diff options
| author | sigoden <sigoden@gmail.com> | 2025-02-08 08:46:34 +0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-02-08 08:46:34 +0800 |
| commit | eb527ea758b49adbc1e311638cc6999190ff1124 (patch) | |
| tree | e2396eb8c92adb7cd236bbe1b6c0fa4ec26f6d05 /src/utils | |
| parent | a491820ab5ff9bb58e62fd25d88f31ab53a81db9 (diff) | |
| download | aichat-eb527ea758b49adbc1e311638cc6999190ff1124.tar.gz | |
feat: enhance `.file` for loading resources from diverse sources (#1155)
Diffstat (limited to 'src/utils')
| -rw-r--r-- | src/utils/command.rs | 5 | ||||
| -rw-r--r-- | src/utils/loader.rs | 47 | ||||
| -rw-r--r-- | src/utils/path.rs | 15 |
3 files changed, 62 insertions, 5 deletions
diff --git a/src/utils/command.rs b/src/utils/command.rs index 32e470d..3008028 100644 --- a/src/utils/command.rs +++ b/src/utils/command.rs @@ -107,9 +107,8 @@ pub fn run_command_with_output<T: AsRef<OsStr>>( } pub fn run_loader_command(path: &str, extension: &str, loader_command: &str) -> Result<String> { - let cmd_args = shell_words::split(loader_command).with_context(|| { - anyhow!("Invalid rag document loader '{extension}': `{loader_command}`") - })?; + let cmd_args = shell_words::split(loader_command) + .with_context(|| anyhow!("Invalid document loader '{extension}': `{loader_command}`"))?; let mut use_stdout = true; let outpath = temp_file("-output-", "").display().to_string(); let cmd_args: Vec<_> = cmd_args diff --git a/src/utils/loader.rs b/src/utils/loader.rs index 2ac671d..b36bf4a 100644 --- a/src/utils/loader.rs +++ b/src/utils/loader.rs @@ -1,17 +1,19 @@ use super::*; -use anyhow::{Context, Result}; +use anyhow::{anyhow, Context, Result}; use indexmap::IndexMap; +use serde::{Deserialize, Serialize}; use std::collections::HashMap; pub const EXTENSION_METADATA: &str = "__extension__"; pub type DocumentMetadata = IndexMap<String, String>; -#[derive(Debug, Clone)] +#[derive(Debug, Clone, Serialize, Deserialize)] pub struct LoadedDocument { pub path: String, pub contents: String, + #[serde(default)] pub metadata: DocumentMetadata, } @@ -80,3 +82,44 @@ fn load_with_command(path: &str, extension: &str, loader_command: &str) -> Resul metadata.insert(EXTENSION_METADATA.into(), DEFAULT_EXTENSION.to_string()); Ok(LoadedDocument::new(path.into(), contents, metadata)) } + +pub fn is_loader_protocol(loaders: &HashMap<String, String>, path: &str) -> bool { + match path.split_once(':') { + Some((protocol, _)) => loaders.contains_key(protocol), + None => false, + } +} + +pub fn load_protocol_path( + loaders: &HashMap<String, String>, + path: &str, +) -> Result<Vec<LoadedDocument>> { + let (protocol, loader_command, new_path) = path + .split_once(':') + .and_then(|(protocol, path)| { + let loader_command = loaders.get(protocol)?; + Some((protocol, loader_command, path)) + }) + .ok_or_else(|| anyhow!("No document loader for '{}'", path))?; + let contents = run_loader_command(new_path, protocol, loader_command)?; + let output = if let Ok(list) = serde_json::from_str::<Vec<LoadedDocument>>(&contents) { + list.into_iter() + .map(|mut v| { + if v.path.starts_with(path) { + } else if v.path.starts_with(new_path) { + v.path = format!("{}:{}", protocol, v.path); + } else { + v.path = format!("{}/{}", path, v.path); + } + v + }) + .collect() + } else { + vec![LoadedDocument::new( + path.into(), + contents, + Default::default(), + )] + }; + Ok(output) +} diff --git a/src/utils/path.rs b/src/utils/path.rs index bf3d936..20a365e 100644 --- a/src/utils/path.rs +++ b/src/utils/path.rs @@ -2,6 +2,7 @@ use std::path::{Component, Path, PathBuf}; use anyhow::{bail, Result}; use indexmap::IndexSet; +use path_absolutize::Absolutize; pub fn safe_join_path<T1: AsRef<Path>, T2: AsRef<Path>>( base_path: T1, @@ -75,6 +76,20 @@ pub fn get_patch_extension(path: &str) -> Option<String> { .map(|v| v.to_string_lossy().to_lowercase()) } +pub fn to_absolute_path(path: &str) -> Result<String> { + Ok(Path::new(&path).absolutize()?.display().to_string()) +} + +pub fn resolve_home_dir(path: &str) -> String { + let mut path = path.to_string(); + if path.starts_with("~/") || path.starts_with("~\\") { + if let Some(home_dir) = dirs::home_dir() { + path.replace_range(..1, &home_dir.display().to_string()); + } + } + path +} + fn parse_glob(path_str: &str) -> Result<(String, Vec<String>)> { if let Some(start) = path_str.find("/**/*.").or_else(|| path_str.find(r"\**\*.")) { let base_path = path_str[..start].to_string(); |
