use std::cell::RefCell;
use std::env;
use std::fs::File;
use std::io::{self, BufRead, Read};
use std::path::{Path, PathBuf};
use std::str;
use std::sync::Arc;
use globset::{Candidate, GlobBuilder, GlobSet, GlobSetBuilder};
use regex::bytes::Regex;
use thread_local::ThreadLocal;
use pathutil::{is_file_name, strip_prefix};
use {Error, Match, PartialErrorBuilder};
#[derive(Clone, Debug)]
pub struct Glob {
from: Option<PathBuf>,
original: String,
actual: String,
is_whitelist: bool,
is_only_dir: bool,
}
impl Glob {
pub fn from(&self) -> Option<&Path> {
self.from.as_ref().map(|p| &**p)
}
pub fn original(&self) -> &str {
&self.original
}
pub fn actual(&self) -> &str {
&self.actual
}
pub fn is_whitelist(&self) -> bool {
self.is_whitelist
}
pub fn is_only_dir(&self) -> bool {
self.is_only_dir
}
fn has_doublestar_prefix(&self) -> bool {
self.actual.starts_with("**/") || self.actual == "**"
}
}
#[derive(Clone, Debug)]
pub struct Gitignore {
set: GlobSet,
root: PathBuf,
globs: Vec<Glob>,
num_ignores: u64,
num_whitelists: u64,
matches: Option<Arc<ThreadLocal<RefCell<Vec<usize>>>>>,
}
impl Gitignore {
pub fn new<P: AsRef<Path>>(
gitignore_path: P,
) -> (Gitignore, Option<Error>) {
let path = gitignore_path.as_ref();
let parent = path.parent().unwrap_or(Path::new("/"));
let mut builder = GitignoreBuilder::new(parent);
let mut errs = PartialErrorBuilder::default();
errs.maybe_push_ignore_io(builder.add(path));
match builder.build() {
Ok(gi) => (gi, errs.into_error_option()),
Err(err) => {
errs.push(err);
(Gitignore::empty(), errs.into_error_option())
}
}
}
pub fn global() -> (Gitignore, Option<Error>) {
GitignoreBuilder::new("").build_global()
}
pub fn empty() -> Gitignore {
Gitignore {
set: GlobSet::empty(),
root: PathBuf::from(""),
globs: vec![],
num_ignores: 0,
num_whitelists: 0,
matches: None,
}
}
pub fn path(&self) -> &Path {
&*self.root
}
pub fn is_empty(&self) -> bool {
self.set.is_empty()
}
pub fn len(&self) -> usize {
self.set.len()
}
pub fn num_ignores(&self) -> u64 {
self.num_ignores
}
pub fn num_whitelists(&self) -> u64 {
self.num_whitelists
}
pub fn matched<P: AsRef<Path>>(
&self,
path: P,
is_dir: bool,
) -> Match<&Glob> {
if self.is_empty() {
return Match::None;
}
self.matched_stripped(self.strip(path.as_ref()), is_dir)
}
pub fn matched_path_or_any_parents<P: AsRef<Path>>(
&self,
path: P,
is_dir: bool,
) -> Match<&Glob> {
if self.is_empty() {
return Match::None;
}
let mut path = self.strip(path.as_ref());
assert!(!path.has_root(), "path is expected to be under the root");
match self.matched_stripped(path, is_dir) {
Match::None => (),
a_match => return a_match,
}
while let Some(parent) = path.parent() {
match self.matched_stripped(parent, true) {
Match::None => path = parent,
a_match => return a_match,
}
}
Match::None
}
fn matched_stripped<P: AsRef<Path>>(
&self,
path: P,
is_dir: bool,
) -> Match<&Glob> {
if self.is_empty() {
return Match::None;
}
let path = path.as_ref();
let _matches = self.matches.as_ref().unwrap().get_or_default();
let mut matches = _matches.borrow_mut();
let candidate = Candidate::new(path);
self.set.matches_candidate_into(&candidate, &mut *matches);
for &i in matches.iter().rev() {
let glob = &self.globs[i];
if !glob.is_only_dir() || is_dir {
return if glob.is_whitelist() {
Match::Whitelist(glob)
} else {
Match::Ignore(glob)
};
}
}
Match::None
}
fn strip<'a, P: 'a + AsRef<Path> + ?Sized>(
&'a self,
path: &'a P,
) -> &'a Path {
let mut path = path.as_ref();
if let Some(p) = strip_prefix("./", path) {
path = p;
}
if self.root != Path::new(".") && !is_file_name(path) {
if let Some(p) = strip_prefix(&self.root, path) {
path = p;
if let Some(p) = strip_prefix("/", path) {
path = p;
}
}
}
path
}
}
#[derive(Clone, Debug)]
pub struct GitignoreBuilder {
builder: GlobSetBuilder,
root: PathBuf,
globs: Vec<Glob>,
case_insensitive: bool,
}
impl GitignoreBuilder {
pub fn new<P: AsRef<Path>>(root: P) -> GitignoreBuilder {
let root = root.as_ref();
GitignoreBuilder {
builder: GlobSetBuilder::new(),
root: strip_prefix("./", root).unwrap_or(root).to_path_buf(),
globs: vec![],
case_insensitive: false,
}
}
pub fn build(&self) -> Result<Gitignore, Error> {
let nignore = self.globs.iter().filter(|g| !g.is_whitelist()).count();
let nwhite = self.globs.iter().filter(|g| g.is_whitelist()).count();
let set = self
.builder
.build()
.map_err(|err| Error::Glob { glob: None, err: err.to_string() })?;
Ok(Gitignore {
set: set,
root: self.root.clone(),
globs: self.globs.clone(),
num_ignores: nignore as u64,
num_whitelists: nwhite as u64,
matches: Some(Arc::new(ThreadLocal::default())),
})
}
pub fn build_global(mut self) -> (Gitignore, Option<Error>) {
match gitconfig_excludes_path() {
None => (Gitignore::empty(), None),
Some(path) => {
if !path.is_file() {
(Gitignore::empty(), None)
} else {
let mut errs = PartialErrorBuilder::default();
errs.maybe_push_ignore_io(self.add(path));
match self.build() {
Ok(gi) => (gi, errs.into_error_option()),
Err(err) => {
errs.push(err);
(Gitignore::empty(), errs.into_error_option())
}
}
}
}
}
}
pub fn add<P: AsRef<Path>>(&mut self, path: P) -> Option<Error> {
let path = path.as_ref();
let file = match File::open(path) {
Err(err) => return Some(Error::Io(err).with_path(path)),
Ok(file) => file,
};
let rdr = io::BufReader::new(file);
let mut errs = PartialErrorBuilder::default();
for (i, line) in rdr.lines().enumerate() {
let lineno = (i + 1) as u64;
let line = match line {
Ok(line) => line,
Err(err) => {
errs.push(Error::Io(err).tagged(path, lineno));
break;
}
};
if let Err(err) = self.add_line(Some(path.to_path_buf()), &line) {
errs.push(err.tagged(path, lineno));
}
}
errs.into_error_option()
}
#[cfg(test)]
fn add_str(
&mut self,
from: Option<PathBuf>,
gitignore: &str,
) -> Result<&mut GitignoreBuilder, Error> {
for line in gitignore.lines() {
self.add_line(from.clone(), line)?;
}
Ok(self)
}
pub fn add_line(
&mut self,
from: Option<PathBuf>,
mut line: &str,
) -> Result<&mut GitignoreBuilder, Error> {
#![allow(deprecated)]
if line.starts_with("#") {
return Ok(self);
}
if !line.ends_with("\\ ") {
line = line.trim_right();
}
if line.is_empty() {
return Ok(self);
}
let mut glob = Glob {
from: from,
original: line.to_string(),
actual: String::new(),
is_whitelist: false,
is_only_dir: false,
};
let mut is_absolute = false;
if line.starts_with("\\!") || line.starts_with("\\#") {
line = &line[1..];
is_absolute = line.chars().nth(0) == Some('/');
} else {
if line.starts_with("!") {
glob.is_whitelist = true;
line = &line[1..];
}
if line.starts_with("/") {
line = &line[1..];
is_absolute = true;
}
}
if let Some((i, c)) = line.char_indices().rev().nth(0) {
if c == '/' {
glob.is_only_dir = true;
line = &line[..i];
}
}
glob.actual = line.to_string();
if !is_absolute && !line.chars().any(|c| c == '/') {
if !glob.has_doublestar_prefix() {
glob.actual = format!("**/{}", glob.actual);
}
}
if glob.actual.ends_with("/**") {
glob.actual = format!("{}/*", glob.actual);
}
let parsed = GlobBuilder::new(&glob.actual)
.literal_separator(true)
.case_insensitive(self.case_insensitive)
.backslash_escape(true)
.build()
.map_err(|err| Error::Glob {
glob: Some(glob.original.clone()),
err: err.kind().to_string(),
})?;
self.builder.add(parsed);
self.globs.push(glob);
Ok(self)
}
pub fn case_insensitive(
&mut self,
yes: bool,
) -> Result<&mut GitignoreBuilder, Error> {
self.case_insensitive = yes;
Ok(self)
}
}
fn gitconfig_excludes_path() -> Option<PathBuf> {
match gitconfig_home_contents().and_then(|x| parse_excludes_file(&x)) {
Some(path) => return Some(path),
None => {}
}
match gitconfig_xdg_contents().and_then(|x| parse_excludes_file(&x)) {
Some(path) => return Some(path),
None => {}
}
excludes_file_default()
}
fn gitconfig_home_contents() -> Option<Vec<u8>> {
let home = match home_dir() {
None => return None,
Some(home) => home,
};
let mut file = match File::open(home.join(".gitconfig")) {
Err(_) => return None,
Ok(file) => io::BufReader::new(file),
};
let mut contents = vec![];
file.read_to_end(&mut contents).ok().map(|_| contents)
}
fn gitconfig_xdg_contents() -> Option<Vec<u8>> {
let path = env::var_os("XDG_CONFIG_HOME")
.and_then(|x| if x.is_empty() { None } else { Some(PathBuf::from(x)) })
.or_else(|| home_dir().map(|p| p.join(".config")))
.map(|x| x.join("git/config"));
let mut file = match path.and_then(|p| File::open(p).ok()) {
None => return None,
Some(file) => io::BufReader::new(file),
};
let mut contents = vec![];
file.read_to_end(&mut contents).ok().map(|_| contents)
}
fn excludes_file_default() -> Option<PathBuf> {
env::var_os("XDG_CONFIG_HOME")
.and_then(|x| if x.is_empty() { None } else { Some(PathBuf::from(x)) })
.or_else(|| home_dir().map(|p| p.join(".config")))
.map(|x| x.join("git/ignore"))
}
fn parse_excludes_file(data: &[u8]) -> Option<PathBuf> {
lazy_static! {
static ref RE: Regex =
Regex::new(r"(?im)^\s*excludesfile\s*=\s*(.+)\s*$").unwrap();
};
let caps = match RE.captures(data) {
None => return None,
Some(caps) => caps,
};
str::from_utf8(&caps[1]).ok().map(|s| PathBuf::from(expand_tilde(s)))
}
fn expand_tilde(path: &str) -> String {
let home = match home_dir() {
None => return path.to_string(),
Some(home) => home.to_string_lossy().into_owned(),
};
path.replace("~", &home)
}
fn home_dir() -> Option<PathBuf> {
#![allow(deprecated)]
env::home_dir()
}
#[cfg(test)]
mod tests {
use super::{Gitignore, GitignoreBuilder};
use std::path::Path;
fn gi_from_str<P: AsRef<Path>>(root: P, s: &str) -> Gitignore {
let mut builder = GitignoreBuilder::new(root);
builder.add_str(None, s).unwrap();
builder.build().unwrap()
}
macro_rules! ignored {
($name:ident, $root:expr, $gi:expr, $path:expr) => {
ignored!($name, $root, $gi, $path, false);
};
($name:ident, $root:expr, $gi:expr, $path:expr, $is_dir:expr) => {
#[test]
fn $name() {
let gi = gi_from_str($root, $gi);
assert!(gi.matched($path, $is_dir).is_ignore());
}
};
}
macro_rules! not_ignored {
($name:ident, $root:expr, $gi:expr, $path:expr) => {
not_ignored!($name, $root, $gi, $path, false);
};
($name:ident, $root:expr, $gi:expr, $path:expr, $is_dir:expr) => {
#[test]
fn $name() {
let gi = gi_from_str($root, $gi);
assert!(!gi.matched($path, $is_dir).is_ignore());
}
};
}
const ROOT: &'static str = "/home/foobar/rust/rg";
ignored!(ig1, ROOT, "months", "months");
ignored!(ig2, ROOT, "*.lock", "Cargo.lock");
ignored!(ig3, ROOT, "*.rs", "src/main.rs");
ignored!(ig4, ROOT, "src/*.rs", "src/main.rs");
ignored!(ig5, ROOT, "/*.c", "cat-file.c");
ignored!(ig6, ROOT, "/src/*.rs", "src/main.rs");
ignored!(ig7, ROOT, "!src/main.rs\n*.rs", "src/main.rs");
ignored!(ig8, ROOT, "foo/", "foo", true);
ignored!(ig9, ROOT, "**/foo", "foo");
ignored!(ig10, ROOT, "**/foo", "src/foo");
ignored!(ig11, ROOT, "**/foo/**", "src/foo/bar");
ignored!(ig12, ROOT, "**/foo/**", "wat/src/foo/bar/baz");
ignored!(ig13, ROOT, "**/foo/bar", "foo/bar");
ignored!(ig14, ROOT, "**/foo/bar", "src/foo/bar");
ignored!(ig15, ROOT, "abc/**", "abc/x");
ignored!(ig16, ROOT, "abc/**", "abc/x/y");
ignored!(ig17, ROOT, "abc/**", "abc/x/y/z");
ignored!(ig18, ROOT, "a/**/b", "a/b");
ignored!(ig19, ROOT, "a/**/b", "a/x/b");
ignored!(ig20, ROOT, "a/**/b", "a/x/y/b");
ignored!(ig21, ROOT, r"\!xy", "!xy");
ignored!(ig22, ROOT, r"\#foo", "#foo");
ignored!(ig23, ROOT, "foo", "./foo");
ignored!(ig24, ROOT, "target", "grep/target");
ignored!(ig25, ROOT, "Cargo.lock", "./tabwriter-bin/Cargo.lock");
ignored!(ig26, ROOT, "/foo/bar/baz", "./foo/bar/baz");
ignored!(ig27, ROOT, "foo/", "xyz/foo", true);
ignored!(ig28, "./src", "/llvm/", "./src/llvm", true);
ignored!(ig29, ROOT, "node_modules/ ", "node_modules", true);
ignored!(ig30, ROOT, "**/", "foo/bar", true);
ignored!(ig31, ROOT, "path1/*", "path1/foo");
ignored!(ig32, ROOT, ".a/b", ".a/b");
ignored!(ig33, "./", ".a/b", ".a/b");
ignored!(ig34, ".", ".a/b", ".a/b");
ignored!(ig35, "./.", ".a/b", ".a/b");
ignored!(ig36, "././", ".a/b", ".a/b");
ignored!(ig37, "././.", ".a/b", ".a/b");
ignored!(ig38, ROOT, "\\[", "[");
ignored!(ig39, ROOT, "\\?", "?");
ignored!(ig40, ROOT, "\\*", "*");
ignored!(ig41, ROOT, "\\a", "a");
ignored!(ig42, ROOT, "s*.rs", "sfoo.rs");
ignored!(ig43, ROOT, "**", "foo.rs");
ignored!(ig44, ROOT, "**/**/*", "a/foo.rs");
not_ignored!(ignot1, ROOT, "amonths", "months");
not_ignored!(ignot2, ROOT, "monthsa", "months");
not_ignored!(ignot3, ROOT, "/src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot4, ROOT, "/*.c", "mozilla-sha1/sha1.c");
not_ignored!(ignot5, ROOT, "/src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot6, ROOT, "*.rs\n!src/main.rs", "src/main.rs");
not_ignored!(ignot7, ROOT, "foo/", "foo", false);
not_ignored!(ignot8, ROOT, "**/foo/**", "wat/src/afoo/bar/baz");
not_ignored!(ignot9, ROOT, "**/foo/**", "wat/src/fooa/bar/baz");
not_ignored!(ignot10, ROOT, "**/foo/bar", "foo/src/bar");
not_ignored!(ignot11, ROOT, "#foo", "#foo");
not_ignored!(ignot12, ROOT, "\n\n\n", "foo");
not_ignored!(ignot13, ROOT, "foo/**", "foo", true);
not_ignored!(
ignot14,
"./third_party/protobuf",
"m4/ltoptions.m4",
"./third_party/protobuf/csharp/src/packages/repositories.config"
);
not_ignored!(ignot15, ROOT, "!/bar", "foo/bar");
not_ignored!(ignot16, ROOT, "*\n!**/", "foo", true);
not_ignored!(ignot17, ROOT, "src/*.rs", "src/grep/src/main.rs");
not_ignored!(ignot18, ROOT, "path1/*", "path2/path1/foo");
not_ignored!(ignot19, ROOT, "s*.rs", "src/foo.rs");
fn bytes(s: &str) -> Vec<u8> {
s.to_string().into_bytes()
}
fn path_string<P: AsRef<Path>>(path: P) -> String {
path.as_ref().to_str().unwrap().to_string()
}
#[test]
fn parse_excludes_file1() {
let data = bytes("[core]\nexcludesFile = /foo/bar");
let got = super::parse_excludes_file(&data).unwrap();
assert_eq!(path_string(got), "/foo/bar");
}
#[test]
fn parse_excludes_file2() {
let data = bytes("[core]\nexcludesFile = ~/foo/bar");
let got = super::parse_excludes_file(&data).unwrap();
assert_eq!(path_string(got), super::expand_tilde("~/foo/bar"));
}
#[test]
fn parse_excludes_file3() {
let data = bytes("[core]\nexcludeFile = /foo/bar");
assert!(super::parse_excludes_file(&data).is_none());
}
#[test]
fn regression_106() {
gi_from_str("/", " ");
}
#[test]
fn case_insensitive() {
let gi = GitignoreBuilder::new(ROOT)
.case_insensitive(true)
.unwrap()
.add_str(None, "*.html")
.unwrap()
.build()
.unwrap();
assert!(gi.matched("foo.html", false).is_ignore());
assert!(gi.matched("foo.HTML", false).is_ignore());
assert!(!gi.matched("foo.htm", false).is_ignore());
assert!(!gi.matched("foo.HTM", false).is_ignore());
}
ignored!(cs1, ROOT, "*.html", "foo.html");
not_ignored!(cs2, ROOT, "*.html", "foo.HTML");
not_ignored!(cs3, ROOT, "*.html", "foo.htm");
not_ignored!(cs4, ROOT, "*.html", "foo.HTM");
}