Files
ripgrep/src/worker.rs
Andrew Gallant 8bbe58d623 Add support for additional text encodings.
This includes, but is not limited to, UTF-16, latin-1, GBK, EUC-JP and
Shift_JIS. (Courtesy of the `encoding_rs` crate.)

Specifically, this feature enables ripgrep to search files that are
encoded in an encoding other than UTF-8. The list of available encodings
is tied directly to what the `encoding_rs` crate supports, which is in
turn tied to the Encoding Standard. The full list of available encodings
can be found here: https://encoding.spec.whatwg.org/#concept-encoding-get

This pull request also introduces the notion that text encodings can be
automatically detected on a best effort basis. Currently, the only
support for this is checking for a UTF-16 bom. In all other cases, a
text encoding of `auto` (the default) implies a UTF-8 or ASCII
compatible source encoding. When a text encoding is otherwise specified,
it is unconditionally used for all files searched.

Since ripgrep's regex engine is fundamentally built on top of UTF-8,
this feature works by transcoding the files to be searched from their
source encoding to UTF-8. This transcoding only happens when:

1. `auto` is specified and a non-UTF-8 encoding is detected.
2. A specific encoding is given by end users (including UTF-8).

When transcoding occurs, errors are handled by automatically inserting
the Unicode replacement character. In this case, ripgrep's output is
guaranteed to be valid UTF-8 (excluding non-UTF-8 file paths, if they
are printed).

In all other cases, the source text is searched directly, which implies
an assumption that it is at least ASCII compatible, but where UTF-8 is
most useful. In this scenario, encoding errors are not detected. In this
case, ripgrep's output will match the input exactly, byte-for-byte.

This design may not be optimal in all cases, but it has some advantages:

1. In the happy path ("UTF-8 everywhere") remains happy. I have not been
   able to witness any performance regressions.
2. In the non-UTF-8 path, implementation complexity is kept relatively
   low. The cost here is transcoding itself. A potentially superior
   implementation might build decoding of any encoding into the regex
   engine itself. In particular, the fundamental problem with
   transcoding everything first is that literal optimizations are nearly
   negated.

Future work should entail improving the user experience. For example, we
might want to auto-detect more text encodings. A more elaborate UX
experience might permit end users to specify multiple text encodings,
although this seems hard to pull off in an ergonomic way.

Fixes #1
2017-03-12 19:54:48 -04:00

314 lines
9.1 KiB
Rust

use std::fs::File;
use std::io;
use std::path::Path;
use encoding_rs::Encoding;
use grep::Grep;
use ignore::DirEntry;
use memmap::{Mmap, Protection};
use termcolor::WriteColor;
use decoder::DecodeReader;
use pathutil::strip_prefix;
use printer::Printer;
use search_buffer::BufferSearcher;
use search_stream::{InputBuffer, Searcher};
use Result;
pub enum Work {
Stdin,
DirEntry(DirEntry),
}
pub struct WorkerBuilder {
grep: Grep,
opts: Options,
}
#[derive(Clone, Debug)]
struct Options {
mmap: bool,
encoding: Option<&'static Encoding>,
after_context: usize,
before_context: usize,
count: bool,
files_with_matches: bool,
files_without_matches: bool,
eol: u8,
invert_match: bool,
line_number: bool,
max_count: Option<u64>,
no_messages: bool,
quiet: bool,
text: bool,
}
impl Default for Options {
fn default() -> Options {
Options {
mmap: false,
encoding: None,
after_context: 0,
before_context: 0,
count: false,
files_with_matches: false,
files_without_matches: false,
eol: b'\n',
invert_match: false,
line_number: false,
max_count: None,
no_messages: false,
quiet: false,
text: false,
}
}
}
impl WorkerBuilder {
/// Create a new builder for a worker.
///
/// A reusable input buffer and a grep matcher are required, but there
/// are numerous additional options that can be configured on this builder.
pub fn new(grep: Grep) -> WorkerBuilder {
WorkerBuilder {
grep: grep,
opts: Options::default(),
}
}
/// Create the worker from this builder.
pub fn build(self) -> Worker {
let mut inpbuf = InputBuffer::new();
inpbuf.eol(self.opts.eol);
Worker {
grep: self.grep,
inpbuf: inpbuf,
decodebuf: vec![0; 8 * (1<<10)],
opts: self.opts,
}
}
/// The number of contextual lines to show after each match. The default
/// is zero.
pub fn after_context(mut self, count: usize) -> Self {
self.opts.after_context = count;
self
}
/// The number of contextual lines to show before each match. The default
/// is zero.
pub fn before_context(mut self, count: usize) -> Self {
self.opts.before_context = count;
self
}
/// If enabled, searching will print a count instead of each match.
///
/// Disabled by default.
pub fn count(mut self, yes: bool) -> Self {
self.opts.count = yes;
self
}
/// Set the encoding to use to read each file.
///
/// If the encoding is `None` (the default), then the encoding is
/// automatically detected on a best-effort per-file basis.
pub fn encoding(mut self, enc: Option<&'static Encoding>) -> Self {
self.opts.encoding = enc;
self
}
/// If enabled, searching will print the path instead of each match.
///
/// Disabled by default.
pub fn files_with_matches(mut self, yes: bool) -> Self {
self.opts.files_with_matches = yes;
self
}
/// If enabled, searching will print the path of files without any matches.
///
/// Disabled by default.
pub fn files_without_matches(mut self, yes: bool) -> Self {
self.opts.files_without_matches = yes;
self
}
/// Set the end-of-line byte used by this searcher.
pub fn eol(mut self, eol: u8) -> Self {
self.opts.eol = eol;
self
}
/// If enabled, matching is inverted so that lines that *don't* match the
/// given pattern are treated as matches.
pub fn invert_match(mut self, yes: bool) -> Self {
self.opts.invert_match = yes;
self
}
/// If enabled, compute line numbers and prefix each line of output with
/// them.
pub fn line_number(mut self, yes: bool) -> Self {
self.opts.line_number = yes;
self
}
/// Limit the number of matches to the given count.
///
/// The default is None, which corresponds to no limit.
pub fn max_count(mut self, count: Option<u64>) -> Self {
self.opts.max_count = count;
self
}
/// If enabled, try to use memory maps for searching if possible.
pub fn mmap(mut self, yes: bool) -> Self {
self.opts.mmap = yes;
self
}
/// If enabled, error messages are suppressed.
///
/// This is disabled by default.
pub fn no_messages(mut self, yes: bool) -> Self {
self.opts.no_messages = yes;
self
}
/// If enabled, don't show any output and quit searching after the first
/// match is found.
pub fn quiet(mut self, yes: bool) -> Self {
self.opts.quiet = yes;
self
}
/// If enabled, search binary files as if they were text.
pub fn text(mut self, yes: bool) -> Self {
self.opts.text = yes;
self
}
}
/// Worker is responsible for executing searches on file paths, while choosing
/// streaming search or memory map search as appropriate.
pub struct Worker {
grep: Grep,
inpbuf: InputBuffer,
decodebuf: Vec<u8>,
opts: Options,
}
impl Worker {
/// Execute the worker with the given printer and work item.
///
/// A work item can either be stdin or a file path.
pub fn run<W: WriteColor>(
&mut self,
printer: &mut Printer<W>,
work: Work,
) -> u64 {
let result = match work {
Work::Stdin => {
let stdin = io::stdin();
let stdin = stdin.lock();
self.search(printer, Path::new("<stdin>"), stdin)
}
Work::DirEntry(dent) => {
let mut path = dent.path();
let file = match File::open(path) {
Ok(file) => file,
Err(err) => {
if !self.opts.no_messages {
eprintln!("{}: {}", path.display(), err);
}
return 0;
}
};
if let Some(p) = strip_prefix("./", path) {
path = p;
}
if self.opts.mmap {
self.search_mmap(printer, path, &file)
} else {
self.search(printer, path, file)
}
}
};
match result {
Ok(count) => {
count
}
Err(err) => {
if !self.opts.no_messages {
eprintln!("{}", err);
}
0
}
}
}
fn search<R: io::Read, W: WriteColor>(
&mut self,
printer: &mut Printer<W>,
path: &Path,
rdr: R,
) -> Result<u64> {
let rdr = DecodeReader::new(
rdr, &mut self.decodebuf, self.opts.encoding);
let searcher = Searcher::new(
&mut self.inpbuf, printer, &self.grep, path, rdr);
searcher
.after_context(self.opts.after_context)
.before_context(self.opts.before_context)
.count(self.opts.count)
.files_with_matches(self.opts.files_with_matches)
.files_without_matches(self.opts.files_without_matches)
.eol(self.opts.eol)
.line_number(self.opts.line_number)
.invert_match(self.opts.invert_match)
.max_count(self.opts.max_count)
.quiet(self.opts.quiet)
.text(self.opts.text)
.run()
.map_err(From::from)
}
fn search_mmap<W: WriteColor>(
&mut self,
printer: &mut Printer<W>,
path: &Path,
file: &File,
) -> Result<u64> {
if try!(file.metadata()).len() == 0 {
// Opening a memory map with an empty file results in an error.
// However, this may not actually be an empty file! For example,
// /proc/cpuinfo reports itself as an empty file, but it can
// produce data when it's read from. Therefore, we fall back to
// regular read calls.
return self.search(printer, path, file);
}
let mmap = try!(Mmap::open(file, Protection::Read));
let buf = unsafe { mmap.as_slice() };
if buf.len() >= 3 && Encoding::for_bom(buf).is_some() {
// If we have a UTF-16 bom in our memory map, then we need to fall
// back to the stream reader, which will do transcoding.
return self.search(printer, path, file);
}
let searcher = BufferSearcher::new(printer, &self.grep, path, buf);
Ok(searcher
.count(self.opts.count)
.files_with_matches(self.opts.files_with_matches)
.files_without_matches(self.opts.files_without_matches)
.eol(self.opts.eol)
.line_number(self.opts.line_number)
.invert_match(self.opts.invert_match)
.max_count(self.opts.max_count)
.quiet(self.opts.quiet)
.text(self.opts.text)
.run())
}
}