From 82da6dd1ef9dbda9ac5e0894aef0a71750f7bdcd Mon Sep 17 00:00:00 2001
From: Michael Aaron Murphy <mmstickman@gmail.com>
Date: Sun, 30 Jul 2017 21:25:04 -0400
Subject: [PATCH] Eliminate PEG Dependency
---
Cargo.lock | 16 ----
Cargo.toml | 1 -
build.rs | 3 -
examples/builtin_piping.ion | 5 --
examples/builtin_piping.out | 2 -
src/parser/grammar.rustpeg | 86 ------------------
src/parser/peg.rs | 173 ++++++++++++++++++++----------------
7 files changed, 97 insertions(+), 189 deletions(-)
delete mode 100644 src/parser/grammar.rustpeg
diff --git a/Cargo.lock b/Cargo.lock
index d5003ae2..fb67f68a 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -12,7 +12,6 @@ dependencies = [
"libc 0.2.28 (registry+https://github.com/rust-lang/crates.io-index)",
"liner 0.4.0 (git+https://github.com/MovingtoMars/liner/)",
"nix 0.8.1 (registry+https://github.com/rust-lang/crates.io-index)",
- "peg 0.5.4 (registry+https://github.com/rust-lang/crates.io-index)",
"permutate 0.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
"redox_syscall 0.1.28 (registry+https://github.com/rust-lang/crates.io-index)",
"regex 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -130,24 +129,11 @@ dependencies = [
"winapi-build 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
]
-[[package]]
-name = "peg"
-version = "0.5.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "quote 0.3.15 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
[[package]]
name = "permutate"
version = "0.3.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
-[[package]]
-name = "quote"
-version = "0.3.15"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
[[package]]
name = "redox_syscall"
version = "0.1.28"
@@ -297,9 +283,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
"checksum memchr 1.0.1 (registry+https://github.com/rust-lang/crates.io-index)" = "1dbccc0e46f1ea47b9f17e6d67c5a96bd27030519c519c9c91327e31275a47b4"
"checksum nix 0.8.1 (registry+https://github.com/rust-lang/crates.io-index)" = "47e49f6982987135c5e9620ab317623e723bd06738fd85377e8d55f57c8b6487"
"checksum ole32-sys 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)" = "5d2c49021782e5233cd243168edfa8037574afed4eba4bbaf538b3d8d1789d8c"
-"checksum peg 0.5.4 (registry+https://github.com/rust-lang/crates.io-index)" = "36a474cba42744afe0f223e9d4263594b3387f172e512259c72d2011e477c4fb"
"checksum permutate 0.3.2 (registry+https://github.com/rust-lang/crates.io-index)" = "53b7d5b19a715ffab38693a9dd44b067fdfa2b18eef65bd93562dfe507022fae"
-"checksum quote 0.3.15 (registry+https://github.com/rust-lang/crates.io-index)" = "7a6e920b65c65f10b2ae65c831a81a073a89edd28c7cce89475bff467ab4167a"
"checksum redox_syscall 0.1.28 (registry+https://github.com/rust-lang/crates.io-index)" = "ddab7acd8e7bf3e49dfdf78ac1209b992329eb2f66e0bf672ab49c70a76d1d68"
"checksum redox_termios 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)" = "bc495930de8d330f14856cface52561b7d79a072c76e438cf8f34d7233a35fa7"
"checksum regex 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "1731164734096285ec2a5ec7fea5248ae2f5485b3feeb0115af4fda2183b2d1b"
diff --git a/Cargo.toml b/Cargo.toml
index e56116bf..e172c336 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -65,7 +65,6 @@ users = "0.5.1"
[build-dependencies]
ansi_term = "0.9"
-peg = "0.5"
version_check = "0.1.3"
[profile.release]
diff --git a/build.rs b/build.rs
index 525c5949..d74a7c3f 100644
--- a/build.rs
+++ b/build.rs
@@ -1,5 +1,4 @@
extern crate ansi_term;
-extern crate peg;
extern crate version_check;
use ansi_term::Color::{Red, Yellow, Blue, White};
@@ -46,6 +45,4 @@ fn main() {
println!("cargo:warning={}", "Build may fail due to incompatible rustc version.");
}
}
-
- peg::cargo_build("src/parser/grammar.rustpeg");
}
diff --git a/examples/builtin_piping.ion b/examples/builtin_piping.ion
index 8d161a80..30a70e91 100644
--- a/examples/builtin_piping.ion
+++ b/examples/builtin_piping.ion
@@ -1,9 +1,4 @@
matches Foo '([A-Z])\w+' && echo true
matches foo '([A-Z])\w+' || echo false
-
-fn foobar x; end
-
-fn | tr '[a-z]' '[A-Z]'
-
read foo <<< $(echo bar)
echo $foo
diff --git a/examples/builtin_piping.out b/examples/builtin_piping.out
index 5a56044f..cd0c5dc5 100644
--- a/examples/builtin_piping.out
+++ b/examples/builtin_piping.out
@@ -1,5 +1,3 @@
true
false
-# FUNCTIONS
- FOOBAR
bar
diff --git a/src/parser/grammar.rustpeg b/src/parser/grammar.rustpeg
deleted file mode 100644
index f22b2131..00000000
--- a/src/parser/grammar.rustpeg
+++ /dev/null
@@ -1,86 +0,0 @@
-use parser::pipelines;
-use parser::peg::get_function_args;
-use shell::flow_control::*;
-
-#[pub]
-parse_ -> Statement
- = fn_
- / match_
- / case_
- / pipelines
-
-#[pub]
-fn_ -> Statement
- = "fn " n:_name whitespace* args:_fn_args whitespace* description:_description? {?
- get_function_args(args).map(|args| Statement::Function {
- description: description.unwrap_or("".into()),
- name: n.into(),
- args: args,
- statements: Vec::new(),
- }).ok_or("ion: invalid function argument\n")
- }
-
-_description -> String
- = "--" whitespace* description:$([^\r\n]*) { description.into() }
-
-_name -> String
- = n:$([A-z0-9_]+) { n.into() }
-
-_fn_args -> Vec<String>
- = _fn_arg ** " "
-
-_args -> Vec<String>
- = _arg ** " "
-
-_arg -> String
- = n:$([A-z0-9_]+) { n.into() }
-
-_fn_arg -> String
- = n:$([A-z0-9_:]+) { n.into()}
-
-wildcard_ -> Option<String> = "_" { None }
-value_ -> Option<String> = contents:$(.*) { Some(contents.into())}
-
-pattern_ -> Option<String> = wildcard_ / value_
-
-case_ -> Statement
- = "case" whitespace+ p:pattern_ {
- Statement::Case(Case { value: p, statements: Vec::new() })
- }
-
-#[pub]
-match_ -> Statement = "match" whitespace+ expression:$(.*) {
- Statement::Match {
- expression: expression.into(),
- cases: Vec::new()
- }
-}
-
-pub pipelines -> Statement
- = (unused* newline)* [#] .* { Statement::Default }
- / [ \n\t\r]* pipeline:_pipelines { pipeline }
- / (unused*) ** newline { Statement::Default }
-
-
-// Converts the pipeline string into a statement, handling redirection, piping, and backgrounds.
-_pipelines -> Statement
- = command:$(.+) {?
- pipelines::Collector::run(command).map(Statement::Pipeline)
-}
-
-unused -> ()
- = whitespace comment? { () }
- / comment { () }
-
-comment -> ()
- = [#] [^\r\n]*
-
-whitespace -> ()
- = [ \t]+
-
-job_ending -> ()
- = [;]
- / newline
-
-newline -> ()
- = [\r\n]
diff --git a/src/parser/peg.rs b/src/parser/peg.rs
index 93e2ecb3..3ca5e69f 100644
--- a/src/parser/peg.rs
+++ b/src/parser/peg.rs
@@ -1,14 +1,12 @@
use std::char;
use std::fmt;
-use std::io::{Write, stderr};
-use self::grammar::parse_;
use super::{ArgumentSplitter, pipelines};
use super::{ExpanderFunctions, Select, expand_string};
use super::assignments::parse_assignment;
use shell::{Job, JobKind};
use shell::directory_stack::DirectoryStack;
-use shell::flow_control::{ElseIf, FunctionArgument, Statement, Type};
+use shell::flow_control::{Case, ElseIf, FunctionArgument, Statement, Type};
use shell::variables::Variables;
#[derive(Debug, PartialEq, Clone, Copy)]
@@ -128,12 +126,20 @@ fn collect<F>(arguments: &str, statement: F) -> Statement
}
}
+fn is_valid_name(name: &str) -> bool {
+ !name.chars().any(|c| !(c.is_alphanumeric() || c == '_'))
+}
+
pub fn parse(code: &str) -> Statement {
let cmd = code.trim();
match cmd {
"end" => return Statement::End,
"break" => return Statement::Break,
"continue" => return Statement::Continue,
+ "for" | "match" | "case" => {
+ eprintln!("ion: syntax error: incomplete control flow statement");
+ return Statement::Default;
+ }
_ if cmd.starts_with("let ") => return Statement::Let { expression: parse_assignment(cmd[4..].trim_left()) },
_ if cmd.starts_with("export ") => return Statement::Export(parse_assignment(cmd[7..].trim_left())),
_ if cmd.starts_with("if ") => {
@@ -181,30 +187,106 @@ pub fn parse(code: &str) -> Statement {
let variable = &cmd[..pos];
cmd = &cmd[pos..].trim_left();
- if cmd.starts_with("in ") {
- cmd = cmd[3..].trim_left();
- } else {
+ if !cmd.starts_with("in ") {
eprintln!("ion: syntax error: incorrect for loop syntax");
return Statement::Default;
}
return Statement::For {
variable: variable.into(),
- values: ArgumentSplitter::new(cmd).map(String::from).collect(),
+ values: ArgumentSplitter::new(cmd[3..].trim_left()).map(String::from).collect(),
statements: Vec::new(),
};
}
- _ => (),
- }
+ _ if cmd.starts_with("case ") => {
+ let value = match cmd[5..].trim_left() {
+ "_" => None,
+ value @ _ => Some(value.into())
+ };
+ return Statement::Case(Case { value: value, statements: Vec::new() });
+ }
+ _ if cmd.starts_with("match ") => {
+ return Statement::Match {
+ expression: cmd[6..].trim_left().into(),
+ cases: Vec::new()
+ };
+ }
+ _ if cmd.starts_with("fn ") => {
+ let cmd = cmd[3..].trim_left();
+ let pos = cmd.find(char::is_whitespace).unwrap_or(cmd.len());
+ let name = &cmd[..pos];
+ if !is_valid_name(name) {
+ eprintln!("ion: syntax error: {} is not a valid function name\n \
+ Function names may only contain alphanumeric characters", name);
+ return Statement::Default;
+ }
- match parse_(cmd) {
- Ok(code_ok) => code_ok,
- Err(err) => {
- let stderr = stderr();
- let _ = writeln!(stderr.lock(), "ion: Syntax {}", err);
- Statement::Default
+ let mut args_iter = cmd[pos..].split_whitespace();
+ let mut args = Vec::new();
+ let mut description = String::new();
+ let mut description_flag = 0u8;
+
+ while let Some(arg) = args_iter.next() {
+ if arg.starts_with("--") {
+ if arg.len() > 2 {
+ description.push_str(&arg[2..]);
+ description_flag |= 1;
+ }
+ description_flag |= 2;
+ break
+ } else {
+ args.push(arg.to_owned());
+ }
+ }
+
+ if description_flag & 2 != 0 {
+ if description_flag & 1 != 0 {
+ if let Some(arg) = args_iter.next() {
+ description.push(' ');
+ description.push_str(&arg);
+ }
+
+ for argument in args_iter {
+ description.push(' ');
+ description.push_str(&argument);
+ }
+ } else {
+ if let Some(arg) = args_iter.next() {
+ description.push_str(&arg);
+ }
+
+ for argument in args_iter {
+ description.push(' ');
+ description.push_str(&argument);
+ }
+ }
+ }
+
+ match get_function_args(args) {
+ Some(args) => {
+ return Statement::Function {
+ description: description,
+ name: name.into(),
+ args: args,
+ statements: Vec::new(),
+ };
+ }
+ None => {
+ eprintln!("ion: syntax error: invalid arguments");
+ return Statement::Default;
+ }
+ }
}
+ _ => ()
+ }
+
+
+ if cmd.is_empty() || cmd.starts_with('#') {
+ Statement::Default
+ } else {
+ collect(cmd, Statement::Pipeline)
}
+
}
pub fn get_function_args(args: Vec<String>) -> Option<Vec<FunctionArgument>> {
@@ -247,73 +329,12 @@ pub fn get_function_args(args: Vec<String>) -> Option<Vec<FunctionArgument>> {
Some(fn_args)
}
-mod grammar {
- include!(concat!(env!("OUT_DIR"), "/grammar.rs"));
-}
-
#[cfg(test)]
mod tests {
use super::*;
- use super::grammar::*;
use shell::JobKind;
use shell::flow_control::Statement;
- #[test]
- fn full_script() {
- pipelines(
- r#"if a == a
- echo true a == a
-
- if b != b
- echo true b != b
- else
- echo false b != b
-
- if 3 > 2
- echo true 3 > 2
- else
- echo false 3 > 2
- fi
- fi
-else
- echo false a == a
-fi
-"#,
- ).unwrap(); // Make sure it parses
- }
-
- #[test]
- fn leading_and_trailing_junk() {
- pipelines(
- r#"
-
-# comment
- # comment
-
-
- if a == a
- echo true a == a # Line ending commment
-
- if b != b
- echo true b != b
- else
- echo false b != b
-
- if 3 > 2
- echo true 3 > 2
- else
- echo false 3 > 2
- fi
- fi
-else
- echo false a == a
- fi
-
-# comment
-
-"#,
- ).unwrap(); // Make sure it parses
- }
#[test]
fn parsing_ifs() {
// Default case where spaced normally
--
GitLab