diff --git a/Cargo.lock b/Cargo.lock index f5e14a1d76..41dde0fecb 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -2196,7 +2196,6 @@ dependencies = [ "nu_plugin_fetch", "nu_plugin_inc", "nu_plugin_match", - "nu_plugin_parse", "nu_plugin_post", "nu_plugin_ps", "nu_plugin_start", @@ -2497,19 +2496,6 @@ dependencies = [ "regex", ] -[[package]] -name = "nu_plugin_parse" -version = "0.14.1" -dependencies = [ - "futures 0.3.5", - "nu-build", - "nu-errors", - "nu-plugin", - "nu-protocol", - "nu-source", - "regex", -] - [[package]] name = "nu_plugin_post" version = "0.14.1" diff --git a/Cargo.toml b/Cargo.toml index e6ef044ad2..6a273e027f 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -29,7 +29,6 @@ nu_plugin_binaryview = { version = "0.14.1", path = "./crates/nu_plugin_binaryvi nu_plugin_fetch = { version = "0.14.1", path = "./crates/nu_plugin_fetch", optional=true } nu_plugin_inc = { version = "0.14.1", path = "./crates/nu_plugin_inc", optional=true } nu_plugin_match = { version = "0.14.1", path = "./crates/nu_plugin_match", optional=true } -nu_plugin_parse = { version = "0.14.1", path = "./crates/nu_plugin_parse", optional=true } nu_plugin_post = { version = "0.14.1", path = "./crates/nu_plugin_post", optional=true } nu_plugin_ps = { version = "0.14.1", path = "./crates/nu_plugin_ps", optional=true } nu_plugin_start = { version = "0.1.0", path = "./crates/nu_plugin_start", optional=true } @@ -59,7 +58,7 @@ nu-build = { version = "0.14.1", path = "./crates/nu-build" } [features] default = ["sys", "ps", "textview", "inc"] -stable = ["default", "starship-prompt", "binaryview", "match", "tree", "parse", "post", "fetch", "clipboard-cli", "trash-support", "start"] +stable = ["default", "starship-prompt", "binaryview", "match", "tree", "post", "fetch", "clipboard-cli", "trash-support", "start"] # Default textview = ["crossterm", "syntect", "url", "nu_plugin_textview"] @@ -71,7 +70,6 @@ inc = ["semver", "nu_plugin_inc"] binaryview = ["nu_plugin_binaryview"] fetch = ["nu_plugin_fetch"] match = ["nu_plugin_match"] -parse = ["nu_plugin_parse"] post = ["nu_plugin_post"] trace = ["nu-parser/trace"] tree = ["nu_plugin_tree"] @@ -120,11 +118,6 @@ name = "nu_plugin_stable_match" path = "src/plugins/nu_plugin_stable_match.rs" required-features = ["match"] -[[bin]] -name = "nu_plugin_stable_parse" -path = "src/plugins/nu_plugin_stable_parse.rs" -required-features = ["parse"] - [[bin]] name = "nu_plugin_stable_post" path = "src/plugins/nu_plugin_stable_post.rs" diff --git a/crates/nu-cli/src/cli.rs b/crates/nu-cli/src/cli.rs index 24d1ed9fde..bd5097812e 100644 --- a/crates/nu-cli/src/cli.rs +++ b/crates/nu-cli/src/cli.rs @@ -291,6 +291,7 @@ pub fn create_default_context( whole_stream_command(Lines), whole_stream_command(Trim), whole_stream_command(Echo), + whole_stream_command(Parse), whole_stream_command(Str), whole_stream_command(StrToDecimal), whole_stream_command(StrToInteger), diff --git a/crates/nu-cli/src/commands.rs b/crates/nu-cli/src/commands.rs index 87b8e8834b..b9e416d348 100644 --- a/crates/nu-cli/src/commands.rs +++ b/crates/nu-cli/src/commands.rs @@ -74,6 +74,7 @@ pub(crate) mod mv; pub(crate) mod next; pub(crate) mod nth; pub(crate) mod open; +pub(crate) mod parse; pub(crate) mod pivot; pub(crate) mod plugin; pub(crate) mod prepend; @@ -203,6 +204,7 @@ pub(crate) use mv::Move; pub(crate) use next::Next; pub(crate) use nth::Nth; pub(crate) use open::Open; +pub(crate) use parse::Parse; pub(crate) use pivot::Pivot; pub(crate) use prepend::Prepend; pub(crate) use prev::Previous; diff --git a/crates/nu-cli/src/commands/parse/command.rs b/crates/nu-cli/src/commands/parse/command.rs new file mode 100644 index 0000000000..80d55bfaaf --- /dev/null +++ b/crates/nu-cli/src/commands/parse/command.rs @@ -0,0 +1,175 @@ +use crate::commands::WholeStreamCommand; +use crate::prelude::*; +use nu_errors::ShellError; +use nu_protocol::{Signature, SyntaxShape, TaggedDictBuilder, UntaggedValue, Value}; +use nu_source::Tagged; + +use regex::Regex; + +#[derive(Deserialize)] +struct Arguments { + pattern: Tagged, + regex: Tagged, +} + +pub struct Command; + +#[async_trait] +impl WholeStreamCommand for Command { + fn name(&self) -> &str { + "parse" + } + + fn signature(&self) -> Signature { + Signature::build("parse") + .required( + "pattern", + SyntaxShape::String, + "the pattern to match. Eg) \"{foo}: {bar}\"", + ) + .switch("regex", "use full regex syntax for patterns", Some('r')) + } + + fn usage(&self) -> &str { + "Parse columns from string data using a simple pattern." + } + + async fn run( + &self, + args: CommandArgs, + registry: &CommandRegistry, + ) -> Result { + operate(args, registry).await + } +} + +pub async fn operate( + args: CommandArgs, + registry: &CommandRegistry, +) -> Result { + let name_tag = args.call_info.name_tag.clone(); + let (Arguments { regex, pattern }, mut input) = args.process(®istry).await?; + + let regex_pattern = if let Tagged { item: true, tag } = regex { + Regex::new(&pattern.item) + .map_err(|_| ShellError::labeled_error("Invalid regex", "invalid regex", tag.span))? + } else { + let parse_regex = build_regex(&pattern.item, name_tag.clone())?; + + Regex::new(&parse_regex).map_err(|_| { + ShellError::labeled_error("Invalid pattern", "invalid pattern", name_tag.span) + })? + }; + + let columns = column_names(®ex_pattern); + let mut parsed: VecDeque = VecDeque::new(); + + while let Some(v) = input.next().await { + match v.as_string() { + Ok(s) => { + let results = regex_pattern.captures_iter(&s); + + for c in results { + let mut dict = TaggedDictBuilder::new(&v.tag); + + for (column_name, cap) in columns.iter().zip(c.iter().skip(1)) { + let cap_string = cap.map(|v| v.as_str()).unwrap_or("").to_string(); + dict.insert_untagged(column_name, UntaggedValue::string(cap_string)); + } + + parsed.push_back(dict.into_value()); + } + } + Err(_) => { + return Err(ShellError::labeled_error_with_secondary( + "Expected string input", + "expected string input", + &name_tag, + "value originated here", + v.tag, + )) + } + } + } + + Ok(futures::stream::iter(parsed).to_output_stream()) +} + +fn build_regex(input: &str, tag: Tag) -> Result { + let mut output = "(?s)\\A".to_string(); + + //let mut loop_input = input; + let mut loop_input = input.chars().peekable(); + loop { + let mut before = String::new(); + while let Some(c) = loop_input.next() { + if c == '{' { + // If '{{', still creating a plaintext parse command, but just for a single '{' char + if loop_input.peek() == Some(&'{') { + let _ = loop_input.next(); + } else { + break; + } + } + before.push(c); + } + + if !before.is_empty() { + output.push_str(®ex::escape(&before)); + } + + // Look for column as we're now at one + let mut column = String::new(); + while let Some(c) = loop_input.next() { + if c == '}' { + break; + } + column.push(c); + + if loop_input.peek().is_none() { + return Err(ShellError::labeled_error( + "Found opening `{` without an associated closing `}`", + "invalid parse pattern", + tag, + )); + } + } + + if !column.is_empty() { + output.push_str("(?P<"); + output.push_str(&column); + output.push_str(">.*?)"); + } + + if before.is_empty() && column.is_empty() { + break; + } + } + + output.push_str("\\z"); + Ok(output) +} + +fn column_names(regex: &Regex) -> Vec { + regex + .capture_names() + .enumerate() + .skip(1) + .map(|(i, name)| { + name.map(String::from) + .unwrap_or_else(|| format!("Capture{}", i)) + }) + .collect() +} + +#[cfg(test)] +mod tests { + use super::Command; + + #[test] + fn examples_work_as_expected() { + use crate::examples::test as test_examples; + + test_examples(Command {}) + } +} diff --git a/crates/nu-cli/src/commands/parse/mod.rs b/crates/nu-cli/src/commands/parse/mod.rs new file mode 100644 index 0000000000..e7c608da74 --- /dev/null +++ b/crates/nu-cli/src/commands/parse/mod.rs @@ -0,0 +1,3 @@ +mod command; + +pub use command::Command as Parse; diff --git a/crates/nu_plugin_parse/Cargo.toml b/crates/nu_plugin_parse/Cargo.toml deleted file mode 100644 index 258526d688..0000000000 --- a/crates/nu_plugin_parse/Cargo.toml +++ /dev/null @@ -1,21 +0,0 @@ -[package] -name = "nu_plugin_parse" -version = "0.14.1" -authors = ["The Nu Project Contributors"] -edition = "2018" -description = "A string parsing plugin for Nushell" -license = "MIT" - -[lib] -doctest = false - -[dependencies] -nu-plugin = { path = "../nu-plugin", version = "0.14.1" } -nu-protocol = { path = "../nu-protocol", version = "0.14.1" } -nu-source = { path = "../nu-source", version = "0.14.1" } -nu-errors = { path = "../nu-errors", version = "0.14.1" } -futures = { version = "0.3", features = ["compat", "io-compat"] } -regex = "1" - -[build-dependencies] -nu-build = { version = "0.14.1", path = "../nu-build" } diff --git a/crates/nu_plugin_parse/build.rs b/crates/nu_plugin_parse/build.rs deleted file mode 100644 index b7511cfc6a..0000000000 --- a/crates/nu_plugin_parse/build.rs +++ /dev/null @@ -1,3 +0,0 @@ -fn main() -> Result<(), Box> { - nu_build::build() -} diff --git a/crates/nu_plugin_parse/src/lib.rs b/crates/nu_plugin_parse/src/lib.rs deleted file mode 100644 index b5bd5842d0..0000000000 --- a/crates/nu_plugin_parse/src/lib.rs +++ /dev/null @@ -1,4 +0,0 @@ -mod nu; -mod parse; - -pub use parse::Parse; diff --git a/crates/nu_plugin_parse/src/main.rs b/crates/nu_plugin_parse/src/main.rs deleted file mode 100644 index 4a7d5ee3c1..0000000000 --- a/crates/nu_plugin_parse/src/main.rs +++ /dev/null @@ -1,7 +0,0 @@ -use nu_plugin::serve_plugin; -use nu_plugin_parse::Parse; - -fn main() -> Result<(), Box> { - serve_plugin(&mut Parse::new()?); - Ok(()) -} diff --git a/crates/nu_plugin_parse/src/nu/mod.rs b/crates/nu_plugin_parse/src/nu/mod.rs deleted file mode 100644 index 357572e7c2..0000000000 --- a/crates/nu_plugin_parse/src/nu/mod.rs +++ /dev/null @@ -1,159 +0,0 @@ -use regex::{self, Regex}; - -use nu_errors::ShellError; -use nu_plugin::Plugin; -use nu_protocol::{ - CallInfo, Primitive, ReturnSuccess, ReturnValue, ShellTypeName, Signature, SyntaxShape, - TaggedDictBuilder, UntaggedValue, Value, -}; -use nu_source::Tag; - -use crate::Parse; - -impl Plugin for Parse { - fn config(&mut self) -> Result { - Ok(Signature::build("parse") - .switch("regex", "use full regex syntax for patterns", Some('r')) - .required( - "pattern", - SyntaxShape::String, - "the pattern to match. Eg) \"{foo}: {bar}\"", - ) - .filter()) - } - - fn begin_filter(&mut self, call_info: CallInfo) -> Result, ShellError> { - if let Some(ref args) = call_info.args.positional { - let value = &args[0]; - match value { - Value { - value: UntaggedValue::Primitive(Primitive::String(s)), - tag, - } => { - self.pattern_tag = tag.clone(); - self.regex = if call_info.args.has("regex") { - Regex::new(&s).map_err(|_| { - ShellError::labeled_error("Invalid regex", "invalid regex", tag.span) - })? - } else { - let parse_regex = build_regex(&s, tag.clone())?; - Regex::new(&parse_regex).map_err(|_| { - ShellError::labeled_error( - "Invalid pattern", - "invalid pattern", - tag.span, - ) - })? - }; - - self.column_names = column_names(&self.regex); - } - Value { tag, .. } => { - return Err(ShellError::labeled_error( - format!( - "Unexpected type in params (found `{}`, expected `String`)", - value.type_name() - ), - "unexpected type", - tag, - )); - } - } - } - - Ok(vec![]) - } - - fn filter(&mut self, input: Value) -> Result, ShellError> { - if let Ok(s) = input.as_string() { - Ok(self - .regex - .captures_iter(&s) - .map(|caps| { - let mut dict = TaggedDictBuilder::new(&input.tag); - for (column_name, cap) in self.column_names.iter().zip(caps.iter().skip(1)) { - let cap_string = cap.map(|v| v.as_str()).unwrap_or("").to_string(); - dict.insert_untagged(column_name, UntaggedValue::string(cap_string)); - } - - Ok(ReturnSuccess::Value(dict.into_value())) - }) - .collect()) - } else { - Err(ShellError::labeled_error_with_secondary( - "Expected string input", - "expected string input", - &self.name, - "value originated here", - input.tag, - )) - } - } -} - -fn build_regex(input: &str, tag: Tag) -> Result { - let mut output = "(?s)\\A".to_string(); - - //let mut loop_input = input; - let mut loop_input = input.chars().peekable(); - loop { - let mut before = String::new(); - while let Some(c) = loop_input.next() { - if c == '{' { - // If '{{', still creating a plaintext parse command, but just for a single '{' char - if loop_input.peek() == Some(&'{') { - let _ = loop_input.next(); - } else { - break; - } - } - before.push(c); - } - - if !before.is_empty() { - output.push_str(®ex::escape(&before)); - } - - // Look for column as we're now at one - let mut column = String::new(); - while let Some(c) = loop_input.next() { - if c == '}' { - break; - } - column.push(c); - - if loop_input.peek().is_none() { - return Err(ShellError::labeled_error( - "Found opening `{` without an associated closing `}`", - "invalid parse pattern", - tag, - )); - } - } - - if !column.is_empty() { - output.push_str("(?P<"); - output.push_str(&column); - output.push_str(">.*?)"); - } - - if before.is_empty() && column.is_empty() { - break; - } - } - - output.push_str("\\z"); - Ok(output) -} - -fn column_names(regex: &Regex) -> Vec { - regex - .capture_names() - .enumerate() - .skip(1) - .map(|(i, name)| { - name.map(String::from) - .unwrap_or_else(|| format!("Capture{}", i)) - }) - .collect() -} diff --git a/crates/nu_plugin_parse/src/parse.rs b/crates/nu_plugin_parse/src/parse.rs deleted file mode 100644 index ba82632442..0000000000 --- a/crates/nu_plugin_parse/src/parse.rs +++ /dev/null @@ -1,21 +0,0 @@ -use nu_source::Tag; -use regex::Regex; - -pub struct Parse { - pub regex: Regex, - pub name: Tag, - pub pattern_tag: Tag, - pub column_names: Vec, -} - -impl Parse { - #[allow(clippy::trivial_regex)] - pub fn new() -> Result> { - Ok(Parse { - regex: Regex::new("")?, - name: Tag::unknown(), - pattern_tag: Tag::unknown(), - column_names: vec![], - }) - } -}