frank/sharkdp--bat
1
0
Fork 0
mirror of https://github.com/sharkdp/bat synced 2026-08-10 20:01:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' into fix/3697-bat-auto-detect-language-from-stdin

Browse Source
This commit is contained in:
Keith Hall
2026-08-10 19:32:11 +03:00
committed by GitHub
parent 530e83a455 2fb2415731
commit a5deaf63ff
72 changed files with 3776 additions and 1490 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.cargo/config.toml
+11
View File
@@ -0,0 +1,11 @@
# On Windows MSVC, statically link the C runtime so that the resulting EXE does
# not depend on the vcruntime DLL.
#
# See: https://github.com/sharkdp/bat/issues/3634
[target.x86_64-pc-windows-msvc]
rustflags = ["-C", "target-feature=+crt-static"]
[target.i686-pc-windows-msvc]
rustflags = ["-C", "target-feature=+crt-static"]
[target.aarch64-pc-windows-msvc]
rustflags = ["-C", "target-feature=+crt-static"]
.github/workflows/CICD.yml
+4 -6
View File
@@ -161,12 +161,12 @@ jobs:
fail-fast: false
matrix:
job:
- { target: aarch64-unknown-linux-musl , os: ubuntu-latest , dpkg_arch: arm64, use-cross: true }
- { target: aarch64-unknown-linux-musl , os: ubuntu-latest , dpkg_arch: musl-linux-arm64, use-cross: true }
- { target: aarch64-unknown-linux-gnu , os: ubuntu-latest , dpkg_arch: arm64, use-cross: true }
- { target: arm-unknown-linux-gnueabihf , os: ubuntu-latest , dpkg_arch: armhf, use-cross: true }
- { target: arm-unknown-linux-musleabihf, os: ubuntu-latest , dpkg_arch: musl-linux-armhf, use-cross: true }
- { target: i686-pc-windows-msvc , os: windows-2025 , }
- { target: i686-unknown-linux-gnu , os: ubuntu-latest , dpkg_arch: i686, use-cross: true }
- { target: i686-unknown-linux-gnu , os: ubuntu-latest , dpkg_arch: i386, use-cross: true }
- { target: i686-unknown-linux-musl , os: ubuntu-latest , dpkg_arch: musl-linux-i686, use-cross: true }
- { target: x86_64-apple-darwin , os: macos-15-intel, }
- { target: aarch64-apple-darwin , os: macos-latest , }
@@ -195,9 +195,7 @@ jobs:
- name: Install cross
if: matrix.job.use-cross
uses: taiki-e/install-action@v2
with:
tool: cross
run: cargo install cross --git https://github.com/cross-rs/cross --rev 588b3c99db52b5a9c5906fab96cfadcf1bde7863
- name: Overwrite build command env variable
if: matrix.job.use-cross
@@ -335,7 +333,7 @@ jobs:
DPKG_BASENAME=${{ needs.crate_metadata.outputs.name }}
DPKG_CONFLICTS=${{ needs.crate_metadata.outputs.name }}-musl
case ${{ matrix.job.target }} in *-musl) DPKG_BASENAME=${{ needs.crate_metadata.outputs.name }}-musl ; DPKG_CONFLICTS=${{ needs.crate_metadata.outputs.name }} ;; esac;
case ${{ matrix.job.target }} in *-musl*) DPKG_BASENAME=${{ needs.crate_metadata.outputs.name }}-musl ; DPKG_CONFLICTS=${{ needs.crate_metadata.outputs.name }} ;; esac;
DPKG_VERSION=${{ needs.crate_metadata.outputs.version }}
DPKG_ARCH="${{ matrix.job.dpkg_arch }}"
DPKG_NAME="${DPKG_BASENAME}_${DPKG_VERSION}_${DPKG_ARCH}.deb"
.gitmodules
+6 -3
View File
@@ -49,9 +49,6 @@
[submodule "assets/themes/zenburn"]
path = assets/themes/zenburn
url = https://github.com/colinta/zenburn.git
[submodule "assets/syntaxes/Kotlin"]
path = assets/syntaxes/02_Extra/Kotlin
url = https://github.com/vkostyukov/kotlin-sublime-package
[submodule "assets/syntaxes/Elm"]
path = assets/syntaxes/02_Extra/Elm
url = https://github.com/elm-community/SublimeElmLanguageSupport
@@ -281,3 +278,9 @@
[submodule "assets/syntaxes/02_Extra/COBOL"]
path = assets/syntaxes/02_Extra/COBOL
url = https://github.com/adukhan99/sublime_cobol.git
[submodule "assets/syntaxes/02_Extra/Kotlin"]
path = assets/syntaxes/02_Extra/Kotlin
url = https://github.com/guille/sublime-kotlin
[submodule "assets/syntaxes/02_Extra/Caddy"]
path = assets/syntaxes/02_Extra/Caddy
url = https://github.com/caddyserver/sublimetext.git
CHANGELOG.md
+44 -3
View File
@@ -1,19 +1,44 @@
# unreleased
- Fixed bug caused by using `--plain` and `--terminal-width=N` flags simultaneously, see #3529 (@H4k1l)
- Fixed syntax tests path, see #3610 (@foxfromworld)
## Other
- Update Cargo dependencies to resolve current RustSec advisories, see #3861 (@TyceHerrman)
- Add instructions for removing fish help abbreviations to README, see #3655 (@claw-explorer). Closes #3536
- Add .NET slnx extension, see #3682 (@ltrzesniewski)
## Features
- Add a `--sanitize=<auto|always|never>` flag for safe display of untrusted input. It implies `--strip-ansi` at the same value and additionally substitutes terminal-active control bytes (cursor moves, charset switches, beep, etc.) and Unicode bidi / zero-width formatting characters with the Unicode replacement character (U+FFFD). Mitigates Trojan-Source-style spoofing (CVE-2021-42574). See #3729 (@curious-rabbit)
- Map justfile, Justfile, .justfile, and *.justfile to Makefile syntax highlighting, see #3623 (@zachvalenta)
- Preserve `--diff` change markers and snip separators when `--plain` is set. Closes #3630, see #3643 (@mvanhorn)
- Added support for `hidden_file_extensions` from `.sublime-syntax` files, see #3613 (@Matei02355)
- Add word wrapping mode via `--wrap=word`, see #3597 (@veeceey)
- Support configuring `--terminal-width` via `BAT_WIDTH`, see #3679 (@officialasishkumar)
- Implement `--unbuffered` mode for streaming input, allowing partial lines to display immediately (e.g. `tail -f | bat -u`). Closes #3555, see #3583 (@mainnebula)
- Added an initial `flake.nix` for a ready made development environment; see #3578 (@vorburger)
- Add `--quiet-empty` (`-E`) flag to suppress output when input is empty. Closes #1936, see #3563 (@NORMAL-EX)
- Improve native man pages and command help syntax highlighting by stripping overstriking, see #3517 (@akirk)
- Add `--fallback-syntax`/`--fallback-language` to apply syntax highlighting only when auto-detection fails, see #1341 (@Xavrir)
- Map `BUILD` case sensitively to Python (Starlark) for Bazel, see #3576 (@vorburger)
- Syntax highlighting for Python files using uv as script runner in shebang #3689 (@janlarres)
## Bugfixes
- Fix `--list-languages` respecting `--paging=never`, see #3828 (@cyphercodes)
- Fix `--sanitize` passing through the bidi control characters U+200E, U+200F and U+061C, see #3862 (@lenamonj)
- `--strip-ansi`: also strip 8-bit C1 introducers (U+0090, U+0098, U+009B, U+009D, U+009E, U+009F) and DCS/SOS/PM/APC sequence bodies, which previously passed through. See #3729 (@curious-rabbit)
- Fix `--ignored-suffix` not falling back to first-line/shebang detection when the ignored suffix is also a registered extension (e.g. `--ignored-suffix .txt` on a shebang script), see #2745 and #3816 (@adnrivera)
- Fix `capacity overflow` panic when printing a snip separator at `--terminal-width=1` with multiple line ranges. Closes #3803, see #3804 (@leeewee)
- Pass `--no-paging` to `bat` invocations inside the bash / zsh / fish / PowerShell shell completion scripts so that shell-level pager wiring (e.g. `LESSOPEN='|-bat -f -pp %s'`) cannot inject ANSI escape sequences into the completion candidates. Closes #3760 (@mvanhorn)
- Quote filenames before substituting them into `$LESSOPEN` / `$LESSCLOSE` templates, preventing shell injection when a filename contains shell metacharacters, see #3726 (@curious-rabbit)
- Fix `--list-themes` unconditionally probing the terminal via OSC 10/11 even when `--theme` was set to an explicit value, see #3700 (regression introduced in bc42149a). (@optimistiCli)
- Fix inverted `$LESSCLOSE` warning so bat warns on nonzero exit, not on success. See #3654 (@cuiweixie)
- Sanitize control characters in filenames before displaying them in the file header, error messages, and the terminal title, preventing ANSI escape injection via crafted filenames. Closes #3054, see #3691 (@curious-rabbit)
- Report initial input read errors instead of treating them as empty input. Closes #3002, see #3706 (@lawrence3699)
- Treat ZIP archives as binary content based on their magic header, see #3686 (@officialasishkumar)
- Fix i686 `.deb` package using incorrect architecture name (`i686` instead of `i386`), preventing installation on Debian. Closes #3611, see #3650 (@Sim-hu)
- Fix inconsistent `.deb` MUSL package names (aarch64-musl used `arm64` instead of `musl-linux-arm64`, and `musleabihf` target missed `bat-musl` prefix). Closes #3482, see #3642 (@mvanhorn)
- Fix incorrect text width computation when using `--binary=as-text` with non-printable characters in caret notation, see #3640 and #3631 (@eyupcanakman)
- Fix `BAT_CONFIG_DIR` pointing at system config directory causing duplicate flag errors. Closes #3589, see #3620 (@Xavrir)
- Fix syntax highlighting for symlinked files when the symlink name has no extension but the target does. Closes #1001, see #3621 (@Xavrir)
- Report error when pager is missing instead of silently falling back, see #3588 (@IMaloney)
@@ -23,19 +48,35 @@
- `--help` now correctly honors `--pager=builtin`. See #3516 (@keith-hall)
- `--help` now correctly honors custom themes. See #3524 (@keith-hall)
- Fixed test compatibility with future Cargo build directory changes, see #3550 (@nmacl)
- Fixed bug caused by using `--plain` and `--terminal-width=N` flags simultaneously, see #3529 (@H4k1l)
- Fixed syntax tests path, see #3610 (@foxfromworld)
- Fix zsh tab completion word-splitting language names containing spaces (e.g. `HTML (Jinja2)`, `Apache Conf`), see #3693 (@YoshKoz)
- Fix zsh tab completion offering invalid `-l` arguments (file globs, paths, hidden filenames) sourced from the second column of `--list-languages`. Closes #3735, see #3737 (@truffle-dev)
- Fix `usize` underflow in `--list-languages` when `--terminal-width` is smaller than the longest language name, see #3812 (@greymoth-jp)
## Other
- Use git version of cross. See #3533 (@OctopusET)
- Bump MSRV to 1.88, update `time` crate to 0.3.47 to fix RUSTSEC-2026-0009, see #3581 (@NORMAL-EX)
- Allow home and end keys to be used with builtin pager, see #3651 (@keith-hall)
- Builtin syntax mapping: cleanup matcher glob parsing logic #3652 (@cyqsimon)
- Statically link the CRT for MSVC builds via Cargo config to avoid runtime DLL dependencies. Closes #3634, see #3692 (@barry3406)
- Replace `libgit2` with a pure Rust implementation of git called `gitoxide`, see PR #3703 (@blinxen)
## Syntaxes
- Add shebang-based detection for Tcl (`tclsh`, `wish`) and Expect (`expect`) scripts, see #3647 (@mvanhorn)
- Change the URL of Zig submodule from GitHub to Codeberg, see #3519 (@sorairolake)
- Don't color strings inside CSV files, to make it easier to tell which column they belong to, see #3521 (@keith-hall)
- Add syntax highlighting support for COBOL, see #3584 (@adukhan99)
- Fixed manpage syntax so that ANSI escape codes don't get incorrectly highlighted and thus broken, see #3586 (@BlueElectivire)
- Map several Google Cloud CLI config files to their appropriate syntax #3635 (@victor-gp)
- Map all ignore dotfiles to Git Ignore syntax #3636 (@victor-gp)
- Improved Kotlin syntax, see #3699 (@guille)
- Include subdirectories in SSH Config syntax mapping, see #3758 (@injust)
- Add Ghostty syntax mapping, see #3759 (@injust)
- Add syntax highlighting for `Caddyfile` #3789 (@CosmicHorrorDev)
- Include `.code-workspace` as a JSON extension #3809 (@dhruvkb)
- Add syntax mapping for DNF repo configuration files, see #3814 (@injust)
## Themes
Cargo.lock
Generated
+1591 -691
View File
File diff suppressed because it is too large Load Diff
Cargo.toml
+19 -19
View File
@@ -31,8 +31,7 @@ minimal-application = [
"regex-onig",
"wild",
]
git = ["git2"] # Support indicating git modifications
vendored-libgit2 = ["git2/vendored-libgit2"]
git = ["gix"] # Support indicating git modifications
paging = [ "shell-words", "grep-cli", "minus"] # Support applying a pager on the output
lessopen = ["execute"] # Support $LESSOPEN preprocessor
build-assets = ["syntect/yaml-load", "syntect/plist-load", "regex", "walkdir"]
@@ -52,11 +51,11 @@ thiserror = "2.0"
wild = { version = "2.2", optional = true }
content_inspector = "0.2.4"
shell-words = { version = "1.1.1", optional = true }
minus = { version = "5.6", optional = true, features = [
minus = { version = "5.7", optional = true, features = [
"dynamic_output",
"search",
] }
unicode-width = "0.2.1"
unicode-width = "0.2.2"
globset = "0.4"
serde = "1.0"
serde_derive = "1.0"
@@ -64,22 +63,23 @@ serde_yaml = "0.9.28"
semver = "1.0"
path_abs = { version = "0.5", default-features = false }
clircle = { version = "0.6.1", default-features = false }
bugreport = { version = "0.5.0", optional = true }
bugreport = { version = "0.6.0", optional = true }
etcetera = { version = "0.11.0", optional = true }
grep-cli = { version = "0.1.12", optional = true }
regex = { version = "1.12.2", optional = true }
regex = { version = "1.12.3", optional = true }
walkdir = { version = "2.5", optional = true }
bytesize = { version = "2.3.1" }
encoding_rs = "0.8.35"
execute = { version = "0.2.15", optional = true }
execute = { version = "0.3.0", optional = true }
terminal-colorsaurus = "1.0"
unicode-segmentation = "1.12.0"
unicode-segmentation = "1.13.2"
itertools = "0.14.0"
[dependencies.git2]
version = "0.20"
[dependencies.gix]
version = "0.86"
optional = true
default-features = false
features = ["sha1", "blob-diff"]
[dependencies.syntect]
version = "5.3.0"
@@ -87,20 +87,20 @@ default-features = false
features = ["parsing"]
[dependencies.clap]
version = "4.5.60"
version = "4.6.1"
optional = true
features = ["wrap_help", "cargo"]
[target.'cfg(target_os = "macos")'.dependencies]
plist = "1.7.0"
plist = "1.9.0"
[dev-dependencies]
assert_cmd = "2.0.16"
expect-test = "1.5.0"
serial_test = { version = "2.0.0", default-features = false }
predicates = "3.1.3"
predicates = "3.1.4"
wait-timeout = "0.2.1"
tempfile = "3.23.0"
tempfile = "3.27.0"
serde = { version = "1.0", features = ["derive"] }
[target.'cfg(unix)'.dev-dependencies]
@@ -108,22 +108,22 @@ nix = { version = "0.31", default-features = false, features = ["term"] }
[build-dependencies]
anyhow = "1.0.97"
indexmap = { version = "2.13.0", features = ["serde"] }
indexmap = { version = "2.14.0", features = ["serde"] }
itertools = "0.14.0"
once_cell = "1.20"
prettyplease = "0.2.37"
proc-macro2 = "1.0.106"
quote = "1.0.40"
regex = "1.12.2"
quote = "1.0.45"
regex = "1.12.3"
serde = "1.0"
serde_derive = "1.0"
serde_with = { version = "3.17.0", default-features = false, features = ["macros"] }
syn = { version = "2.0.104", features = ["full"] }
toml = { version = "0.9.8", features = ["preserve_order"] }
toml = { version = "1.1.1", features = ["preserve_order"] }
walkdir = "2.5"
[build-dependencies.clap]
version = "4.5.60"
version = "4.6.1"
optional = true
features = ["wrap_help", "cargo"]
README.md
+38 -22
View File
@@ -19,20 +19,6 @@
[<a href="doc/README-ru.md">Русский</a>]
</p>
### Sponsors
A special *thank you* goes to our biggest <a href="doc/sponsors.md">sponsors</a>:<br>
<p>
<a href="https://www.warp.dev/bat">
<img src="doc/sponsors/warp-logo.png" width="200" alt="Warp">
<br>
<strong>Warp, the intelligent terminal</strong>
<br>
<sub>Available on MacOS, Linux, Windows</sub>
</a>
</p>
### Syntax highlighting
`bat` supports syntax highlighting for a large number of programming and markup
@@ -208,12 +194,6 @@ Note that the [Manpage syntax](assets/syntaxes/02_Extra/Manpage.sublime-syntax)
The [`prettybat`](https://github.com/eth-p/bat-extras/blob/master/doc/prettybat.md) script is a wrapper that will format code and print it with `bat`.
#### `Warp`
<a href="https://app.warp.dev/drive/folder/-Bat-Warp-Pack-lxhe7HrEwgwpG17mvrFSz1">
<img src="doc/sponsors/warp-pack-header.png" alt="Warp">
</a>
#### Highlighting `--help` messages
You can use `bat` to colorize help text: `$ cp --help | bat -plhelp`
@@ -246,6 +226,14 @@ abbr -a --position anywhere -- -h '-h | bat -plhelp'
This way, you can keep on using `cp --help`, but get colorized help pages.
> [!TIP]
> To remove these abbreviations later, run:
> ```fish
> abbr -e -- --help
> abbr -e -- -h
> ```
> The `--` before the abbreviation name is required because `--help` and `-h` start with dashes, which would otherwise be interpreted as flags to `abbr` itself.
Be aware that in some cases, `-h` may not be a shorthand of `--help` (for example with `ls`). In cases where you need to use `-h`
as a command argument you can prepend `\` to the argument (eg. `ls \-h`) to escape the aliasing defined above.
@@ -319,7 +307,7 @@ pacman -S bat
### On Fedora
You can install [the `bat` package](https://koji.fedoraproject.org/koji/packageinfo?packageID=27506) from the official [Fedora Modular](https://docs.fedoraproject.org/en-US/modularity/using-modules/) repository.
You can install [the `bat` package](https://packages.fedoraproject.org/pkgs/rust-bat/bat/) from the official sources:
```bash
dnf install bat
@@ -519,6 +507,30 @@ and line numbers but no grid and no file header. Set the `BAT_STYLE` environment
variable to make these changes permanent or use `bat`'s
[configuration file](#configuration-file).
By default, `bat` enables `changes`, `grid`, `header-filename`, `numbers`, and `snip`.
The available pre-defined styles are:
| Style | Description |
|-------|-------------|
| `default` | Enables the recommended style components listed above. |
| `full` | Enables all available components. |
| `auto` | Same as `default`, unless the output is piped. |
| `plain` | Disables all available components. |
The available individual components are:
| Component | Description |
|-----------|-------------|
| `changes` | Show Git modification markers. |
| `header` | Alias for `header-filename`. |
| `header-filename` | Show filenames before the content. |
| `header-filesize` | Show file sizes before the content. |
| `grid` | Vertical/horizontal lines to separate the side bar and header from the content. |
| `rule` | Horizontal lines to delimit files. |
| `numbers` | Show line numbers in the side bar. |
| `snip` | Draw separation lines between distinct line ranges. |
>[!tip]
> If you specify a default style in `bat`'s config file, you can change which components
> are displayed during a single run of `bat` using the `--style` command-line argument.
@@ -831,6 +843,10 @@ If your version of `bat` supports the `--strip-ansi=auto` option, it can be used
before syntax highlighting. Alternatively, you may disable both syntax highlighting and wrapping by
passing the `--color=never --wrap=never` options to `bat`.
For untrusted input, the `--sanitize=auto|always|never` option additionally replaces terminal-active
control bytes and Unicode bidi / zero-width formatting characters with the Unicode replacement
character. It implies `--strip-ansi` at the same value.
> [!NOTE]
> The `auto` option of `--strip-ansi` avoids removing escape sequences when the syntax is plain text.
@@ -886,7 +902,7 @@ cargo install --path . --locked --force
```
If you want to build an application that uses `bat`'s pretty-printing
features as a library, check out the [the API documentation](https://docs.rs/bat/).
features as a library, check out the [API documentation](https://docs.rs/bat/).
Note that you have to use either `regex-onig` or `regex-fancy` as a feature
when you depend on `bat` as a library.
assets/completions/_bat.ps1.in
Vendored
+8 -2
View File
@@ -14,12 +14,12 @@ Register-ArgumentCompleter -Native -CommandName '{{PROJECT_EXECUTABLE}}' -Script
$ArrayPrint = @('unicode', 'caret')
function Get-MyThemes(){
$themes = {{PROJECT_EXECUTABLE}} --list-themes | ForEach-Object {$_ -replace "^(.*)$", '''$1'''} | select-object
$themes = {{PROJECT_EXECUTABLE}} --no-paging --list-themes | ForEach-Object {$_ -replace "^(.*)$", '''$1'''} | select-object
return $themes
}
function Get-MyLanguages(){
$themes = {{PROJECT_EXECUTABLE}} --list-languages | ForEach-Object{[pscustomobject]@{MyParameter=$_.Substring(0,$_.IndexOf(":")).Trim();MyDescription=$_.Substring($_.IndexOf(":")+1)}} | select-object
$themes = {{PROJECT_EXECUTABLE}} --no-paging --list-languages | ForEach-Object{[pscustomobject]@{MyParameter=$_.Substring(0,$_.IndexOf(":")).Trim();MyDescription=$_.Substring($_.IndexOf(":")+1)}} | select-object
return $themes
}
@@ -100,6 +100,11 @@ Register-ArgumentCompleter -Native -CommandName '{{PROJECT_EXECUTABLE}}' -Script
ForEach-Object {[System.Management.Automation.CompletionResult]::new($_, $_, [CompletionResultType]::ParameterValue, $_)}
break
}
'*;--sanitize' {
$ArrayWhen |
ForEach-Object {[System.Management.Automation.CompletionResult]::new($_, $_, [CompletionResultType]::ParameterValue, $_)}
break
}
'*;--strip-ansi' {
$ArrayWhen |
ForEach-Object {[System.Management.Automation.CompletionResult]::new($_, $_, [CompletionResultType]::ParameterValue, $_)}
@@ -158,6 +163,7 @@ Register-ArgumentCompleter -Native -CommandName '{{PROJECT_EXECUTABLE}}' -Script
[CompletionResult]::new('--ignored-suffix' , 'ignored-suffix' , [CompletionResultType]::ParameterName, 'Ignore extension. For example: ''bat --ignored-suffix ".dev" my_file.json.dev'' will use JSON syntax, and ignore ''.dev''')
[CompletionResult]::new('--squeeze-blank' , 'squeeze-blank' , [CompletionResultType]::ParameterName, 'Squeeze consecutive empty lines into a single empty line.')
[CompletionResult]::new('--squeeze-limit' , 'squeeze-limit' , [CompletionResultType]::ParameterName, 'Set the maximum number of consecutive empty lines to be printed.')
[CompletionResult]::new('--sanitize' , 'sanitize' , [CompletionResultType]::ParameterName, 'Specify when to sanitize untrusted input for safe display. Implies --strip-ansi and also replaces terminal-active and bidi / zero-width bytes. (auto, always, *never*).')
[CompletionResult]::new('--strip-ansi' , 'strip-ansi' , [CompletionResultType]::ParameterName, 'Specify when to strip ANSI escape sequences from the input. The automatic mode will remove escape sequences unless the syntax highlighting language is plain text. (auto, always, *never*).')
# [CompletionResult]::new('-p' , 'p' , [CompletionResultType]::ParameterName, 'Show plain style (alias for ''--style=plain'').')
[CompletionResult]::new('--plain' , 'plain' , [CompletionResultType]::ParameterName, 'Show plain style (alias for ''--style=plain'').')
assets/completions/bat.bash.in
Vendored
+8 -3
View File
@@ -80,7 +80,7 @@ _bat() {
-l | --language)
local IFS=$'\n'
COMPREPLY=($(compgen -W "$(
"$1" --list-languages | while IFS=: read -r lang _; do
"$1" --no-paging --list-languages | while IFS=: read -r lang _; do
printf "%s\n" "$lang"
done
)" -- "$cur"))
@@ -132,6 +132,10 @@ _bat() {
COMPREPLY=($(compgen -W "auto never always" -- "$cur"))
return 0
;;
--sanitize)
COMPREPLY=($(compgen -W "auto never always" -- "$cur"))
return 0
;;
--completion)
COMPREPLY=($(compgen -W "bash fish zsh ps1" -- "$cur"))
return 0
@@ -150,14 +154,14 @@ _bat() {
;;
--theme)
local IFS=$'\n'
COMPREPLY=($(compgen -W "auto${IFS}auto:always${IFS}auto:system${IFS}dark${IFS}light${IFS}$("$1" --list-themes)" -- "$cur"))
COMPREPLY=($(compgen -W "auto${IFS}auto:always${IFS}auto:system${IFS}dark${IFS}light${IFS}$("$1" --no-paging --list-themes)" -- "$cur"))
__bat_escape_completions
return 0
;;
--theme-dark | \
--theme-light)
local IFS=$'\n'
COMPREPLY=($(compgen -W "$("$1" --list-themes)" -- "$cur"))
COMPREPLY=($(compgen -W "$("$1" --no-paging --list-themes)" -- "$cur"))
__bat_escape_completions
return 0
;;
@@ -221,6 +225,7 @@ _bat() {
--list-themes
--squeeze-blank
--squeeze-limit
--sanitize
--strip-ansi
--style
--line-range
assets/completions/bat.fish.in
Vendored
+6 -5
View File
@@ -15,11 +15,11 @@ function __bat_complete_files -a token
end
function __bat_complete_one_language -a comp
command $bat --list-languages | string split -f1 : | string match -e "$comp"
command $bat --no-paging --list-languages | string split -f1 : | string match -e "$comp"
end
function __bat_complete_list_languages
for spec in (command $bat --list-languages)
for spec in (command $bat --no-paging --list-languages)
set -l name (string split -f1 : $spec)
for ext in (string split -f2 : $spec | string split ,)
test -n "$ext"; or continue
@@ -221,6 +221,7 @@ complete -c $bat -s s -l squeeze-blank -d "Squeeze consecutive empty lines into
complete -c $bat -l squeeze-limit -x -d "Set the maximum number of consecutive empty lines to be printed" -n __bat_no_excl_args
complete -c $bat -l strip-ansi -x -a "auto never always" -d "Specify when to strip ANSI escape sequences from the input" -n __bat_no_excl_args
complete -c $bat -l sanitize -x -a "auto never always" -d "Specify when to sanitize untrusted input for safe display" -n __bat_no_excl_args
complete -c $bat -s p -l plain -d "Disable decorations" -n __bat_no_excl_args
@@ -234,11 +235,11 @@ complete -c $bat -l tabs -x -a "$tabs_opts" -d "Set tab width" -n __bat_no_excl_
complete -c $bat -l terminal-width -x -d "Set terminal <width>, +<offset>, or -<offset>" -n __bat_no_excl_args
complete -c $bat -l theme -x -a "$special_themes(command $bat --list-themes | command cat)" -d "Set the syntax highlighting theme" -n __bat_no_excl_args
complete -c $bat -l theme -x -a "$special_themes(command $bat --no-paging --list-themes | command cat)" -d "Set the syntax highlighting theme" -n __bat_no_excl_args
complete -c $bat -l theme-dark -x -a "(command $bat --list-themes | command cat)" -d "Set the syntax highlighting theme for dark backgrounds" -n __bat_no_excl_args
complete -c $bat -l theme-dark -x -a "(command $bat --no-paging --list-themes | command cat)" -d "Set the syntax highlighting theme for dark backgrounds" -n __bat_no_excl_args
complete -c $bat -l theme-light -x -a "(command $bat --list-themes | command cat)" -d "Set the syntax highlighting theme for light backgrounds" -n __bat_no_excl_args
complete -c $bat -l theme-light -x -a "(command $bat --no-paging --list-themes | command cat)" -d "Set the syntax highlighting theme for light backgrounds" -n __bat_no_excl_args
complete -c $bat -s u -l unbuffered -d "Enable unbuffered input reading for streaming use cases" -n __bat_no_excl_args
assets/completions/bat.zsh.in
Vendored
+9 -3
View File
@@ -54,6 +54,7 @@ _{{PROJECT_EXECUTABLE}}_main() {
--squeeze-blank'[squeeze consecutive empty lines into a single empty line]'
--squeeze-limit='[set the maximum number of consecutive empty lines]:limit:'
--strip-ansi='[specify when to strip ANSI escape sequences]:when:(auto never always)'
--sanitize='[specify when to sanitize untrusted input for safe display]:when:(auto never always)'
--style='[comma-separated list of style elements to display]: : _values "style [default]"
default auto full plain changes header header-filename header-filesize grid rule numbers snip'
\*{-r+,--line-range=}'[only print the specified line range]:start\:end'
@@ -90,20 +91,25 @@ _{{PROJECT_EXECUTABLE}}_main() {
languages)
local IFS=$'\n'
local -a languages
languages=( $({{PROJECT_EXECUTABLE}} --list-languages | awk -F':|,' '{ for (i = 1; i <= NF; ++i) printf("%s:%s\n", $i, $1) }') )
# `--list-languages` emits one `name:matchers` line per language,
# which `_describe` parses as `value:description`. Only the
# language name is offered as the completion value; the matchers
# show up as the menu description. See
# https://github.com/sharkdp/bat/issues/3735.
languages=( ${(f)"$({{PROJECT_EXECUTABLE}} --no-paging --color=never --decorations=never --list-languages)"} )
_describe 'language' languages && ret=0
;;
themes)
local -a themes expl
themes=(${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --list-themes)"} )
themes=(${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --no-paging --color=never --decorations=never --list-themes)"} )
_wanted themes expl 'theme' compadd -a themes && ret=0
;;
theme_preferences)
local -a themes expl
themes=(auto dark light auto:always auto:system ${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --list-themes)"} )
themes=(auto dark light auto:always auto:system ${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --no-paging --color=never --decorations=never --list-themes)"} )
_wanted themes expl 'theme' compadd -a themes && ret=0
;;
assets/patches/Python.sublime-syntax.patch
Vendored
+9
View File
@@ -2,6 +2,15 @@ diff --git syntaxes/01_Packages/Python/Python.sublime-syntax syntaxes/01_Package
index 2acd86d8..86257f7b 100644
--- syntaxes/01_Packages/Python/Python.sublime-syntax
+++ syntaxes/01_Packages/Python/Python.sublime-syntax
@@ -25,7 +31,7 @@ file_extensions:
- wscript
- bazel
- bzl
-first_line_match: ^#!\s*/.*\bpython(\d(\.\d)?)?\b
+first_line_match: ^#!\s*/.*\b(python(\d(\.\d)?)?|uv)\b
scope: source.python
variables:
@@ -988,10 +988,6 @@ contexts:
- match: \}
scope: punctuation.section.mapping-or-set.end.python
assets/patches/Tcl.sublime-syntax.patch
Vendored
+12
View File
@@ -0,0 +1,12 @@
diff --git syntaxes/01_Packages/TCL/Tcl.sublime-syntax syntaxes/01_Packages/TCL/Tcl.sublime-syntax
index 1234567..abcdefg 100644
--- syntaxes/01_Packages/TCL/Tcl.sublime-syntax
+++ syntaxes/01_Packages/TCL/Tcl.sublime-syntax
@@ -3,6 +3,7 @@
# http://www.sublimetext.com/docs/3/syntax.html
name: Tcl
file_extensions:
- tcl
+first_line_match: ^\#!.*\b(tclsh|wish|expect)\b
scope: source.tcl
variables:
assets/syntaxes/02_Extra/Caddy
Vendored Submodule
+1
Submodule assets/syntaxes/02_Extra/Caddy added at cd7d132da3
assets/syntaxes/02_Extra/Idris2
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/Idris2 updated: bbfe50e023...4d8eb35a38
assets/syntaxes/02_Extra/Kotlin
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/Kotlin updated: aeeed2780b...c353694169
assets/syntaxes/02_Extra/Kotlin.sublime-syntax
Vendored
-398
View File
@@ -1,398 +0,0 @@
%YAML 1.2
---
# http://www.sublimetext.com/docs/3/syntax.html
name: Kotlin
file_extensions:
- kt
- kts
scope: source.Kotlin
contexts:
main:
- include: comments
- match: '^\s*(package)\b(?:\s*([^ ;$]+)\s*)?'
captures:
1: keyword.other.kotlin
2: entity.name.package.kotlin
- include: imports
- include: statements
classes:
- match: (?<!::)(?=\b(?:companion|class|object|interface)\b)
push:
- match: '(?=$|\})'
pop: true
- include: comments
- match: \b(companion\s*)?(class|object|interface)\b
captures:
1: storage.modifier.kotlin
2: storage.modifier.kotlin
push:
- match: '(?=<|\{|\(|:|$)'
pop: true
- include: comments
- match: \w+
scope: entity.name.type.class.kotlin
- match: <
push:
- match: ">"
pop: true
- include: generics
- match: \(
push:
- match: \)
pop: true
- include: parameters
- match: (:)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: "(?={|$)"
pop: true
- match: \w+
scope: entity.other.inherited-class.kotlin
- match: \(
push:
- match: \)
pop: true
- include: expressions
- match: '\{'
push:
- match: '\}'
pop: true
- include: statements
comments:
- match: /\*
captures:
0: punctuation.definition.comment.kotlin
push:
- meta_scope: comment.block.kotlin
- match: \*/
captures:
0: punctuation.definition.comment.kotlin
pop: true
- match: \s*((//).*$\n?)
captures:
1: comment.line.double-slash.kotlin
2: punctuation.definition.comment.kotlin
constants:
- match: \b(true|false|null|this|super)\b
scope: constant.language.kotlin
- match: '\b((0(x|X)[0-9a-fA-F]*)|(([0-9]+\.?[0-9]*)|(\.[0-9]+))((e|E)(\+|-)?[0-9]+)?)([LlFf])?\b'
scope: constant.numeric.kotlin
- match: '\b([A-Z][A-Z0-9_]+)\b'
scope: constant.other.kotlin
expressions:
- match: \(
push:
- match: \)
pop: true
- include: expressions
- include: types
- include: strings
- include: constants
- include: comments
- include: keywords
functions:
- match: (?=\s*\b(?:fun)\b)
push:
- match: '(?=$|\})'
pop: true
- match: \b(fun)\b
captures:
1: keyword.other.kotlin
push:
- match: (?=\()
pop: true
- match: <
push:
- match: ">"
pop: true
- include: generics
- match: '([\.<\?>\w]+\.)?(\w+)'
captures:
2: entity.name.function.kotlin
- match: \(
push:
- match: \)
pop: true
- include: parameters
- match: (:)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: "(?={|=|$)"
pop: true
- include: types
- match: '\{'
push:
- match: '(?=\})'
pop: true
- include: statements
- match: (=)
captures:
1: keyword.operator.assignment.kotlin
push:
- match: (?=$)
pop: true
- include: expressions
generics:
- match: (:)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: (?=,|>)
pop: true
- include: types
- include: keywords
- match: \w+
scope: storage.type.generic.kotlin
getters-and-setters:
- match: \b(get)\b\s*\(\s*\)
captures:
1: entity.name.function.kotlin
push:
- match: '\}|(?=\bset\b)|$'
pop: true
- match: (=)
captures:
1: keyword.operator.assignment.kotlin
push:
- match: (?=$|\bset\b)
pop: true
- include: expressions
- match: '\{'
push:
- match: '\}'
pop: true
- include: expressions
- match: \b(set)\b\s*(?=\()
captures:
1: entity.name.function.kotlin
push:
- match: '\}|(?=\bget\b)|$'
pop: true
- match: \(
push:
- match: \)
pop: true
- include: parameters
- match: (=)
captures:
1: keyword.operator.assignment.kotlin
push:
- match: (?=$|\bset\b)
pop: true
- include: expressions
- match: '\{'
push:
- match: '\}'
pop: true
- include: expressions
imports:
- match: '^\s*(import)\s+[^ $]+\s+(as)?'
captures:
1: keyword.other.kotlin
2: keyword.other.kotlin
keywords:
- match: \b(var|val|public|private|protected|abstract|final|sealed|enum|open|attribute|annotation|override|inline|vararg|in|out|internal|data|tailrec|operator|infix|const|yield|typealias|typeof|reified|suspend)\b
scope: storage.modifier.kotlin
- match: \b(try|catch|finally|throw)\b
scope: keyword.control.catch-exception.kotlin
- match: \b(if|else|while|for|do|return|when|where|break|continue)\b
scope: keyword.control.kotlin
- match: \b(in|is|!in|!is|as|as\?|assert)\b
scope: keyword.operator.kotlin
- match: (==|!=|===|!==|<=|>=|<|>)
scope: keyword.operator.comparison.kotlin
- match: (=)
scope: keyword.operator.assignment.kotlin
- match: (::)
scope: keyword.operator.kotlin
- match: (:)
scope: keyword.operator.declaration.kotlin
- match: \b(by)\b
scope: keyword.other.by.kotlin
- match: (\?\.)
scope: keyword.operator.safenav.kotlin
- match: (\.)
scope: keyword.operator.dot.kotlin
- match: (\?:)
scope: keyword.operator.elvis.kotlin
- match: (\-\-|\+\+)
scope: keyword.operator.increment-decrement.kotlin
- match: (\+=|\-=|\*=|\/=)
scope: keyword.operator.arithmetic.assign.kotlin
- match: (\.\.)
scope: keyword.operator.range.kotlin
- match: (\-|\+|\*|\/|%)
scope: keyword.operator.arithmetic.kotlin
- match: (!|&&|\|\|)
scope: keyword.operator.logical.kotlin
- match: (;)
scope: punctuation.terminator.kotlin
namespaces:
- match: \b(namespace)\b
scope: keyword.other.kotlin
- match: '\{'
push:
- match: '\}'
pop: true
- include: statements
parameters:
- match: (:)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: (?=,|\)|=)
pop: true
- include: types
- match: (=)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: (?=,|\))
pop: true
- include: expressions
- include: keywords
- match: \w+
scope: variable.parameter.function.kotlin
statements:
- include: namespaces
- include: typedefs
- include: classes
- include: functions
- include: variables
- include: getters-and-setters
- include: expressions
strings:
- match: '"""'
captures:
0: punctuation.definition.string.begin.kotlin
push:
- meta_scope: string.quoted.third.kotlin
- match: '"""'
captures:
0: punctuation.definition.string.end.kotlin
pop: true
- match: '(\$\w+|\$\{[^\}]+\})'
scope: variable.parameter.template.kotlin
- match: \\.
scope: constant.character.escape.kotlin
- match: '"'
captures:
0: punctuation.definition.string.begin.kotlin
push:
- meta_scope: string.quoted.double.kotlin
- match: '"'
captures:
0: punctuation.definition.string.end.kotlin
pop: true
- match: '(\$\w+|\$\{[^\}]+\})'
scope: variable.parameter.template.kotlin
- match: \\.
scope: constant.character.escape.kotlin
- match: "'"
captures:
0: punctuation.definition.string.begin.kotlin
push:
- meta_scope: string.quoted.single.kotlin
- match: "'"
captures:
0: punctuation.definition.string.end.kotlin
pop: true
- match: \\.
scope: constant.character.escape.kotlin
- match: "`"
captures:
0: punctuation.definition.string.begin.kotlin
push:
- meta_scope: string.quoted.single.kotlin
- match: "`"
captures:
0: punctuation.definition.string.end.kotlin
pop: true
typedefs:
- match: (?=\s*(?:type))
push:
- match: (?=$)
pop: true
- match: \b(type)\b
scope: keyword.other.kotlin
- match: <
push:
- match: ">"
pop: true
- include: generics
- include: expressions
types:
- match: \b(Nothing|Any|Unit|String|CharSequence|Int|Boolean|Char|Long|Double|Float|Short|Byte|dynamic)\b
scope: storage.type.buildin.kotlin
- match: \b(IntArray|BooleanArray|CharArray|LongArray|DoubleArray|FloatArray|ShortArray|ByteArray)\b
scope: storage.type.buildin.array.kotlin
- match: \b(Array|Collection|List|Map|Set|MutableList|MutableMap|MutableSet|Sequence)<\b
captures:
1: storage.type.buildin.collection.kotlin
push:
- match: ">"
pop: true
- include: types
- include: keywords
- match: \w+<
push:
- match: ">"
pop: true
- include: types
- include: keywords
- match: '\{'
push:
- match: '\}'
pop: true
- include: statements
- match: \(
push:
- match: \)
pop: true
- include: types
- match: (->)
scope: keyword.operator.declaration.kotlin
variables:
- match: (?=\s*\b(?:var|val)\b)
push:
- match: (?=:|=|(\b(by)\b)|$)
pop: true
- match: \b(var|val)\b
captures:
1: keyword.other.kotlin
push:
- match: (?=:|=|(\b(by)\b)|$)
pop: true
- match: <
push:
- match: ">"
pop: true
- include: generics
- match: '([\.<\?>\w]+\.)?(\w+)'
captures:
2: entity.name.variable.kotlin
- match: (:)
captures:
1: keyword.operator.declaration.kotlin
push:
- match: (?==|$)
pop: true
- include: types
- include: getters-and-setters
- match: \b(by)\b
captures:
1: keyword.other.kotlin
push:
- match: (?=$)
pop: true
- include: expressions
- match: (=)
captures:
1: keyword.operator.assignment.kotlin
push:
- match: (?=$)
pop: true
- include: expressions
- include: getters-and-setters
assets/syntaxes/02_Extra/LESS
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/LESS updated: 836b47ec61...d076a4dd41
assets/syntaxes/02_Extra/PureScript
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/PureScript updated: 5acebc1850...1773f4fddb
assets/syntaxes/02_Extra/typst-syntax-highlight
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/typst-syntax-highlight updated: 363f0e767c...5f71d12fa1
assets/syntaxes/02_Extra/vscode-wgsl
Vendored
+1 -1
Submodule assets/syntaxes/02_Extra/vscode-wgsl updated: acf26718d7...a285c38f74
build/syntax_mapping.rs
+118 -53
View File
@@ -47,12 +47,49 @@ impl ToTokens for MappingTarget {
}
}
#[derive(Clone, Debug, PartialEq, Eq, Hash, DeserializeFromStr)]
/// Helper type for deserializing a `Matcher` from either a plain string or a
/// `{ glob = "...", case_sensitive = true }` struct.
#[derive(Deserialize)]
#[serde(untagged)]
enum RawMatcher {
Simple(String),
Full {
glob: String,
case_sensitive: Option<bool>,
},
}
/// Whether a glob pattern should be matched case-sensitively or case-insensitively.
///
/// Mirrors the runtime `Case` type in `src/syntax_mapping.rs`.
#[derive(Clone, Debug, PartialEq, Eq, Hash)]
enum Case {
Sensitive,
Insensitive,
}
impl ToTokens for Case {
fn to_tokens(&self, tokens: &mut TokenStream) {
let t = match self {
Self::Sensitive => quote! { Case::Sensitive },
Self::Insensitive => quote! { Case::Insensitive },
};
tokens.append_all(t);
}
}
/// A single matcher.
///
/// Codegen converts this into a `Lazy<Option<GlobMatcher>>`.
struct Matcher(Vec<MatcherSegment>);
/// Parse a matcher.
#[derive(Clone, Debug, PartialEq, Eq, Hash, Deserialize)]
#[serde(try_from = "RawMatcher")]
struct Matcher {
segments: Vec<MatcherSegment>,
/// Whether the glob pattern should be matched case-sensitively.
///
/// Defaults to `Case::Insensitive` for backwards compatibility.
case: Case,
}
/// Parse the glob pattern of a matcher.
///
/// Note that this implementation is rather strict: it will greedily interpret
/// every valid environment variable replacement as such, then immediately
@@ -66,68 +103,95 @@ struct Matcher(Vec<MatcherSegment>);
///
/// Revision history:
/// - 2024-02-20: allow `{` and `}` (glob brace expansion)
impl FromStr for Matcher {
type Err = anyhow::Error;
fn from_str(s: &str) -> Result<Self, Self::Err> {
use MatcherSegment as Seg;
static VAR_REGEX: Lazy<Regex> = Lazy::new(|| Regex::new(r"\$\{([\w\d_]+)\}").unwrap());
fn parse_glob(s: &str) -> Result<Vec<MatcherSegment>, anyhow::Error> {
use MatcherSegment as Seg;
static VAR_REGEX: Lazy<Regex> = Lazy::new(|| Regex::new(r"\$\{([\w\d_]+)\}").unwrap());
let mut segments = vec![];
let mut text_start = 0;
for capture in VAR_REGEX.captures_iter(s) {
let match_0 = capture.get(0).unwrap();
let mut segments = vec![];
let mut text_start = 0;
for capture in VAR_REGEX.captures_iter(s) {
let match_0 = capture.get(0).unwrap();
// text before this var
let text_end = match_0.start();
segments.push(Seg::Text(s[text_start..text_end].into()));
text_start = match_0.end();
// text before this var
let text_end = match_0.start();
segments.push(Seg::Text(s[text_start..text_end].into()));
text_start = match_0.end();
// this var
segments.push(Seg::Env(capture.get(1).unwrap().as_str().into()));
// this var
segments.push(Seg::Env(capture.get(1).unwrap().as_str().into()));
}
// possible trailing text
segments.push(Seg::Text(s[text_start..].into()));
// cleanup empty text segments
let non_empty_segments = segments
.into_iter()
.filter(|seg| seg.text().map(|t| !t.is_empty()).unwrap_or(true))
.collect_vec();
// sanity check
if non_empty_segments
.windows(2)
.any(|segs| segs[0].is_text() && segs[1].is_text())
{
unreachable!("Parsed into consecutive text segments: {non_empty_segments:?}");
}
// guard empty case
if non_empty_segments.is_empty() {
bail!(r#"Parsed an empty matcher: "{s}""#);
}
// guard variable syntax leftover fragments
if non_empty_segments
.iter()
.filter_map(Seg::text)
.any(|t| t.contains('$'))
{
bail!(r#"Invalid matcher: "{s}""#);
}
Ok(non_empty_segments)
}
impl TryFrom<RawMatcher> for Matcher {
type Error = anyhow::Error;
fn try_from(raw: RawMatcher) -> Result<Self, Self::Error> {
const DEFAULT_CASE: Case = Case::Insensitive;
match &raw {
RawMatcher::Simple(s) => {
let segments = parse_glob(s)?;
Ok(Self {
segments,
case: DEFAULT_CASE,
})
}
RawMatcher::Full {
glob,
case_sensitive,
} => {
let segments = parse_glob(glob)?;
let case = match case_sensitive {
None => DEFAULT_CASE,
Some(false) => Case::Insensitive,
Some(true) => Case::Sensitive,
};
Ok(Self { segments, case })
}
}
// possible trailing text
segments.push(Seg::Text(s[text_start..].into()));
// cleanup empty text segments
let non_empty_segments = segments
.into_iter()
.filter(|seg| seg.text().map(|t| !t.is_empty()).unwrap_or(true))
.collect_vec();
// sanity check
if non_empty_segments
.windows(2)
.any(|segs| segs[0].is_text() && segs[1].is_text())
{
unreachable!("Parsed into consecutive text segments: {non_empty_segments:?}");
}
// guard empty case
if non_empty_segments.is_empty() {
bail!(r#"Parsed an empty matcher: "{s}""#);
}
// guard variable syntax leftover fragments
if non_empty_segments
.iter()
.filter_map(Seg::text)
.any(|t| t.contains('$'))
{
bail!(r#"Invalid matcher: "{s}""#);
}
Ok(Self(non_empty_segments))
}
}
impl ToTokens for Matcher {
fn to_tokens(&self, tokens: &mut TokenStream) {
let t = match self.0.as_slice() {
let case = &self.case;
let t = match self.segments.as_slice() {
[] => unreachable!("0-length matcher should never be created"),
[MatcherSegment::Text(text)] => {
quote! { Lazy::new(|| Some(build_matcher_fixed(#text))) }
quote! { Lazy::new(|| Some(build_matcher_fixed(#text, #case))) }
}
// parser logic ensures that this case can only happen when there are dynamic segments
segs @ [_, ..] => quote! { Lazy::new(|| build_matcher_dynamic(&[ #(#segs),* ])) },
segs @ [_, ..] => {
quote! { Lazy::new(|| build_matcher_dynamic(&[ #(#segs),* ], #case)) }
}
};
tokens.append_all(t);
}
@@ -175,6 +239,7 @@ impl MatcherSegment {
/// A struct that models a single .toml file in /src/syntax_mapping/builtins/.
#[derive(Clone, Debug, Deserialize)]
struct MappingDefModel {
#[serde(default)]
mappings: IndexMap<MappingTarget, Vec<Matcher>>,
}
impl MappingDefModel {
doc/assets.md
+1 -1
View File
@@ -97,7 +97,7 @@ The following files have been manually modified after converting from a `.tmLang
as it is not kept in a standalone repository. The file is generated from
https://github.com/open-policy-agent/opa/blob/master/misc/syntax/textmate/Rego.tmLanguage
* `SML.sublime_syntax` has been added manually from
https://github.com/seanjames777/SML-Language-Definitiona as it is not
https://github.com/seanjames777/SML-Language-Definition as it is not
kept in a standalone repository. The file generated is from
https://github.com/seanjames777/SML-Language-Definition/blob/master/sml.tmLanguage
* `Cabal.sublime_syntax` has been added manually from
doc/long-help.txt
+10 -1
View File
@@ -77,7 +77,8 @@ Options:
--terminal-width <width>
Explicitly set the width of the terminal instead of determining it automatically. If
prefixed with '+' or '-', the value will be treated as an offset to the actual terminal
width. See also: '--wrap'.
width. This can also be configured via the BAT_WIDTH environment variable (e.g. export
BAT_WIDTH="100"). See also: '--wrap'.
-n, --number
Only show line numbers, no other decorations. This is an alias for '--style=numbers'
@@ -163,6 +164,14 @@ Options:
escape sequences unless the syntax highlighting language is plain text. Possible values:
auto, always, *never*.
--sanitize <when>
Specify when to sanitize input bytes for safe terminal display. Implies --strip-ansi to
the same value, and additionally substitutes terminal-active control bytes (cursor moves,
charset switches, beep, etc.) and Unicode bidi / zero-width formatting characters with the
Unicode replacement character (U+FFFD). Tab, LF, FF, and CRLF pass through. Useful for
displaying untrusted file content (e.g. file-manager preview panes). Possible values:
auto, always, *never*.
--style <components>
Configure which elements (line numbers, file headers, grid borders, Git modifications, ..)
to display in addition to the file contents. The argument is a comma-separated list of
doc/sponsors.md
-2
View File
@@ -10,5 +10,3 @@ No issue will have a different priority based on sponsorship status of the
reporter.
Contributions from anybody are most welcomed, please see our [`CONTRIBUTING.md`](../CONTRIBUTING.md) guide.
If you want to see our biggest sponsors, check the top of [`README.md`](../README.md#sponsors).
doc/sponsors/graphite-logo.jpeg
BIN
View File
Binary file not shown.
Side by Side

Before

Width:  |  Height:  |  Size: 28 KiB

doc/sponsors/warp-logo.png
BIN
View File
Binary file not shown.
Side by Side

Before

Width:  |  Height:  |  Size: 130 KiB

doc/sponsors/warp-pack-header.png
BIN
View File
Binary file not shown.
Side by Side

Before

Width:  |  Height:  |  Size: 37 KiB

src/assets.rs
+61 -13
View File
@@ -158,7 +158,9 @@ impl HighlightingAssets {
let syntax_match = mapping.get_syntax_for(path);
if let Some(MappingTarget::MapToUnknown) = syntax_match {
return Err(Error::UndetectedSyntax(path.to_string_lossy().into()));
return Err(Error::UndetectedSyntax(
crate::preprocessor::sanitize_for_terminal(&path.to_string_lossy()),
));
}
if let Some(MappingTarget::MapTo(syntax_name)) = syntax_match {
@@ -175,13 +177,17 @@ impl HighlightingAssets {
) {
(Some(syntax), _) => Ok(syntax),
(_, Some(MappingTarget::MapExtensionToUnknown)) => {
Err(Error::UndetectedSyntax(path.to_string_lossy().into()))
}
(_, Some(MappingTarget::MapExtensionToUnknown)) => Err(Error::UndetectedSyntax(
crate::preprocessor::sanitize_for_terminal(&path.to_string_lossy()),
)),
_ => self
.get_syntax_for_file_extension(file_name, &mapping.ignored_suffixes)?
.ok_or_else(|| Error::UndetectedSyntax(path.to_string_lossy().into())),
.ok_or_else(|| {
Error::UndetectedSyntax(crate::preprocessor::sanitize_for_terminal(
&path.to_string_lossy(),
))
}),
}
}
@@ -341,15 +347,15 @@ impl HighlightingAssets {
file_name: &OsStr,
ignored_suffixes: &IgnoredSuffixes,
) -> Result<Option<SyntaxReferenceInSet<'_>>> {
let mut syntax = self.find_syntax_by_extension(Path::new(file_name).extension())?;
if syntax.is_none() {
syntax =
ignored_suffixes.try_with_stripped_suffix(file_name, |stripped_file_name| {
// Note: recursion
self.get_syntax_for_file_extension(stripped_file_name, ignored_suffixes)
})?;
let stripped =
ignored_suffixes.try_with_stripped_suffix(file_name, |stripped_file_name| {
self.get_syntax_for_file_extension(stripped_file_name, ignored_suffixes)
.map(Some)
})?;
match stripped {
Some(syntax) => Ok(syntax),
None => self.find_syntax_by_extension(Path::new(file_name).extension()),
}
Ok(syntax)
}
fn get_first_line_syntax(
@@ -680,6 +686,48 @@ mod tests {
);
}
#[test]
fn syntax_detection_ignored_suffix_falls_back_to_first_line() {
let mut test = SyntaxDetectionTest::new();
// By default a `.txt` file uses Plain Text, even with a shebang: the
// `.txt` extension wins and first-line detection is not reached.
assert_eq!(
test.syntax_for_file_with_content("test.txt", "#!/usr/bin/env bash"),
"Plain Text"
);
// Once `.txt` is an ignored suffix it is stripped before detection. The
// stripped name (`test`) has no extension, so detection falls back to the
// first line -- even though `.txt` is itself a registered extension that
// would otherwise match as Plain Text. See #2745.
test.syntax_mapping.insert_ignored_suffix(".txt");
assert_eq!(
test.syntax_for_file_with_content("test.txt", "#!/usr/bin/env bash"),
"Bourne Again Shell (bash)"
);
assert_eq!(
test.syntax_for_file_with_content("test.txt", "<?php"),
"PHP"
);
// A `.txt` file without a recognizable first line keeps no syntax (the
// caller renders it as plain text); we must not resurrect the shadowed
// `.txt` -> Plain Text match here.
assert_eq!(
test.syntax_for_file_with_content("notes.txt", "just some prose"),
"!no syntax!"
);
// Stripping that exposes a real extension still works: `.dev` is ignored,
// and the remaining `.json` extension is detected as usual.
test.syntax_mapping.insert_ignored_suffix(".dev");
assert_eq!(
test.syntax_for_file_with_content("config.json.dev", ""),
"JSON"
);
}
#[test]
fn syntax_detection_is_case_insensitive() {
let mut test = SyntaxDetectionTest::new();
src/bin/bat/app.rs
+46 -11
View File
@@ -38,6 +38,15 @@ pub fn env_no_color() -> bool {
env::var_os("NO_COLOR").is_some_and(|x| !x.is_empty())
}
fn parse_strip_ansi_value(raw: Option<&str>, flag_name: &str) -> StripAnsiMode {
match raw {
Some("never") | None => StripAnsiMode::Never,
Some("always") => StripAnsiMode::Always,
Some("auto") => StripAnsiMode::Auto,
_ => unreachable!("other values for {flag_name} are not allowed"),
}
}
enum HelpType {
Short,
Long,
@@ -458,16 +467,32 @@ impl App {
4
},
),
strip_ansi: match self
.matches
.get_one::<String>("strip-ansi")
.map(|s| s.as_str())
{
Some("never") => StripAnsiMode::Never,
Some("always") => StripAnsiMode::Always,
Some("auto") => StripAnsiMode::Auto,
_ => unreachable!("other values for --strip-ansi are not allowed"),
strip_ansi: {
let sanitize = parse_strip_ansi_value(
self.matches
.get_one::<String>("sanitize")
.map(|s| s.as_str()),
"--sanitize",
);
let strip_ansi = parse_strip_ansi_value(
self.matches
.get_one::<String>("strip-ansi")
.map(|s| s.as_str()),
"--strip-ansi",
);
// --sanitize implies --strip-ansi to the same value.
if sanitize != StripAnsiMode::Never {
sanitize
} else {
strip_ansi
}
},
sanitize: parse_strip_ansi_value(
self.matches
.get_one::<String>("sanitize")
.map(|s| s.as_str()),
"--sanitize",
),
quiet_empty: self.matches.get_flag("quiet-empty"),
unbuffered: self.matches.get_flag("unbuffered"),
theme: theme(self.theme_options()).to_string(),
@@ -590,7 +615,17 @@ impl App {
// Plain if `--plain` is specified at least once.
if self.matches.get_count("plain") > 0 {
return Some(StyleComponents(HashSet::from([StyleComponent::Plain])));
let mut components = HashSet::from([StyleComponent::Plain]);
// When --diff is active, preserve change markers and snip separators
// so that diff output remains visually useful.
if self.matches.try_contains_id("diff").unwrap_or_default()
&& self.matches.get_flag("diff")
{
#[cfg(feature = "git")]
components.insert(StyleComponent::Changes);
components.insert(StyleComponent::Snip);
}
return Some(StyleComponents(components));
}
// Default behavior.
@@ -635,7 +670,7 @@ impl App {
Ok(styled_components)
}
fn theme_options(&self) -> ThemeOptions {
pub(crate) fn theme_options(&self) -> ThemeOptions {
Self::theme_options_from_matches(&self.matches)
}
src/bin/bat/clap_app.rs
+24 -2
View File
@@ -240,6 +240,7 @@ pub fn build_app(interactive_output: bool) -> Command {
.arg(
Arg::new("terminal-width")
.long("terminal-width")
.overrides_with("terminal-width")
.value_name("width")
.hide_short_help(true)
.allow_hyphen_values(true)
@@ -255,10 +256,13 @@ pub fn build_app(interactive_output: bool) -> Command {
})
.map_err(|e| e.to_string())
})
.help(
.help("Explicitly set the width of the terminal instead of determining it automatically.")
.long_help(
"Explicitly set the width of the terminal instead of determining it \
automatically. If prefixed with '+' or '-', the value will be treated \
as an offset to the actual terminal width. See also: '--wrap'.",
as an offset to the actual terminal width. This can also be configured \
via the BAT_WIDTH environment variable (e.g. export BAT_WIDTH=\"100\"). \
See also: '--wrap'.",
),
)
.arg(
@@ -476,6 +480,24 @@ pub fn build_app(interactive_output: bool) -> Command {
language is plain text. Possible values: auto, always, *never*.")
.hide_short_help(true)
)
.arg(
Arg::new("sanitize")
.long("sanitize")
.overrides_with("sanitize")
.value_name("when")
.value_parser(["auto", "always", "never"])
.default_value("never")
.hide_default_value(true)
.help("Sanitize untrusted input for safe display (auto, always, *never*)")
.long_help("Specify when to sanitize input bytes for safe terminal display. \
Implies --strip-ansi to the same value, and additionally substitutes \
terminal-active control bytes (cursor moves, charset switches, beep, etc.) \
and Unicode bidi / zero-width formatting characters with the Unicode \
replacement character (U+FFFD). Tab, LF, FF, and CRLF pass through. Useful \
for displaying untrusted file content (e.g. file-manager preview panes). \
Possible values: auto, always, *never*.")
.hide_short_help(true)
)
.arg(
Arg::new("style")
.long("style")
src/bin/bat/config.rs
+1
View File
@@ -153,6 +153,7 @@ fn get_args_from_str(content: &str) -> Result<Vec<OsString>, shell_words::ParseE
pub fn get_args_from_env_vars() -> Vec<OsString> {
[
("--tabs", "BAT_TABS"),
("--terminal-width", "BAT_WIDTH"),
("--theme", bat::theme::env::BAT_THEME),
("--theme-dark", bat::theme::env::BAT_THEME_DARK),
("--theme-light", bat::theme::env::BAT_THEME_LIGHT),
src/bin/bat/main.rs
+44 -13
View File
@@ -17,7 +17,6 @@ use std::path::Path;
use std::process;
use bat::output::{OutputHandle, OutputType};
use bat::theme::DetectColorScheme;
use nu_ansi_term::Color::Green;
use nu_ansi_term::Style;
@@ -39,7 +38,7 @@ use bat::{
error::*,
input::Input,
style::{StyleComponent, StyleComponents},
theme::{color_scheme, default_theme, ColorScheme},
theme::{default_theme, theme, ColorScheme, ThemeOptions},
MappingTarget, PagingMode,
};
@@ -153,7 +152,14 @@ pub fn get_languages(config: &Config, cache_dir: &Path) -> Result<String> {
let comma_separator = ", ";
let separator = " ";
// Line-wrapping for the possible file extension overflow.
let desired_width = config.term_width - longest - separator.len();
// Clamp instead of subtracting: a tiny `--terminal-width` (smaller than the
// longest language name) would otherwise underflow `usize` and wrap to a huge
// value, silently disabling wrapping (and panicking in debug/overflow-checked
// builds). Mirrors the `saturating_sub` fix applied to `print_snip` in #3804.
let desired_width = config
.term_width
.saturating_sub(longest)
.saturating_sub(separator.len());
let style = if config.colored_output {
Green.normal()
@@ -197,7 +203,7 @@ pub fn list_themes(
cfg: &Config,
config_dir: &Path,
cache_dir: &Path,
detect_color_scheme: DetectColorScheme,
theme_options: ThemeOptions,
) -> Result<()> {
let assets = assets_from_cache_or_binary(cfg.use_custom_assets, cache_dir)?;
let mut config = cfg.clone();
@@ -206,7 +212,7 @@ pub fn list_themes(
config.language = Some("Rust");
config.style_components = StyleComponents(style);
let default_theme_name = default_theme(color_scheme(detect_color_scheme).unwrap_or_default());
let default_theme_name = theme(theme_options).to_string();
let mut buf = String::new();
let mut handle = OutputHandle::FmtWrite(&mut buf);
@@ -259,7 +265,9 @@ pub fn list_themes(
fn set_terminal_title_to(new_terminal_title: String) {
let osc_command_for_setting_terminal_title = "\x1b]0;";
let osc_end_command = "\x07";
print!("{osc_command_for_setting_terminal_title}{new_terminal_title}{osc_end_command}");
// Prevent BEL/ESC/C1 bytes in the title from terminating or nesting the OSC.
let safe_title = bat::sanitize_for_terminal(&new_terminal_title);
print!("{osc_command_for_setting_terminal_title}{safe_title}{osc_end_command}");
io::stdout().flush().unwrap();
}
@@ -285,7 +293,28 @@ fn run_controller(inputs: Vec<Input>, config: &Config, cache_dir: &Path) -> Resu
#[cfg(feature = "bugreport")]
fn invoke_bugreport(app: &App, cache_dir: &Path) {
use bugreport::{bugreport, collector::*, format::Markdown};
use bugreport::{bugreport, collector::*, format::Markdown, report::ReportEntry};
struct ColorSchemeCollector;
impl Collector for ColorSchemeCollector {
fn description(&self) -> &str {
"Detected terminal color scheme"
}
fn collect(
&mut self,
_: &bugreport::CrateInfo,
) -> std::result::Result<ReportEntry, CollectionError> {
let color_scheme = bat::theme::color_scheme(bat::theme::DetectColorScheme::Always);
let text = match color_scheme {
Some(bat::theme::ColorScheme::Dark) => "dark",
Some(bat::theme::ColorScheme::Light) => "light",
None => "not detected",
};
Ok(ReportEntry::Text(text.to_string()))
}
}
let pager = bat::config::get_pager_executable(
app.matches.get_one::<String>("pager").map(|s| s.as_str()),
)
@@ -307,6 +336,9 @@ fn invoke_bugreport(app: &App, cache_dir: &Path) {
"BAT_STYLE",
"BAT_TABS",
"BAT_THEME",
"BAT_WIDTH",
bat::theme::env::BAT_THEME_DARK,
bat::theme::env::BAT_THEME_LIGHT,
"COLORTERM",
"LANG",
"LC_ALL",
@@ -326,6 +358,7 @@ fn invoke_bugreport(app: &App, cache_dir: &Path) {
custom_assets_metadata,
))
.info(DirectoryEntries::new("Custom assets", cache_dir))
.info(ColorSchemeCollector)
.info(CompileTimeInformation::default());
#[cfg(feature = "paging")]
@@ -394,14 +427,12 @@ fn run() -> Result<bool> {
if app.matches.get_flag("list-languages") {
let languages: String = get_languages(&config, cache_dir)?;
let inputs: Vec<Input> = vec![Input::from_reader(Box::new(languages.as_bytes()))];
let plain_config = Config {
style_components: StyleComponents::new(StyleComponent::Plain.components(false)),
paging_mode: PagingMode::QuitIfOneScreen,
..Default::default()
};
let mut plain_config = config.clone();
plain_config.style_components =
StyleComponents::new(StyleComponent::Plain.components(false));
run_controller(inputs, &plain_config, cache_dir)
} else if app.matches.get_flag("list-themes") {
list_themes(&config, config_dir, cache_dir, DetectColorScheme::default())?;
list_themes(&config, config_dir, cache_dir, app.theme_options())?;
Ok(true)
} else if app.matches.get_flag("config-file") {
println!("{}", config_file().to_string_lossy());
src/config.rs
+3
View File
@@ -111,6 +111,9 @@ pub struct Config<'a> {
// Whether or not to strip ANSI escape codes from the input
pub strip_ansi: StripAnsiMode,
// Substitute terminal-active and spoofing-relevant bytes; implies strip_ansi.
pub sanitize: StripAnsiMode,
/// Whether or not to produce no output when input is empty
pub quiet_empty: bool,
src/diff.rs
+176 -62
View File
@@ -1,12 +1,13 @@
#![cfg(feature = "git")]
use gix::diff::blob::pipeline::{Mode, WorktreeRoots};
use gix::diff::blob::{Algorithm, HunkIter, ResourceKind};
use gix::object::tree::EntryKind;
use path_abs::PathInfo;
use std::collections::HashMap;
use std::fs;
use std::path::Path;
use git2::{DiffOptions, IntoCString, Repository};
#[derive(Copy, Clone, Debug)]
#[derive(Copy, Clone, Debug, PartialEq, Eq)]
pub enum LineChange {
Added,
RemovedAbove,
@@ -16,68 +17,181 @@ pub enum LineChange {
pub type LineChanges = HashMap<u32, LineChange>;
pub fn get_git_diff(filename: &Path) -> Option<LineChanges> {
let repo = Repository::discover(filename).ok()?;
let repo_path_absolute = fs::canonicalize(repo.workdir()?).ok()?;
let filepath_absolute = fs::canonicalize(filename).ok()?;
let filepath_relative_to_repo = filepath_absolute.strip_prefix(&repo_path_absolute).ok()?;
let mut diff_options = DiffOptions::new();
let pathspec = filepath_relative_to_repo.into_c_string().ok()?;
diff_options.pathspec(pathspec);
diff_options.context_lines(0);
let diff = repo
.diff_index_to_workdir(None, Some(&mut diff_options))
.ok()?;
let mut line_changes: LineChanges = HashMap::new();
let mark_section =
|line_changes: &mut LineChanges, start: u32, end: i32, change: LineChange| {
for line in start..=end as u32 {
line_changes.insert(line, change);
fn collect_changes_from_hunks(hunks: HunkIter) -> Option<LineChanges> {
let mut changes: LineChanges = HashMap::new();
for hunk in hunks {
if hunk.before.is_empty() && !hunk.after.is_empty() {
for line in hunk.after {
changes.insert(line + 1, LineChange::Added);
}
};
let _ = diff.foreach(
&mut |_, _| true,
None,
Some(&mut |delta, hunk| {
let path = delta.new_file().path().unwrap_or_else(|| Path::new(""));
if filepath_relative_to_repo != path {
return false;
}
let old_lines = hunk.old_lines();
let new_start = hunk.new_start();
let new_lines = hunk.new_lines();
let new_end = (new_start + new_lines) as i32 - 1;
if old_lines == 0 && new_lines > 0 {
mark_section(&mut line_changes, new_start, new_end, LineChange::Added);
} else if new_lines == 0 && old_lines > 0 {
if new_start == 0 {
mark_section(&mut line_changes, 1, 1, LineChange::RemovedAbove);
} else {
mark_section(
&mut line_changes,
new_start,
new_start as i32,
LineChange::RemovedBelow,
);
}
} else if hunk.after.is_empty() && !hunk.before.is_empty() {
if hunk.after.start == 0 {
changes.insert(1, LineChange::RemovedAbove);
} else {
mark_section(&mut line_changes, new_start, new_end, LineChange::Modified);
changes.insert(hunk.after.start, LineChange::RemovedBelow);
}
} else {
for line in hunk.after {
changes.insert(line + 1, LineChange::Modified);
}
}
}
true
}),
None,
Some(changes)
}
pub fn get_git_diff(filename: &Path) -> Option<LineChanges> {
let filepath_absolute = filename.canonicalize().ok()?;
let repository = gix::discover(filepath_absolute.parent().ok()?).ok()?;
let repo_path_absolute = repository.workdir()?.canonicalize().ok()?;
let filepath_relative_to_repo = gix::path::to_unix_separators_on_windows(gix::path::into_bstr(
filepath_absolute.strip_prefix(&repo_path_absolute).ok()?,
));
let index = repository.index_or_load_from_head_or_empty().ok()?;
let index_entry = index.entry_by_path(filepath_relative_to_repo.as_ref())?;
let mut cache = repository
.diff_resource_cache(
Mode::ToGit,
WorktreeRoots {
old_root: None,
new_root: repository.workdir().map(Path::to_path_buf),
},
)
.ok()?;
cache
.set_resource(
index_entry.id,
index_entry.mode.to_tree_entry_mode()?.kind(),
filepath_relative_to_repo.as_ref(),
ResourceKind::OldOrSource,
&repository,
)
.ok()?;
cache
.set_resource(
repository.object_hash().null(),
EntryKind::Blob,
filepath_relative_to_repo.as_ref(),
ResourceKind::NewOrDestination,
&repository,
)
.ok()?;
let diff = gix::diff::blob::diff_with_slider_heuristics(
Algorithm::Histogram,
&cache.prepare_diff().ok()?.interned_input(),
);
Some(line_changes)
collect_changes_from_hunks(diff.hunks())
}
#[cfg(test)]
mod tests {
use super::{get_git_diff, LineChange};
use std::path::{Path, PathBuf};
use std::process::Command;
fn git(repo: &Path, args: &[&str]) {
let output = Command::new("git")
.args(args)
.current_dir(repo)
.output()
.expect("git command can run");
assert!(
output.status.success(),
"git {args:?} failed\nstdout: {}\nstderr: {}",
String::from_utf8_lossy(&output.stdout),
String::from_utf8_lossy(&output.stderr)
);
}
fn setup_repo() -> tempfile::TempDir {
let dir = tempfile::tempdir().expect("can create temporary directory");
let repo = dir.path();
git(repo, &["init"]);
git(repo, &["config", "user.email", "test@test.test"]);
git(repo, &["config", "user.name", "Test"]);
dir
}
fn create_and_track_file(repo: &tempfile::TempDir, filename: &str) -> PathBuf {
let filepath = repo.path().join(filename);
std::fs::write(&filepath, "file\n").expect("can write file");
git(repo.path(), &["add", filename]);
git(repo.path(), &["commit", "-m", "initial"]);
filepath.to_owned()
}
#[test]
fn not_a_git_repository() {
let dir = tempfile::tempdir().expect("can create temporary directory");
let file = dir.path().join("file.txt");
std::fs::write(&file, "line 1\n").expect("can write file");
assert_eq!(get_git_diff(&file), None);
}
#[test]
fn diff_is_calculated_against_index() {
let repo = setup_repo();
let file = create_and_track_file(&repo, "file.txt");
std::fs::write(&file, "line 1\nline 2 modified\n").expect("can write file");
// Add modified file to index -> should find 0 changes but not none
git(repo.path(), &["add", "file.txt"]);
assert_eq!(get_git_diff(&file).expect("empty map").len(), 0);
}
#[test]
fn diff_is_calculated_against_index_2() {
let repo = setup_repo();
let file = create_and_track_file(&repo, "file.txt");
std::fs::write(&file, "line 1\nline 2 modified\n").expect("can write file");
git(repo.path(), &["add", "file.txt"]);
// modify the second line again which should show a single change
std::fs::write(&file, "line 1\nline 2 modified again\n").expect("can write file");
assert_eq!(get_git_diff(&file).expect("one change").len(), 1);
}
#[test]
fn diff_is_calculated_correctly() {
let repo = setup_repo();
create_and_track_file(&repo, "committed.txt");
let filename = "committed2.txt";
let file = create_and_track_file(&repo, filename);
std::fs::write(&file, "file\nline 2 added\n").expect("can write file");
let mut changes = get_git_diff(&file).expect("multiple changes");
assert_eq!(changes.get(&2), Some(&LineChange::Added));
git(repo.path(), &["add", filename]);
std::fs::write(&file, "file\nline 2 modified\n").expect("can write file");
changes = get_git_diff(&file).expect("multiple changes");
assert_eq!(changes.get(&2), Some(&LineChange::Modified));
git(repo.path(), &["add", filename]);
std::fs::write(&file, "line 2 modified\nline 3 added").expect("can write file");
changes = get_git_diff(&file).expect("multiple changes");
assert_eq!(changes.get(&1), Some(&LineChange::RemovedAbove));
assert_eq!(changes.get(&2), Some(&LineChange::Added));
}
#[test]
fn faulty_git_version_does_not_panic() {
let repo = setup_repo();
let file = create_and_track_file(&repo, "file.txt");
std::fs::write(&file, "line 1\nline 2 modified\n").expect("can write file");
// changes are detected
assert_eq!(get_git_diff(&file).expect("one change").len(), 2);
// write invalid repositoryformatversion
git(
repo.path(),
&["config", "core.repositoryformatversion", "one"],
);
// changes are no longer detected
assert_eq!(get_git_diff(&file), None);
}
}
src/input.rs
+80 -23
View File
@@ -207,7 +207,7 @@ impl<'a> Input<'a> {
kind: OpenedInputKind::StdIn,
description,
metadata: self.metadata,
reader: InputReader::new(stdin),
reader: InputReader::try_new(stdin)?,
})
}
@@ -216,34 +216,34 @@ impl<'a> Input<'a> {
description,
metadata: self.metadata,
reader: {
let mut file = File::open(&path)
.map_err(|e| format!("'{}': {e}", path.to_string_lossy()))?;
let path_display =
crate::preprocessor::sanitize_for_terminal(&path.to_string_lossy());
let mut file =
File::open(&path).map_err(|e| format!("'{path_display}': {e}"))?;
if file.metadata()?.is_dir() {
return Err(format!("'{}' is a directory.", path.to_string_lossy()).into());
return Err(format!("'{path_display}' is a directory.").into());
}
if let Some(stdout) = stdout_identifier {
let input_identifier = Identifier::try_from(file).map_err(|e| {
format!("{}: Error identifying file: {e}", path.to_string_lossy())
})?;
let input_identifier = Identifier::try_from(file)
.map_err(|e| format!("{path_display}: Error identifying file: {e}"))?;
if stdout.surely_conflicts_with(&input_identifier) {
return Err(format!(
"IO circle detected. The input from '{}' is also an output. Aborting to avoid infinite loop.",
path.to_string_lossy()
"IO circle detected. The input from '{path_display}' is also an output. Aborting to avoid infinite loop.",
)
.into());
}
file = input_identifier.into_inner().expect("The file was lost in the clircle::Identifier, this should not have happened...");
}
InputReader::new(BufReader::new(file))
InputReader::try_new(BufReader::new(file))?
},
}),
InputKind::CustomReader(reader) => Ok(OpenedInput {
description,
kind: OpenedInputKind::CustomReader,
metadata: self.metadata,
reader: InputReader::new(BufReader::new(reader)),
reader: InputReader::try_new(BufReader::new(reader))?,
}),
}
}
@@ -257,28 +257,29 @@ pub(crate) struct InputReader<'a> {
}
impl<'a> InputReader<'a> {
pub(crate) fn new<R: BufRead + 'a>(mut reader: R) -> InputReader<'a> {
let mut first_line = vec![];
reader.read_until(b'\n', &mut first_line).ok();
#[cfg(test)]
pub(crate) fn new<R: BufRead + 'a>(reader: R) -> InputReader<'a> {
Self::try_new(reader).expect("reading the first line failed")
}
let content_type = if first_line.is_empty() {
None
} else {
Some(content_inspector::inspect(&first_line[..]))
};
pub(crate) fn try_new<R: BufRead + 'a>(mut reader: R) -> io::Result<InputReader<'a>> {
let mut first_line = vec![];
reader.read_until(b'\n', &mut first_line)?;
let content_type = inspect_content_type(&first_line);
if content_type == Some(ContentType::UTF_16LE) {
read_utf16_line(&mut reader, &mut first_line, 0x00, 0x0A).ok();
read_utf16_line(&mut reader, &mut first_line, 0x00, 0x0A)?;
} else if content_type == Some(ContentType::UTF_16BE) {
read_utf16_line(&mut reader, &mut first_line, 0x0A, 0x00).ok();
read_utf16_line(&mut reader, &mut first_line, 0x0A, 0x00)?;
}
InputReader {
Ok(InputReader {
inner: Box::new(reader),
first_line,
content_type,
unbuffered: false,
}
})
}
pub(crate) fn read_line(&mut self, buf: &mut Vec<u8>) -> io::Result<bool> {
@@ -319,6 +320,25 @@ impl<'a> InputReader<'a> {
}
}
fn inspect_content_type(first_line: &[u8]) -> Option<ContentType> {
if first_line.is_empty() {
return None;
}
let content_type = content_inspector::inspect(first_line);
if content_type == ContentType::UTF_8 && has_zip_signature(first_line) {
Some(ContentType::BINARY)
} else {
Some(content_type)
}
}
fn has_zip_signature(bytes: &[u8]) -> bool {
[b"PK\x03\x04", b"PK\x05\x06", b"PK\x07\x08"]
.into_iter()
.any(|signature| bytes.starts_with(signature))
}
fn read_utf16_line<R: BufRead>(
reader: &mut R,
buf: &mut Vec<u8>,
@@ -374,6 +394,43 @@ fn basic() {
assert!(buffer.is_empty());
}
#[test]
fn zip_magic_headers_are_treated_as_binary() {
for content in [b"PK\x03\x04hello", b"PK\x05\x06hello", b"PK\x07\x08hello"] {
let reader = InputReader::new(&content[..]);
assert_eq!(Some(ContentType::BINARY), reader.content_type);
}
}
#[test]
fn non_zip_pk_prefix_is_not_treated_as_binary() {
assert_eq!(
Some(ContentType::UTF_8),
inspect_content_type(b"PK\x03\x03hello")
);
}
#[test]
fn input_open_returns_initial_read_errors() {
struct FailingRead;
impl Read for FailingRead {
fn read(&mut self, _buf: &mut [u8]) -> io::Result<usize> {
Err(io::Error::other("initial read failed"))
}
}
let input = Input::from_reader(Box::new(FailingRead));
let result = input.open(io::empty(), None);
assert!(result.is_err());
assert!(result
.err()
.unwrap()
.to_string()
.contains("initial read failed"));
}
#[test]
fn utf16le() {
let content = b"\xFF\xFE\x73\x00\x0A\x00\x64\x00";
src/lessopen.rs
+115 -7
View File
@@ -14,6 +14,53 @@ use crate::{
input::{Input, InputKind, InputReader, OpenedInput, OpenedInputKind},
};
/// Wrap `s` in POSIX single quotes. An embedded `'` is encoded as `'\''`
/// (close, escaped quote, reopen) since `'...'` has no escape character.
fn shell_quote(s: &str) -> String {
if !s.contains('\'') {
let mut quoted = String::with_capacity(s.len() + 2);
quoted.push('\'');
quoted.push_str(s);
quoted.push('\'');
return quoted;
}
let mut quoted = String::with_capacity(s.len() + 2);
quoted.push('\'');
for c in s.chars() {
if c == '\'' {
quoted.push_str("'\\''");
} else {
quoted.push(c);
}
}
quoted.push('\'');
quoted
}
/// Substitute the first `%s` in a $LESSOPEN/$LESSCLOSE template with a quoted value.
fn shell_substitute(template: &str, replacement: &str) -> String {
template.replacen("%s", &shell_quote(replacement), 1)
}
/// Substitute the first two `%s` occurrences positionally in one pass; chaining
/// `shell_substitute` twice would mis-target if `first` itself contains `%s`.
fn shell_substitute_two(template: &str, first: &str, second: &str) -> String {
let mut out = String::with_capacity(template.len() + first.len() + second.len() + 4);
let mut remaining = template;
let mut to_substitute = [Some(first), Some(second)];
for slot in &mut to_substitute {
if let Some(idx) = remaining.find("%s") {
out.push_str(&remaining[..idx]);
out.push_str(&shell_quote(slot.take().unwrap()));
remaining = &remaining[idx + 2..];
} else {
break;
}
}
out.push_str(remaining);
out
}
/// Preprocess files and/or stdin using $LESSOPEN and $LESSCLOSE
pub(crate) struct LessOpenPreprocessor {
lessopen: String,
@@ -87,7 +134,7 @@ impl LessOpenPreprocessor {
None => return input.open(stdin, stdout_identifier),
};
let mut lessopen_command = shell(self.lessopen.replacen("%s", path_str, 1));
let mut lessopen_command = shell(shell_substitute(&self.lessopen, path_str));
lessopen_command.stdout(Stdio::piped());
let lessopen_output = match lessopen_command.execute_output() {
@@ -122,7 +169,7 @@ impl LessOpenPreprocessor {
let mut stdin_buffer = Vec::new();
stdin.read_to_end(&mut stdin_buffer)?;
let mut lessopen_command = shell(self.lessopen.replacen("%s", "-", 1));
let mut lessopen_command = shell(shell_substitute(&self.lessopen, "-"));
lessopen_command.stdout(Stdio::piped());
let lessopen_output = match lessopen_command.execute_input_output(&stdin_buffer)
@@ -155,7 +202,7 @@ impl LessOpenPreprocessor {
Ok(OpenedInput {
kind,
reader: InputReader::new(BufReader::new(
reader: InputReader::try_new(BufReader::new(
if matches!(self.kind, LessOpenKind::TempFile) {
let lessopen_string = match String::from_utf8(lessopen_stdout) {
Ok(string) => string,
@@ -181,7 +228,7 @@ impl LessOpenPreprocessor {
lessclose: self
.lessclose
.as_ref()
.map(|s| s.replacen("%s", &path_str, 1).replacen("%s", &stdout, 1)),
.map(|s| shell_substitute_two(s, &path_str, &stdout)),
}
} else {
Preprocessed {
@@ -189,10 +236,10 @@ impl LessOpenPreprocessor {
lessclose: self
.lessclose
.as_ref()
.map(|s| s.replacen("%s", &path_str, 1).replacen("%s", "-", 1)),
.map(|s| shell_substitute_two(s, &path_str, "-")),
}
},
)),
))?,
metadata: input.metadata,
description: input.description,
})
@@ -261,7 +308,7 @@ impl Drop for Preprocessed {
}
};
if lessclose_output.status.success() {
if !lessclose_output.status.success() {
bat_warning!("$LESSCLOSE exited with nonzero exit code",)
};
}
@@ -384,4 +431,65 @@ mod tests {
Ok(())
}
#[test]
fn shell_quote_plain_filename() {
assert_eq!(super::shell_quote("file.txt"), "'file.txt'");
assert_eq!(super::shell_quote("with space.txt"), "'with space.txt'");
assert_eq!(super::shell_quote(""), "''");
assert_eq!(super::shell_quote("-"), "'-'");
}
#[test]
fn shell_quote_metacharacters_neutralised() {
assert_eq!(super::shell_quote("; rm -rf ~ ;"), "'; rm -rf ~ ;'");
assert_eq!(super::shell_quote("$(payload)"), "'$(payload)'");
assert_eq!(super::shell_quote("`payload`"), "'`payload`'");
assert_eq!(super::shell_quote("a|b&c;d"), "'a|b&c;d'");
assert_eq!(super::shell_quote("..>/dev/null"), "'..>/dev/null'");
assert_eq!(super::shell_quote("\n\t"), "'\n\t'");
}
#[test]
fn shell_quote_embedded_single_quote() {
assert_eq!(super::shell_quote("it's"), "'it'\\''s'");
assert_eq!(super::shell_quote("'"), "''\\'''");
assert_eq!(super::shell_quote("a'b'c"), "'a'\\''b'\\''c'");
}
#[test]
fn shell_substitute_only_replaces_first_percent_s() {
assert_eq!(super::shell_substitute("echo %s", "x"), "echo 'x'");
assert_eq!(super::shell_substitute("echo %s %s", "x"), "echo 'x' %s");
}
#[test]
fn shell_substitute_protects_against_filename_injection() {
let template = "|preproc %s";
let attacker_filename = "; rm -rf ~ ;";
let result = super::shell_substitute(template, attacker_filename);
assert_eq!(result, "|preproc '; rm -rf ~ ;'");
}
#[test]
fn shell_substitute_two_replaces_first_two_placeholders() {
assert_eq!(
super::shell_substitute_two("echo %s and %s done", "a", "b"),
"echo 'a' and 'b' done"
);
// A third %s is left alone.
assert_eq!(
super::shell_substitute_two("echo %s %s %s", "a", "b"),
"echo 'a' 'b' %s"
);
}
#[test]
fn shell_substitute_two_handles_percent_s_in_first_argument() {
// The chained-substitute approach would replace the injected `%s` here.
assert_eq!(
super::shell_substitute_two("a %s b %s c", "%s", "second"),
"a '%s' b 'second' c"
);
}
}
src/lib.rs
+1
View File
@@ -54,6 +54,7 @@ mod vscreen;
pub(crate) mod wrapping;
pub use nonprintable_notation::{BinaryBehavior, NonprintableNotation};
pub use preprocessor::sanitize_for_terminal;
pub use preprocessor::StripAnsiMode;
pub use pretty_printer::{Input, PrettyPrinter, Syntax};
pub use syntax_mapping::{MappingTarget, SyntaxMapping};
src/output.rs
+12
View File
@@ -23,6 +23,18 @@ pub struct BuiltinPager {
impl BuiltinPager {
fn new() -> Self {
let pager = minus::Pager::new();
let mut input_register = minus::input::HashedEventRegister::default();
input_register.add_key_events(&["home"], |_, _| {
minus::input::InputEvent::UpdateUpperMark(0)
});
input_register.add_key_events(&["end"], |_, _| {
minus::input::InputEvent::UpdateUpperMark(usize::MAX)
});
pager
.set_input_classifier(Box::new(input_register))
.expect("failed to set input classifier on newly created pager");
let handle = {
let pager = pager.clone();
Some(spawn(move || {
src/preprocessor.rs
+270 -2
View File
@@ -139,16 +139,140 @@ pub fn replace_nonprintable(
/// Strips ANSI escape sequences from the input.
pub fn strip_ansi(line: &str) -> String {
let mut buffer = String::with_capacity(line.len());
for seq in EscapeSequenceOffsetsIterator::new(line) {
if let EscapeSequenceOffsets::Text { .. } = seq {
buffer.push_str(&line[seq.index_of_start()..seq.index_past_end()]);
}
}
buffer
}
/// Strips ANSI escape sequences and substitutes terminal-active control bytes
/// and visual-spoofing Unicode codepoints (bidi, zero-width) with U+FFFD.
pub fn sanitize(line: &str) -> String {
let stripped = strip_ansi(line);
let mut buffer = String::with_capacity(stripped.len());
let bytes = stripped.as_bytes();
let mut start = 0;
let mut i = 0;
// Skip directly to the next trigger byte instead of testing each one.
while let Some(off) = bytes[i..].iter().position(|&b| is_sanitize_trigger(b)) {
i += off;
let len = sanitize_at(bytes, i, &stripped, &mut buffer, &mut start);
i += len;
}
buffer.push_str(&stripped[start..]);
buffer
}
#[inline]
fn is_sanitize_trigger(b: u8) -> bool {
// C0 controls minus \t \n \f; DEL; UTF-8 leads with dangerous codepoints.
matches!(b, 0x00..=0x08 | 0x0B | 0x0D..=0x1F | 0x7F | 0xC2 | 0xD8 | 0xE2 | 0xEF)
}
/// Substitutes the byte/sequence at `bytes[i]` (or passes it through on
/// false-alarm trigger), flushing the prefix from `start`. Returns bytes consumed.
fn sanitize_at(
bytes: &[u8],
i: usize,
full: &str,
buffer: &mut String,
start: &mut usize,
) -> usize {
buffer.push_str(&full[*start..i]);
let consumed = match bytes[i] {
b'\r' if bytes.get(i + 1) == Some(&b'\n') => {
buffer.push_str("\r\n");
2
}
// 0xC2 leads U+0080..U+00FF; filter the C1 range.
0xC2 if matches!(bytes.get(i + 1), Some(0x80..=0x9F)) => {
buffer.push('\u{FFFD}');
2
}
0xE2 if is_dangerous_e2(bytes, i) => {
buffer.push('\u{FFFD}');
3
}
// 0xEF 0xBB 0xBF = U+FEFF (BOM / zero-width no-break space).
0xEF if bytes.get(i + 1) == Some(&0xBB) && bytes.get(i + 2) == Some(&0xBF) => {
buffer.push('\u{FFFD}');
3
}
// 0xD8 0x9C = U+061C (Arabic letter mark, a bidi control). The rest of
// the 0xD8 block is ordinary Arabic text.
0xD8 if bytes.get(i + 1) == Some(&0x9C) => {
buffer.push('\u{FFFD}');
2
}
// False-alarm trigger: pass the full UTF-8 sequence through.
lead @ (0xC2 | 0xD8 | 0xE2 | 0xEF) => {
let n = utf8_len_from_lead(lead);
buffer.push_str(&full[i..i + n]);
n
}
_ => {
buffer.push('\u{FFFD}');
1
}
};
*start = i + consumed;
consumed
}
#[inline]
fn is_dangerous_e2(bytes: &[u8], i: usize) -> bool {
// U+200B..D (zero-width), U+200E..F (LRM/RLM), U+202A..E (bidi embedding
// and override), U+2066..9 (bidi isolates).
matches!(
(bytes.get(i + 1), bytes.get(i + 2)),
(Some(0x80), Some(0x8B..=0x8F | 0xAA..=0xAE)) | (Some(0x81), Some(0xA6..=0xA9))
)
}
#[inline]
fn utf8_len_from_lead(lead: u8) -> usize {
if lead < 0x80 {
1
} else if lead < 0xE0 {
2
} else if lead < 0xF0 {
3
} else {
4
}
}
/// Escape C0, DEL, and C1 control characters so a string from an untrusted
/// filename or path can be safely written to the terminal.
pub fn sanitize_for_terminal(input: &str) -> String {
if !input
.chars()
.any(|c| matches!(c, '\x00'..='\x08' | '\x0A'..='\x1F' | '\x7F'..='\u{9F}'))
{
return input.to_owned();
}
let mut out = String::with_capacity(input.len() + 8);
for c in input.chars() {
match c {
'\t' => out.push('\t'),
'\x00'..='\x1F' => {
out.push('^');
out.push(char::from_u32(0x40 + c as u32).unwrap_or('?'));
}
'\x7F' => out.push_str("^?"),
'\u{80}'..='\u{9F}' => {
use std::fmt::Write as _;
let _ = write!(out, "\\u{{{:x}}}", c as u32);
}
other => out.push(other),
}
}
out
}
/// Strips overstrike sequences (backspace formatting) from input.
///
/// Overstrike formatting is used by man pages and some help output:
@@ -241,6 +365,113 @@ fn test_strip_ansi() {
);
}
#[test]
fn test_strip_ansi_8bit_c1_introducers() {
assert_eq!(strip_ansi("a\u{9B}31mRED\u{9B}0mb"), "aREDb");
assert_eq!(strip_ansi("a\x1bP1;0|payload\x1b\\b"), "ab");
assert_eq!(strip_ansi("a\u{90}body\u{9C}b"), "ab");
}
#[test]
fn test_strip_ansi_single_char_esc() {
// strip_ansi must consume both bytes of single-byte ESC sequences (RIS, DECSC, keypad, VT52).
assert_eq!(strip_ansi("a\x1bcb"), "ab");
assert_eq!(strip_ansi("a\x1b7b\x1b8c"), "abc");
assert_eq!(strip_ansi("a\x1b=b\x1b>c"), "abc");
assert_eq!(strip_ansi("a\x1bZb"), "ab");
}
#[test]
fn test_strip_ansi_preserves_control_bytes() {
// strip_ansi removes only ANSI escape sequences; control bytes pass through.
assert_eq!(strip_ansi("safe\rEVIL"), "safe\rEVIL");
assert_eq!(strip_ansi("a\x08b\x07c\x0E\x0Fd"), "a\x08b\x07c\x0E\x0Fd");
}
#[test]
fn test_sanitize_substitutes_dangerous_bytes() {
let r = '\u{FFFD}';
assert_eq!(sanitize("safe\rEVIL"), format!("safe{r}EVIL"));
assert_eq!(sanitize("a\x08b"), format!("a{r}b"));
assert_eq!(sanitize("a\x07b"), format!("a{r}b"));
assert_eq!(sanitize("a\x0Bb"), format!("a{r}b"));
assert_eq!(sanitize("a\x0Eb\x0Fc"), format!("a{r}b{r}c"));
assert_eq!(sanitize("a\u{8D}b"), format!("a{r}b"));
assert_eq!(sanitize("a\u{85}b"), format!("a{r}b"));
assert_eq!(sanitize("trailing\r"), format!("trailing{r}"));
assert_eq!(sanitize("a\x7Fb"), format!("a{r}b"));
}
#[test]
fn test_sanitize_substitutes_bidi_and_zero_width() {
let r = '\u{FFFD}';
// Trojan-Source bidi formatting: U+202A..U+202E
assert_eq!(sanitize("a\u{202A}b"), format!("a{r}b"));
assert_eq!(sanitize("a\u{202E}b"), format!("a{r}b"));
// Bidi isolates: U+2066..U+2069
assert_eq!(sanitize("a\u{2066}b"), format!("a{r}b"));
assert_eq!(sanitize("a\u{2069}b"), format!("a{r}b"));
// Zero-width: U+200B..U+200D
assert_eq!(sanitize("a\u{200B}b"), format!("a{r}b"));
assert_eq!(sanitize("a\u{200D}b"), format!("a{r}b"));
// BOM in middle of file: U+FEFF
assert_eq!(sanitize("a\u{FEFF}b"), format!("a{r}b"));
}
#[test]
fn test_sanitize_substitutes_every_bidi_control() {
// Unicode Bidi_Control is exactly these 12 codepoints. Covering only some
// of them leaves a reordering attack available through the rest.
let r = '\u{FFFD}';
for c in [
'\u{061C}', '\u{200E}', '\u{200F}', '\u{202A}', '\u{202B}', '\u{202C}', '\u{202D}',
'\u{202E}', '\u{2066}', '\u{2067}', '\u{2068}', '\u{2069}',
] {
assert_eq!(
sanitize(&format!("a{c}b")),
format!("a{r}b"),
"U+{:04X} was not substituted",
c as u32
);
}
}
#[test]
fn test_sanitize_preserves_arabic_sharing_the_alm_lead_byte() {
// U+061C is reached via lead byte 0xD8, which also leads ordinary Arabic.
assert_eq!(
sanitize("\u{0600}\u{061F}\u{06FF}"),
"\u{0600}\u{061F}\u{06FF}"
);
assert_eq!(sanitize("مرحبا بالعالم"), "مرحبا بالعالم");
}
#[test]
fn test_sanitize_preserves_legitimate_bytes() {
assert_eq!(sanitize("crlf\r\nline\r\n"), "crlf\r\nline\r\n");
assert_eq!(sanitize("a\tb\nc"), "a\tb\nc");
assert_eq!(sanitize("plain ascii"), "plain ascii");
assert_eq!(sanitize("üñíçödé"), "üñíçödé");
// FF (U+000C) passes through; section separator in C source / Emacs Lisp.
assert_eq!(sanitize("section1\x0Csection2"), "section1\x0Csection2");
// Common Unicode that shares a UTF-8 lead byte with dangerous codepoints
// must pass through unchanged.
assert_eq!(sanitize("snowman ☃ moon ☾"), "snowman ☃ moon ☾");
assert_eq!(sanitize("emoji 🎉 ☃"), "emoji 🎉 ☃");
assert_eq!(sanitize("0xC2 lead: ÿ ñ ç"), "0xC2 lead: ÿ ñ ç");
assert_eq!(sanitize("CJK 漢字 emoji 🦀"), "CJK 漢字 emoji 🦀");
}
#[test]
fn test_sanitize_strips_ansi() {
let r = '\u{FFFD}';
// sanitize is a strict superset of strip_ansi.
assert_eq!(sanitize("a\x1B[31mb\x1B[0mc"), "abc");
assert_eq!(sanitize("a\u{9B}31mb\u{9B}0mc"), "abc");
// ANSI then dangerous byte: ANSI gone, byte substituted.
assert_eq!(sanitize("\x1B[31mhello\rEVIL"), format!("hello{r}EVIL"));
}
#[test]
fn test_strip_overstrike() {
// Bold: X\x08X (same char repeated)
@@ -261,3 +492,40 @@ fn test_strip_overstrike() {
// Unicode with overstrike
assert_eq!(strip_overstrike("ä\x08äöü", 2), "äöü");
}
#[test]
fn test_sanitize_for_terminal_passthrough() {
assert_eq!(sanitize_for_terminal(""), "");
assert_eq!(sanitize_for_terminal("hello.txt"), "hello.txt");
assert_eq!(sanitize_for_terminal("résumé.pdf"), "résumé.pdf");
assert_eq!(sanitize_for_terminal("日本語.md"), "日本語.md");
assert_eq!(
sanitize_for_terminal("path/with spaces/file.log"),
"path/with spaces/file.log"
);
assert_eq!(sanitize_for_terminal("a\tb"), "a\tb");
}
#[test]
fn test_sanitize_for_terminal_c0_controls() {
assert_eq!(
sanitize_for_terminal("\x1b[31mINJECTED\x1b[0m.txt"),
"^[[31mINJECTED^[[0m.txt"
);
assert_eq!(sanitize_for_terminal("bad\x07rest"), "bad^Grest");
assert_eq!(sanitize_for_terminal("\x00\x08\n\r\x7F"), "^@^H^J^M^?");
assert_eq!(sanitize_for_terminal("\u{9b}31m"), "\\u{9b}31m");
assert_eq!(
sanitize_for_terminal("\u{9d}0;pwned\x07"),
"\\u{9d}0;pwned^G"
);
}
#[test]
fn test_sanitize_for_terminal_idempotent_on_sanitized() {
let dirty = "\x1b]0;pwned\x07file.txt";
let clean = sanitize_for_terminal(dirty);
assert_eq!(sanitize_for_terminal(&clean), clean);
assert!(!clean.contains('\x1b'));
assert!(!clean.contains('\x07'));
}
src/pretty_printer.rs
+9
View File
@@ -192,6 +192,15 @@ impl<'a> PrettyPrinter<'a> {
self
}
/// Whether to sanitize untrusted input for safe display (default: never)
///
/// Strips ANSI escape sequences and additionally substitutes terminal-active
/// control bytes and bidi / zero-width codepoints with U+FFFD.
pub fn sanitize(&mut self, mode: StripAnsiMode) -> &mut Self {
self.config.sanitize = mode;
self
}
/// Text wrapping mode (default: do not wrap)
pub fn wrapping_mode(&mut self, mode: WrappingMode) -> &mut Self {
self.config.wrapping_mode = mode;
src/printer.rs
+54 -15
View File
@@ -29,7 +29,10 @@ use crate::error::*;
use crate::input::OpenedInput;
use crate::line_range::{MaxBufferedLineNumber, RangeCheckResult};
use crate::output::OutputHandle;
use crate::preprocessor::{expand_tabs, replace_nonprintable, strip_ansi, strip_overstrike};
use crate::preprocessor::{
expand_tabs, replace_nonprintable, sanitize, sanitize_for_terminal, strip_ansi,
strip_overstrike,
};
use crate::style::StyleComponent;
use crate::terminal::{as_terminal_escaped, to_ansi_color};
use crate::vscreen::{AnsiStyle, EscapeSequence, EscapeSequenceIterator};
@@ -37,6 +40,16 @@ use crate::wrapping::WrappingMode;
use crate::BinaryBehavior;
use crate::StripAnsiMode;
// Return the displayed width of a character.
//
// Control characters (0x00..=0x1F and 0x7F) are rendered by the terminal
// in caret notation (e.g. ^@, ^A, ..., ^?), which occupies two columns.
// UnicodeWidthChar::width() returns None for these, so we map them to 2
// here instead of the previous default of 0.
fn char_width(c: char) -> usize {
c.width().unwrap_or(if c.is_control() { 2 } else { 0 })
}
const ANSI_UNDERLINE_ENABLE: EscapeSequence = EscapeSequence::CSI {
raw_sequence: "\x1B[4m",
parameters: "4",
@@ -198,6 +211,7 @@ pub(crate) struct InteractivePrinter<'a> {
background_color_highlight: Option<Color>,
consecutive_empty_lines: usize,
strip_ansi: bool,
sanitize: bool,
strip_overstrike: bool,
}
@@ -261,7 +275,9 @@ impl<'a> InteractivePrinter<'a> {
let needs_to_match_syntax = (!is_printing_binary
|| matches!(config.binary, BinaryBehavior::AsText))
&& (config.colored_output || config.strip_ansi == StripAnsiMode::Auto);
&& (config.colored_output
|| config.strip_ansi == StripAnsiMode::Auto
|| config.sanitize == StripAnsiMode::Auto);
let (is_plain_text, strip_overstrike, highlighter_from_set) = if needs_to_match_syntax {
// Determine the type of syntax for highlighting
@@ -307,6 +323,14 @@ impl<'a> InteractivePrinter<'a> {
_ => false,
};
let sanitize = match config.sanitize {
_ if config.show_nonprintable => false,
StripAnsiMode::Always => true,
StripAnsiMode::Auto if is_plain_text => false,
StripAnsiMode::Auto => true,
_ => false,
};
Ok(InteractivePrinter {
panel_width,
colors,
@@ -320,6 +344,7 @@ impl<'a> InteractivePrinter<'a> {
background_color_highlight,
consecutive_empty_lines: 0,
strip_ansi,
sanitize,
strip_overstrike,
})
}
@@ -479,7 +504,7 @@ impl Printer for InteractivePrinter<'_> {
(but will be present if the output of 'bat' is piped). You can use 'bat -A' \
to show the binary file contents.",
Yellow.paint("[bat warning]"),
input.description.summary(),
sanitize_for_terminal(&input.description.summary()),
)?;
} else if self.config.style_components.grid() {
self.print_horizontal_line(handle, '┬')?;
@@ -533,9 +558,11 @@ impl Printer for InteractivePrinter<'_> {
"{}{}{mode}",
description
.kind()
.map(|kind| format!("{kind}: "))
.map(|kind| format!("{}: ", sanitize_for_terminal(kind)))
.unwrap_or_else(|| "".into()),
self.colors.header_value.paint(description.title()),
self.colors
.header_value
.paint(sanitize_for_terminal(description.title())),
);
self.print_header_multiline_component(handle, &header_filename)
}
@@ -589,11 +616,23 @@ impl Printer for InteractivePrinter<'_> {
let title = "8<";
let title_count = title.chars().count();
let snip_left = "─ ".repeat((self.config.term_width - panel_count - (title_count / 2)) / 4);
let snip_left = "─ ".repeat(
self.config
.term_width
.saturating_sub(panel_count)
.saturating_sub(title_count / 2)
/ 4,
);
let snip_left_count = snip_left.chars().count(); // Can't use .len() with Unicode.
let snip_right =
" ─".repeat((self.config.term_width - panel_count - snip_left_count - title_count) / 2);
let snip_right = " ─".repeat(
self.config
.term_width
.saturating_sub(panel_count)
.saturating_sub(snip_left_count)
.saturating_sub(title_count)
/ 2,
);
writeln!(
handle,
@@ -649,8 +688,10 @@ impl Printer for InteractivePrinter<'_> {
}
}
// If ANSI escape sequences are supposed to be stripped, do it before syntax highlighting.
if self.strip_ansi {
// Sanitize is the strict superset; otherwise strip-ansi alone.
if self.sanitize {
line = sanitize(&line).into()
} else if self.strip_ansi {
line = strip_ansi(&line).into()
}
@@ -793,7 +834,7 @@ impl Printer for InteractivePrinter<'_> {
for c in text.chars() {
// calculate the displayed width for next character
let cw = c.width().unwrap_or(0);
let cw = char_width(c);
current_width += cw;
// Track whitespace positions for word wrapping.
@@ -868,10 +909,8 @@ impl Printer for InteractivePrinter<'_> {
if let Some(rs) = rest_start {
// Word wrap: carry remainder to next line.
let remainder = line_buf[rs..].to_string();
let rem_width: usize = remainder
.chars()
.map(|ch| ch.width().unwrap_or(0))
.sum();
let rem_width: usize =
remainder.chars().map(char_width).sum();
line_buf.clear();
line_buf.push_str(&remainder);
current_width = rem_width + cw;
src/syntax_mapping.rs
+54 -3
View File
@@ -17,9 +17,16 @@ use ignored_suffixes::IgnoredSuffixes;
mod builtin;
pub mod ignored_suffixes;
fn make_glob_matcher(from: &str) -> Result<GlobMatcher> {
/// Whether a glob pattern should be matched case-sensitively or case-insensitively.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub(crate) enum Case {
Sensitive,
Insensitive,
}
fn make_glob_matcher(from: &str, case: Case) -> Result<GlobMatcher> {
let matcher = GlobBuilder::new(from)
.case_insensitive(true)
.case_insensitive(matches!(case, Case::Insensitive))
.literal_separator(true)
.build()?
.compile_matcher();
@@ -97,7 +104,14 @@ impl<'a> SyntaxMapping<'a> {
}
pub fn insert(&mut self, from: &str, to: MappingTarget<'a>) -> Result<()> {
let matcher = make_glob_matcher(from)?;
let matcher = make_glob_matcher(from, Case::Insensitive)?;
self.custom_mappings.push((matcher, to));
Ok(())
}
/// Like [`Self::insert`], but the glob pattern is matched case-sensitively.
pub fn insert_case_sensitive(&mut self, from: &str, to: MappingTarget<'a>) -> Result<()> {
let matcher = make_glob_matcher(from, Case::Sensitive)?;
self.custom_mappings.push((matcher, to));
Ok(())
}
@@ -261,4 +275,41 @@ mod tests {
Some(MappingTarget::MapTo("alpha"))
);
}
#[test]
fn case_sensitive_custom_mappings_work() {
let mut map = SyntaxMapping::new();
map.insert_case_sensitive("MY_SPECIAL_FILE", MappingTarget::MapTo("Python"))
.ok();
// Exact case matches
assert_eq!(
map.get_syntax_for("/path/to/MY_SPECIAL_FILE"),
Some(MappingTarget::MapTo("Python"))
);
// Different case should NOT match the case-sensitive rule
assert_eq!(map.get_syntax_for("/path/to/my_special_file"), None);
assert_eq!(map.get_syntax_for("/path/to/My_Special_File"), None);
}
#[test]
fn builtin_mappings_build_is_case_sensitive() {
let map = SyntaxMapping::new();
// "BUILD" (uppercase) should map to Python via case-sensitive builtin
assert_eq!(
map.get_syntax_for("/path/to/BUILD"),
Some(MappingTarget::MapTo("Python"))
);
// "build" (lowercase) should still map to MapToUnknown
assert_eq!(
map.get_syntax_for("/path/to/build"),
Some(MappingTarget::MapToUnknown)
);
// Mixed case should NOT match the Python rule
assert_eq!(
map.get_syntax_for("/path/to/Build"),
Some(MappingTarget::MapToUnknown)
);
}
}
src/syntax_mapping/builtin.rs
+5 -5
View File
@@ -3,7 +3,7 @@ use std::env;
use globset::GlobMatcher;
use once_cell::sync::Lazy;
use crate::syntax_mapping::{make_glob_matcher, MappingTarget};
use crate::syntax_mapping::{make_glob_matcher, Case, MappingTarget};
// Static syntax mappings generated from /src/syntax_mapping/builtins/ by the
// build script (/build/syntax_mapping.rs).
@@ -53,8 +53,8 @@ include!(concat!(
/// A failure to compile is a fatal error.
///
/// Used internally by `Lazy<Option<GlobMatcher>>`'s lazy evaluation closure.
fn build_matcher_fixed(from: &str) -> GlobMatcher {
make_glob_matcher(from).expect("A builtin fixed glob matcher failed to compile")
fn build_matcher_fixed(from: &str, case: Case) -> GlobMatcher {
make_glob_matcher(from, case).expect("A builtin fixed glob matcher failed to compile")
}
/// Join a list of matcher segments to create a glob string, replacing all
@@ -64,7 +64,7 @@ fn build_matcher_fixed(from: &str) -> GlobMatcher {
/// to compile.
///
/// Used internally by `Lazy<Option<GlobMatcher>>`'s lazy evaluation closure.
fn build_matcher_dynamic(segs: &[MatcherSegment]) -> Option<GlobMatcher> {
fn build_matcher_dynamic(segs: &[MatcherSegment], case: Case) -> Option<GlobMatcher> {
// join segments
let mut buf = String::new();
for seg in segs {
@@ -77,7 +77,7 @@ fn build_matcher_dynamic(segs: &[MatcherSegment]) -> Option<GlobMatcher> {
}
}
// compile glob matcher
let matcher = make_glob_matcher(&buf).ok()?;
let matcher = make_glob_matcher(&buf, case).ok()?;
Some(matcher)
}
src/syntax_mapping/builtins/README.md
+12 -5
View File
@@ -20,12 +20,10 @@ syntax mappings defined by all TOML files, and embed them into the binary.
## File syntax
Each TOML file should contain a single section named `mappings`, with each of
its keys being a language identifier (first column of `bat -L`; also referred to
as "target").
Each TOML file should contain a single section named `mappings`, with each of its keys being a language
identifier (first column of `bat -L`; also referred to as "target").
The value of each key should be an array of strings, with each item being a glob
matcher. We will call each of these items a "rule".
The value of each key should be an array of "rules". The rules are expected to be objects with a `glob` string and a `case_sensitive` boolean. For simplification, a rule can be just a glob string, which is shorthand for the default case insensitive mode.
For example, if `foo-application` uses both TOML and YAML configuration files,
we could write something like this:
@@ -98,6 +96,15 @@ like this:
]
```
### Case sensitivity
By default, all glob patterns are matched case-insensitively. To match a pattern case-sensitively, use the object form of the rule with the `case_sensitive` option:
```toml
[mappings]
"Python" = [{ glob = "BUILD", case_sensitive = true }]
```
## Ordering
At compile time, all TOML files applicable to the target are processed in
src/syntax_mapping/builtins/common/50-bazel.toml
+2
View File
@@ -0,0 +1,2 @@
[mappings]
"Python" = [{ glob = "BUILD", case_sensitive = true }]
src/syntax_mapping/builtins/common/50-dotnet-xml.toml
+1 -1
View File
@@ -1,2 +1,2 @@
[mappings]
"XML" = ["*.csproj", "*.vbproj", "*.props", "*.targets"]
"XML" = ["*.csproj", "*.vbproj", "*.props", "*.targets", "*.slnx"]
src/syntax_mapping/builtins/common/50-json.toml
+1 -1
View File
@@ -1,3 +1,3 @@
# JSON Lines is a simple variation of JSON #2535
[mappings]
"JSON" = ["*.jsonl", "*.jsonc", "*.jsonld", "*.geojson", "*.ndjson"]
"JSON" = ["*.jsonl", "*.jsonc", "*.jsonld", "*.geojson", "*.ndjson", "*.code-workspace"]
src/syntax_mapping/builtins/common/50-just.toml
+2
View File
@@ -0,0 +1,2 @@
[mappings]
"Makefile" = ["justfile", "Justfile", ".justfile", "*.justfile"]
src/syntax_mapping/builtins/common/50-ssh.toml
+1 -1
View File
@@ -1,2 +1,2 @@
[mappings]
"SSH Config" = ["**/.ssh/config"]
"SSH Config" = ["**/.ssh/**/config"]
src/syntax_mapping/builtins/linux/50-dnf.toml
+2
View File
@@ -0,0 +1,2 @@
[mappings]
"INI" = ["/etc/dnf/dnf.conf", "/etc/yum.repos.d/*.repo"]
src/syntax_mapping/builtins/unix-family/50-ghostty.toml
+2
View File
@@ -0,0 +1,2 @@
[mappings]
"INI" = ["**/ghostty/**/*.ghostty", "**/ghostty/themes/*"]
src/vscreen.rs
+203 -13
View File
@@ -386,12 +386,23 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
}
fn next_text(&mut self) -> Option<EscapeSequenceOffsets> {
self.chars_take_while(|c| c != '\x1B')
self.chars_take_while(|c| !is_sequence_introducer(c))
.map(|(start, end)| EscapeSequenceOffsets::Text { start, end })
}
fn next_sequence(&mut self) -> Option<EscapeSequenceOffsets> {
let (start_sequence, c) = self.chars.next().expect("to not be finished");
// Handle 8-bit C1 introducers as their 7-bit `ESC <x>` equivalents.
match c {
'\u{9B}' => return self.next_csi_body(start_sequence),
'\u{9D}' => return self.next_osc_body(start_sequence),
'\u{90}' | '\u{98}' | '\u{9E}' | '\u{9F}' => {
return self.next_string_terminated_body(start_sequence)
}
_ => {}
}
match self.chars.peek() {
None => Some(EscapeSequenceOffsets::Unknown {
start: start_sequence,
@@ -400,12 +411,24 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
Some((_, ']')) => self.next_osc(start_sequence),
Some((_, '[')) => self.next_csi(start_sequence),
Some((i, c)) => match c {
// 7-bit DCS/SOS/PM/APC: `ESC P/X/^/_` introduces a string body.
Some((_, 'P' | 'X' | '^' | '_')) => {
self.chars.next();
self.next_string_terminated_body(start_sequence)
}
Some(&(i, c)) => match c {
'\x20'..='\x2F' => self.next_nf(start_sequence),
c => Some(EscapeSequenceOffsets::Unknown {
start: start_sequence,
end: i + c.len_utf8(),
}),
c => {
// Single-byte ESC sequence (RIS, DECSC/DECRC, keypad, VT52 etc.).
let end = match self.chars.next() {
Some((j, fc)) => j + fc.len_utf8(),
None => i + c.len_utf8(),
};
Some(EscapeSequenceOffsets::Unknown {
start: start_sequence,
end,
})
}
},
}
}
@@ -413,12 +436,27 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
fn next_osc(&mut self, start_sequence: usize) -> Option<EscapeSequenceOffsets> {
let (osc_open_index, osc_open_char) = self.chars.next().expect("to not be finished");
debug_assert_eq!(osc_open_char, ']');
let start_command = osc_open_index + osc_open_char.len_utf8();
Some(self.read_osc_body(start_sequence, start_command))
}
/// OSC body parser entered after the 8-bit introducer U+009D was consumed.
fn next_osc_body(&mut self, start_sequence: usize) -> Option<EscapeSequenceOffsets> {
let start_command = start_sequence + '\u{9D}'.len_utf8();
Some(self.read_osc_body(start_sequence, start_command))
}
fn read_osc_body(
&mut self,
start_sequence: usize,
start_command: usize,
) -> EscapeSequenceOffsets {
let mut start_terminator: usize;
let mut end_sequence: usize;
loop {
match self.chars_take_while(|c| !matches!(c, '\x07' | '\x1B')) {
// ST is BEL, ESC `\\`, or U+009C.
match self.chars_take_while(|c| !matches!(c, '\x07' | '\x1B' | '\u{9C}')) {
None => {
start_terminator = self.text.len();
end_sequence = start_terminator;
@@ -437,6 +475,11 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
break;
}
Some((ti, '\u{9C}')) => {
end_sequence = ti + '\u{9C}'.len_utf8();
break;
}
Some((ti, '\x1B')) => {
match self.chars.next() {
Some((i, '\\')) => {
@@ -466,20 +509,82 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
}
}
Some(EscapeSequenceOffsets::OSC {
EscapeSequenceOffsets::OSC {
start_sequence,
start_command: osc_open_index + osc_open_char.len_utf8(),
start_command,
start_terminator,
end: end_sequence,
}
}
/// DCS/SOS/PM/APC body parser. Emitted as `Unknown` so the body is stripped.
fn next_string_terminated_body(
&mut self,
start_sequence: usize,
) -> Option<EscapeSequenceOffsets> {
let mut end_sequence: usize;
loop {
match self.chars_take_while(|c| !matches!(c, '\x07' | '\x1B' | '\u{9C}')) {
None => {
end_sequence = self.text.len();
break;
}
Some((_, end)) => {
end_sequence = end;
}
}
match self.chars.next() {
Some((ti, '\x07')) => {
end_sequence = ti + '\x07'.len_utf8();
break;
}
Some((ti, '\u{9C}')) => {
end_sequence = ti + '\u{9C}'.len_utf8();
break;
}
Some((ti, '\x1B')) => match self.chars.next() {
Some((i, '\\')) => {
end_sequence = i + '\\'.len_utf8();
break;
}
None => {
end_sequence = ti + '\x1B'.len_utf8();
break;
}
_ => {}
},
None => break,
Some((_, tc)) => {
panic!("this should not be reached: char {tc:?}")
}
}
}
Some(EscapeSequenceOffsets::Unknown {
start: start_sequence,
end: end_sequence,
})
}
fn next_csi(&mut self, start_sequence: usize) -> Option<EscapeSequenceOffsets> {
let (csi_open_index, csi_open_char) = self.chars.next().expect("to not be finished");
debug_assert_eq!(csi_open_char, '[');
Some(self.read_csi_body(start_sequence, csi_open_index + csi_open_char.len_utf8()))
}
let start_parameters: usize = csi_open_index + csi_open_char.len_utf8();
/// CSI body parser entered after the 8-bit introducer U+009B was consumed.
fn next_csi_body(&mut self, start_sequence: usize) -> Option<EscapeSequenceOffsets> {
let start_parameters = start_sequence + '\u{9B}'.len_utf8();
Some(self.read_csi_body(start_sequence, start_parameters))
}
fn read_csi_body(
&mut self,
start_sequence: usize,
start_parameters: usize,
) -> EscapeSequenceOffsets {
// Keep iterating while within the range of `0x30-0x3F`.
let mut start_intermediates: usize = start_parameters;
if let Some((_, end)) = self.chars_take_while(|c| matches!(c, '\x30'..='\x3F')) {
@@ -498,13 +603,13 @@ impl<'a> EscapeSequenceOffsetsIterator<'a> {
Some((i, c)) => i + c.len_utf8(),
};
Some(EscapeSequenceOffsets::CSI {
EscapeSequenceOffsets::CSI {
start_sequence,
start_parameters,
start_intermediates,
start_final_byte,
end: end_of_sequence,
})
}
}
fn next_nf(&mut self, start_sequence: usize) -> Option<EscapeSequenceOffsets> {
@@ -543,13 +648,25 @@ impl Iterator for EscapeSequenceOffsetsIterator<'_> {
type Item = EscapeSequenceOffsets;
fn next(&mut self) -> Option<Self::Item> {
match self.chars.peek() {
Some((_, '\x1B')) => self.next_sequence(),
Some((_, c)) if is_sequence_introducer(*c) => self.next_sequence(),
Some((_, _)) => self.next_text(),
None => None,
}
}
}
/// True for ESC and the 8-bit C1 sequence introducers (DCS/SOS/CSI/OSC/PM/APC).
#[inline]
fn is_sequence_introducer(c: char) -> bool {
if (c as u32) < 0x80 {
return c == '\x1B';
}
matches!(
c,
'\u{90}' | '\u{98}' | '\u{9B}' | '\u{9D}' | '\u{9E}' | '\u{9F}'
)
}
/// An iterator over ANSI/VT escape sequences within a string.
///
/// ## Example
@@ -717,6 +834,79 @@ mod tests {
);
}
#[test]
fn test_escape_sequence_offsets_iterator_parses_8bit_csi() {
let mut iter = EscapeSequenceOffsetsIterator::new("\u{9B}31m");
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::CSI {
start_sequence: 0,
start_parameters: 2,
start_intermediates: 4,
start_final_byte: 4,
end: 5,
})
);
assert_eq!(iter.next(), None);
}
#[test]
fn test_escape_sequence_offsets_iterator_parses_8bit_osc_with_8bit_st() {
let mut iter = EscapeSequenceOffsetsIterator::new("\u{9D}0;title\u{9C}");
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::OSC {
start_sequence: 0,
start_command: 2,
start_terminator: 9,
end: 11,
})
);
assert_eq!(iter.next(), None);
}
#[test]
fn test_escape_sequence_offsets_iterator_parses_7bit_dcs_consumes_body() {
let mut iter = EscapeSequenceOffsetsIterator::new("\x1BP1;0;|payload\x1B\\rest");
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::Unknown { start: 0, end: 16 })
);
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::Text { start: 16, end: 20 })
);
assert_eq!(iter.next(), None);
}
#[test]
fn test_escape_sequence_offsets_iterator_8bit_dcs_consumes_body() {
let mut iter = EscapeSequenceOffsetsIterator::new("\u{90}body\u{9C}rest");
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::Unknown { start: 0, end: 8 })
);
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::Text { start: 8, end: 12 })
);
}
#[test]
fn test_escape_sequence_offsets_iterator_truncated_dcs_consumes_to_eof() {
// Unterminated DCS must still be consumed, not emitted as `Text`.
let input = "\x1BPno-terminator";
let mut iter = EscapeSequenceOffsetsIterator::new(input);
assert_eq!(
iter.next(),
Some(EscapeSequenceOffsets::Unknown {
start: 0,
end: input.len(),
})
);
assert_eq!(iter.next(), None);
}
#[test]
fn test_escape_sequence_offsets_iterator_parses_csi() {
let mut iter = EscapeSequenceOffsetsIterator::new("\x1B[m");
tests/examples/regression_tests/issue_2745.txt
Vendored
+5
View File
@@ -0,0 +1,5 @@
#!/usr/bin/env bash
NAME="${0##*/}"
for i in {1..3}; do
echo "hello $NAME $i"
done
tests/examples/regression_tests/issue_3631.txt
Vendored
BIN
View File
Binary file not shown.
tests/examples/regression_tests/issue_3647_expect
Vendored
+2
View File
@@ -0,0 +1,2 @@
#!/usr/bin/expect -f
set timeout 30
tests/examples/regression_tests/issue_3647_tclsh
Vendored
+2
View File
@@ -0,0 +1,2 @@
#!/usr/bin/env tclsh
puts "Hello from tclsh"
tests/examples/regression_tests/issue_3647_wish
Vendored
+2
View File
@@ -0,0 +1,2 @@
#!/usr/bin/wish
button .b -text "Click"
tests/integration_tests.rs
+480 -5
View File
@@ -305,6 +305,19 @@ fn line_range_multiple() {
.stdout("line 1\nline 2\nline 4\n");
}
#[test]
fn snip_at_terminal_width_one_does_not_panic() {
bat()
.arg("multiline.txt")
.arg("--style=snip")
.arg("--color=always")
.arg("--terminal-width=1")
.arg("--line-range=1:2")
.arg("--line-range=4:4")
.assert()
.success();
}
#[test]
fn line_range_multiple_with_context() {
bat()
@@ -613,12 +626,18 @@ fn list_themes_to_piped_output() {
}
#[test]
#[serial]
fn list_languages() {
bat()
.arg("--list-languages")
.assert()
.success()
.stdout(predicate::str::contains("Rust").normalize());
mocked_pagers::with_mocked_versions_of_more_and_most_in_path(|| {
bat()
.env("PAGER", mocked_pagers::from("echo pager-output"))
.arg("--list-languages")
.arg("--paging=never")
.assert()
.success()
.stdout(predicate::str::contains("Rust").normalize())
.stdout(predicate::str::contains("pager-output").not());
});
}
#[test]
@@ -1047,6 +1066,57 @@ fn tabs_4_arg_overrides_env_noconfig() {
);
}
#[test]
fn terminal_width_env_var_is_respected() {
let tmp_dir = tempdir().expect("can create temporary directory");
let tmp_path = tmp_dir.path().join("long.txt");
std::fs::write(
&tmp_path,
"0123456789abcdef0123456789abcdef0123456789abcdef\n",
)
.expect("can write temporary file");
bat()
.env("BAT_WIDTH", "20")
.arg(&tmp_path)
.arg("--paging=never")
.arg("--color=never")
.arg("--style=numbers")
.arg("--decorations=always")
.arg("--wrap=character")
.assert()
.success()
.stdout(" 1 0123456789abcde\n f0123456789abcd\n ef0123456789abc\n def\n")
.stderr("");
}
#[test]
fn terminal_width_arg_overrides_env() {
let tmp_dir = tempdir().expect("can create temporary directory");
let tmp_path = tmp_dir.path().join("long.txt");
std::fs::write(
&tmp_path,
"0123456789abcdef0123456789abcdef0123456789abcdef\n",
)
.expect("can write temporary file");
bat()
.env("BAT_WIDTH", "20")
.arg(&tmp_path)
.arg("--paging=never")
.arg("--color=never")
.arg("--style=numbers")
.arg("--decorations=always")
.arg("--wrap=character")
.arg("--terminal-width=10")
.assert()
.success()
.stdout(
" 1 01234\n 56789\n abcde\n f0123\n 45678\n 9abcd\n ef012\n 34567\n 89abc\n def\n",
)
.stderr("");
}
#[test]
fn fail_non_existing() {
bat().arg("non-existing-file").assert().failure();
@@ -1474,6 +1544,7 @@ fn diagnostic_sanity_check() {
.assert()
.success()
.stdout(predicate::str::contains("BAT_PAGER="))
.stdout(predicate::str::contains("BAT_WIDTH="))
.stderr("");
}
@@ -2093,6 +2164,24 @@ fn header_binary() {
.stderr("");
}
#[test]
fn header_zip_file_is_binary() {
let tmp_dir = tempdir().expect("can create temporary directory");
let tmp_path = tmp_dir.path().join("test.zip");
std::fs::write(&tmp_path, b"PK\x03\x04hello").expect("can write temporary file");
bat()
.arg(&tmp_path)
.arg("--decorations=always")
.arg("--style=header")
.arg("-r=0:0")
.arg("--file-name=test.zip")
.assert()
.success()
.stdout("File: test.zip <BINARY>\n")
.stderr("");
}
#[test]
fn header_full_binary() {
bat()
@@ -2820,6 +2909,25 @@ fn binary_as_text() {
.stderr("");
}
#[test]
fn binary_as_text_control_char_width() {
// Control characters are displayed as caret notation (e.g. ^@) by the
// terminal, occupying 2 columns each. With 20 NUL bytes (40 columns) +
// "END" (3 columns) = 43 columns, wrapping at terminal width 40 must
// produce 2 lines, not 1. See #3631.
bat()
.arg("--binary=as-text")
.arg("--wrap=character")
.arg("--terminal-width=40")
.arg("--decorations=always")
.arg("--style=plain")
.arg("--color=never")
.arg("regression_tests/issue_3631.txt")
.assert()
.success()
.stdout(predicate::function(|s: &str| s.lines().count() == 2));
}
#[test]
fn no_strip_overstrike_for_plain_text() {
// Overstrike is preserved for plain text files (no syntax highlighting)
@@ -3881,6 +3989,158 @@ fn strip_ansi_auto_does_not_strip_ansi_when_plain_text_by_option() {
assert!(output.contains("\x1B[33mYellow"))
}
#[test]
fn sanitize_implies_strip_ansi() {
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("\x1B[33mYellow\x1B[m")
.assert()
.success()
.stdout("Yellow");
}
#[test]
fn sanitize_strips_osc_clipboard_hijack() {
// OSC 52 sets the system clipboard. A file containing this would silently
// overwrite the user's clipboard if displayed unfiltered.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("safe\x1B]52;c;cm0=\x07payload")
.assert()
.success()
.stdout("safepayload");
}
#[test]
fn sanitize_strips_osc_8_hyperlink_spoof() {
// OSC 8 hyperlinks let displayed text point to an arbitrary URL.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("\x1B]8;;https://evil.example\x07click here\x1B]8;;\x07")
.assert()
.success()
.stdout("click here");
}
#[test]
fn sanitize_strips_window_title_injection() {
// OSC 0/1/2 set the terminal window title.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("hello\x1B]0;evil-title\x07world")
.assert()
.success()
.stdout("helloworld");
}
#[test]
fn sanitize_strips_8bit_csi() {
// 8-bit CSI introducer (U+009B) is the single-codepoint equivalent of ESC [.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("a\u{9B}31mRED\u{9B}0mb")
.assert()
.success()
.stdout("aREDb");
}
#[test]
fn sanitize_substitutes_bare_cr() {
// Bare CR (not part of CRLF) is the line-overwrite forgery vector.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("safe\rEVIL")
.assert()
.success()
.stdout("safe\u{FFFD}EVIL");
}
#[test]
fn sanitize_preserves_crlf() {
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("line1\r\nline2\r\n")
.assert()
.success()
.stdout("line1\r\nline2\r\n");
}
#[test]
fn sanitize_substitutes_bidi_controls() {
// Trojan-Source attack (CVE-2021-42574): U+202E (RLO) reorders display.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("admin\u{202E}check")
.assert()
.success()
.stdout("admin\u{FFFD}check");
}
#[test]
fn sanitize_substitutes_zero_width() {
// Zero-width chars allow invisible content / identifier confusion.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("ad\u{200B}min")
.assert()
.success()
.stdout("ad\u{FFFD}min");
}
#[test]
fn sanitize_preserves_form_feed_in_source() {
// FF (U+000C) is used as a section separator in C source and Emacs Lisp.
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("section1\x0Csection2")
.assert()
.success()
.stdout("section1\x0Csection2");
}
#[test]
fn sanitize_preserves_unicode_text() {
bat()
.arg("--style=plain")
.arg("--decorations=always")
.arg("--color=never")
.arg("--sanitize=always")
.write_stdin("snowman ☃ CJK 漢字 emoji 🦀")
.assert()
.success()
.stdout("snowman ☃ CJK 漢字 emoji 🦀");
}
// Tests that style components can be removed with `-component`.
#[test]
fn style_components_can_be_removed() {
@@ -4104,3 +4364,218 @@ fn word_wrap_short_line_no_wrap() {
.success()
.stdout("Single Line\n");
}
#[cfg(unix)]
#[cfg(feature = "git")]
fn setup_diff_test_repo() -> tempfile::TempDir {
use std::process::Command;
let dir = tempfile::tempdir().expect("can create temporary directory");
let repo = dir.path();
// Initialize a git repo and commit a file
Command::new("git")
.args(["init"])
.current_dir(repo)
.output()
.expect("git init");
Command::new("git")
.args(["config", "user.email", "test@test.com"])
.current_dir(repo)
.output()
.expect("git config email");
Command::new("git")
.args(["config", "user.name", "Test"])
.current_dir(repo)
.output()
.expect("git config name");
std::fs::write(repo.join("test.txt"), "line 1\nline 2\nline 3\n").expect("can write test file");
Command::new("git")
.args(["add", "test.txt"])
.current_dir(repo)
.output()
.expect("git add");
Command::new("git")
.args(["commit", "-m", "initial"])
.current_dir(repo)
.output()
.expect("git commit");
// Modify the file so --diff has something to show
std::fs::write(
repo.join("test.txt"),
"line 1\nline 2 modified\nline 3\nline 4 added\n",
)
.expect("can write modified test file");
dir
}
#[cfg(unix)]
#[cfg(feature = "git")]
#[test]
fn diff_plain_preserves_change_markers() {
let repo = setup_diff_test_repo();
// With --diff --plain, output should contain the change marker column
// but not other decorations like line numbers or grid
let output = bat()
.current_dir(repo.path())
.arg("--diff")
.arg("--plain")
.arg("--color=never")
.arg("--decorations=always")
.arg("test.txt")
.assert()
.success()
.get_output()
.stdout
.clone();
let stdout = std::str::from_utf8(&output).expect("valid utf-8");
// The output should contain the modified and added lines
assert!(
stdout.contains("line 2 modified"),
"diff plain output should contain modified line, got: {stdout}"
);
assert!(
stdout.contains("line 4 added"),
"diff plain output should contain added line, got: {stdout}"
);
// Should NOT contain line numbers (a decoration that --plain disables)
assert!(
!stdout.contains(" 1"),
"diff plain output should not contain line numbers, got: {stdout}"
);
}
#[cfg(unix)]
#[cfg(feature = "git")]
#[test]
fn diff_plain_does_not_show_grid_or_header() {
let repo = setup_diff_test_repo();
let output = bat()
.current_dir(repo.path())
.arg("--diff")
.arg("--plain")
.arg("--color=never")
.arg("--decorations=always")
.arg("--terminal-width=80")
.arg("test.txt")
.assert()
.success()
.get_output()
.stdout
.clone();
let stdout = std::str::from_utf8(&output).expect("valid utf-8");
// Grid lines use box-drawing characters
assert!(
!stdout.contains('─'),
"diff plain output should not contain grid lines, got: {stdout}"
);
assert!(
!stdout.contains('│'),
"diff plain output should not contain grid separators, got: {stdout}"
);
// Header shows "File: <filename>"
assert!(
!stdout.contains("File:"),
"diff plain output should not contain file header, got: {stdout}"
);
}
#[cfg(unix)]
#[cfg(feature = "git")]
#[test]
fn plain_without_diff_still_works() {
let repo = setup_diff_test_repo();
// --plain without --diff should output file content with no decorations at all
bat()
.current_dir(repo.path())
.arg("--plain")
.arg("--color=never")
.arg("--decorations=always")
.arg("test.txt")
.assert()
.success()
.stdout("line 1\nline 2 modified\nline 3\nline 4 added\n");
}
#[test]
fn tcl_shebang_detection_tclsh() {
bat()
.arg("--color=always")
.arg("--style=plain")
.arg("--decorations=always")
.arg("regression_tests/issue_3647_tclsh")
.assert()
.success();
}
#[test]
fn tcl_shebang_detection_wish() {
bat()
.arg("--color=always")
.arg("--style=plain")
.arg("--decorations=always")
.arg("regression_tests/issue_3647_wish")
.assert()
.success();
}
#[test]
fn tcl_shebang_detection_expect() {
bat()
.arg("--color=always")
.arg("--style=plain")
.arg("--decorations=always")
.arg("regression_tests/issue_3647_expect")
.assert()
.success();
}
#[test]
fn ignored_suffix_enables_first_line_detection() {
// A shebang shell script saved with a `.txt` extension is Plain Text by
// default (the extension wins). With `--ignored-suffix .txt` the suffix is
// stripped before detection, so it falls back to the first line and is
// highlighted exactly as if its language were forced to bash. See #2745.
let fixture = "regression_tests/issue_2745.txt";
let common = ["--color=always", "--decorations=never", "--style=plain"];
let stdout = |args: &[&str]| -> Vec<u8> {
let assert = bat()
.args(common)
.args(args)
.arg(fixture)
.assert()
.success();
assert.get_output().stdout.clone()
};
let forced_bash = stdout(&["--language", "bash"]);
let with_ignored_suffix = stdout(&["--ignored-suffix", ".txt"]);
let default = stdout(&[]);
// The fixture really is being highlighted (forcing bash is not a no-op).
assert!(
forced_bash.windows(2).any(|w| w == b"\x1b["),
"forced-bash output should contain ANSI color codes"
);
// With the ignored suffix, detection matches forced bash highlighting...
assert_eq!(with_ignored_suffix, forced_bash);
// ...while the default (extension wins) stays plain and differs.
assert_ne!(default, forced_bash);
}
tests/snapshots/generate_snapshots.py
Vendored Executable → Regular
+12 -10
View File
@@ -4,6 +4,7 @@ import itertools
import subprocess
import pathlib
import shutil
from typing import Iterable
def generate_snapshots():
@@ -19,22 +20,23 @@ def generate_snapshots():
def generate_style_snapshot(style):
generate_snapshot(style.replace(",", "_"), "--style={}".format(style))
generate_snapshot(style.replace(",", "_"), ["--style={}".format(style)])
def generate_snapshot(name, arguments):
command = "cargo run -- --paging=never --color=never --decorations=always "
command += "{args} sample.rs > output/{name}.snapshot.txt".format(
name=name,
args=arguments
)
def generate_snapshot(name: str, arguments: Iterable[str]):
output_file = "output/{name}.snapshot.txt".format(name=name)
command = [
"cargo", "run", "--", "--paging=never", "--color=never",
"--decorations=always", *arguments, "sample.rs"
]
print("generating snapshot for {}".format(name))
subprocess.call(command, shell=True)
with open(output_file, "w") as f:
subprocess.call(command, stdout=f)
def build_bat():
print("building bat")
subprocess.call("cargo build", cwd="../..", shell=True)
subprocess.call(["cargo", "build"], cwd="../..")
def prepare_output_dir():
@@ -49,7 +51,7 @@ def modify_sample_file():
def undo_sample_file_modification():
print("undoing sample.rs modifications")
subprocess.call("git checkout -- sample.rs", shell=True)
subprocess.call(["git", "checkout", "--", "sample.rs"])
build_bat()
tests/syntax-tests/highlighted/Caddyfile/Caddyfile
Vendored
+25
View File
@@ -0,0 +1,25 @@
(logging) {
 log {
 output file /var/log/caddy.log
 }
}
# a comment
localhost:8080, example.com {
 root * /var/www/site
 file_server
 reverse_proxy /.well-known/matrix/* localhost:8008 {
 header_up Host {upstream_hostport}
 }
 import logging
}
www.example.com {
 redir https://example.com{uri} permanent
 import logging
}
status.example.com {
 reverse_proxy localhost:3002
 import logging
}
tests/syntax-tests/highlighted/Kotlin/test.kt
Vendored
+52 -52
View File
@@ -1,85 +1,85 @@
import kotlin.math.*
import kotlin.math.*
data class Example(
 val name: String,
 val numbers: List<Int?>
data class Example(
 val name: String,
 val numbers: List<Int?>
)
fun interface JokeInterface {
 fun isFunny(): Boolean
fun interface JokeInterface {
 fun isFunny(): Boolean
}
abstract class AbstractJoke : JokeInterface {
 override fun isFunny() = false
 abstract fun content(): String
abstract class AbstractJoke : JokeInterface {
 override fun isFunny() = false
 abstract fun content(): String
}
class Joke : AbstractJoke() {
 override fun isFunny(): Boolean {
class Joke : AbstractJoke() {
 override fun isFunny(): Boolean {
 return true
 }
 override fun content(): String = "content of joke here, haha"
 }
 override fun content(): String = "content of joke here, haha"
}
class DelegatedJoke(val joke: Joke) : JokeInterface by joke {
 val number: Long = 123L
class DelegatedJoke(val joke: Joke) : JokeInterface by joke {
 val number: Long = 123L
 companion object {
 const val someConstant = "some constant text"
 }
 companion object {
 const val someConstant = "some constant text"
 }
}
object SomeSingleton
object SomeSingleton
sealed class Shape {
 abstract fun area(): Double
sealed class Shape {
 abstract fun area(): Double
}
data class Square(val sideLength: Double) : Shape() {
 override fun area(): Double = sideLength.pow(2)
data class Square(val sideLength: Double) : Shape() {
 override fun area(): Double = sideLength.pow(2)
}
object Point : Shape() {
 override fun area() = .0
object Point : Shape() {
 override fun area() = .0
}
class Circle(val radius: Double) : Shape() {
 override fun area(): Double {
 return PI * radius * radius
 }
class Circle(val radius: Double) : Shape() {
 override fun area(): Double {
 return PI * radius * radius
 }
}
fun String.extensionMethod() = "test"
fun String.extensionMethod() = "test"
fun main() {
 val name = """
fun main() {
 val name = """
 multiline
 string
 
 some numbers: 123123 42
 """.trimIndent()
 val example = Example(name = name, numbers = listOf(512, 42, null, -1))
 """.trimIndent()
 val example = Example(name = name, numbers = listOf(512, 42, null, -1))
 example.numbers
 .filterNotNull()
 .forEach { println(it) }
 example.numbers
 .filterNotNull()
 .forEach { println(it) }
 setOf(Joke(), DelegatedJoke(Joke()).joke)
 .filter(JokeInterface::isFunny)
 .map(AbstractJoke::content)
 .forEachIndexed { index: Int, joke ->
 println("I heard a funny joke(#${index + 1}): $joke")
 }
 setOf(Joke(), DelegatedJoke(Joke()).joke)
 .filter(JokeInterface::isFunny)
 .map(AbstractJoke::content)
 .forEachIndexed { index: Int, joke ->
 println("I heard a funny joke(#${index + 1}): $joke")
 }
 listOf(Square(12.3), Point, Circle(5.2))
 .associateWith(Shape::area)
 .toList()
 .sortedBy { it.second }
 .forEach {
 println("${it.first}: ${it.second}")
 }
 listOf(Square(12.3), Point, Circle(5.2))
 .associateWith(Shape::area)
 .toList()
 .sortedBy { it.second }
 .forEach {
 println("${it.first}: ${it.second}")
 }
 println("some string".extensionMethod())
 println("some string".extensionMethod())
 require(SomeSingleton::class.simpleName == "SomeSingletonName") { "something does not seem right..." }
 require(SomeSingleton::class.simpleName == "SomeSingletonName") { "something does not seem right..." }
}
tests/syntax-tests/highlighted/Tcl/expect_shebang
Vendored
+7
View File
@@ -0,0 +1,7 @@
#!/usr/bin/expect -f
# Expect script detected via expect shebang
set timeout 30
spawn ssh user@host
expect "password:"
send "secret\r"
expect eof
tests/syntax-tests/highlighted/Tcl/tclsh_shebang
Vendored
+7
View File
@@ -0,0 +1,7 @@
#!/usr/bin/env tclsh
# Tcl script detected via tclsh shebang
puts "Hello from tclsh"
set x 42
if {$x > 0} {
 puts "positive"
}
tests/syntax-tests/highlighted/Tcl/wish_shebang
Vendored
+5
View File
@@ -0,0 +1,5 @@
#!/usr/bin/wish
# Tk script detected via wish shebang
package require Tk
button .b -text "Click" -command {puts "clicked"}
pack .b
tests/syntax-tests/highlighted/XML/solution.slnx
Vendored
+7
View File
@@ -0,0 +1,7 @@
<Solution>
 <Folder Name="/Build/">
 <File Path="Directory.Build.props" />
 <File Path="projectname.targets" />
 </Folder>
 <Project Path="console.csproj" />
</Solution>
tests/syntax-tests/source/Caddyfile/Caddyfile
Vendored
+25
View File
@@ -0,0 +1,25 @@
(logging) {
log {
output file /var/log/caddy.log
}
}
# a comment
localhost:8080, example.com {
root * /var/www/site
file_server
reverse_proxy /.well-known/matrix/* localhost:8008 {
header_up Host {upstream_hostport}
}
import logging
}
www.example.com {
redir https://example.com{uri} permanent
import logging
}
status.example.com {
reverse_proxy localhost:3002
import logging
}
tests/syntax-tests/source/Tcl/expect_shebang
Vendored
+7
View File
@@ -0,0 +1,7 @@
#!/usr/bin/expect -f
# Expect script detected via expect shebang
set timeout 30
spawn ssh user@host
expect "password:"
send "secret\r"
expect eof
tests/syntax-tests/source/Tcl/tclsh_shebang
Vendored
+7
View File
@@ -0,0 +1,7 @@
#!/usr/bin/env tclsh
# Tcl script detected via tclsh shebang
puts "Hello from tclsh"
set x 42
if {$x > 0} {
puts "positive"
}
tests/syntax-tests/source/Tcl/wish_shebang
Vendored
+5
View File
@@ -0,0 +1,5 @@
#!/usr/bin/wish
# Tk script detected via wish shebang
package require Tk
button .b -text "Click" -command {puts "clicked"}
pack .b
tests/syntax-tests/source/XML/solution.slnx
Vendored
+7
View File
@@ -0,0 +1,7 @@
<Solution>
<Folder Name="/Build/">
<File Path="Directory.Build.props" />
<File Path="projectname.targets" />
</Folder>
<Project Path="console.csproj" />
</Solution>
tests/tester/mod.rs
+37 -28
View File
@@ -1,15 +1,17 @@
use gix::actor::SignatureRef;
use gix::bstr::BString;
use gix::bstr::ByteSlice;
use gix::date::time::Format;
use gix::date::Time;
use gix::objs::tree;
use std::env;
use std::fs::{self, File};
use std::io::Read;
use std::path::{Path, PathBuf};
use std::path::PathBuf;
use std::process::Command;
use tempfile::TempDir;
use git2::build::CheckoutBuilder;
use git2::Repository;
use git2::Signature;
pub struct BatTester {
/// Temporary working directory
temp_dir: TempDir,
@@ -59,35 +61,42 @@ impl Default for BatTester {
}
}
fn create_sample_directory() -> Result<TempDir, git2::Error> {
fn create_sample_directory() -> Result<TempDir, Box<dyn std::error::Error>> {
// Create temp directory and initialize repository
let temp_dir = TempDir::new().expect("Temp directory");
let repo = Repository::init(&temp_dir)?;
let repo = gix::init(&temp_dir)?;
let mut tree = gix::objs::Tree::empty();
// Copy over `sample.rs`
let sample_path = temp_dir.path().join("sample.rs");
println!("{sample_path:?}");
fs::copy("tests/snapshots/sample.rs", &sample_path).expect("successful copy");
// Create sample.rs from snapshot file
let blob_id = repo.write_blob_stream(File::open("tests/snapshots/sample.rs")?)?;
let entry = tree::Entry {
mode: tree::EntryMode::from(tree::EntryKind::Blob),
oid: blob_id.object()?.id,
filename: BString::from("sample.rs"),
};
tree.entries.push(entry);
let tree_id = repo.write_object(tree)?;
// Commit
let mut index = repo.index()?;
index.add_path(Path::new("sample.rs"))?;
let oid = index.write_tree()?;
let signature = Signature::now("bat test runner", "bat@test.runner")?;
let tree = repo.find_tree(oid)?;
let _ = repo.commit(
Some("HEAD"), // point HEAD to our new commit
&signature, // author
&signature, // committer
let author = SignatureRef {
name: "test".as_bytes().as_bstr(),
email: "test@test.test".as_bytes().as_bstr(),
time: &Time::now_local_or_utc().format_or_unix(Format::Raw),
};
let commit_id = repo.commit_as(
author,
author,
"HEAD",
"initial commit",
&tree,
&[],
);
let mut opts = CheckoutBuilder::new();
repo.checkout_head(Some(opts.force()))?;
tree_id,
gix::commit::NO_PARENT_IDS,
)?;
assert_eq!(commit_id, repo.head_id()?);
fs::copy("tests/snapshots/sample.modified.rs", &sample_path).expect("successful copy");
fs::copy(
"tests/snapshots/sample.modified.rs",
temp_dir.path().join("sample.rs"),
)
.expect("successful copy");
Ok(temp_dir)
}