[fn::This footnote does not have balanced [ brackets inside it]

View File

@ -0,0 +1,6 @@
$foo
bar
baz
lorem
ipsum
dolar$

View File

@ -0,0 +1,17 @@
foo *bar
baz* lorem
text *markup
can
span* more
than *three
lines.
foo
bar* baz
foo *bar \\
baz \\
lorem \\
ipsum \\
dolar* cat

View File

@ -11,17 +11,16 @@ use nom::multi::many_till;
use nom::multi::separated_list1;
use nom::sequence::tuple;
use super::citation_reference::must_balance_bracket;
use super::org_source::OrgSource;
use super::Context;
use crate::error::CustomError;
use crate::error::Res;
use crate::parser::citation_reference::citation_reference;
use crate::parser::citation_reference::citation_reference_key;
use crate::parser::citation_reference::get_bracket_depth;
use crate::parser::exiting::ExitClass;
use crate::parser::object::Citation;
use crate::parser::object_parser::standard_set_object;
use crate::parser::parser_context::CitationBracket;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_with_context::parser_with_context;
@ -38,13 +37,15 @@ pub fn citation<'r, 's>(
let (remaining, _) = tag_no_case("[cite")(input)?;
let (remaining, _) = opt(citestyle)(remaining)?;
let (remaining, _) = tag(":")(remaining)?;
let (remaining, _prefix) = opt(parser_with_context!(global_prefix)(context))(remaining)?;
let (remaining, _prefix) =
must_balance_bracket(opt(parser_with_context!(global_prefix)(context)))(remaining)?;
let (remaining, _references) =
separated_list1(tag(";"), parser_with_context!(citation_reference)(context))(remaining)?;
let (remaining, _suffix) = opt(tuple((
let (remaining, _suffix) = must_balance_bracket(opt(tuple((
tag(";"),
parser_with_context!(global_suffix)(context),
)))(remaining)?;
))))(remaining)?;
let (remaining, _) = tag("]")(remaining)?;
let (remaining, _) = space0(remaining)?;
let source = get_consumed(input, remaining);
@ -83,15 +84,11 @@ fn global_prefix<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Vec<Object<'s>>> {
// TODO: I could insert CitationBracket entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let parser_context = context
.with_additional_node(ContextElement::CitationBracket(CitationBracket {
position: input,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = global_prefix_end(input.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &global_prefix_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, (children, _exit_contents)) = verify(
many_till(
@ -104,28 +101,24 @@ fn global_prefix<'r, 's>(
Ok((remaining, children))
}
fn global_prefix_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_global_prefix_end(context, input, starting_bracket_depth)
}
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn global_prefix_end<'r, 's>(
fn _global_prefix_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a citation.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded citation global prefix bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the citation.
unreachable!("Exceeded citation global prefix bracket depth.")
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("]")(input);
@ -144,15 +137,11 @@ fn global_suffix<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Vec<Object<'s>>> {
// TODO: I could insert CitationBracket entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let parser_context = context
.with_additional_node(ContextElement::CitationBracket(CitationBracket {
position: input,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = global_suffix_end(input.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &global_suffix_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, (children, _exit_contents)) = verify(
many_till(
@ -164,28 +153,24 @@ fn global_suffix<'r, 's>(
Ok((remaining, children))
}
fn global_suffix_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_global_suffix_end(context, input, starting_bracket_depth)
}
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn global_suffix_end<'r, 's>(
fn _global_suffix_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a citation.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded citation global suffix bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the citation.
unreachable!("Exceeded citation global suffix bracket depth.")
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("]")(input);

View File

@ -13,11 +13,11 @@ use nom::sequence::tuple;
use super::org_source::OrgSource;
use super::Context;
use crate::error::CustomError;
use crate::error::MyError;
use crate::error::Res;
use crate::parser::exiting::ExitClass;
use crate::parser::object::CitationReference;
use crate::parser::object_parser::minimal_set_object;
use crate::parser::parser_context::CitationBracket;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_with_context::parser_with_context;
@ -31,9 +31,11 @@ pub fn citation_reference<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, CitationReference<'s>> {
let (remaining, _prefix) = opt(parser_with_context!(key_prefix)(context))(input)?;
let (remaining, _prefix) =
must_balance_bracket(opt(parser_with_context!(key_prefix)(context)))(input)?;
let (remaining, _key) = parser_with_context!(citation_reference_key)(context)(remaining)?;
let (remaining, _suffix) = opt(parser_with_context!(key_suffix)(context))(remaining)?;
let (remaining, _suffix) =
must_balance_bracket(opt(parser_with_context!(key_suffix)(context)))(remaining)?;
let source = get_consumed(input, remaining);
Ok((
@ -69,15 +71,11 @@ fn key_prefix<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Vec<Object<'s>>> {
// TODO: I could insert CitationBracket entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let parser_context = context
.with_additional_node(ContextElement::CitationBracket(CitationBracket {
position: input,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = key_prefix_end(input.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &key_prefix_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, (children, _exit_contents)) = verify(
many_till(
@ -94,15 +92,11 @@ fn key_suffix<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Vec<Object<'s>>> {
// TODO: I could insert CitationBracket entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let parser_context = context
.with_additional_node(ContextElement::CitationBracket(CitationBracket {
position: input,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = key_suffix_end(input.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &key_suffix_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, (children, _exit_contents)) = verify(
many_till(
@ -114,39 +108,24 @@ fn key_suffix<'r, 's>(
Ok((remaining, children))
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
pub fn get_bracket_depth<'r, 's>(context: Context<'r, 's>) -> Option<&'r CitationBracket<'s>> {
for node in context.iter() {
match node.get_data() {
ContextElement::CitationBracket(depth) => return Some(depth),
_ => {}
}
fn key_prefix_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_key_prefix_end(context, input, starting_bracket_depth)
}
None
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn key_prefix_end<'r, 's>(
fn _key_prefix_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a citation reference.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded citation reference key prefix bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the citation.
unreachable!("Exceeded citation key prefix bracket depth.")
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("]")(input);
@ -160,28 +139,24 @@ fn key_prefix_end<'r, 's>(
))(input)
}
fn key_suffix_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_key_suffix_end(context, input, starting_bracket_depth)
}
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn key_suffix_end<'r, 's>(
context: Context<'r, 's>,
fn _key_suffix_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a citation reference.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded citation reference key prefix bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the citation.
unreachable!("Exceeded citation key suffix bracket depth.")
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("]")(input);
@ -191,3 +166,21 @@ fn key_suffix_end<'r, 's>(
}
tag(";")(input)
}
pub fn must_balance_bracket<'s, F, O>(
mut inner: F,
) -> impl FnMut(OrgSource<'s>) -> Res<OrgSource<'s>, O>
where
F: FnMut(OrgSource<'s>) -> Res<OrgSource<'s>, O>,
{
move |input: OrgSource<'_>| {
let pre_bracket_depth = input.get_bracket_depth();
let (remaining, output) = inner(input)?;
if remaining.get_bracket_depth() - pre_bracket_depth != 0 {
return Err(nom::Err::Error(CustomError::MyError(MyError(
"UnbalancedBrackets".into(),
))));
}
Ok((remaining, output))
}
}

View File

@ -15,7 +15,6 @@ use crate::parser::exiting::ExitClass;
use crate::parser::footnote_definition::label;
use crate::parser::object_parser::standard_set_object;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_context::FootnoteReferenceDefinition;
use crate::parser::parser_with_context::parser_with_context;
use crate::parser::util::exit_matcher_parser;
use crate::parser::util::get_consumed;
@ -39,18 +38,12 @@ fn anonymous_footnote<'r, 's>(
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, FootnoteReference<'s>> {
let (remaining, _) = tag_no_case("[fn::")(input)?;
let parser_context = context
.with_additional_node(ContextElement::FootnoteReferenceDefinition(
FootnoteReferenceDefinition {
position: remaining,
depth: 0,
},
))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = footnote_definition_end(remaining.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Beta,
exit_matcher: &footnote_definition_end,
exit_matcher: &exit_with_depth,
}));
// TODO: I could insert FootnoteReferenceDefinition entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let (remaining, (children, _exit_contents)) = verify(
many_till(
parser_with_context!(standard_set_object)(&parser_context),
@ -80,18 +73,12 @@ fn inline_footnote<'r, 's>(
let (remaining, _) = tag_no_case("[fn:")(input)?;
let (remaining, label_contents) = label(remaining)?;
let (remaining, _) = tag(":")(remaining)?;
let parser_context = context
.with_additional_node(ContextElement::FootnoteReferenceDefinition(
FootnoteReferenceDefinition {
position: remaining,
depth: 0,
},
))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = footnote_definition_end(remaining.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Beta,
exit_matcher: &footnote_definition_end,
exit_matcher: &exit_with_depth,
}));
// TODO: I could insert FootnoteReferenceDefinition entries in the context after each matched object to reduce the scanning done for counting brackets which should be more efficient.
let (remaining, (children, _exit_contents)) = verify(
many_till(
parser_with_context!(standard_set_object)(&parser_context),
@ -133,47 +120,30 @@ fn footnote_reference_only<'r, 's>(
))
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn footnote_definition_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a footnote definition.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded footnote reference definition bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
fn footnote_definition_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_footnote_definition_end(context, input, starting_bracket_depth)
}
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn _footnote_definition_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth > 0 {
// Its impossible for the next character to end the footnote reference definition if we're any amount of brackets deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"NoFootnoteReferenceDefinitionEnd".into(),
))));
}
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the footnote definition.
unreachable!("Exceeded footnote reference definition bracket depth.")
}
tag("]")(input)
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn get_bracket_depth<'r, 's>(
context: Context<'r, 's>,
) -> Option<&'r FootnoteReferenceDefinition<'s>> {
for node in context.iter() {
match node.get_data() {
ContextElement::FootnoteReferenceDefinition(depth) => return Some(depth),
_ => {}
}
}
None
}

View File

@ -12,9 +12,10 @@ use nom::multi::many_till;
use super::org_source::OrgSource;
use super::Context;
use crate::error::CustomError;
use crate::error::MyError;
use crate::error::Res;
use crate::parser::exiting::ExitClass;
use crate::parser::parser_context::BabelHeaderBracket;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_with_context::parser_with_context;
@ -74,14 +75,11 @@ fn header<'r, 's>(
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let (remaining, _) = tag("[")(input)?;
let parser_context = context
.with_additional_node(ContextElement::BabelHeaderBracket(BabelHeaderBracket {
position: remaining,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = header_end(remaining.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &header_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, name) = recognize(many_till(
@ -92,28 +90,30 @@ fn header<'r, 's>(
Ok((remaining, name))
}
fn header_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_header_end(context, input, starting_bracket_depth)
}
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn header_end<'r, 's>(
context: Context<'r, 's>,
fn _header_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside an inline babel call header.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'(' => {
current_depth += 1;
}
')' if current_depth == 0 => {
panic!("Exceeded inline babel call header bracket depth.")
}
')' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth > 0 {
// Its impossible for the next character to end the header if we're any amount of bracket deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"NoHeaderEnd".into(),
))));
}
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the header.
unreachable!("Exceeded header bracket depth.")
}
alt((tag("]"), line_ending))(input)
}
@ -125,14 +125,11 @@ fn argument<'r, 's>(
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let (remaining, _) = tag("(")(input)?;
let parser_context = context
.with_additional_node(ContextElement::BabelHeaderBracket(BabelHeaderBracket {
position: remaining,
depth: 0,
}))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = argument_end(remaining.get_parenthesis_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &argument_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, name) = recognize(many_till(
@ -143,39 +140,30 @@ fn argument<'r, 's>(
Ok((remaining, name))
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn argument_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside an inline babel call argument.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded inline babel call argument bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
fn argument_end(
starting_parenthesis_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_argument_end(context, input, starting_parenthesis_depth)
}
alt((tag(")"), line_ending))(input)
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
pub fn get_bracket_depth<'r, 's>(context: Context<'r, 's>) -> Option<&'r BabelHeaderBracket<'s>> {
for node in context.iter() {
match node.get_data() {
ContextElement::BabelHeaderBracket(depth) => return Some(depth),
_ => {}
}
fn _argument_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_parenthesis_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let current_depth = input.get_parenthesis_depth() - starting_parenthesis_depth;
if current_depth > 0 {
// Its impossible for the next character to end the argument if we're any amount of parenthesis deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"NoArgumentEnd".into(),
))));
}
None
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing parenthesis should end the argument.
unreachable!("Exceeded argument parenthesis depth.")
}
alt((tag(")"), line_ending))(input)
}

View File

@ -1,3 +1,4 @@
use nom::branch::alt;
use nom::bytes::complete::tag;
use nom::bytes::complete::tag_no_case;
use nom::character::complete::anychar;
@ -14,11 +15,11 @@ use tracing::span;
use super::org_source::OrgSource;
use super::Context;
use crate::error::CustomError;
use crate::error::MyError;
use crate::error::Res;
use crate::parser::exiting::ExitClass;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_context::InlineSourceBlockBracket;
use crate::parser::parser_with_context::parser_with_context;
use crate::parser::util::exit_matcher_parser;
use crate::parser::util::get_consumed;
@ -75,16 +76,11 @@ fn header<'r, 's>(
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let (remaining, _) = tag("[")(input)?;
let parser_context = context
.with_additional_node(ContextElement::InlineSourceBlockBracket(
InlineSourceBlockBracket {
position: remaining,
depth: 0,
},
))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = header_end(remaining.get_bracket_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Beta,
exit_matcher: &header_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, header_contents) = recognize(many_till(
@ -95,37 +91,32 @@ fn header<'r, 's>(
Ok((remaining, header_contents))
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn header_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside an inline source block header.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'[' => {
current_depth += 1;
}
']' if current_depth == 0 => {
panic!("Exceeded inline source block header bracket depth.")
}
']' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("]")(input);
if close_bracket.is_ok() {
return close_bracket;
}
fn header_end(
starting_bracket_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_header_end(context, input, starting_bracket_depth)
}
}
line_ending(input)
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn _header_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_bracket_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let current_depth = input.get_bracket_depth() - starting_bracket_depth;
if current_depth > 0 {
// Its impossible for the next character to end the header if we're any amount of bracket deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"NoHeaderEnd".into(),
))));
}
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing bracket should end the header.
unreachable!("Exceeded header bracket depth.")
}
alt((tag("]"), line_ending))(input)
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
@ -135,16 +126,11 @@ fn body<'r, 's>(
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let (remaining, _) = tag("{")(input)?;
let parser_context = context
.with_additional_node(ContextElement::InlineSourceBlockBracket(
InlineSourceBlockBracket {
position: remaining,
depth: 0,
},
))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = body_end(remaining.get_brace_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Beta,
exit_matcher: &body_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, body_contents) = recognize(many_till(
@ -165,60 +151,28 @@ fn body<'r, 's>(
Ok((remaining, body_contents))
}
fn body_end(
starting_brace_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| _body_end(context, input, starting_brace_depth)
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn body_end<'r, 's>(
context: Context<'r, 's>,
fn _body_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_brace_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside an inline source block body.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'{' => {
current_depth += 1;
}
'}' if current_depth == 0 => {
panic!("Exceeded inline source block body bracket depth.")
}
'}' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
let current_depth = input.get_brace_depth() - starting_brace_depth;
if current_depth > 0 {
// Its impossible for the next character to end the body if we're any amount of brace deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"NoBodyEnd".into(),
))));
}
{
#[cfg(feature = "tracing")]
let span = span!(
tracing::Level::DEBUG,
"inside end body",
remaining = Into::<&str>::into(input),
current_depth = current_depth
);
#[cfg(feature = "tracing")]
let _enter = span.enter();
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("}")(input);
if close_bracket.is_ok() {
return close_bracket;
}
}
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing brace should end the body.
unreachable!("Exceeded body brace depth.")
}
line_ending(input)
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
pub fn get_bracket_depth<'r, 's>(
context: Context<'r, 's>,
) -> Option<&'r InlineSourceBlockBracket<'s>> {
for node in context.iter() {
match node.get_data() {
ContextElement::InlineSourceBlockBracket(depth) => return Some(depth),
_ => {}
}
}
None
alt((tag("}"), line_ending))(input)
}

View File

@ -202,17 +202,13 @@ fn bordered_dollar_fragment<'r, 's>(
// TODO: I'm assuming I should be peeking at the borders but the documentation is not clear. Test to figure out.
let (_, _) = peek(parser_with_context!(open_border)(context))(remaining)?;
// TODO: As an optimization it would be nice to exit early upon hitting the 3rd line break
let (remaining, _) = verify(
recognize(many_till(
anychar,
peek(alt((
parser_with_context!(exit_matcher_parser)(context),
tag("$"),
))),
)),
|body: &OrgSource<'_>| Into::<&str>::into(body).lines().take(4).count() <= 3,
)(remaining)?;
let (remaining, _) = recognize(many_till(
anychar,
peek(alt((
parser_with_context!(exit_matcher_parser)(context),
tag("$"),
))),
))(remaining)?;
let (_, _) = peek(parser_with_context!(close_border)(context))(remaining)?;
let (remaining, _) = tag("$")(remaining)?;

View File

@ -18,6 +18,10 @@ pub struct OrgSource<'s> {
end: usize, // exclusive
start_of_line: usize,
preceding_character: Option<char>,
line_number: usize,
bracket_depth: isize, // []
brace_depth: isize, // {}
parenthesis_depth: isize, // ()
}
impl<'s> OrgSource<'s> {
@ -31,6 +35,10 @@ impl<'s> OrgSource<'s> {
end: input.len(),
start_of_line: 0,
preceding_character: None,
line_number: 1,
bracket_depth: 0,
brace_depth: 0,
parenthesis_depth: 0,
}
}
@ -56,6 +64,22 @@ impl<'s> OrgSource<'s> {
assert!(other.end <= self.end);
self.slice(..(other.start - self.start))
}
pub fn get_line_number(&self) -> usize {
self.line_number
}
pub fn get_bracket_depth(&self) -> isize {
self.bracket_depth
}
pub fn get_brace_depth(&self) -> isize {
self.brace_depth
}
pub fn get_parenthesis_depth(&self) -> isize {
self.parenthesis_depth
}
}
impl<'s> InputTake for OrgSource<'s> {
@ -119,10 +143,38 @@ where
}
let skipped_text = &self.full_source[self.start..new_start];
let start_of_line = skipped_text
.rfind('\n')
.map(|idx| self.start + idx + 1)
.unwrap_or(self.start_of_line);
let mut start_of_line = self.start_of_line;
let mut line_number = self.line_number;
let mut bracket_depth = self.bracket_depth;
let mut brace_depth = self.brace_depth;
let mut parenthesis_depth = self.parenthesis_depth;
for (offset, byte) in skipped_text.bytes().enumerate() {
match byte {
b'\n' => {
start_of_line = self.start + offset + 1;
line_number += 1;
}
b'[' => {
bracket_depth += 1;
}
b']' => {
bracket_depth -= 1;
}
b'{' => {
brace_depth += 1;
}
b'}' => {
brace_depth -= 1;
}
b'(' => {
parenthesis_depth += 1;
}
b')' => {
parenthesis_depth -= 1;
}
_ => {}
};
}
OrgSource {
full_source: self.full_source,
@ -130,6 +182,10 @@ where
end: new_end,
start_of_line,
preceding_character: skipped_text.chars().last(),
line_number,
bracket_depth,
brace_depth,
parenthesis_depth,
}
}
}
@ -369,4 +425,28 @@ mod tests {
assert_eq!(input.get_preceding_character(), None);
assert_eq!(input.slice(8..).get_preceding_character(), Some('💛'));
}
#[test]
fn line_number() {
let input = OrgSource::new("lorem\nfoo\nbar\nbaz\nipsum");
assert_eq!(input.get_line_number(), 1);
assert_eq!(input.slice(5..).get_line_number(), 1);
assert_eq!(input.slice(6..).get_line_number(), 2);
assert_eq!(input.slice(6..).slice(10..).get_line_number(), 4);
}
#[test]
fn depth() {
let input = OrgSource::new("[][()][({)]}}}}");
assert_eq!(input.get_bracket_depth(), 0);
assert_eq!(input.get_brace_depth(), 0);
assert_eq!(input.get_parenthesis_depth(), 0);
assert_eq!(input.slice(4..).get_bracket_depth(), 1);
assert_eq!(input.slice(4..).get_brace_depth(), 0);
assert_eq!(input.slice(4..).get_parenthesis_depth(), 1);
assert_eq!(input.slice(4..).slice(6..).get_bracket_depth(), 1);
assert_eq!(input.slice(4..).slice(6..).get_brace_depth(), 1);
assert_eq!(input.slice(4..).slice(6..).get_parenthesis_depth(), 0);
assert_eq!(input.slice(14..).get_brace_depth(), -2);
}
}

View File

@ -123,71 +123,6 @@ pub enum ContextElement<'r, 's> {
/// org-mode document since text needs to be re-parsed to look for
/// radio links matching the contents of radio targets.
RadioTarget(Vec<&'r Vec<Object<'s>>>),
/// Stores the current bracket depth inside a footnote reference's definition.
///
/// The definition inside a footnote reference must have balanced
/// brackets [] inside the definition, so this stores the amount
/// of opening brackets subtracted by the amount of closing
/// brackets within the definition must equal zero.
///
/// A reference to the position in the string is also included so
/// unbalanced brackets can be detected in the middle of an
/// object.
FootnoteReferenceDefinition(FootnoteReferenceDefinition<'s>),
/// Stores the current bracket depth inside a citation.
///
/// The global prefix, global suffix, key prefix, and key suffix
/// inside a footnote reference must have balanced brackets []
/// inside the definition, so this stores the amount of opening
/// brackets subtracted by the amount of closing brackets within
/// the definition must equal zero. None of the prefixes or
/// suffixes can be nested inside each other so we can use a
/// single type for this without conflict.
///
/// A reference to the position in the string is also included so
/// unbalanced brackets can be detected in the middle of an
/// object.
CitationBracket(CitationBracket<'s>),
/// Stores the current bracket or parenthesis depth inside an inline babel call.
///
/// Inside an inline babel call the headers must have balanced
/// parentheses () and the arguments must have balanced brackets
/// [], so this stores the amount of opening brackets subtracted
/// by the amount of closing brackets within the definition must
/// equal zero.
///
/// A reference to the position in the string is also included so
/// unbalanced brackets can be detected in the middle of an
/// object.
BabelHeaderBracket(BabelHeaderBracket<'s>),
/// Stores the current bracket or parenthesis depth inside an inline babel call.
///
/// Inside an inline babel call the headers must have balanced
/// parentheses () and the arguments must have balanced brackets
/// [], so this stores the amount of opening brackets subtracted
/// by the amount of closing brackets within the definition must
/// equal zero.
///
/// A reference to the position in the string is also included so
/// unbalanced brackets can be detected in the middle of an
/// object.
InlineSourceBlockBracket(InlineSourceBlockBracket<'s>),
/// Stores the current bracket or parenthesis depth inside a
/// superscript or superscript.
///
/// Inside the braces of a subscript or superscript there must be
/// balanced braces {}, so this stores the amount of opening
/// braces subtracted by the amount of closing braces within the
/// definition must equal zero.
///
/// A reference to the position in the string is also included so
/// unbalanced braces can be detected in the middle of an object.
SubscriptSuperscriptBrace(SubscriptSuperscriptBrace<'s>),
}
pub struct ExitMatcherNode<'r> {
@ -195,36 +130,6 @@ pub struct ExitMatcherNode<'r> {
pub class: ExitClass,
}
#[derive(Debug)]
pub struct FootnoteReferenceDefinition<'s> {
pub position: OrgSource<'s>,
pub depth: usize,
}
#[derive(Debug)]
pub struct CitationBracket<'s> {
pub position: OrgSource<'s>,
pub depth: usize,
}
#[derive(Debug)]
pub struct BabelHeaderBracket<'s> {
pub position: OrgSource<'s>,
pub depth: usize,
}
#[derive(Debug)]
pub struct InlineSourceBlockBracket<'s> {
pub position: OrgSource<'s>,
pub depth: usize,
}
#[derive(Debug)]
pub struct SubscriptSuperscriptBrace<'s> {
pub position: OrgSource<'s>,
pub depth: usize,
}
impl<'r> std::fmt::Debug for ExitMatcherNode<'r> {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
let mut formatter = f.debug_struct("ExitMatcherNode");

View File

@ -21,7 +21,6 @@ use crate::parser::exiting::ExitClass;
use crate::parser::object_parser::standard_set_object;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_context::SubscriptSuperscriptBrace;
use crate::parser::parser_with_context::parser_with_context;
use crate::parser::util::exit_matcher_parser;
use crate::parser::util::get_consumed;
@ -154,16 +153,11 @@ fn script_with_braces<'r, 's>(
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Vec<Object<'s>>> {
let (remaining, _) = tag("{")(input)?;
let parser_context = context
.with_additional_node(ContextElement::SubscriptSuperscriptBrace(
SubscriptSuperscriptBrace {
position: remaining.into(),
depth: 0,
},
))
.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
let exit_with_depth = script_with_braces_end(remaining.get_brace_depth());
let parser_context =
context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
class: ExitClass::Gamma,
exit_matcher: &script_with_braces_end,
exit_matcher: &exit_with_depth,
}));
let (remaining, (children, _exit_contents)) = many_till(
@ -175,49 +169,30 @@ fn script_with_braces<'r, 's>(
Ok((remaining, children))
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn script_with_braces_end<'r, 's>(
context: Context<'r, 's>,
input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let context_depth = get_bracket_depth(context)
.expect("This function should only be called from inside a subscript or superscript.");
let text_since_context_entry = get_consumed(context_depth.position, input);
let mut current_depth = context_depth.depth;
for c in Into::<&str>::into(text_since_context_entry).chars() {
match c {
'{' => {
current_depth += 1;
}
'}' if current_depth == 0 => {
panic!("Exceeded subscript or superscript brace depth.")
}
'}' if current_depth > 0 => {
current_depth -= 1;
}
_ => {}
}
fn script_with_braces_end(
starting_brace_depth: isize,
) -> impl for<'r, 's> Fn(Context<'r, 's>, OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
move |context: Context, input: OrgSource<'_>| {
_script_with_braces_end(context, input, starting_brace_depth)
}
if current_depth == 0 {
let close_bracket = tag::<&str, OrgSource<'_>, CustomError<OrgSource<'_>>>("}")(input);
if close_bracket.is_ok() {
return close_bracket;
}
}
return Err(nom::Err::Error(CustomError::MyError(MyError(
"Not a valid end for subscript or superscript.".into(),
))));
}
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn get_bracket_depth<'r, 's>(
context: Context<'r, 's>,
) -> Option<&'r SubscriptSuperscriptBrace<'s>> {
for node in context.iter() {
match node.get_data() {
ContextElement::SubscriptSuperscriptBrace(depth) => return Some(depth),
_ => {}
}
fn _script_with_braces_end<'r, 's>(
_context: Context<'r, 's>,
input: OrgSource<'s>,
starting_brace_depth: isize,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
let current_depth = input.get_brace_depth() - starting_brace_depth;
if current_depth > 0 {
// Its impossible for the next character to end the subscript or superscript if we're any amount of braces deep
return Err(nom::Err::Error(CustomError::MyError(MyError(
"Not a valid end for subscript or superscript.".into(),
))));
}
None
if current_depth < 0 {
// This shouldn't be possible because if depth is 0 then a closing brace should end the subscript or superscript.
unreachable!("Exceeded subscript or superscript brace depth.")
}
tag("}")(input)
}