### Install direnv with Nix

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Installs the direnv utility using Nix. direnv helps manage environment variables for development projects.

```bash
nix profile install nixpkgs#direnv
```

--------------------------------

### Install rga with Scoop on Windows

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Use Scoop to install the rga package on Windows. This is a supported download method that includes necessary dependencies.

```bash
scoop install rga
```

--------------------------------

### Install rga with Homebrew on macOS/Linux

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Install rga using Homebrew on macOS or Linuxbrew. Optional useful dependencies like pandoc, poppler, and ffmpeg can also be installed.

```bash
brew install rga
```

```bash
brew install pandoc poppler ffmpeg
```

--------------------------------

### Launch `rga-fzf` with a Pre-filled Query

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Initiates an interactive fuzzy search with `rga-fzf`, starting with a specific query already entered.

```bash
# Start with a pre-filled query
rga-fzf "invoice 2024"
```

--------------------------------

### Custom Adapter Configuration for Ripgrep-All

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Illustrates the JSONC configuration for defining a custom adapter in ripgrep-all. This example shows how to configure pandoc to convert reStructuredText files to plain text.

```jsonc
{
  "name": "pandoc_rst",          // unique a-z0-9_ identifier
  "version": 2,                  // bump when output format changes (invalidates cache)
  "description": "Convert reStructuredText to plain text via pandoc",
  "extensions": ["rst"],
  "mimetypes": ["text/x-rst"],
  "match_only_by_mime": false,   // if true, extensions are ignored when --rga-accurate
  "disabled_by_default": false,
  "binary": "pandoc",
  "args": [
    "--from=rst",
    "--to=plain",
    "--wrap=none",
    "$input_virtual_path"        // full virtual path placeholder
  ],
  // output_path_hint controls what the next adapter sees as the filename.
  // Default is "${input_virtual_path}.txt". Set to another extension to chain adapters.
  // e.g. output a .tar so the tar adapter processes it next:
  "output_path_hint": "${input_virtual_path}.txt"
}
```

--------------------------------

### Install rga with Chocolatey on Windows

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Use Chocolatey to install the ripgrep-all package on Windows. This method ensures dependencies are also handled.

```bash
choco install ripgrep-all
```

--------------------------------

### Programmatic Adapter Filtering in Ripgrep-All (Rust API)

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Demonstrates using the `get_adapters_filtered` function in the Rust API to select specific file adapters. Examples show how to include, exclude, and prioritize adapters, including custom ones.

```rust
use ripgrep_all::adapters::{get_adapters_filtered, CustomAdapterConfig};

// Use only the poppler (PDF) and sqlite adapters
let adapters = get_adapters_filtered(None, &["poppler", "sqlite"])?;
println!("Selected: {:?}", adapters.iter().map(|a| &a.metadata().name).collect::<Vec<_>>());
// Selected: ["poppler", "sqlite"]

// Use all default adapters except ffmpeg
let adapters = get_adapters_filtered(None, &["-ffmpeg"])?;

// Add the mail adapter on top of all defaults (highest priority)
let adapters = get_adapters_filtered(None, &["+mail"])?;

// Include a custom adapter config alongside defaults
let custom = vec![CustomAdapterConfig {
    name: "my_converter".to_string(),
    version: 1,
    description: "Custom text extractor".to_string(),
    extensions: vec!["myext".to_string()],
    mimetypes: None,
    match_only_by_mime: None,
    binary: "/usr/local/bin/myconverter".to_string(),
    args: vec!["$input_virtual_path".to_string()],
    disabled_by_default: Some(false),
    output_path_hint: None,
}];
let adapters = get_adapters_filtered(Some(custom), &[])?;
```

--------------------------------

### Launch Interactive Search with `rga-fzf`

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Starts an interactive fuzzy search session using `rga-fzf`. The fzf UI displays matching files and provides a preview of matches within the selected file.

```bash
# Launch interactive search in the current directory
rga-fzf
```

--------------------------------

### Configure ripgrep-all Cache Directory

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Specify a custom directory for the rga cache, for example, to use a faster SSD.

```bash
rga --rga-cache-path=/mnt/fast-ssd/rga-cache "query" ./docs
```

--------------------------------

### Configure Custom PDF Adapter for ripgrep-all

Source: https://github.com/phiresky/ripgrep-all/wiki/Home

Example configuration for a custom adapter that uses 'pdftotext' to extract text from PDF files. This adapter is useful for integrating external tools into the search process.

```json
    "custom_adapters": [
        {
            "name": "poppler",
            "version": 1,
            "description": "Uses pdftotext (from poppler-utils) to extract plain text from PDF files",

            "extensions": ["pdf"],
            "mimetypes": ["application/pdf"],

            "binary": "pdftotext",
            "args": ["-", "-"],
            "disabled_by_default": false,
            "match_only_by_mime": false,
            "output_path_hint": "${input_virtual_path}.txt.asciipagebreaks"
        }
    ]
```

--------------------------------

### Install rga with MacPorts on macOS

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Install rga using MacPorts on macOS. This command installs the ripgrep-all port.

```bash
sudo port install ripgrep-all
```

--------------------------------

### Compile rga from source

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Compile rga from source using Cargo. Ensure you have a compatible Rust version (v1.75.0+) and necessary build dependencies installed.

```bash
apt install build-essential pandoc poppler-utils ffmpeg ripgrep cargo
```

```bash
cargo install --locked ripgrep_all
```

```bash
rga --version    # this should work now
```

--------------------------------

### Search Across All Built-in Formats with ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Perform a search across all enabled built-in file formats, including mail archives. This example also enables accurate MIME-sniffing and sets a maximum recursion depth for archives.

```bash
rga --rga-adapters=+mail \
    --rga-accurate \
    --rga-max-archive-recursion=5 \
    "confidential" \
    ~/Documents ~/Downloads ~/Mail
```

--------------------------------

### Print Help Information

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Displays a concise overview of available commands and options for rga.

```bash
-h
```

```bash
--help
```

--------------------------------

### List Available Adapters

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Use this command to see all the file format adapters that ripgrep-all recognizes and can use for searching.

```bash
rga --rga-list-adapters
```

--------------------------------

### Verifying and Using Custom Ripgrep-All Adapters

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Provides bash commands to verify if a custom adapter is loaded and how to use it in searches. It also shows how to explicitly enable adapters that are disabled by default.

```bash
# Verify the custom adapter is loaded
rga --rga-list-adapters | grep pandoc_rst

# Use it (enabled by default since disabled_by_default is false)
rga "introduction" ./docs/guide.rst

# Enable a disabled-by-default custom adapter explicitly
rga --rga-adapters=+pandoc_rst "conclusion" ./reports/
```

--------------------------------

### Run Appropriate Adapter with rga_preproc (Rust API)

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Use `rga_preproc` to select and run the correct adapter for a given file. It returns an `AsyncRead` stream of plain text, handling recursive archive entries internally. Configure matching behavior using `RgaConfig`.

```rust
use ripgrep_all::{
    adapters::{AdaptInfo, ReadBox},
    config::RgaConfig,
    preproc::rga_preproc,
};
use std::path::PathBuf;
use tokio::fs::File;
use tokio::io::AsyncReadExt;

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    let path = PathBuf::from("./exampledir/short.pdf");
    let file = File::open(&path).await?;

    let config = RgaConfig {
        accurate: false,        // use extension-based matching
        ..RgaConfig::default()
    };

    let ai = AdaptInfo {
        filepath_hint: path.clone(),
        inp: Box::pin(file),
        is_real_file: true,
        archive_recursion_depth: 0,
        line_prefix: String::new(),
        postprocess: true,      // prefix each line with "Page N: "
        config,
    };

    let mut output: ReadBox = rga_preproc(ai).await?;

    let mut text = String::new();
    output.read_to_string(&mut text).await?;
    println!("{text}");
    // Page 1: hello world
    // Page 1: this is just a test.
    // Page 1:
    // Page 1:
    // 1

    Ok(())
}
```

--------------------------------

### Specify Adapters to Use

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Control which adapters are used and their priority. You can include, exclude, or add adapters to the default set.

```bash
--rga-adapters=<adapters>...
```

--------------------------------

### Specify Configuration File

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Provide a custom path to the rga configuration file.

```bash
--rga-config-file=<config-file-path>
```

--------------------------------

### Debug `rga-preproc` with Logging

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Enables debug logging for `rga-preproc` to show adapter selection and timing, useful for troubleshooting.

```bash
# Debug mode: show which adapter was chosen and timing
RUST_LOG=debug rga-preproc --rga-no-cache ./exampledir/wasteland.epub 2>&1 | head -30
```

--------------------------------

### Match File to Adapter with adapter_matcher (Rust API)

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Build a closure that maps `FileMeta` to the best matching adapter using `adapter_matcher`. Set `slow = true` for accurate MIME-type matching, or `false` for faster extension-based matching.

```rust
use ripgrep_all::{
    adapters::{get_all_adapters, get_adapters_filtered},
    matching::{adapter_matcher, FileMeta},
};

let (enabled, _disabled) = get_all_adapters(None);
let adapters = get_adapters_filtered(None, &[])?;

// Build a fast extension-only matcher (slow = false)
let matcher = adapter_matcher(&adapters, false)?;

// Match a PDF by extension
let result = matcher(FileMeta {
    lossy_filename: "report.pdf".to_string(),
    mimetype: None,
});
if let Some((adapter, reason)) = result {
    println!("Matched adapter: {}", adapter.metadata().name);
    println!("Match reason: {:?}", reason);
}
// Matched adapter: poppler
// Match reason: Fast(FileExtension("pdf"))

// Build an accurate MIME-type matcher (slow = true)
let slow_matcher = adapter_matcher(&adapters, true)?;
let result = slow_matcher(FileMeta {
    lossy_filename: "datafile",          // no extension
    mimetype: Some("application/x-sqlite3"),
});
if let Some((adapter, _)) = result {
    println!("MIME match: {}", adapter.metadata().name);
}
// MIME match: sqlite
```

--------------------------------

### ripgrep-all Configuration File (`config.jsonc`)

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Defines persistent configuration settings for ripgrep-all. Supports CLI options and custom adapters. Configuration is merged with environment variables and CLI arguments.

```jsonc
// Location: ~/.config/ripgrep-all/config.jsonc  (Linux/macOS XDG)
//           ~/Library/Application Support/ripgrep-all/config.jsonc  (macOS)
//           %APPDATA%\ripgrep-all\config.jsonc  (Windows)
{
  "$schema": "./config.v1.schema.json",

  // Enable MIME-based matching globally
  "accurate": true,

  // Increase max archive nesting (default: 5)
  "max_archive_recursion": 10,

  // Adjust cache settings
  "cache": {
    "disabled": false,
    "max_blob_len": 5000000,          // 5 MB
    "compression_level": 6,           // zstd level 1-22, default 12
    "path": "/home/user/.cache/rga"
  },

  // Define custom subprocess-based adapters
  "custom_adapters": [
    {
      "name": "djvu_text",
      "version": 1,
      "description": "Extracts text from DjVu files using djvutxt",
      "extensions": ["djvu"],
      "binary": "djvutxt",
      "args": ["$input_virtual_path", "-"],
      "disabled_by_default": false
    },
    {
      "name": "rtf_to_text",
      "version": 1,
      "description": "Converts RTF to plain text via unrtf",
      "extensions": ["rtf"],
      "binary": "unrtf",
      "args": ["--text", "$input_virtual_path"],
      "disabled_by_default": false
    },
    {
      "name": "image_ocr",
      "version": 1,
      "description": "OCR images using tesseract",
      "extensions": ["png", "jpg", "jpeg", "tiff"],
      "mimetypes": ["image/png", "image/jpeg", "image/tiff"],
      "match_only_by_mime": false,
      "binary": "tesseract",
      "args": ["$input_virtual_path", "stdout", "-l", "eng"],
      "disabled_by_default": true
    }
  ]
}

```

--------------------------------

### Print ripgrep Help

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Shows the help information specific to the underlying ripgrep utility.

```bash
--rg-help
```

--------------------------------

### CustomAdapterConfig

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Defines a custom adapter for ripgrep-all, specifying the binary to run, its arguments, and the file types it handles. This allows ripgrep-all to process files using external tools.

```APIDOC
## CustomAdapterConfig

### Description
Struct that describes a subprocess-spawning adapter for use in the config file or Rust API. Each custom adapter specifies a binary to run, its arguments (with `$input_virtual_path`, `$input_file_stem`, and `$input_file_extension` placeholders), and the file extensions or MIME types it handles. The adapter pipes the input file to the binary's stdin and reads the output from stdout.

### Configuration Fields

- **name** (string) - Required - A unique a-z0-9_ identifier for the adapter.
- **version** (integer) - Required - Bump when output format changes to invalidate the cache.
- **description** (string) - Required - A human-readable description of the adapter.
- **extensions** (array of strings) - Optional - File extensions the adapter handles.
- **mimetypes** (array of strings) - Optional - MIME types the adapter handles.
- **match_only_by_mime** (boolean) - Optional - If true, extensions are ignored when `--rga-accurate` is used. Defaults to false.
- **disabled_by_default** (boolean) - Optional - If true, the adapter is disabled by default. Defaults to false.
- **binary** (string) - Required - The path to the binary to execute.
- **args** (array of strings) - Required - Arguments to pass to the binary. Supports placeholders like `$input_virtual_path`, `$input_file_stem`, and `$input_file_extension`.
- **output_path_hint** (string) - Optional - Controls what the next adapter sees as the filename. Defaults to `${input_virtual_path}.txt`. Can be set to another extension to chain adapters.
```

--------------------------------

### `adapter_matcher` — File-to-Adapter Matching

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Builds a closure that maps `FileMeta` (filename + optional MIME type) to the best matching adapter. It compiles all adapter matchers into `RegexSet` instances for fast dispatch. When `slow = true` (accurate mode), MIME type matching takes precedence over extension matching.

```APIDOC
## `adapter_matcher` — File-to-Adapter Matching (Rust API)

**Builds a closure that maps `FileMeta` (filename + optional MIME type) to the best matching adapter**

Compiles all adapter matchers into `RegexSet` instances for fast dispatch. When `slow = true` (accurate mode), MIME type matching takes precedence over extension matching.

```rust
use ripgrep_all::adapters::{get_all_adapters, get_adapters_filtered};
use ripgrep_all::matching::{adapter_matcher, FileMeta};

let (enabled, _disabled) = get_all_adapters(None);
let adapters = get_adapters_filtered(None, &[])?;

// Build a fast extension-only matcher (slow = false)
let matcher = adapter_matcher(&adapters, false)?;

// Match a PDF by extension
let result = matcher(FileMeta {
    lossy_filename: "report.pdf".to_string(),
    mimetype: None,
});
if let Some((adapter, reason)) = result {
    println!("Matched adapter: {}", adapter.metadata().name);
    println!("Match reason: {:?}", reason);
}
// Matched adapter: poppler
// Match reason: Fast(FileExtension("pdf"))

// Build an accurate MIME-type matcher (slow = true)
let slow_matcher = adapter_matcher(&adapters, true)?;
let result = slow_matcher(FileMeta {
    lossy_filename: "datafile",          // no extension
    mimetype: Some("application/x-sqlite3"),
});
if let Some((adapter, _)) = result {
    println!("MIME match: {}", adapter.metadata().name);
}
// MIME match: sqlite
```
```

--------------------------------

### Print Configuration Schema

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Outputs the JSON Schema for the rga configuration file, useful for validation and understanding configuration options.

```bash
--rga-print-config-schema
```

--------------------------------

### Print Version Information

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Displays the version of the ripgrep-all tool.

```bash
-V
```

```bash
--version
```

--------------------------------

### Basic Search with ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Perform a basic search for a pattern across all supported file types in a directory tree. Options prefixed with --rga- are consumed by rga, while others are passed to ripgrep.

```bash
# Search for "hello world" in every supported file under ./docs
rga "hello world" ./docs
```

```bash
# Case-insensitive search (-i forwarded to rg), show 3 lines of context
rga -i --context 3 "invoice" ~/Downloads
```

```bash
# Search only inside PDF and EPUB files (reduce scope)
rga --rga-adapters=poppler,pandoc "chapter one" ~/books
```

```bash
# Disable caching (useful when debugging adapters)
rga --rga-no-cache "TODO" ./project
```

```bash
# Use accurate MIME-based matching instead of file extension matching
# Useful for SQLite files with unusual extensions (e.g. .db3, no extension)
rga --rga-accurate "CREATE TABLE" ~/data
```

```bash
# Limit archive recursion to 2 levels deep (default is 5)
rga --rga-max-archive-recursion=2 "secret" ./archives
```

```bash
# Do not prefix inner archive paths on matched lines
rga --rga-no-prefix-filenames "error" ./logs.tar.gz
```

```bash
# Show only filenames that contain a match (--files-with-matches forwarded to rg)
rga --files-with-matches "password" ~/backup.zip
```

```bash
# Use a custom config file path
rga --rga-config-file=/etc/rga/config.jsonc "pattern" ./dir
```

```bash
# Enable all adapters including disabled-by-default ones (e.g. mail)
rga --rga-adapters=+mail "From:" ~/mail.mbox
```

```bash
# Use all adapters EXCEPT pandoc and ffmpeg
rga --rga-adapters=-pandoc,ffmpeg "keyword" ./docs
```

```bash
# Enable debug logging
RUST_LOG=debug rga --rga-no-cache "test" ./exampledir 2>&1
```

--------------------------------

### Enable Debug Logging for ripgrep-all

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Set these environment variables to enable debug logging and backtraces for ripgrep-all. This is useful for diagnosing issues during development.

```bash
export RUST_LOG=debug
export RUST_BACKTRACE=1
```

--------------------------------

### Print ripgrep Version

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Shows the version of the underlying ripgrep utility.

```bash
--rg-version
```

--------------------------------

### Specify Cache Database Path

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Set a custom location for the rga cache database.

```bash
--rga-cache-path=<path>
```

--------------------------------

### Inspect `rga-preproc` Output for a Zip Archive

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Examines the preprocessing of a zip archive using `rga-preproc`, which will recursively process its contents.

```bash
# Inspect preprocessing of a zip archive (will recurse into contents)
rga-preproc ./exampledir/demo/somearchive.zip
```

--------------------------------

### Print Configuration Schema with rga

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Output the JSON Schema for the rga configuration file, which is useful for IDE auto-completion and validation. The schema can be redirected to a file for use with editors like VS Code.

```bash
rga --rga-print-config-schema | python3 -m json.tool | head -60
```

```bash
# Redirect to a schema file for VS Code IntelliSense
rga --rga-print-config-schema > ~/.config/ripgrep-all/config.v1.schema.json
```

--------------------------------

### Preprocess File with MIME Detection using `rga-preproc`

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Uses `rga-preproc` with the `--rga-accurate` flag to preprocess a file without an extension, relying on accurate MIME detection.

```bash
# Preprocess a file with no extension using accurate MIME detection
rga-preproc --rga-accurate ./exampledir/exampledir/sqlitedb
```

--------------------------------

### Bash/Zsh fzf Integration for rga

Source: https://github.com/phiresky/ripgrep-all/wiki/fzf-Integration

Add this function to your ~/.bashrc or ~/.zshrc to enable interactive searching with fzf. It allows fuzzy finding files that match a query and opens the selected file.

```bash
rga-fzf() {
	RG_PREFIX="rga --files-with-matches"
	local file
	file=(
		FZF_DEFAULT_COMMAND="$RG_PREFIX '$1'" \
			fzf --sort --preview="[[ ! -z {} ]] && rga --pretty --context 5 {q} {}" \
			--phony -q "$1" \
			--bind "change:reload:$RG_PREFIX {q}" \
			--preview-window="70%:wrap"
	)
	&& 
	echo "opening $file" && 
	xdg-open "$file"
}
```

--------------------------------

### Basic Usage of ripgrep-all

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

The fundamental command structure for using rga, combining rga options with ripgrep's pattern and path arguments.

```bash
rga [RGA OPTIONS] [RG OPTIONS] PATTERN [PATH ...]
```

--------------------------------

### Bash/Zsh fzf Integration for Ripgrep-All

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Defines a bash function for integrating ripgrep-all with fzf, enabling custom preview options for search results. It sets up environment variables for fzf and uses rga for file searching and previewing.

```bash
rga-fzf-manual() {
    local RG_PREFIX="rga --files-with-matches --rga-cache-max-blob-len=10M"
    local file
    file=(
        FZF_DEFAULT_COMMAND="$RG_PREFIX '$1'" \
        fzf --preview="rga --pretty --context 5 {q} --rga-fzf-path=_{}" \
            --preview-window=70%:wrap \
            --phony --query "$1" --print-query \
            --bind="change:reload:$RG_PREFIX {q}"
    )
    echo "selected: $file"
}
rga-fzf-manual "search term"
```

--------------------------------

### Fish Shell fzf Integration for rga

Source: https://github.com/phiresky/ripgrep-all/wiki/fzf-Integration

This function is for ~/.config/fish/config.fish to integrate rga with fzf. It supports passing arguments to rga and opens the selected file.

```fish
function rga-fzf
    set RG_PREFIX 'rga --files-with-matches'
    if test (count $argv) -gt 1
        set RG_PREFIX "$RG_PREFIX $argv[1..-2]"
    end
    set -l file $file
    set file (
        FZF_DEFAULT_COMMAND="$RG_PREFIX '$argv[-1]'" \
        fzf --sort \
            --preview='test ! -z {} && \
                rga --pretty --context 5 {q} {}' \
            --phony -q "$argv[-1]" \
            --bind "change:reload:$RG_PREFIX {q}" \
            --preview-window='50%:wrap'
    ) && \
    echo "opening $file" && \
    open "$file"
end
```

--------------------------------

### `parse_args` / `split_args` — Configuration Parsing

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Merges CLI arguments, the `RGA_CONFIG` env var, and the config file into a single `RgaConfig`. `split_args` partitions `std::env::args_os()` into rga-owned args (prefixed `--rga-` or `--rg-`) and passthrough args forwarded to ripgrep. `parse_args` accepts an explicit iterator (useful for testing or embedding) and performs the three-way JSON merge.

```APIDOC
## `parse_args` / `split_args` — Configuration Parsing (Rust API)

**Merges CLI arguments, the `RGA_CONFIG` env var, and the config file into a single `RgaConfig`**

`split_args` partitions `std::env::args_os()` into rga-owned args (prefixed `--rga-` or `--rg-`) and passthrough args forwarded to ripgrep. `parse_args` accepts an explicit iterator (useful for testing or embedding) and performs the three-way JSON merge.

```rust
use ripgrep_all::config::{parse_args, split_args};

// ── Embedded usage with explicit args ──────────────────────────────────────
let config = parse_args(
    ["rga", "--rga-accurate", "--rga-no-cache", "--rga-max-archive-recursion=3"],
    false, // is_rga_preproc = false → also reads config file
)?;
assert!(config.accurate);
assert!(config.cache.disabled);
assert_eq!(config.max_archive_recursion.0, 3);

// ── Typical use in the rga binary (reads from real env args) ──────────────
let (rga_config, passthrough_args) = split_args(false)?;
// rga_config  → all --rga-* options, config file, and RGA_CONFIG merged
// passthrough_args → everything else (PATTERN, PATH, -i, --context, etc.)

println!("accurate: {}", rga_config.accurate);
println!("cache path: {}", rga_config.cache.path);
println!("rg args to forward: {:?}", passthrough_args);

// ── Config priorities (low → high) ─────────────────────────────────────────
// 1. ~/.config/ripgrep-all/config.jsonc    (file, lowest priority)
// 2. RGA_CONFIG env var                    (JSON string)
// 3. CLI --rga-* arguments                 (highest priority)
```
```

--------------------------------

### `rga_preproc` — Core Preprocessing Function

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

This is the central function in the preprocessing pipeline. It auto-selects an adapter by matching the file's extension (or MIME type if `config.accurate` is true), runs it, and wraps the output in a zstd-cached `AsyncRead` stream. Recursive archive entries are handled internally.

```APIDOC
## `rga_preproc` — Core Preprocessing Function (Rust API)

**Selects and runs the appropriate adapter for an `AdaptInfo`, returning an `AsyncRead` of plain text**

This is the central function in the preprocessing pipeline. It auto-selects an adapter by matching the file's extension (or MIME type if `config.accurate` is true), runs it, and wraps the output in a zstd-cached `AsyncRead` stream. Recursive archive entries are handled internally.

```rust
use ripgrep_all::preproc::rga_preproc;
use ripgrep_all::adapters::AdaptInfo;
use ripgrep_all::config::RgaConfig;
use tokio::fs::File;
use tokio::io::AsyncReadExt;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    let path = PathBuf::from("./exampledir/short.pdf");
    let file = File::open(&path).await?;

    let config = RgaConfig {
        accurate: false,        // use extension-based matching
        ..RgaConfig::default()
    };

    let ai = AdaptInfo {
        filepath_hint: path.clone(),
        inp: Box::pin(file),
        is_real_file: true,
        archive_recursion_depth: 0,
        line_prefix: String::new(),
        postprocess: true,      // prefix each line with "Page N: "
        config,
    };

    let mut output = rga_preproc(ai).await?;

    let mut text = String::new();
    output.read_to_string(&mut text).await?;
    println!("{text}");
    // Page 1: hello world
    // Page 1: this is just a test.
    // Page 1:
    // Page 1:
    // 1

    Ok(())
}
```
```

--------------------------------

### Search Video Subtitle Tracks with ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Use the ffmpeg adapter to specifically search within video subtitle tracks.

```bash
rga --rga-adapters=ffmpeg "previously on" ~/Videos
```

--------------------------------

### Parse Configuration with parse_args and split_args (Rust API)

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Merge CLI arguments, `RGA_CONFIG` env var, and config file into `RgaConfig`. `split_args` separates rga-owned args from passthrough args for ripgrep. `parse_args` allows explicit argument iterators for testing.

```rust
use ripgrep_all::config::{parse_args, split_args};

// ── Embedded usage with explicit args ──────────────────────────────────────
let config = parse_args(
    ["rga", "--rga-accurate", "--rga-no-cache", "--rga-max-archive-recursion=3"],
    false, // is_rga_preproc = false → also reads config file
)?;
assert!(config.accurate);
assert!(config.cache.disabled);
assert_eq!(config.max_archive_recursion.0, 3);

// ── Typical use in the rga binary (reads from real env args) ──────────────
let (rga_config, passthrough_args) = split_args(false)?;
// rga_config  → all --rga-* options, config file, and RGA_CONFIG merged
// passthrough_args → everything else (PATTERN, PATH, -i, --context, etc.)

println!("accurate: {}", rga_config.accurate);
println!("cache path: {}", rga_config.cache.path);
println!("rg args to forward: {:?}", passthrough_args);

// ── Config priorities (low → high) ─────────────────────────────────────────
// 1. ~/.config/ripgrep-all/config.jsonc    (file, lowest priority)
// 2. RGA_CONFIG env var                    (JSON string)
// 3. CLI --rga-* arguments                 (highest priority)
```

--------------------------------

### Set Maximum Blob Size for Caching

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Define the maximum compressed size for adapter outputs to be stored in the cache. Larger outputs will not be cached.

```bash
--rga-cache-max-blob-len=<max-blob-len>
```

--------------------------------

### Set Zstd Compression Level for ripgrep-all Cache

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Adjust the Zstandard compression level for the cache. Higher levels result in a smaller cache but slower write times, while lower levels result in a larger cache but faster write times.

```bash
rga --rga-cache-compression-level=22 "term" ./books
```

```bash
rga --rga-cache-compression-level=1 "term" ./books
```

--------------------------------

### Search SQLite Databases with ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Search SQLite databases using the sqlite adapter, without relying on file extensions. This command also enables accurate MIME-sniffing.

```bash
rga --rga-accurate --rga-adapters=sqlite "INSERT INTO users" ~/data
```

--------------------------------

### Windows Batch Script fzf Integration for rga

Source: https://github.com/phiresky/ripgrep-all/wiki/fzf-Integration

A batch script for Windows to achieve similar functionality to the Unix-like shells. Place this script in your %PATH% for easy access.

```bat
@echo off
setlocal

set RG_PREFIX=rga --files-with-matches 
set FZF_DEFAULT_COMMAND=%RG_PREFIX% %1 
for /f  "delims=" %%i in ('fzf --sort --preview "rga --pretty --context 5 {q} {}" --phony -q %1 --bind "change:reload:%RG_PREFIX% {q}" --preview-window="70%:wrap"') do (
    set FILE=%%i
)

if not "%FILE%" == "" (
    echo Opening %FILE%
    start "" "%FILE%"
) else (
    echo No file selected.
)

endlocal


```

--------------------------------

### Inspect `rga-preproc` Output for a PDF

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Uses `rga-preproc` to inspect the preprocessing output for a specific file, useful for debugging adapter behavior.

```bash
# Inspect what rga-preproc produces for a PDF
rga-preproc ./exampledir/short.pdf
```

--------------------------------

### Override Configuration with `RGA_CONFIG` Environment Variable

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Supplements or overrides the `config.jsonc` file using a JSON string in the environment. Merged on top of file config but below CLI arguments.

```bash
# Disable cache for a one-off search via environment
RGA_CONFIG='{"cache":{"disabled":true}}' rga "TODO" ./src
```

```bash
# Force accurate mode and limit recursion via env
RGA_CONFIG='{"accurate":true,"max_archive_recursion":3}' rga "error" ./logs
```

```bash
# Enable the mail adapter via env without changing config file
RGA_CONFIG='{"adapters":"+mail"}' rga "Re:" ~/Maildir
```

--------------------------------

### Enable Accurate Matching by Mime Type

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

This flag enables more precise file type detection using magic bytes, which can be slower but more accurate than relying on file extensions.

```bash
--rga-accurate
```

--------------------------------

### Set Cache Compression Level

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Adjust the ZSTD compression level for adapter outputs stored in the cache. Higher levels mean more compression but potentially slower processing.

```bash
--rga-cache-compression-level=<compression-level>
```

--------------------------------

### Limit Cache Entries Size in ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Set a maximum size for cached blobs, useful for excluding large files like PDFs from the cache.

```bash
rga --rga-cache-max-blob-len=500k "keyword" ./archive
```

--------------------------------

### get_adapters_filtered

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Programmatically filters the full adapter list based on a user-supplied selector string. This Rust API function allows for dynamic selection of adapters, including custom ones.

```APIDOC
## get_adapters_filtered

### Description
Filters the full adapter list according to a user-supplied selector string. Returns an ordered `Vec<Arc<dyn FileAdapter>>` based on the selector rules: plain names mean "use only these", a leading `-` means "exclude from defaults", a leading `+` means "add to defaults with higher priority". Custom adapter configs are merged before filtering.

### Parameters

- **custom_adapters** (Option<Vec<CustomAdapterConfig>>) - Optional - A vector of `CustomAdapterConfig` structs to include.
- **selectors** (slice of strings) - Required - A slice of strings used to filter adapters. Examples:
  - `"poppler"`: Use only the poppler adapter.
  - `"-ffmpeg"`: Use all default adapters except ffmpeg.
  - `"+mail"`: Add the mail adapter to the defaults with higher priority.

### Returns

- `Result<Vec<Arc<dyn FileAdapter>>, Error>` - An ordered vector of selected file adapters or an error.

### Example (Rust)
```rust
use ripgrep_all::adapters::{get_adapters_filtered, CustomAdapterConfig};

// Use only the poppler (PDF) and sqlite adapters
let adapters = get_adapters_filtered(None, &["poppler", "sqlite"])?;
println!("Selected: {:?}", adapters.iter().map(|a| &a.metadata().name).collect::<Vec<_>>());
// Selected: ["poppler", "sqlite"]

// Use all default adapters except ffmpeg
let adapters = get_adapters_filtered(None, &["-ffmpeg"])?;

// Add the mail adapter on top of all defaults (highest priority)
let adapters = get_adapters_filtered(None, &["+mail"])?;

// Include a custom adapter config alongside defaults
let custom = vec![CustomAdapterConfig {
    name: "my_converter".to_string(),
    version: 1,
    description: "Custom text extractor".to_string(),
    extensions: vec!["myext".to_string()],
    mimetypes: None,
    match_only_by_mime: None,
    binary: "/usr/local/bin/myconverter".to_string(),
    args: vec!["$input_virtual_path".to_string()],
    disabled_by_default: Some(false),
    output_path_hint: None,
}];
let adapters = get_adapters_filtered(Some(custom), &[])?;
```
```

--------------------------------

### JSON Schema Mimetypes Definition

Source: https://github.com/phiresky/ripgrep-all/blob/master/doc/notes.md

This JSON schema snippet defines the 'mimetypes' property. When not null and the '--rga-accurate' flag is enabled, it specifies that MIME type matching will be used instead of file name matching. The property can be an array of strings or null.

```json
    "mimetypes": {
        "description": "if not null and --rga-accurate is enabled, mime type matching is used instead of file name matching",
        "type": [
            "array",
            "null"
        ],
        "items": {
            "type": "string"
        }
    },
```

--------------------------------

### Disable Cache for a ripgrep-all Run

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Completely disable the cache for a specific command execution.

```bash
rga --rga-no-cache "debug query" ./src
```

--------------------------------

### Recursively Search Nested Archives with ripgrep-all

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Search for a pattern within a deeply nested archive file (e.g., a .tar.gz containing a .zip which contains a .pdf). rga automatically handles decompression and extraction.

```bash
rga "needle" ./haystack.tar.gz
```

--------------------------------

### Clear ripgrep-all Cache Manually

Source: https://context7.com/phiresky/ripgrep-all/llms.txt

Manually remove the ripgrep-all cache directory. This command is for Linux systems.

```bash
rm -rf ~/.cache/ripgrep-all
```

--------------------------------

### Set Maximum Archive Recursion Depth

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Limit how many levels deep rga will recurse into nested archives.

```bash
--rga-max-archive-recursion=<max-archive-recursion>
```

--------------------------------

### Disable Caching

Source: https://github.com/phiresky/ripgrep-all/blob/master/README.md

Use this flag to prevent rga from caching extracted text, ensuring that all content is re-processed on each search.

```bash
--rga-no-cache
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.