Jesús Pérez b99dcc83c3 feat: major repository modernization and tracking cleanup
## Summary

Comprehensive repository cleanup focusing on plugin dependency management, documentation improvements, and git tracking optimization.

## Key Changes

### 🔧 Core Infrastructure
- Synchronized all nu-* dependencies across plugins for version consistency
- Enhanced upstream tracking and automation systems
- Removed nushell directory from git tracking for cleaner repository management

### 📚 Documentation
- Significantly expanded README.md with comprehensive development guides
- Added detailed workflow documentation and command references
- Improved plugin collection overview and usage examples

### 🧹 Repository Cleanup
- Removed legacy bash scripts (build-all.sh, collect-install.sh, make_plugin.sh)
- Streamlined automation through unified justfile and nushell script approach
- Updated .gitignore with nushell directory and archive patterns
- Removed nushell directory from git tracking to prevent unwanted changes

### 🔌 Plugin Updates
- **nu_plugin_image**: Major refactoring with modular architecture improvements
- **nu_plugin_hashes**: Enhanced functionality and build system improvements
- **nu_plugin_highlight**: Updated for new plugin API compatibility
- **nu_plugin_clipboard**: Dependency synchronization
- **nu_plugin_desktop_notifications**: Version alignment
- **nu_plugin_port_extension & nu_plugin_qr_maker**: Consistency updates
- **nu_plugin_kcl & nu_plugin_tera**: Submodule synchronization

### 🏗️ Git Tracking Optimization
- Removed nushell directory from version control for cleaner repository management
- Added comprehensive .gitignore patterns for build artifacts and archives

## Statistics
- 2,082 files changed
- 2,373 insertions, 339,936 deletions
- Net reduction of 337,563 lines (primarily from removing nushell directory tracking)

## Benefits
- Complete version consistency across all plugins
- Cleaner repository with optimized git tracking
- Improved developer experience with streamlined workflows
- Enhanced documentation and automation
- Reduced repository size and complexity

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-20 15:18:58 +01:00

174 lines
4.1 KiB
Rust

//! Contains a generic trait for hashers that implement `Digest`.
//! This implementation is stolen with minimal changes from *generic_digest.rs*
//! source code file of nushell v0.101.0 which can be found at
//! https://github.com/nushell/nushell/blob/0.101.0/crates/nu-command/src/hash/generic_digest.rs
//! The *hash* module is private, so I had no choice.
use std::{io::Write, marker::PhantomData, ops::Not};
use digest::{Digest, Output};
use nu_cmd_base::input_handler::{operate, CmdArgument};
use nu_plugin::{EngineInterface, EvaluatedCall, PluginCommand};
use nu_protocol::{
ast::CellPath,
Category,
Example,
IntoPipelineData,
LabeledError,
PipelineData,
ShellError,
Signature,
Span,
SyntaxShape,
Type,
Value,
};
use crate::HashesPlugin;
pub trait Hasher: Digest + Clone {
fn name() -> &'static str;
fn examples() -> Vec<Example<'static>>;
}
#[derive(Clone)]
pub struct GenericHasher<H: Hasher> {
name: String,
description: String,
_hasher: PhantomData<H>,
}
impl<H: Hasher> Default for GenericHasher<H> {
fn default() -> Self {
Self {
name: format!("hash {}", H::name()),
description: format!(
"Hash a value using the {} hash algorithm.",
H::name()
),
_hasher: PhantomData,
}
}
}
struct Arguments {
cell_paths: Option<Vec<CellPath>>,
binary: bool,
}
impl CmdArgument for Arguments {
fn take_cell_paths(&mut self) -> Option<Vec<CellPath>> {
self.cell_paths.take()
}
}
impl<H> PluginCommand for GenericHasher<H>
where
H: Hasher + Write + Send + Sync + 'static,
Output<H>: core::fmt::LowerHex,
{
type Plugin = HashesPlugin;
fn name(&self) -> &str {
&self.name
}
fn signature(&self) -> Signature {
Signature::build(self.name())
.category(Category::Hash)
.input_output_types(vec![
(Type::Binary, Type::Any),
(Type::String, Type::Any),
(Type::table(), Type::table()),
(Type::record(), Type::record()),
])
.allow_variants_without_examples(true)
.switch(
"binary",
"Output binary instead of hexadecimal representation",
Some('b'),
)
.rest(
"rest",
SyntaxShape::CellPath,
format!("Optionally {} hash data by cell path.", H::name()),
)
}
fn description(&self) -> &str {
&self.description
}
fn examples(&self) -> Vec<Example<'_>> {
H::examples()
}
fn run(
&self,
_plugin: &HashesPlugin,
engine: &EngineInterface,
call: &EvaluatedCall,
input: PipelineData,
) -> Result<PipelineData, LabeledError> {
let head = call.head;
let binary = call.has_flag("binary")?;
let cell_paths: Vec<CellPath> = call.rest(0)?;
let cell_paths = cell_paths.is_empty().not().then_some(cell_paths);
if let PipelineData::ByteStream(stream, ..) = input {
let mut hasher = H::new();
stream.write_to(&mut hasher)?;
let digest = hasher.finalize();
if binary {
Ok(Value::binary(digest.to_vec(), head).into_pipeline_data())
} else {
Ok(Value::string(format!("{digest:x}"), head).into_pipeline_data())
}
} else {
operate(
action::<H>,
Arguments { binary, cell_paths },
input,
head,
engine.signals(),
)
.map_err(Into::into)
}
}
}
fn action<H>(input: &Value, args: &Arguments, _span: Span) -> Value
where
H: Hasher,
Output<H>: core::fmt::LowerHex,
{
let span = input.span();
let (bytes, span) = match input {
Value::String { val, .. } => (val.as_bytes(), span),
Value::Binary { val, .. } => (val.as_slice(), span),
// Propagate existing errors
Value::Error { .. } => return input.clone(),
other => {
let span = input.span();
return Value::error(
ShellError::OnlySupportsThisInputType {
exp_input_type: "string or binary".into(),
wrong_type: other.get_type().to_string(),
dst_span: span,
src_span: other.span(),
},
span,
);
}
};
let digest = H::digest(bytes);
if args.binary {
Value::binary(digest.to_vec(), span)
} else {
Value::string(format!("{digest:x}"), span)
}
}