Compare commits

...

7 Commits

12 changed files with 1759 additions and 13 deletions

8
Cargo.lock generated
View File

@ -1270,9 +1270,9 @@ dependencies = [
[[package]]
name = "indexmap"
version = "2.0.2"
version = "2.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8adf3ddd720272c6ea8bf59463c04e0f93d0bbf7c5439b691bca2987e0270897"
checksum = "d530e1a18b1cb4c484e6e34556a0d948706958449fca0cab753d649f2bce3d1f"
dependencies = [
"equivalent",
"hashbrown 0.14.2",
@ -2678,7 +2678,7 @@ version = "0.19.15"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1b5bb770da30e5cbfde35a2d7b9b8a2c4b8ef89548a7a6aeab5c9a576e3e7421"
dependencies = [
"indexmap 2.0.2",
"indexmap 2.1.0",
"toml_datetime",
"winnow",
]
@ -2729,7 +2729,7 @@ dependencies = [
"cli-log",
"crossterm",
"directories",
"indexmap 2.0.2",
"indexmap 2.1.0",
"language_macros",
"matrix-sdk",
"mlua",

View File

@ -28,7 +28,7 @@ tui-textarea = { version = "0.2", features = ["crossterm"], optional = true }
crossterm = { version = "0.25", optional = true }
tokio-util = { version = "0.7", optional = true }
serde = { version = "1.0", optional = true }
indexmap = { version = "2.0.2", optional = true }
indexmap = { version = "2.1.0", optional = true }
directories = "5.0.1"
[dev-dependencies]

View File

@ -89,11 +89,11 @@
]
},
"locked": {
"lastModified": 1698726852,
"narHash": "sha256-V1S4TTzg++GzPc96i/yy4jib+7/xU0LXHcggm9MllMM=",
"lastModified": 1698804896,
"narHash": "sha256-vSms7A9bWHC00343qyXuNVm65LZDagDkukpkpwC2VxY=",
"owner": "oxalica",
"repo": "rust-overlay",
"rev": "ec19bd20af08f3b004089cc12ab54c823ed899b7",
"rev": "c5f8326c668b78275eccce90839861a3c8e1d3b2",
"type": "github"
},
"original": {

View File

@ -42,7 +42,7 @@
overlays = [(import rust-overlay)];
};
nightly = true;
nightly = false;
rust =
if nightly
then

6
trinitry/.gitignore vendored Normal file
View File

@ -0,0 +1,6 @@
# build
/target
/result
# It is a library
Cargo.lock

14
trinitry/Cargo.toml Normal file
View File

@ -0,0 +1,14 @@
[package]
name = "trinitry"
version = "0.1.0"
edition = "2021"
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
[dependencies]
pest = "2.7.5"
pest_derive = {version = "2.7.5", features = ["grammar-extras"]}
# The header imports katex, a js latex parser, into the doc comments
[package.metadata.docs.rs]
rustdoc-args = [ "--html-in-header", "./docs-header.html" ]

45
trinitry/docs-header.html Normal file
View File

@ -0,0 +1,45 @@
<!doctype html>
<!-- KaTeX requires the use of the HTML5 doctype. Without it, KaTeX may not render properly -->
<html>
<head>
<link
rel="stylesheet"
href="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.css"
integrity="sha384-n8MVd4RsNIU0tAv4ct0nTaAbDJwPJzDEaqSD1odI+WdtXRGWt2kTvGFasHpSy3SV"
crossorigin="anonymous"
/>
<!-- The loading of KaTeX is deferred to speed up page rendering -->
<script
defer
src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.js"
integrity="sha384-XjKyOOlGwcjNTAIQHIpgOno0Hl1YQqzUOEleOLALmuqehneUG+vnGctmUb0ZY0l8"
crossorigin="anonymous"
></script>
<!-- To automatically render math in text elements, include the auto-render extension: -->
<script
defer
src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/contrib/auto-render.min.js"
integrity="sha384-+VBxd3r6XgURycqtZ117nYw44OOcIax56Z4dCRWbxyPt0Koah1uHoK0o4+/RRE05"
crossorigin="anonymous"
></script>
<script>
document.addEventListener("DOMContentLoaded", function () {
renderMathInElement(document.body, {
delimiters: [
{ left: "$$", right: "$$", display: true },
{ left: "\\(", right: "\\)", display: false },
{ left: "$", right: "$", display: false },
{ left: "\\[", right: "\\]", display: true },
]
// FIXME(@soispha): This removes the quotes completely <2023-10-31>
// macros: {
// "”": "\\noexpand ”",
// "“": "\\noexpand “",
// },
});
});
</script>
</head>
</html>

192
trinitry/src/lib.rs Normal file
View File

@ -0,0 +1,192 @@
//! This crate is a parser for the 'Trinitry' (not 'Trinity') language, used to map all sort of
//! Functions to a memorable command.
//!
//! This parser is more of a validator, as Trinitry does not support any language features besides
//! the aforementioned commands and arguments. That includes some simple constructs like: '||' (OR)
//! or '&&' (AND). If you need these features, simple write them in the language, you've written your
//! Function in.
//!
//! # General specification
//! ## Command
//! Basically every command can be a series of alphanumeric ASCII values.
//!
//! Correctly spoken, the Language, containing all valid command names, is just the Kleene closure
//! over an Alphabet $\Sigma$, which contains all alphanumeric characters:
//! $$ \Sigma_{cmd} = \\{x | 0 \leqslant x \leqslant 9\\} \cup \\{x | "a" \leqslant x \leqslant "z"\\} \cup \\{x | "A" \leqslant x \leqslant "Z"\\} \cup \\{"\\_", "\text{-}", "."\\} $$
//!
//! ## Argument
//! Arguments are similar to the command, although they can also contain spaces and quotes,
//! if it's quoted and additional characters (here notated as "$\\dots{}$"):
//! $$ \Sigma_{args-quoted} = \Sigma_{cmd} \cup \\{"\\text{"}", "\\ ", \\dots{}\\} $$
//! $$ \Sigma_{args-single-quoted} = \Sigma_{cmd} \cup \\{"'", "\\ ", \\dots{}\\} $$
//! $$ \Sigma_{args} = \Sigma_{cmd} \cup \\{\\dots{}\\} $$
//! Look at the [trinitry.pest](../../../src/trinitry.pest) file for a full list of the additional
//! allowed characters.
//!
//! # Examples
//! ## Command
//! A valid command would be something like that:
//! ```text
//! quit
//! ```
//! something like that would not be valid however, as Trinitry does not support these 'complex'
//! language features:
//! ```text
//! write && quit
//! ```
//! ## Arguments
//! A valid argumented command would be:
//! ```text
//! lua "function() print('Hi!') end"
//! ```
//! Whilst this would not be valid (that is, it would very likely not be what you want):
//! ```text
//! lua "function() print("Hi!") end"
//! ```
//! as the double quotes in the print statement actually unquote the argument, leaving you with
//! three arguments:
//! 1. `function() print(`
//! 1. `Hi!`
//! 1. `) end`
use std::fmt::Display;
use pest::{error::Error, Parser};
use pest_derive::Parser;
#[derive(Parser)]
#[grammar = "trinitry.pest"]
pub struct Trinitry {
command: String,
arguments: Vec<String>,
}
impl Trinitry {
pub fn new(input: &str) -> Result<Self, Error<Rule>> {
let parsed = Self::parse(Rule::trinitry, input)?;
let command = {
let command: Vec<_> = parsed.clone().find_tagged("command").collect();
// Ensure that we have only one command
// This should be ensured by the grammar, thus the 'debug_assert'
debug_assert_eq!(command.len(), 1);
// PERFORMANCE(@soispha): Replace this with `mem::take` (when pairs implements Default)
// <2023-11-01>
command
.first()
.expect("This should contain exactly one element")
.to_owned()
};
let arguments: Vec<_> = parsed.clone().find_tagged("argument").collect();
Ok(Trinitry {
command: command.as_str().to_owned(),
arguments: arguments
.iter()
.map(|arg| {
let mut arg = arg.as_str().trim();
arg = if let Some(new_arg) = arg.strip_prefix("\"") {
new_arg
} else {
arg
};
arg = if let Some(new_arg) = arg.strip_suffix("\"") {
new_arg
} else {
arg
};
arg = if let Some(new_arg) = arg.strip_prefix("'") {
new_arg
} else {
arg
};
arg = if let Some(new_arg) = arg.strip_suffix("'") {
new_arg
} else {
arg
};
arg.to_owned()
})
.collect(),
})
}
}
impl Display for Trinitry {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
if self.arguments.is_empty() {
f.write_str(&self.command)
} else {
f.write_fmt(format_args!(
"{} {}",
&self.command,
&self.arguments.join(" ")
))
}
}
}
mod tests;
#[cfg(test)]
mod test {
use crate::Trinitry;
#[test]
fn parse_cmd() {
let string = "quit";
let p = Trinitry::new(string).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(&p.command, "quit");
assert!(&p.arguments.is_empty());
}
#[test]
fn parse_arg_clean() {
let string = r##"lua print("Hi")"##;
let p = Trinitry::new(string).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(&p.command, "lua");
assert_eq!(&p.arguments[0], r#"print("Hi")"#);
}
#[test]
fn parse_arg_quote() {
let string = r##"write "some 'file' name""##;
let p = Trinitry::new(string).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(&p.command, "write");
assert_eq!(&p.arguments[0], "some 'file' name");
}
#[test]
fn parse_arg_single_quote() {
let string = r##"write 'some "file" name'"##;
let p = Trinitry::new(string).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(&p.command, "write");
assert_eq!(&p.arguments[0], "some \"file\" name");
}
#[test]
fn parse_arg_multi() {
let string = r##"write 'some "file" name' "other name" last"##;
let p = Trinitry::new(string).unwrap_or_else(|e| {
panic!("{}", e);
});
let expected_args = vec!["some \"file\" name", "other name", "last"]
.iter()
.map(|str| (*str).to_owned())
.collect::<Vec<String>>();
assert_eq!(&p.command, "write");
assert_eq!(&p.arguments, &expected_args);
}
}

1365
trinitry/src/tests.rs Normal file

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,25 @@
chars = { ASCII_ALPHANUMERIC | "_" | "-" | "." }
// TODO(@soispha): Are these all the valid characters? <2023-11-01>
argument_chars = { chars | "(" | ")" | "{" | "}" | "<" | ">" | "?" | "!" | "+" | "^" | "@"
| "&" | "*" | "~" | "|" | "=" | "," | "\\" | "/" }
whitespace = _{ " " } // lower case to avoid special treatment of 'WHITESPACE'
quote = _{ "\"" }
q = _{ quote }
single_quote = _{ "'" }
sq = _{ single_quote }
command = { chars+ }
arg_quoted = { q ~ (!q ~ (argument_chars | " " | "'" ))+ ~ q }
arg_single_quoted = { sq ~ (!sq ~ (argument_chars | " " | "\"" ))+ ~ sq }
arg = { (argument_chars | "\"" | "'")+ }
argument = { whitespace+ ~ (arg_quoted | arg_single_quoted | arg )}
trinitry = { SOI ~ #command = command ~ (#argument = argument)* ~ EOI }

88
trinitry/update.sh Executable file
View File

@ -0,0 +1,88 @@
#!/usr/bin/env sh
# Library {{{
mktmp() {
ensure_tmp_dir
mktemp -p "$LIB_TEMP_DIR_FOR_SCRIPT"
}
ensure_tmp_dir() {
if ! [ -d "$LIB_TEMP_DIR_FOR_SCRIPT" ]; then
LIB_TEMP_DIR_FOR_SCRIPT="$(mktemp -d)"
export LIB_TEMP_DIR_FOR_SCRIPT
fi
}
remove_tmp_dir() {
# The test is here because some scripts still delete this on their own
if [ -d "$LIB_TEMP_DIR_FOR_SCRIPT" ]; then
rm -r "$LIB_TEMP_DIR_FOR_SCRIPT"
fi
}
trap remove_tmp_dir EXIT
ensure_tmp_dir # ensure that the variable has been set, even in subshells
# }}}
tmp="$(mktmp)"
curl https://raw.githubusercontent.com/minimaxir/big-list-of-naughty-strings/master/blns.txt |
awk '!/^#/' |
awk '!/^[\s\t]*$/' |
awk -v s="'" '!/.*s.*/' |
# This entry contains duplicated spaces, just ignore it
awk '!/Power/' |
iconv -c -f utf-8 -t ascii \
>"$tmp"
counter=0
cat <<EOF
// DO NOT EDIT
// This file is automatically generated by the 'update.sh' file, with data from:
// https://raw.githubusercontent.com/minimaxir/big-list-of-naughty-strings/master/blns.txt
#[cfg(test)]
mod test {
use crate::Trinitry;
EOF
while read -r name; do
tmp2="$(mktmp)"
printf "%s" "$name" >"$tmp2"
if rg '^[A-Za-z0-9_.-]+$' "$tmp2" -q; then
cat <<EOF
#[test]
fn parse_$counter() {
let p = Trinitry::new(r##"$name"##).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(r##"$name"##, &p.to_string());
}
EOF
elif rg '^[A-Za-z0-9_.-]+ [A-Za-z0-9_.(){}<>?!+^@&*~|=,/\\ -]*$' "$tmp2" -q; then
cat <<EOF
#[test]
fn parse_$counter() {
let p = Trinitry::new(r##"$name"##).unwrap_or_else(|e| {
panic!("{}", e);
});
assert_eq!(r##"$name"##, &p.to_string());
}
EOF
else
cat <<EOF
#[test]
fn parse_$counter() {
let p = Trinitry::new(r##"$name"##);
assert!(p.is_err());
}
EOF
fi
counter=$((counter + 1))
done <"$tmp"
cat <<EOF
}
EOF
# vim: ft=sh

View File

@ -1,13 +1,24 @@
#!/usr/bin/env bash
#!/usr/bin/env sh
cargo update && cargo upgrade
cd ./language_macros || (echo "No language_macros directory" && exit 1)
cargo update && cargo upgrade
cd ..
git add Cargo.lock Cargo.toml flake.lock ./language_macros/Cargo.toml
cd ./trinitry || (echo "No trinitry directory" && exit 1)
cargo update && cargo upgrade
./update.sh > ./src/tests.rs
cd ..
git add \
Cargo.lock Cargo.toml \
flake.lock \
./language_macros/Cargo.toml \
./trinitry/Cargo.toml \
./trinitry/src/tests.rs
# vim: ft=sh