Release bat v0.25.0

Merge pull request #3166 from sharkdp/release_0_25

.cargo/audit.toml

@@ -0,0 +1,2 @@
+[advisories]
+ignore = ["RUSTSEC-2024-0320", "RUSTSEC-2024-0421"]

.github/ISSUE_TEMPLATE/syntax_request.md

@@ -26,4 +26,4 @@ guidelines for adding new syntaxes:
 [Name or description of the syntax/language here]
 
 **Guideline Criteria:**
-[packagecontro.io link here]
+[packagecontrol.io link here]

.github/workflows/CICD.yml

@@ -152,6 +152,7 @@ jobs:
     name: cargo audit
     runs-on: ubuntu-latest
     steps:
+      - run: cargo install cargo-audit --locked
       - uses: actions/checkout@v4
       - run: cargo audit
 
@@ -163,13 +164,15 @@ jobs:
       fail-fast: false
       matrix:
         job:
+          - { target: aarch64-unknown-linux-musl  , os: ubuntu-20.04, dpkg_arch: arm64,            use-cross: true }
           - { target: aarch64-unknown-linux-gnu   , os: ubuntu-20.04, dpkg_arch: arm64,            use-cross: true }
           - { target: arm-unknown-linux-gnueabihf , os: ubuntu-20.04, dpkg_arch: armhf,            use-cross: true }
           - { target: arm-unknown-linux-musleabihf, os: ubuntu-20.04, dpkg_arch: musl-linux-armhf, use-cross: true }
           - { target: i686-pc-windows-msvc        , os: windows-2019,                                              }
           - { target: i686-unknown-linux-gnu      , os: ubuntu-20.04, dpkg_arch: i686,             use-cross: true }
           - { target: i686-unknown-linux-musl     , os: ubuntu-20.04, dpkg_arch: musl-linux-i686,  use-cross: true }
-          - { target: x86_64-apple-darwin         , os: macos-12,                                                  }
+          - { target: x86_64-apple-darwin         , os: macos-13,                                                  }
+          - { target: aarch64-apple-darwin        , os: macos-14,                                                  }
           - { target: x86_64-pc-windows-gnu       , os: windows-2019,                                              }
           - { target: x86_64-pc-windows-msvc      , os: windows-2019,                                              }
           - { target: x86_64-unknown-linux-gnu    , os: ubuntu-20.04, dpkg_arch: amd64,            use-cross: true }
@@ -444,7 +447,7 @@ jobs:
         echo "IS_RELEASE=${IS_RELEASE}" >> $GITHUB_OUTPUT
 
     - name: Publish archives and packages
-      uses: softprops/action-gh-release@v1
+      uses: softprops/action-gh-release@v2
       if: steps.is-release.outputs.IS_RELEASE
       with:
         files: |

.github/workflows/require-changelog-for-PRs.yml

@@ -29,15 +29,5 @@ jobs:
           ADDED=$(git diff -U0 "origin/${PR_BASE}" HEAD -- CHANGELOG.md | grep -P '^\+[^\+].+$')
           echo "Added lines in CHANGELOG.md:"
           echo "$ADDED"
-
-          escape_regex_meta_chars() {
-            # https://stackoverflow.com/a/16951928/4473405
-            sed 's/[][\.|$(){}?+*^]/\\&/g' <<< "$*"
-          }
-
-          double_escape_slash_for_bash() {
-            sed 's/\\/\\\\/g' <<< "$*"
-          }
-
-          echo "Grepping for PR info:"
-          grep "#${PR_NUMBER}\\b.*[(]@$(quote ${PR_SUBMITTER})[)]" <<< "$ADDED"
+          echo "Grepping for PR info (see CONTRIBUTING.md):"
+          grep "#${PR_NUMBER}\\b.*@${PR_SUBMITTER}\\b" <<< "$ADDED"

.gitignore

@@ -2,6 +2,7 @@
 **/*.rs.bk
 
 # Generated files
+/assets/completions/_bat.ps1
 /assets/completions/bat.bash
 /assets/completions/bat.fish
 /assets/completions/bat.zsh

.gitmodules

@@ -260,3 +260,6 @@
 [submodule "assets/syntaxes/02_Extra/vscode-wgsl"]
 	path = assets/syntaxes/02_Extra/vscode-wgsl
 	url = https://github.com/PolyMeilex/vscode-wgsl.git
+[submodule "assets/syntaxes/02_Extra/CFML"]
+	path = assets/syntaxes/02_Extra/CFML
+	url = https://github.com/jcberquist/sublimetext-cfml.git

CHANGELOG.md

@@ -1,10 +1,29 @@
-# unreleased
+# v0.25.0
 
 ## Features
 
+- Set terminal title to file names when Paging is not Paging::Never #2807 (@Oliver-Looney)
+- `bat --squeeze-blank`/`bat -s` will now squeeze consecutive empty lines, see #1441 (@eth-p) and #2665 (@einfachIrgendwer0815)
+- `bat --squeeze-limit` to set the maximum number of empty consecutive when using `--squeeze-blank`, see #1441 (@eth-p) and #2665 (@einfachIrgendwer0815)
+- `PrettyPrinter::squeeze_empty_lines` to support line squeezing for bat as a library, see #1441 (@eth-p) and #2665 (@einfachIrgendwer0815)
+- Syntax highlighting for JavaScript files that start with `#!/usr/bin/env bun` #2913 (@sharunkumar)
+- `bat --strip-ansi={never,always,auto}` to remove ANSI escape sequences from bat's input, see #2999 (@eth-p)
+- Add or remove individual style components without replacing all styles #2929 (@eth-p)
+- Automatically choose theme based on the terminal's color scheme, see #2896 (@bash)
+- Add option `--binary=as-text` for printing binary content, see issue #2974 and PR #2976 (@einfachIrgendwer0815)
+- Make shell completions available via `--completion <shell>`, see issue #2057 and PR #3126 (@einfachIrgendwer0815)
+- Syntax highlighting for puppet code blocks within Markdown files, see #3152 (@liliwilson)
+
 ## Bugfixes
 
+- Fix long file name wrapping in header, see #2835 (@FilipRazek)
 - Fix `NO_COLOR` support, see #2767 (@acuteenvy)
+- Fix handling of inputs with OSC ANSI escape sequences, see #2541 and #2544 (@eth-p)
+- Fix handling of inputs with combined ANSI color and attribute sequences, see #2185 and #2856 (@eth-p)
+- Fix panel width when line 10000 wraps, see #2854 (@eth-p)
+- Fix compile issue of `time` dependency caused by standard library regression #3045 (@cyqsimon)
+- Fix override behavior of --plain and --paging, see issue #2731 and PR #3108 (@einfachIrgendwer0815)
+- Fix bugs in `$LESSOPEN` support, see #2805 (@Anomalocaridid)
 
 ## Other
 
@@ -15,19 +34,63 @@
 - Minor benchmark script improvements #2768 (@cyqsimon)
 - Update Arch Linux package URL in README files #2779 (@brunobell)
 - Update and improve `zsh` completion, see #2772 (@okapia)
+- More extensible syntax mapping mechanism #2755 (@cyqsimon)
 - Use proper Architecture for Debian packages built for musl, see #2811 (@Enselic)
 - Pull in fix for unsafe-libyaml security advisory, see #2812 (@dtolnay)
 - Update git-version dependency to use Syn v2, see #2816 (@dtolnay)
+- Update git2 dependency to v0.18.2, see #2852 (@eth-p)
+- Improve performance when color output disabled, see #2397 and #2857 (@eth-p)
+- Relax syntax mapping rule restrictions to allow brace expansion #2865 (@cyqsimon)
+- Apply clippy fixes #2864 (@cyqsimon)
+- Faster startup by offloading glob matcher building to a worker thread #2868 (@cyqsimon)
+- Display which theme is the default one in basic output (no colors), see #2937 (@sblondon)
+- Display which theme is the default one in colored output, see #2838 (@sblondon)
+- Add aarch64-apple-darwin ("Apple Silicon") binary tarballs to releases, see #2967 (@someposer)
+- Update the Lisp syntax, see #2970 (@ccqpein)
+- Use bat's ANSI iterator during tab expansion, see #2998 (@eth-p)
+- Support 'statically linked binary' for aarch64 in 'Release' page, see #2992 (@tzq0301)
+- Update options in shell completions and the man page of `bat`, see #2995 (@akinomyoga)
+- Update nix dev-dependency to v0.29.0, see #3112 (@decathorpe)
+- Bump MSRV to [1.74](https://blog.rust-lang.org/2023/11/16/Rust-1.74.0.html), see #3154 (@keith-hall)
+- Update clircle dependency to remove winapi transitive dependency, see #3113 (@niklasmohrin)
 
 ## Syntaxes
 
 - `cmd-help`: scope subcommands followed by other terms, and other misc improvements, see #2819 (@victor-gp)
 - Upgrade JQ syntax, see #2820 (@dependabot[bot])
+- Add syntax mapping for quadman quadlets #2866 (@cyqsimon)
+- Map containers .conf files to TOML syntax #2867 (@cyqsimon)
+- Associate `.xsh` files with `xonsh` syntax that is Python, see #2840 (@anki-code)
+- Associate JSON with Comments `.jsonc` with `json` syntax, see #2795 (@mxaddict)
+- Associate JSON-LD `.jsonld` files with `json` syntax, see #3037 (@vorburger)
+- Associate `.textproto` files with `ProtoBuf` syntax, see #3038 (@vorburger)
+- Associate GeoJSON `.geojson` files with `json` syntax, see #3084 (@mvaaltola)
+- Associate `.aws/{config,credentials}`, see #2795 (@mxaddict)
+- Associate Wireguard config `/etc/wireguard/*.conf`, see #2874 (@cyqsimon)
+- Add support for [CFML](https://www.adobe.com/products/coldfusion-family.html), see #3031 (@brenton-at-pieces)
+- Map `*.mkd` files to `Markdown` syntax, see issue #3060 and PR #3061 (@einfachIrgendwer0815)
+- Add syntax mapping for CITATION.cff, see #3103 (@Ugzuzg)
+- Add syntax mapping for kubernetes config files #3049 (@cyqsimon)
+- Adds support for pipe delimiter for CSV #3115 (@pratik-m)
+- Add syntax mapping for `/etc/pacman.conf` #2961 (@cyqsimon)
+- Associate `uv.lock` with `TOML` syntax, see #3132 (@fepegar)
 
 ## Themes
 
+- Patched/improved themes for better Manpage syntax highlighting support, see #2994 (@keith-hall).
+
 ## `bat` as a library
 
+- Changes to `syntax_mapping::SyntaxMapping` #2755 (@cyqsimon)
+  - `SyntaxMapping::get_syntax_for` is now correctly public
+  - [BREAKING] `SyntaxMapping::{empty,builtin}` are removed; use `SyntaxMapping::new` instead
+  - [BREAKING] `SyntaxMapping::mappings` is replaced by `SyntaxMapping::{builtin,custom,all}_mappings`
+- Make `Controller::run_with_error_handler`'s error handler `FnMut`, see #2831 (@rhysd)
+- Improve compile time by 20%, see #2815 (@dtolnay)
+- Add `theme::theme` for choosing an appropriate theme based on the
+  terminal's color scheme, see #2896 (@bash)
+  - [BREAKING] Remove `HighlightingAssets::default_theme`. Use `theme::default_theme` instead.
+- Add `PrettyPrinter::print_with_writer` for custom output destinations, see #3070 (@kojix2)
 
 # v0.24.0
 
@@ -66,6 +129,7 @@
 - Update `Julia` syntax, see #2553 (@dependabot)
 - add `NSIS` support, see #2577 (@idleberg)
 - Update `ssh-config`, see #2697 (@mrmeszaros)
+- Add syntax mapping `*.debdiff` => `diff`, see #2947 (@jacg)
 
 ## `bat` as a library
 

CONTRIBUTING.md

@@ -6,21 +6,42 @@ Thank you for considering to contribute to `bat`!
 
 ## Add an entry to the changelog
 
-If your contribution changes the behavior of `bat` (as opposed to a typo-fix
-in the documentation), please update the [`CHANGELOG.md`](CHANGELOG.md) file
-and describe your changes. This makes the release process much easier and
-therefore helps to get your changes into a new `bat` release faster.
+Keeping the [`CHANGELOG.md`](CHANGELOG.md) file up-to-date makes the release
+process much easier and therefore helps to get your changes into a new `bat`
+release faster. However, not every change to the repository requires a
+changelog entry. Below are a few examples of that.
+
+Please update the changelog if your contribution contains changes regarding
+any of the following:
+  - the behavior of `bat`
+  - syntax mappings
+  - syntax definitions
+  - themes
+  - the build system, linting, or CI workflows
+
+A changelog entry is not necessary when:
+  - updating documentation
+  - fixing typos
+
+>[!NOTE]
+> For PRs, a CI workflow verifies that a suitable changelog entry is
+> added. If such an entry is missing, the workflow will fail. If your
+> changes do not need an entry to the changelog (see above), that
+> workflow failure can be disregarded.
+
+
+### Changelog entry format
 
 The top of the `CHANGELOG` contains a *"unreleased"* section with a few
 subsections (Features, Bugfixes, …). Please add your entry to the subsection
 that best describes your change.
 
-Entries follow this format:
+Entries must follow this format:
 ```
 - Short description of what has been changed, see #123 (@user)
 ```
-Here, `#123` is the number of the original issue and/or your pull request.
-Please replace `@user` by your GitHub username.
+Please replace `#123` with the number of your pull request (not issue) and
+`@user` by your GitHub username.
 
 
 ## Development

Cargo.toml

@@ -6,11 +6,11 @@ homepage = "https://github.com/sharkdp/bat"
 license = "MIT OR Apache-2.0"
 name = "bat"
 repository = "https://github.com/sharkdp/bat"
-version = "0.24.0"
+version = "0.25.0"
 exclude = ["assets/syntaxes/*", "assets/themes/*"]
 build = "build/main.rs"
 edition = '2021'
-rust-version = "1.70"
+rust-version = "1.74"
 
 [features]
 default = ["application"]
@@ -33,7 +33,7 @@ minimal-application = [
 ]
 git = ["git2"] # Support indicating git modifications
 paging = ["shell-words", "grep-cli"] # Support applying a pager on the output
-lessopen = ["run_script", "os_str_bytes"] # Support $LESSOPEN preprocessor
+lessopen = ["execute"] # Support $LESSOPEN preprocessor
 build-assets = ["syntect/yaml-load", "syntect/plist-load", "regex", "walkdir"]
 
 # You need to use one of these if you depend on bat as a library:
@@ -41,40 +41,41 @@ regex-onig = ["syntect/regex-onig"] # Use the "oniguruma" regex engine
 regex-fancy = ["syntect/regex-fancy"] # Use the rust-only "fancy-regex" engine
 
 [dependencies]
-nu-ansi-term = "0.49.0"
+nu-ansi-term = "0.50.0"
 ansi_colours = "^1.2"
 bincode = "1.0"
-console = "0.15.7"
+console = "0.15.10"
 flate2 = "1.0"
-once_cell = "1.19"
+once_cell = "1.20"
 thiserror = "1.0"
 wild = { version = "2.2", optional = true }
 content_inspector = "0.2.4"
 shell-words = { version = "1.1.0", optional = true }
-unicode-width = "0.1.11"
+unicode-width = "0.1.13"
 globset = "0.4"
-serde = { version = "1.0", features = ["derive"] }
+serde = "1.0"
+serde_derive = "1.0"
 serde_yaml = "0.9.28"
 semver = "1.0"
 path_abs = { version = "0.5", default-features = false }
-clircle = "0.4"
+clircle = { version = "0.6.1", default-features = false }
 bugreport = { version = "0.5.0", optional = true }
 etcetera = { version = "0.8.0", optional = true }
-grep-cli = { version = "0.1.10", optional = true }
-regex = { version = "1.10.2", optional = true }
-walkdir = { version = "2.4", optional = true }
+grep-cli = { version = "0.1.11", optional = true }
+regex = { version = "1.10.6", optional = true }
+walkdir = { version = "2.5", optional = true }
 bytesize = { version = "1.3.0" }
-encoding_rs = "0.8.33"
-os_str_bytes = { version = "~6.6", optional = true }
-run_script = { version = "^0.10.1", optional = true}
+encoding_rs = "0.8.35"
+execute = { version = "0.2.13", optional = true }
+terminal-colorsaurus = "0.4"
 
 [dependencies.git2]
-version = "0.18"
+version = "0.19"
 optional = true
 default-features = false
 
 [dependencies.syntect]
-version = "5.1.0"
+version = "5.2.0"
 default-features = false
 features = ["parsing"]
 
@@ -85,21 +86,31 @@ features = ["wrap_help", "cargo"]
 
 [target.'cfg(target_os = "macos")'.dependencies]
 home = "0.5.9"
-plist = "1.5.1"
+plist = "1.7.0"
 
 [dev-dependencies]
 assert_cmd = "2.0.12"
-expect-test = "1.4.1"
+expect-test = "1.5.0"
 serial_test = { version = "2.0.0", default-features = false }
-predicates = "3.0.4"
+predicates = "3.1.3"
 wait-timeout = "0.2.0"
 tempfile = "3.8.1"
+serde = { version = "1.0", features = ["derive"] }
 
 [target.'cfg(unix)'.dev-dependencies]
-nix = { version = "0.26.4", default-features = false, features = ["term"] }
+nix = { version = "0.29", default-features = false, features = ["term"] }
 
 [build-dependencies]
-anyhow = "1.0.78"
+anyhow = "1.0.86"
+indexmap = { version = "2.3.0", features = ["serde"] }
+itertools = "0.13.0"
+once_cell = "1.20"
+regex = "1.10.6"
+serde = "1.0"
+serde_derive = "1.0"
+serde_with = { version = "3.12.0", default-features = false, features = ["macros"] }
+toml = { version = "0.8.19", features = ["preserve_order"] }
+walkdir = "2.5"
 
 [build-dependencies.clap]
 version = "4.4.12"

README.md

@@ -35,11 +35,11 @@ A special *thank you* goes to our biggest <a href="doc/sponsors.md">sponsors</a>
 <a href="https://www.warp.dev/?utm_source=github&utm_medium=referral&utm_campaign=bat_20231001">
   <img src="doc/sponsors/warp-logo.png" width="200" alt="Warp">
   <br>
-  <strong>Warp is a modern, Rust-based terminal with AI built in<br>so you and your team can build great software, faster.</strong>
+  <strong>Warp, the intelligent terminal</strong>
   <br>
-  <sub>Feel more productive on the command line with parameterized commands,</sub>
+  <sub>Run commands like a power user with AI and your dev team’s</sub>
   <br>
-  <sup>autosuggestions, and an IDE-like text editor.</sup>
+  <sup>knowledge in one fast, intuitive terminal. For MacOS or Linux.</sup>
 </a>
 
 ### Syntax highlighting
@@ -204,19 +204,19 @@ bat main.cpp | xclip
 `MANPAGER` environment variable:
 
 ```bash
-export MANPAGER="sh -c 'col -bx | bat -l man -p'"
+export MANPAGER="sh -c 'sed -u -e \"s/\\x1B\[[0-9;]*m//g; s/.\\x08//g\" | bat -p -lman'"
 man 2 select
 ```
 (replace `bat` with `batcat` if you are on Debian or Ubuntu)
 
-It might also be necessary to set `MANROFFOPT="-c"` if you experience
-formatting problems.
-
 If you prefer to have this bundled in a new command, you can also use [`batman`](https://github.com/eth-p/bat-extras/blob/master/doc/batman.md).
 
-Note that the [Manpage syntax](assets/syntaxes/02_Extra/Manpage.sublime-syntax) is developed in this repository and still needs some work.
+> [!WARNING]  
+> This will [not work](https://github.com/sharkdp/bat/issues/1145) out of the box with Mandoc's `man` implementation.
+>
+> Please either use `batman`, or convert the shell script to a [shebang executable](https://en.wikipedia.org/wiki/Shebang_(Unix)) and point `MANPAGER` to that.
 
-Also, note that this will [not work](https://github.com/sharkdp/bat/issues/1145) with Mandocs `man` implementation.
+Note that the [Manpage syntax](assets/syntaxes/02_Extra/Manpage.sublime-syntax) is developed in this repository and still needs some work.
 
 #### `prettier` / `shfmt` / `rustfmt`
 
@@ -373,6 +373,14 @@ You can install `bat` using the [nix package manager](https://nixos.org/nix):
 nix-env -i bat
 ```
 
+### Via flox
+
+You can install `bat` using [Flox](https://flox.dev)
+
+```bash
+flox install bat
+```
+
 ### On openSUSE
 
 You can install `bat` with zypper:
@@ -447,7 +455,7 @@ binaries are also available: look for archives with `musl` in the file name.
 
 ### From source
 
-If you want to build `bat` from source, you need Rust 1.70.0 or
+If you want to build `bat` from source, you need Rust 1.74.0 or
 higher. You can then use `cargo` to build everything:
 
 ```bash
@@ -457,6 +465,12 @@ cargo install --locked bat
 Note that additional files like the man page or shell completion
 files can not be installed in this way. They will be generated by `cargo` and should be available in the cargo target folder (under `build`).
 
+Shell completions are also available by running:
+```bash
+bat --completion <shell>
+# see --help for supported shells
+```
+
 ## Customization
 
 ### Highlighting theme
@@ -474,8 +488,10 @@ the following command (you need [`fzf`](https://github.com/junegunn/fzf) for thi
 bat --list-themes | fzf --preview="bat --theme={} --color=always /path/to/file"
 ```
 
-`bat` looks good on a dark background by default. However, if your terminal uses a
-light background, some themes like `GitHub` or `OneHalfLight` will work better for you.
+`bat` automatically picks a fitting theme depending on your terminal's background color.
+You can use the `--theme-light` / `--theme-light` options or the `BAT_THEME_DARK` / `BAT_THEME_LIGHT` environment variables
+to customize the themes used. This is especially useful if you frequently switch between dark and light mode.
+
 You can also use a custom theme by following the
 ['Adding new themes' section below](https://github.com/sharkdp/bat#adding-new-themes).
 
@@ -507,6 +523,16 @@ 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](https://github.com/sharkdp/bat#configuration-file).
 
+>[!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.
+> By prefixing a component with `+` or `-`, it can be added or removed from the current style.
+>
+> For example, if your config contains `--style=full,-snip`, you can run bat with
+> `--style=-grid,+snip` to remove the grid and add back the `snip` component.
+> Or, if you want to override the styles completely, you use `--style=numbers` to
+> only show the line numbers.
+
 ### Adding new syntaxes / language definitions
 
 Should you find that a particular syntax is not available within `bat`, you can follow these
@@ -602,7 +628,8 @@ set, `less` is used by default. If you want to use a different pager, you can ei
 `PAGER` variable or set the `BAT_PAGER` environment variable to override what is specified in
 `PAGER`.
 
-**Note**: If `PAGER` is `more` or `most`, `bat` will silently use `less` instead to ensure support for colors.
+>[!NOTE]
+> If `PAGER` is `more` or `most`, `bat` will silently use `less` instead to ensure support for colors.
 
 If you want to pass command-line arguments to the pager, you can also set them via the
 `PAGER`/`BAT_PAGER` variables:
@@ -613,20 +640,37 @@ export BAT_PAGER="less -RF"
 
 Instead of using environment variables, you can also use `bat`s [configuration file](https://github.com/sharkdp/bat#configuration-file) to configure the pager (`--pager` option).
 
-**Note**: By default, if the pager is set to `less` (and no command-line options are specified),
-`bat` will pass the following command line options to the pager: `-R`/`--RAW-CONTROL-CHARS`,
-`-F`/`--quit-if-one-screen` and `-X`/`--no-init`. The last option (`-X`) is only used for `less`
-versions older than 530.
 
-The `-R` option is needed to interpret ANSI colors correctly. The second option (`-F`) instructs
-less to exit immediately if the output size is smaller than the vertical size of the terminal.
-This is convenient for small files because you do not have to press `q` to quit the pager. The
-third option (`-X`) is needed to fix a bug with the `--quit-if-one-screen` feature in old versions
-of `less`. Unfortunately, it also breaks mouse-wheel support in `less`.
+### Using `less` as a pager
 
-If you want to enable mouse-wheel scrolling on older versions of `less`, you can pass just `-R` (as
-in the example above, this will disable the quit-if-one-screen feature). For less 530 or newer,
-it should work out of the box.
+When using `less` as a pager, `bat` will automatically pass extra options along to `less`
+to improve the experience. Specifically, `-R`/`--RAW-CONTROL-CHARS`, `-F`/`--quit-if-one-screen`,
+and under certain conditions, `-X`/`--no-init` and/or `-S`/`--chop-long-lines`.
+
+>[!IMPORTANT]
+> These options will not be added if:
+> - The pager is not named `less`.
+> - The `--pager` argument contains any command-line arguments (e.g. `--pager="less -R"`).
+> - The `BAT_PAGER` environment variable contains any command-line arguments (e.g. `export BAT_PAGER="less -R"`)
+>
+> The `--quit-if-one-screen` option will not be added when:
+> - The `--paging=always` argument is used.
+> - The `BAT_PAGING` environment is set to `always`.
+
+The `-R` option is needed to interpret ANSI colors correctly.
+
+The `-F` option instructs `less` to exit immediately if the output size is smaller than
+the vertical size of the terminal. This is convenient for small files because you do not
+have to press `q` to quit the pager.
+
+The `-X` option is needed to fix a bug with the `--quit-if-one-screen` feature in versions
+of `less` older than version 530. Unfortunately, it also breaks mouse-wheel support in `less`.
+If you want to enable mouse-wheel scrolling on older versions of `less` and do not mind losing
+the quit-if-one-screen feature, you can set the pager (via `--pager` or `BAT_PAGER`) to `less -R`.
+For `less` 530 or newer, it should work out of the box.
+
+The `-S` option is added when `bat`'s `-S`/`--chop-long-lines` option is used. This tells `less`
+to truncate any lines larger than the terminal width.
 
 ### Indentation
 
@@ -657,10 +701,11 @@ on your operating system. To get the default path for your system, call
 bat --config-file
 ```
 
-Alternatively, you can use the `BAT_CONFIG_PATH` environment variable to point `bat` to a
-non-default location of the configuration file:
+Alternatively, you can use `BAT_CONFIG_PATH` or `BAT_CONFIG_DIR` environment variables to point `bat`
+to a non-default location of the configuration file or the configuration directory respectively:
 ```bash
-export BAT_CONFIG_PATH="/path/to/bat.conf"
+export BAT_CONFIG_PATH="/path/to/bat/bat.conf"
+export BAT_CONFIG_DIR="/path/to/bat"
 ```
 
 A default configuration file can be created with the `--generate-config-file` option.
@@ -711,7 +756,7 @@ your `PATH` or [define an environment variable](#using-a-different-pager). The [
 Windows 10 natively supports colors in both `conhost.exe` (Command Prompt) and PowerShell since
 [v1511](https://en.wikipedia.org/wiki/Windows_10_version_history#Version_1511_(November_Update)), as
 well as in newer versions of bash. On earlier versions of Windows, you can use
-[Cmder](http://cmder.net/), which includes [ConEmu](https://conemu.github.io/).
+[Cmder](http://cmder.app/), which includes [ConEmu](https://conemu.github.io/).
 
 **Note:** Old versions of `less` do not correctly interpret colors on Windows. To fix this, you can add the optional Unix tools to your PATH when installing Git. If you don’t have any other pagers installed, you can disable paging entirely by passing `--paging=never` or by setting `BAT_PAGER` to an empty string.
 
@@ -741,9 +786,14 @@ bat() {
 
 If an input file contains color codes or other ANSI escape sequences or control characters, `bat` will have problems
 performing syntax highlighting and text wrapping, and thus the output can become garbled.
-When displaying such files it is recommended to disable both syntax highlighting and wrapping by
+
+If your version of `bat` supports the `--strip-ansi=auto` option, it can be used to remove such sequences
+before syntax highlighting. Alternatively, you may disable both syntax highlighting and wrapping by
 passing the `--color=never --wrap=never` options to `bat`.
 
+> [!NOTE]
+> The `auto` option of `--strip-ansi` avoids removing escape sequences when the syntax is plain text.
+
 ### Terminals & colors
 
 `bat` handles terminals *with* and *without* truecolor support. However, the colors in most syntax

assets/completions/_bat.ps1.in

@@ -37,6 +37,8 @@ Register-ArgumentCompleter -Native -CommandName '{{PROJECT_EXECUTABLE}}' -Script
             [CompletionResult]::new('-m', 'm', [CompletionResultType]::ParameterName, 'Use the specified syntax for files matching the glob pattern (''*.cpp:C++'').')
             [CompletionResult]::new('--map-syntax', 'map-syntax', [CompletionResultType]::ParameterName, 'Use the specified syntax for files matching the glob pattern (''*.cpp:C++'').')
             [CompletionResult]::new('--theme', 'theme', [CompletionResultType]::ParameterName, 'Set the color theme for syntax highlighting.')
+            [CompletionResult]::new('--theme-dark', 'theme', [CompletionResultType]::ParameterName, 'Set the color theme for syntax highlighting for dark backgrounds.')
+            [CompletionResult]::new('--theme-light', 'theme', [CompletionResultType]::ParameterName, 'Set the color theme for syntax highlighting for light backgrounds.')
             [CompletionResult]::new('--style', 'style', [CompletionResultType]::ParameterName, 'Comma-separated list of style elements to display (*default*, auto, full, plain, changes, header, header-filename, header-filesize, grid, rule, numbers, snip).')
             [CompletionResult]::new('-r', 'r', [CompletionResultType]::ParameterName, 'Only print the lines from N to M.')
             [CompletionResult]::new('--line-range', 'line-range', [CompletionResultType]::ParameterName, 'Only print the lines from N to M.')

assets/completions/bat.bash.in

@@ -76,6 +76,7 @@ _bat() {
 	-m | --map-syntax | \
 	--ignored-suffix | \
 	--list-themes | \
+	--squeeze-limit | \
 	--line-range | \
 	-L | --list-languages | \
 	--lessopen | \
@@ -112,6 +113,13 @@ _bat() {
 		return 0
 		;;
 	--theme)
+    	local IFS=$'\n'
+    	COMPREPLY=($(compgen -W "auto${IFS}auto:always${IFS}auto:system${IFS}dark${IFS}light${IFS}$("$1" --list-themes)" -- "$cur"))
+                    __bat_escape_completions
+    	return 0
+    	;;
+	--theme-dark | \
+	--theme-light)
 		local IFS=$'\n'
 		COMPREPLY=($(compgen -W "$("$1" --list-themes)" -- "$cur"))
                 __bat_escape_completions
@@ -157,6 +165,7 @@ _bat() {
 			--diff-context
 			--tabs
 			--wrap
+			--chop-long-lines
 			--terminal-width
 			--number
 			--color
@@ -168,19 +177,27 @@ _bat() {
 			--map-syntax
 			--ignored-suffix
 			--theme
+			--theme-dark
+			--theme-light
 			--list-themes
+			--squeeze-blank
+			--squeeze-limit
 			--style
 			--line-range
 			--list-languages
 			--lessopen
 			--diagnostic
 			--acknowledgements
+			--set-terminal-title
 			--help
 			--version
 			--cache-dir
 			--config-dir
 			--config-file
 			--generate-config-file
+			--no-config
+			--no-custom-assets
+			--no-lessopen
 		" -- "$cur"))
 		return 0
 	fi

assets/completions/bat.fish.in

@@ -129,10 +129,20 @@ set -l tabs_opts '
     8\t
 '
 
+set -l special_themes '
+    auto\tdefault,\ Choose\ a\ theme\ based\ on\ dark\ or\ light\ mode
+    auto:always\tChoose\ a\ theme\ based\ on\ dark\ or\ light\ mode
+    auto:system\tChoose\ a\ theme\ based\ on\ dark\ or\ light\ mode
+    dark\tUse\ the\ theme\ specified\ by\ --theme-dark
+    light\tUse\ the\ theme\ specified\ by\ --theme-light
+'
+
 # Completions:
 
 complete -c $bat -l acknowledgements -d "Print acknowledgements" -n __fish_is_first_arg
 
+complete -c $bat -l cache-dir -f -d "Show bat's cache directory" -n __fish_is_first_arg
+
 complete -c $bat -l color -x -a "$color_opts" -d "When to use colored output" -n __bat_no_excl_args
 
 complete -c $bat -l config-dir -f -d "Display location of configuration directory" -n __fish_is_first_arg
@@ -201,7 +211,11 @@ 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 "(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 --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-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 -s V -l version -f -d "Show version information" -n __fish_is_first_arg
 

assets/completions/bat.zsh.in

@@ -26,7 +26,7 @@ _{{PROJECT_EXECUTABLE}}_main() {
     args=(
         '(-A --show-all)'{-A,--show-all}'[show non-printable characters (space, tab, newline, ..)]'
         --nonprintable-notation='[specify how to display non-printable characters when using --show-all]:notation:(caret unicode)'
-        \*{-p,--plain}'[show plain style (alias for `--style=plain`), repeat twice to disable disable automatic paging (alias for `--paging=never`)]'
+        \*{-p,--plain}'[show plain style (alias for `--style=plain`), repeat twice to disable automatic paging (alias for `--paging=never`)]'
         '(-l --language)'{-l+,--language=}'[set the language for syntax highlighting]:language:->languages'
         \*{-H+,--highlight-line=}'[highlight specified block of lines]:start\:end'
         \*--file-name='[specify the name to display for a file]:name:_files'
@@ -42,12 +42,15 @@ _{{PROJECT_EXECUTABLE}}_main() {
         --decorations='[specify when to show the decorations]:when:(auto never always)'
         --paging='[specify when to use the pager]:when:(auto never always)'
         '(-m --map-syntax)'{-m+,--map-syntax=}'[map a glob pattern to an existing syntax name]: :->syntax-maps'
-        '(--theme)'--theme='[set the color theme for syntax highlighting]:theme:->themes'
+        '(--theme)'--theme='[set the color theme for syntax highlighting]:theme:->theme_preferences'
+        '(--theme-dark)'--theme-dark='[set the color theme for syntax highlighting for dark backgrounds]:theme:->themes'
+        '(--theme-light)'--theme-light='[set the color theme for syntax highlighting for light backgrounds]:theme:->themes'
         '(: --list-themes --list-languages -L)'--list-themes'[show all supported highlighting themes]'
         --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'
         '(* -)'{-L,--list-languages}'[display all supported languages]'
+        -P'[disable paging]'
         "--no-config[don't use the configuration file]"
         "--no-custom-assets[don't load custom assets]"
         '(--no-lessopen)'--lessopen'[enable the $LESSOPEN preprocessor]'
@@ -81,7 +84,13 @@ _{{PROJECT_EXECUTABLE}}_main() {
 
         themes)
             local -a themes expl
-            themes=( ${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --list-themes)"} )
+            themes=(${(f)"$(_call_program themes {{PROJECT_EXECUTABLE}} --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)"} )
 
             _wanted themes expl 'theme' compadd -a themes && ret=0
         ;;

assets/manual/bat.1.in

@@ -87,6 +87,10 @@ Set the tab width to T spaces. Use a width of 0 to pass tabs through directly
 Specify the text\-wrapping mode (*auto*, never, character). The '\-\-terminal\-width' option
 can be used in addition to control the output width.
 .HP
+\fB\-S\fR, \fB\-\-chop\-long\-lines\fR
+.IP
+Truncate all lines longer than screen width. Alias for '\-\-wrap=never'.
+.HP
 \fB\-\-terminal\-width\fR <width>
 .IP
 Explicitly set the width of the terminal instead of determining it automatically. If
@@ -141,16 +145,58 @@ use -m '*.build:Python'. To highlight files named '.myignore' with the Git Ignor
 syntax, use -m '.myignore:Git Ignore'.
 Note that the right-hand side is the *name* of the syntax, not a file extension.
 .HP
+\fB\-\-ignored\-suffix\fR <ignored-suffix>
+.IP
+Ignore extension. For example: 'bat \-\-ignored-suffix ".dev" my_file.json.dev'
+will use JSON syntax, and ignore '.dev'
+.HP
 \fB\-\-theme\fR <theme>
 .IP
-Set the theme for syntax highlighting. Use '\-\-list\-themes' to see all available themes.
-To set a default theme, add the '\-\-theme="..."' option to the configuration file or
-export the BAT_THEME environment variable (e.g.: export BAT_THEME="...").
+Set the theme for syntax highlighting. Use \fB\-\-list\-themes\fP to see all available themes.
+To set a default theme, add the \fB\-\-theme="..."\fP option to the configuration file or
+export the \fBBAT_THEME\fP environment variable (e.g.: \fBexport BAT_THEME="..."\fP).
+
+Special values:
+.RS
+.IP "auto (\fIdefault\fR)"
+Picks a dark or light theme depending on the terminal's colors.
+Use \fB-\-theme\-light\fR and \fB-\-theme\-dark\fR to customize the selected theme.
+.IP "auto:always"
+Variation of \fBauto\fR where where the terminal's colors are detected even when the output is redirected.
+.IP "auto:system (macOS only)"
+Variation of \fBauto\fR where the color scheme is detected from the system-wide preference instead.
+.IP "dark"
+Use the dark theme specified by \fB-\-theme-dark\fR.
+.IP "light"
+Use the light theme specified by \fB-\-theme-light\fR.
+.RE
+.HP
+\fB\-\-theme\-dark\fR <theme>
+.IP
+Sets the theme name for syntax highlighting used when the terminal uses a dark background.
+To set a default theme, add the \fB\-\-theme-dark="..."\fP option to the configuration file or
+export the \fBBAT_THEME_DARK\fP environment variable (e.g. \fBexport BAT_THEME_DARK="..."\fP).
+This option only has an effect when \fB\-\-theme\fP option is set to \fBauto\fR or \fBdark\fR.
+.HP
+\fB\-\-theme\-light\fR <theme>
+.IP
+Sets the theme name for syntax highlighting used when the terminal uses a dark background.
+To set a default theme, add the \fB\-\-theme-dark="..."\fP option to the configuration file or
+export the \fBBAT_THEME_LIGHT\fP environment variable (e.g. \fBexport BAT_THEME_LIGHT="..."\fP).
+This option only has an effect when \fB\-\-theme\fP option is set to \fBauto\fR or \fBlight\fR.
 .HP
 \fB\-\-list\-themes\fR
 .IP
 Display a list of supported themes for syntax highlighting.
 .HP
+\fB\-s\fR, \fB\-\-squeeze\-blank\fR
+.IP
+Squeeze consecutive empty lines into a single empty line.
+.HP
+\fB\-\-squeeze\-limit\fR <squeeze-limit>
+.IP
+Set the maximum number of consecutive empty lines to be printed.
+.HP
 \fB\-\-style\fR <style\-components>
 .IP
 Configure which elements (line numbers, file headers, grid borders, Git modifications,
@@ -184,6 +230,30 @@ Display a list of supported languages for syntax highlighting.
 This option exists for POSIX\-compliance reasons ('u' is for 'unbuffered'). The output is
 always unbuffered \- this option is simply ignored.
 .HP
+\fB\-\-no\-custom\-assets\fR
+.IP
+Do not load custom assets.
+.HP
+\fB\-\-config\-dir\fR
+.IP
+Show bat's configuration directory.
+.HP
+\fB\-\-cache\-dir\fR
+.IP
+Show bat's cache directory.
+.HP
+\fB\-\-diagnostic\fR
+.IP
+Show diagnostic information for bug reports.
+.HP
+\fB\-\-acknowledgements\fR
+.IP
+Show acknowledgements.
+.HP
+\fB\-\-set\-terminal\-title\fR
+.IP
+Sets terminal title to filenames when using a pager.
+.HP
 \fB\-h\fR, \fB\-\-help\fR
 .IP
 Print this help message.
@@ -212,6 +282,20 @@ location of the configuration file.
 To generate a default configuration file, call:
 
 \fB{{PROJECT_EXECUTABLE}} --generate-config-file\fR
+
+These are related options:
+.HP
+\fB\-\-config\-file\fR
+.IP
+Show path to the configuration file.
+.HP
+\fB\-\-generate-config\-file\fR
+.IP
+Generates a default configuration file.
+.HP
+\fB\-\-no\-config\fR
+.IP
+Do not use the configuration file.
 .SH "ADDING CUSTOM LANGUAGES"
 {{PROJECT_EXECUTABLE}} supports Sublime Text \fB.sublime-syntax\fR language files, and can be
 customized to add additional languages to your local installation. To do this, add the \fB.sublime-syntax\fR language
@@ -252,13 +336,23 @@ To use the preprocessor, call:
 
 \fB{{PROJECT_EXECUTABLE}} --lessopen\fR
 
-Alternatively, the preprocessor may be enabled by default by adding the '\-\-lessopen' option to the configuration file. 
+Alternatively, the preprocessor may be enabled by default by adding the '\-\-lessopen' option to the configuration file.
 
 To temporarily disable the preprocessor if it is enabled by default, call:
 
 \fB{{PROJECT_EXECUTABLE}} --no-lessopen\fR
 
-For more information, see the "INPUT PREPROCESSOR" section of less(1).  
+These are related options:
+.HP
+\fB\-\-lessopen\fR
+.IP
+Enable the $LESSOPEN preprocessor.
+.HP
+\fB\-\-no\-lessopen\fR
+.IP
+Disable the $LESSOPEN preprocessor if enabled (overrides --lessopen)
+.PP
+For more information, see the "INPUT PREPROCESSOR" section of less(1).
 
 .SH "MORE INFORMATION"
 

assets/patches/1337.tmTheme.patch

@@ -0,0 +1,22 @@
+diff --git themes/1337-Scheme/1337.tmTheme themes/1337-Scheme/1337.tmTheme
+index fdff5bf..8cfc888 100644
+--- themes/1337-Scheme/1337.tmTheme
++++ themes/1337-Scheme/1337.tmTheme
+@@ -280,7 +280,7 @@ SOFTWARE.
+ 			<key>name</key>
+ 			<string>PHP Namespaces</string>
+ 			<key>scope</key>
+-			<string>support.other.namespace, entity.name.type.namespace</string>
++			<string>support.other.namespace, entity.name.type.namespace, entity.name</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>foreground</key>
+@@ -561,7 +561,7 @@ SOFTWARE.
+ 		    <key>name</key>
+ 		    <string>diff.header</string>
+ 		    <key>scope</key>
+-		    <string>meta.diff, meta.diff.header</string>
++		    <string>meta.diff, meta.diff.header, markup.heading</string>
+ 		    <key>settings</key>
+ 		    <dict>
+ 			<key>foreground</key>

assets/patches/JavaScript.sublime-syntax.patch

@@ -0,0 +1,14 @@
+Submodule assets/syntaxes/01_Packages contains modified content
+diff --git syntaxes/01_Packages/JavaScript/JavaScript.sublime-syntax syntaxes/01_Packages/JavaScript/JavaScript.sublime-syntax
+index 05a4fed6..78a7bf55 100644
+--- syntaxes/01_Packages/JavaScript/JavaScript.sublime-syntax
++++ syntaxes/01_Packages/JavaScript/JavaScript.sublime-syntax
+@@ -5,7 +5,7 @@ name: JavaScript
+ file_extensions:
+   - js
+   - htc
+-first_line_match: ^#!\s*/.*\b(node|js)\b
++first_line_match: ^#!\s*/.*\b(node|bun|js)\b
+ scope: source.js
+ variables:
+   bin_digit: '[01_]'

assets/patches/Markdown.sublime-syntax.patch

@@ -1,5 +1,5 @@
 diff --git syntaxes/01_Packages/Markdown/Markdown.sublime-syntax syntaxes/01_Packages/Markdown/Markdown.sublime-syntax
-index 19dc685d..44440c7f 100644
+index 19dc685d..3a45ea05 100644
 --- syntaxes/01_Packages/Markdown/Markdown.sublime-syntax
 +++ syntaxes/01_Packages/Markdown/Markdown.sublime-syntax
 @@ -24,7 +24,6 @@ variables:
@@ -166,7 +166,29 @@ index 19dc685d..44440c7f 100644
      - match: ^\s*$\n?
        scope: invalid.illegal.non-terminated.bold-italic.markdown
        pop: true
-@@ -1152,7 +1110,7 @@ contexts:
+@@ -1073,6 +1031,21 @@ contexts:
+       escape: '{{code_fence_escape}}'
+       escape_captures:
+         0: meta.code-fence.definition.end.python.markdown-gfm
++        1: punctuation.definition.raw.code-fence.end.markdown
++    - match: |-
++         (?x)
++          {{fenced_code_block_start}}
++          ((?i:puppet))
++          {{fenced_code_block_trailing_infostring_characters}}
++      captures:
++        0: meta.code-fence.definition.begin.puppet.markdown-gfm
++        2: punctuation.definition.raw.code-fence.begin.markdown
++        5: constant.other.language-name.markdown
++      embed: scope:source.puppet
++      embed_scope: markup.raw.code-fence.puppet.markdown-gfm
++      escape: '{{code_fence_escape}}'
++      escape_captures:
++        0: meta.code-fence.definition.end.puppet.markdown-gfm
+         1: punctuation.definition.raw.code-fence.end.markdown
+     - match: |-
+          (?x)
+@@ -1152,7 +1125,7 @@ contexts:
      - match: |-
           (?x)
            {{fenced_code_block_start}}

assets/patches/OneHalfDark.tmTheme.patch

@@ -0,0 +1,47 @@
+diff --git themes/onehalf/sublimetext/OneHalfDark.tmTheme themes/onehalf/sublimetext/OneHalfDark.tmTheme
+index b16050c..b021071 100644
+--- themes/onehalf/sublimetext/OneHalfDark.tmTheme
++++ themes/onehalf/sublimetext/OneHalfDark.tmTheme
+@@ -28,7 +28,7 @@
+ <plist version="1.0">
+ <dict>
+ 	<key>name</key>
+-	<string>OneHalfLight</string>
++	<string>OneHalfDark</string>
+ 	<key>semanticClass</key>
+ 	<string>theme.dark.one_half_dark</string>
+ 	<key>uuid</key>
+@@ -155,7 +155,7 @@
+ 			<key>name</key>
+ 			<string>Classes</string>
+ 			<key>scope</key>
+-			<string>support.class, entity.name.class, entity.name.type.class</string>
++			<string>support.class, entity.name.class, entity.name.type.class, entity.name</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>foreground</key>
+@@ -188,7 +188,7 @@
+ 			<key>name</key>
+ 			<string>Storage</string>
+ 			<key>scope</key>
+-			<string>storage</string>
++			<string>storage, meta.mapping.key string</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>foreground</key>
+@@ -309,7 +309,7 @@
+ 			<key>name</key>
+ 			<string>Markdown: Headings</string>
+ 			<key>scope</key>
+-			<string>markup.heading punctuation.definition.heading, entity.name.section</string>
++			<string>markup.heading punctuation.definition.heading, entity.name.section, markup.heading - text.html.markdown</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>fontStyle</key>
+@@ -660,4 +660,4 @@
+ 		</dict>
+ 	</array>
+ </dict>
+-</plist>
+\ No newline at end of file
++</plist>

assets/patches/TwoDark.tmTheme.patch

@@ -2,6 +2,24 @@ diff --git themes/TwoDark/TwoDark.tmTheme themes/TwoDark/TwoDark.tmTheme
 index 87fd358..56376d3 100644
 --- themes/TwoDark/TwoDark.tmTheme
 +++ themes/TwoDark/TwoDark.tmTheme
+@@ -125,7 +125,7 @@
+ 			<key>name</key>
+ 			<string>Classes</string>
+ 			<key>scope</key>
+-			<string>support.class, entity.name.class, entity.name.type.class</string>
++			<string>support.class, entity.name.class, entity.name.type.class, entity.name</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>foreground</key>
+@@ -290,7 +290,7 @@
+ 			<key>name</key>
+ 			<string>Headings</string>
+ 			<key>scope</key>
+-			<string>markup.heading punctuation.definition.heading, entity.name.section</string>
++			<string>markup.heading punctuation.definition.heading, entity.name.section, markup.heading - text.html.markdown</string>
+ 			<key>settings</key>
+ 			<dict>
+ 				<key>fontStyle</key>
 @@ -533,7 +533,7 @@
  			<key>name</key>
  			<string>Json key</string>

assets/syntaxes/02_Extra/CSV.sublime-syntax

@@ -7,14 +7,14 @@ file_extensions:
   - tsv
 scope: text.csv
 variables:
-  field_separator: (?:[,;\t])
+  field_separator: (?:[,;|\t])
   record_separator: (?:$\n?)
 contexts:
   prototype:
     - match: (?={{record_separator}})
       pop: true
   fields:
-    - match: ''
+    - match: ""
       push:
         - field_or_record_separator
         - field4
@@ -26,15 +26,15 @@ contexts:
         - field1
   main:
     - meta_include_prototype: false
-    - match: '^'
+    - match: "^"
       set: fields
 
   field_or_record_separator:
     - meta_include_prototype: false
-    - match: '{{record_separator}}'
+    - match: "{{record_separator}}"
       scope: punctuation.terminator.record.csv
       pop: true
-    - match: '{{field_separator}}'
+    - match: "{{field_separator}}"
       scope: punctuation.separator.sequence.csv
       pop: true
 
@@ -56,23 +56,22 @@ contexts:
       pop: true
 
   field1:
-    - match: ''
+    - match: ""
       set:
         - meta_content_scope: meta.field-1.csv support.type
         - include: field_contents
   field2:
-    - match: ''
+    - match: ""
       set:
         - meta_content_scope: meta.field-2.csv support.function
         - include: field_contents
   field3:
-    - match: ''
+    - match: ""
       set:
         - meta_content_scope: meta.field-3.csv constant.numeric
         - include: field_contents
   field4:
-    - match: ''
+    - match: ""
       set:
         - meta_content_scope: meta.field-4.csv keyword.operator
         - include: field_contents
-

assets/syntaxes/02_Extra/Manpage.sublime-syntax

@@ -85,6 +85,9 @@ contexts:
 
   options:
     # command-line options like --option=value, --some-flag, or -x
+    - match: '^[ ]{7}(-)(?=\s)'
+      captures:
+        1: entity.name.command-line-option.man
     - match: '^[ ]{7}(?=-|\+)'
       push: expect-command-line-option
     - match: '(?:[^a-zA-Z0-9_-]|^|\s){{command_line_option}}'

assets/syntaxes/02_Extra/syntax_test_man.man

@@ -131,6 +131,12 @@ OPTIONS
 #                                ^^ - variable
               output NUM (default 3) lines of copied context
 
+       -      This is  not really a switch, but indicates that standard input
+#      ^ entity.name.command-line-option.man
+#       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - entity - variable
+              is coming from a file or a pipe and not interactively from the
+              command line.
+
 EXAMPLE
        #include <stdio.h>
 #      ^^^^^^^^ source.c meta.preprocessor.include keyword.control.import.include

assets/themes/ansi.tmTheme

@@ -69,7 +69,7 @@
                 <key>name</key>
                 <string>Labels</string>
                 <key>scope</key>
-                <string>entity.name.label</string>
+                <string>entity.name.label, variable.parameter</string>
                 <key>settings</key>
                 <dict>
                     <key>foreground</key>
@@ -80,7 +80,7 @@
                 <key>name</key>
                 <string>Classes</string>
                 <key>scope</key>
-                <string>support.class, entity.name.class, entity.name.type.class</string>
+                <string>support.class, entity.name.class, entity.name.type.class, entity.name</string>
                 <key>settings</key>
                 <dict>
                     <key>foreground</key>
@@ -234,7 +234,7 @@
                 <key>name</key>
                 <string>Headings</string>
                 <key>scope</key>
-                <string>markup.heading punctuation.definition.heading, entity.name.section</string>
+                <string>markup.heading punctuation.definition.heading, entity.name.section, markup.heading - text.html.markdown</string>
                 <key>settings</key>
                 <dict>
                     <key>fontStyle</key>

assets/themes/base16-256.tmTheme

@@ -257,7 +257,7 @@
                 <key>name</key>
                 <string>Tags</string>
                 <key>scope</key>
-                <string>entity.name.tag</string>
+                <string>entity.name.tag, entity.name</string>
                 <key>settings</key>
                 <dict>
                     <key>foreground</key>
@@ -312,7 +312,7 @@
                 <key>name</key>
                 <string>Headings</string>
                 <key>scope</key>
-                <string>markup.heading punctuation.definition.heading, entity.name.section</string>
+                <string>markup.heading punctuation.definition.heading, entity.name.section, markup.heading - text.html.markdown</string>
                 <key>settings</key>
                 <dict>
                     <key>fontStyle</key>

assets/themes/base16.tmTheme

@@ -256,7 +256,7 @@
                 <key>name</key>
                 <string>Tags</string>
                 <key>scope</key>
-                <string>entity.name.tag</string>
+                <string>entity.name.tag, entity.name</string>
                 <key>settings</key>
                 <dict>
                     <key>foreground</key>
@@ -311,7 +311,7 @@
                 <key>name</key>
                 <string>Headings</string>
                 <key>scope</key>
-                <string>markup.heading punctuation.definition.heading, entity.name.section</string>
+                <string>markup.heading punctuation.definition.heading, entity.name.section, markup.heading - text.html.markdown</string>
                 <key>settings</key>
                 <dict>
                     <key>fontStyle</key>

build/application.rs

@@ -63,5 +63,22 @@ pub fn gen_man_and_comp() -> anyhow::Result<()> {
         out_dir.join("assets/completions/bat.zsh"),
     )?;
 
+    println!(
+        "cargo:rustc-env=BAT_GENERATED_COMPLETION_BASH={}",
+        out_dir.join("assets/completions/bat.bash").display()
+    );
+    println!(
+        "cargo:rustc-env=BAT_GENERATED_COMPLETION_FISH={}",
+        out_dir.join("assets/completions/bat.fish").display()
+    );
+    println!(
+        "cargo:rustc-env=BAT_GENERATED_COMPLETION_PS1={}",
+        out_dir.join("assets/completions/_bat.ps1").display()
+    );
+    println!(
+        "cargo:rustc-env=BAT_GENERATED_COMPLETION_ZSH={}",
+        out_dir.join("assets/completions/bat.zsh").display()
+    );
+
     Ok(())
 }

build/main.rs

@@ -1,5 +1,6 @@
 #[cfg(feature = "application")]
 mod application;
+mod syntax_mapping;
 mod util;
 
 fn main() -> anyhow::Result<()> {
@@ -7,6 +8,8 @@ fn main() -> anyhow::Result<()> {
     // see: https://doc.rust-lang.org/cargo/reference/build-scripts.html#rerun-if-changed
     println!("cargo:rerun-if-changed=build/");
 
+    syntax_mapping::build_static_mappings()?;
+
     #[cfg(feature = "application")]
     application::gen_man_and_comp()?;
 

build/syntax_mapping.rs

@@ -0,0 +1,294 @@
+use std::{
+    convert::Infallible,
+    env, fs,
+    path::{Path, PathBuf},
+    str::FromStr,
+};
+
+use anyhow::{anyhow, bail};
+use indexmap::IndexMap;
+use itertools::Itertools;
+use once_cell::sync::Lazy;
+use regex::Regex;
+use serde_derive::Deserialize;
+use serde_with::DeserializeFromStr;
+use walkdir::WalkDir;
+
+/// Known mapping targets.
+///
+/// Corresponds to `syntax_mapping::MappingTarget`.
+#[allow(clippy::enum_variant_names)]
+#[derive(Clone, Debug, Eq, PartialEq, Hash, DeserializeFromStr)]
+pub enum MappingTarget {
+    MapTo(String),
+    MapToUnknown,
+    MapExtensionToUnknown,
+}
+impl FromStr for MappingTarget {
+    type Err = Infallible;
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        match s {
+            "MappingTarget::MapToUnknown" => Ok(Self::MapToUnknown),
+            "MappingTarget::MapExtensionToUnknown" => Ok(Self::MapExtensionToUnknown),
+            syntax => Ok(Self::MapTo(syntax.into())),
+        }
+    }
+}
+impl MappingTarget {
+    fn codegen(&self) -> String {
+        match self {
+            Self::MapTo(syntax) => format!(r###"MappingTarget::MapTo(r#"{syntax}"#)"###),
+            Self::MapToUnknown => "MappingTarget::MapToUnknown".into(),
+            Self::MapExtensionToUnknown => "MappingTarget::MapExtensionToUnknown".into(),
+        }
+    }
+}
+
+#[derive(Clone, Debug, PartialEq, Eq, Hash, DeserializeFromStr)]
+/// A single matcher.
+///
+/// Codegen converts this into a `Lazy<Option<GlobMatcher>>`.
+struct Matcher(Vec<MatcherSegment>);
+/// Parse a matcher.
+///
+/// Note that this implementation is rather strict: it will greedily interpret
+/// every valid environment variable replacement as such, then immediately
+/// hard-error if it finds a '$' anywhere in the remaining text segments.
+///
+/// The reason for this strictness is I currently cannot think of a valid reason
+/// why you would ever need '$' as plaintext in a glob pattern. Therefore any
+/// such occurrences are likely human errors.
+///
+/// If we later discover some edge cases, it's okay to make it more permissive.
+///
+/// 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());
+
+        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();
+
+            // 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(Self(non_empty_segments))
+    }
+}
+impl Matcher {
+    fn codegen(&self) -> String {
+        match self.0.len() {
+            0 => unreachable!("0-length matcher should never be created"),
+            // if-let guard would be ideal here
+            // see: https://github.com/rust-lang/rust/issues/51114
+            1 if self.0[0].is_text() => {
+                let s = self.0[0].text().unwrap();
+                format!(r###"Lazy::new(|| Some(build_matcher_fixed(r#"{s}"#)))"###)
+            }
+            // parser logic ensures that this case can only happen when there are dynamic segments
+            _ => {
+                let segs = self.0.iter().map(MatcherSegment::codegen).join(", ");
+                format!(r###"Lazy::new(|| build_matcher_dynamic(&[{segs}]))"###)
+            }
+        }
+    }
+}
+
+/// A segment in a matcher.
+///
+/// Corresponds to `syntax_mapping::MatcherSegment`.
+#[derive(Debug, Clone, PartialEq, Eq, Hash)]
+enum MatcherSegment {
+    Text(String),
+    Env(String),
+}
+#[allow(dead_code)]
+impl MatcherSegment {
+    fn is_text(&self) -> bool {
+        matches!(self, Self::Text(_))
+    }
+    fn is_env(&self) -> bool {
+        matches!(self, Self::Env(_))
+    }
+    fn text(&self) -> Option<&str> {
+        match self {
+            Self::Text(t) => Some(t),
+            Self::Env(_) => None,
+        }
+    }
+    fn env(&self) -> Option<&str> {
+        match self {
+            Self::Text(_) => None,
+            Self::Env(t) => Some(t),
+        }
+    }
+    fn codegen(&self) -> String {
+        match self {
+            Self::Text(s) => format!(r###"MatcherSegment::Text(r#"{s}"#)"###),
+            Self::Env(s) => format!(r###"MatcherSegment::Env(r#"{s}"#)"###),
+        }
+    }
+}
+
+/// A struct that models a single .toml file in /src/syntax_mapping/builtins/.
+#[derive(Clone, Debug, Deserialize)]
+struct MappingDefModel {
+    mappings: IndexMap<MappingTarget, Vec<Matcher>>,
+}
+impl MappingDefModel {
+    fn into_mapping_list(self) -> MappingList {
+        let list = self
+            .mappings
+            .into_iter()
+            .flat_map(|(target, matchers)| {
+                matchers
+                    .into_iter()
+                    .map(|matcher| (matcher, target.clone()))
+                    .collect::<Vec<_>>()
+            })
+            .collect();
+        MappingList(list)
+    }
+}
+
+#[derive(Clone, Debug)]
+struct MappingList(Vec<(Matcher, MappingTarget)>);
+impl MappingList {
+    fn codegen(&self) -> String {
+        let array_items: Vec<_> = self
+            .0
+            .iter()
+            .map(|(matcher, target)| {
+                format!("({m}, {t})", m = matcher.codegen(), t = target.codegen())
+            })
+            .collect();
+        let len = array_items.len();
+
+        format!(
+            "/// Generated by build script from /src/syntax_mapping/builtins/.\n\
+            pub(crate) static BUILTIN_MAPPINGS: [(Lazy<Option<GlobMatcher>>, MappingTarget); {len}] = [\n{items}\n];",
+            items = array_items.join(",\n")
+        )
+    }
+}
+
+/// Get the list of paths to all mapping definition files that should be
+/// included for the current target platform.
+fn get_def_paths() -> anyhow::Result<Vec<PathBuf>> {
+    let source_subdirs = [
+        "common",
+        #[cfg(target_family = "unix")]
+        "unix-family",
+        #[cfg(any(
+            target_os = "freebsd",
+            target_os = "netbsd",
+            target_os = "openbsd",
+            target_os = "macos"
+        ))]
+        "bsd-family",
+        #[cfg(target_os = "linux")]
+        "linux",
+        #[cfg(target_os = "macos")]
+        "macos",
+        #[cfg(target_os = "windows")]
+        "windows",
+    ];
+
+    let mut toml_paths = vec![];
+    for subdir in source_subdirs {
+        let wd = WalkDir::new(Path::new("src/syntax_mapping/builtins").join(subdir));
+        let paths = wd
+            .into_iter()
+            .filter_map_ok(|entry| {
+                let path = entry.path();
+                (path.is_file() && path.extension().map(|ext| ext == "toml").unwrap_or(false))
+                    .then(|| path.to_owned())
+            })
+            .collect::<Result<Vec<_>, _>>()?;
+        toml_paths.extend(paths);
+    }
+
+    toml_paths.sort_by_key(|path| {
+        path.file_name()
+            .expect("file name should not terminate in ..")
+            .to_owned()
+    });
+
+    Ok(toml_paths)
+}
+
+fn read_all_mappings() -> anyhow::Result<MappingList> {
+    let mut all_mappings = vec![];
+
+    for path in get_def_paths()? {
+        let toml_string = fs::read_to_string(path)?;
+        let mappings = toml::from_str::<MappingDefModel>(&toml_string)?.into_mapping_list();
+        all_mappings.extend(mappings.0);
+    }
+
+    let duplicates = all_mappings
+        .iter()
+        .duplicates_by(|(matcher, _)| matcher)
+        .collect_vec();
+    if !duplicates.is_empty() {
+        bail!("Rules with duplicate matchers found: {duplicates:?}");
+    }
+
+    Ok(MappingList(all_mappings))
+}
+
+/// Build the static syntax mappings defined in /src/syntax_mapping/builtins/
+/// into a .rs source file, which is to be inserted with `include!`.
+pub fn build_static_mappings() -> anyhow::Result<()> {
+    println!("cargo:rerun-if-changed=src/syntax_mapping/builtins/");
+
+    let mappings = read_all_mappings()?;
+
+    let codegen_path = Path::new(&env::var_os("OUT_DIR").ok_or(anyhow!("OUT_DIR is unset"))?)
+        .join("codegen_static_syntax_mappings.rs");
+
+    fs::write(codegen_path, mappings.codegen())?;
+
+    Ok(())
+}

doc/README-ja.md

@@ -366,7 +366,7 @@ ansible-galaxy install aeimer.install_bat
 ### From source
 
 
-`bat` をソースからビルドしたいならば、Rust 1.70.0 以上の環境が必要です。
+`bat` をソースからビルドしたいならば、Rust 1.74.0 以上の環境が必要です。
 `cargo` を使用してビルドすることができます:
 
 ```bash

doc/README-ko.md

@@ -416,7 +416,7 @@ scoop install bat
 
 ### 소스에서
 
-`bat`의 소스를 빌드하기 위해서는, Rust 1.70.0 이상이 필요합니다.
+`bat`의 소스를 빌드하기 위해서는, Rust 1.74.0 이상이 필요합니다.
 `cargo`를 이용해 전부 빌드할 수 있습니다:
 
 ```bash

doc/README-ru.md

@@ -344,7 +344,7 @@ ansible-galaxy install aeimer.install_bat
 
 ### Из исходников
 
-Если вы желаете установить `bat` из исходников, вам понадобится Rust 1.70.0 или выше. После этого используйте `cargo`, чтобы все скомпилировать:
+Если вы желаете установить `bat` из исходников, вам понадобится Rust 1.74.0 или выше. После этого используйте `cargo`, чтобы все скомпилировать:
 
 ```bash
 cargo install --locked bat

doc/README-zh.md

@@ -372,7 +372,7 @@ scoop install bat
 
 ### 从源码编译
 
-如果你想要自己构建`bat`，那么你需要安装有高于1.70.0版本的 Rust。
+如果你想要自己构建`bat`，那么你需要安装有高于1.74.0版本的 Rust。
 
 使用以下命令编译。
 
@@ -412,7 +412,7 @@ bat --list-themes | fzf --preview="bat --theme={} --color=always /path/to/file"
 
 ### 输出样式
 
-你可以用`--style`参数来控制`bat`输出的样式。使用`--style=numbers,chanegs`可以只开启 Git 修改和行号显示而不添加其他内容。`BAT_STYLE`环境变量具有相同功能。
+你可以用`--style`参数来控制`bat`输出的样式。使用`--style=numbers,changes`可以只开启 Git 修改和行号显示而不添加其他内容。`BAT_STYLE`环境变量具有相同功能。
 
 ### 添加新的语言和语法
 
@@ -616,63 +616,59 @@ iconv -f ISO-8859-1 -t UTF-8 my-file.php | bat
 
 注意： 当`bat`无法识别语言时你可能会需要`-l`/`--language`参数。
 
-## Development
+## 开发
 
 ```bash
-# Recursive clone to retrieve all submodules
+# 递归 clone 以获取所有子模块
 git clone --recursive https://github.com/sharkdp/bat
 
-# Build (debug version)
+# 构建（调试版本）
 cd bat
 cargo build --bins
 
-# Run unit tests and integration tests
+# 运行单元测试和集成测试
 cargo test
 
-# Install (release version)
+# 安装（发布版本）
 cargo install --path . --locked
 
-# Build a bat binary with modified syntaxes and themes
+# 使用修改后的语法和主题构建一个 bat 二进制文件
 bash assets/create.sh
 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/).
-Note that you have to use either `regex-onig` or `regex-fancy` as a feature
-when you depend on `bat` as a library.
+如果你想构建一个使用 `bat` 美化打印功能的应用程序，请查看 [API 文档](https://docs.rs/bat/)。请注意，当你依赖 `bat` 作为库时，必须使用 `regex-onig` 或 `regex-fancy` 作为特性。
 
-## Contributing
+## 贡献指南
 
-Take a look at the [`CONTRIBUTING.md`](CONTRIBUTING.md) guide.
+请查看 [`CONTRIBUTING.md`](CONTRIBUTING.md) 指南。
 
-## Maintainers
+## 维护者
 
 - [sharkdp](https://github.com/sharkdp)
 - [eth-p](https://github.com/eth-p)
 - [keith-hall](https://github.com/keith-hall)
 - [Enselic](https://github.com/Enselic)
 
-## Security vulnerabilities
+## 安全漏洞
 
-Please contact [David Peter](https://david-peter.de/) via email if you want to report a vulnerability in `bat`.
+如果你想报告 `bat` 中的漏洞，请通过邮件联系 [David Peter](https://david-peter.de/)。
 
-## Project goals and alternatives
+## 项目目标和替代方案
 
-`bat` tries to achieve the following goals:
+`bat` 试图实现以下目标：
 
-- Provide beautiful, advanced syntax highlighting
-- Integrate with Git to show file modifications
-- Be a drop-in replacement for (POSIX) `cat`
-- Offer a user-friendly command-line interface
+- 提供美观的高级语法高亮
+- 与 Git 集成以显示文件修改
+- 成为 (POSIX) `cat` 的替代品
+- 提供用户友好的命令行界面
 
-There are a lot of alternatives, if you are looking for similar programs. See
-[this document](doc/alternatives.md) for a comparison.
+如果你在寻找类似的程序，有很多替代方案。请参阅[本文档](doc/alternatives.md)进行比较。
 
-## License
+## 许可证
 
-Copyright (c) 2018-2021 [bat-developers](https://github.com/sharkdp/bat).
+版权所有 (c) 2018-2021 [bat-developers](https://github.com/sharkdp/bat)。
 
-`bat` is made available under the terms of either the MIT License or the Apache License 2.0, at your option.
+`bat` 可根据 MIT 许可证或 Apache 许可证 2.0 的条款使用，任选其一。
 
-See the [LICENSE-APACHE](LICENSE-APACHE) and [LICENSE-MIT](LICENSE-MIT) files for license details.
+有关许可证的详细信息，请参阅 [LICENSE-APACHE](LICENSE-APACHE) 和 [LICENSE-MIT](LICENSE-MIT) 文件。

doc/long-help.txt

@@ -20,6 +20,13 @@ Options:
             * unicode (␇, ␊, ␀, ..)
             * caret   (^G, ^J, ^@, ..)
 
+      --binary <behavior>
+          How to treat binary content. (default: no-printing)
+          
+          Possible values:
+            * no-printing: do not print any binary content
+            * as-text: treat binary content as normal text
+
   -p, --plain...
           Only show plain style, no decorations. This is an alias for '--style=plain'. When '-p' is
           used twice ('-pp'), it also disables automatic paging (alias for '--style=plain
@@ -112,10 +119,42 @@ Options:
           Set the theme for syntax highlighting. Use '--list-themes' to see all available themes. To
           set a default theme, add the '--theme="..."' option to the configuration file or export
           the BAT_THEME environment variable (e.g.: export BAT_THEME="...").
+          
+          Special values:
+          
+            * auto: Picks a dark or light theme depending on the terminal's colors (default).
+                    Use '--theme-light' and '--theme-dark' to customize the selected theme.
+              * auto:always: Detect the terminal's colors even when the output is redirected.
+              * auto:system: Detect the color scheme from the system-wide preference (macOS only).
+            * dark: Use the dark theme specified by '--theme-dark'.
+            * light: Use the light theme specified by '--theme-light'.
+
+      --theme-light <theme>
+          Sets the theme name for syntax highlighting used when the terminal uses a light
+          background. Use '--list-themes' to see all available themes. To set a default theme, add
+          the '--theme-light="..." option to the configuration file or export the BAT_THEME_LIGHT
+          environment variable (e.g. export BAT_THEME_LIGHT="...").
+
+      --theme-dark <theme>
+          Sets the theme name for syntax highlighting used when the terminal uses a dark background.
+          Use '--list-themes' to see all available themes. To set a default theme, add the
+          '--theme-dark="..." option to the configuration file or export the BAT_THEME_DARK
+          environment variable (e.g. export BAT_THEME_DARK="...").
 
       --list-themes
           Display a list of supported themes for syntax highlighting.
 
+  -s, --squeeze-blank
+          Squeeze consecutive empty lines into a single empty line.
+
+      --squeeze-limit <squeeze-limit>
+          Set the maximum number of consecutive empty lines to be printed.
+
+      --strip-ansi <when>
+          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. 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
@@ -123,6 +162,15 @@ Options:
           set a default style, add the '--style=".."' option to the configuration file or export the
           BAT_STYLE environment variable (e.g.: export BAT_STYLE="..").
           
+          When styles are specified in multiple places, the "nearest" set of styles take precedence.
+          The command-line arguments are the highest priority, followed by the BAT_STYLE environment
+          variable, and then the configuration file. If any set of styles consists entirely of
+          components prefixed with "+" or "-", it will modify the previous set of styles instead of
+          replacing them.
+          
+          By default, the following components are enabled:
+            changes, grid, header-filename, numbers, snip
+          
           Possible values:
           
             * default: enables recommended style components (default).
@@ -154,12 +202,18 @@ Options:
           This option exists for POSIX-compliance reasons ('u' is for 'unbuffered'). The output is
           always unbuffered - this option is simply ignored.
 
+      --completion <SHELL>
+          Show shell completion for a certain shell. [possible values: bash, fish, zsh, ps1]
+
       --diagnostic
           Show diagnostic information for bug reports.
 
       --acknowledgements
           Show acknowledgements.
 
+      --set-terminal-title
+          Sets terminal title to filenames when using a pager.
+
   -h, --help
           Print help (see a summary with '-h')
 

doc/short-help.txt

@@ -11,6 +11,8 @@ Options:
           Show non-printable characters (space, tab, newline, ..).
       --nonprintable-notation <notation>
           Set notation for non-printable characters.
+      --binary <behavior>
+          How to treat binary content. (default: no-printing)
   -p, --plain...
           Show plain style (alias for '--style=plain').
   -l, --language <language>
@@ -41,8 +43,14 @@ Options:
           Use the specified syntax for files matching the glob pattern ('*.cpp:C++').
       --theme <theme>
           Set the color theme for syntax highlighting.
+      --theme-light <theme>
+          Sets the color theme for syntax highlighting used for light backgrounds.
+      --theme-dark <theme>
+          Sets the color theme for syntax highlighting used for dark backgrounds.
       --list-themes
           Display all supported highlighting themes.
+  -s, --squeeze-blank
+          Squeeze consecutive empty lines.
       --style <components>
           Comma-separated list of style elements to display (*default*, auto, full, plain, changes,
           header, header-filename, header-filesize, grid, rule, numbers, snip).
@@ -50,6 +58,8 @@ Options:
           Only print the lines from N to M.
   -L, --list-languages
           Display all supported languages.
+      --completion <SHELL>
+          Show shell completion for a certain shell. [possible values: bash, fish, zsh, ps1]
   -h, --help
           Print help (see more with '--help')
   -V, --version

examples/list_syntaxes_and_themes.rs

@@ -13,6 +13,6 @@ fn main() {
 
     println!("Themes:");
     for theme in printer.themes() {
-        println!("- {}", theme);
+        println!("- {theme}");
     }
 }

src/assets.rs

@@ -13,6 +13,7 @@ use crate::error::*;
 use crate::input::{InputReader, OpenedInput};
 use crate::syntax_mapping::ignored_suffixes::IgnoredSuffixes;
 use crate::syntax_mapping::MappingTarget;
+use crate::theme::{default_theme, ColorScheme};
 use crate::{bat_warning, SyntaxMapping};
 
 use lazy_theme_set::LazyThemeSet;
@@ -69,57 +70,6 @@ impl HighlightingAssets {
         }
     }
 
-    /// The default theme.
-    ///
-    /// ### Windows and Linux
-    ///
-    /// Windows and most Linux distributions has a dark terminal theme by
-    /// default. On these platforms, this function always returns a theme that
-    /// looks good on a dark background.
-    ///
-    /// ### macOS
-    ///
-    /// On macOS the default terminal background is light, but it is common that
-    /// Dark Mode is active, which makes the terminal background dark. On this
-    /// platform, the default theme depends on
-    /// ```bash
-    /// defaults read -globalDomain AppleInterfaceStyle
-    /// ```
-    /// To avoid the overhead of the check on macOS, simply specify a theme
-    /// explicitly via `--theme`, `BAT_THEME`, or `~/.config/bat`.
-    ///
-    /// See <https://github.com/sharkdp/bat/issues/1746> and
-    /// <https://github.com/sharkdp/bat/issues/1928> for more context.
-    pub fn default_theme() -> &'static str {
-        #[cfg(not(target_os = "macos"))]
-        {
-            Self::default_dark_theme()
-        }
-        #[cfg(target_os = "macos")]
-        {
-            if macos_dark_mode_active() {
-                Self::default_dark_theme()
-            } else {
-                Self::default_light_theme()
-            }
-        }
-    }
-
-    /**
-     * The default theme that looks good on a dark background.
-     */
-    fn default_dark_theme() -> &'static str {
-        "Monokai Extended"
-    }
-
-    /**
-     * The default theme that looks good on a light background.
-     */
-    #[cfg(target_os = "macos")]
-    fn default_light_theme() -> &'static str {
-        "Monokai Extended Light"
-    }
-
     pub fn from_cache(cache_path: &Path) -> Result<Self> {
         Ok(HighlightingAssets::new(
             SerializedSyntaxSet::FromFile(cache_path.join("syntaxes.bin")),
@@ -248,7 +198,10 @@ impl HighlightingAssets {
                     bat_warning!("Unknown theme '{}', using default.", theme)
                 }
                 self.get_theme_set()
-                    .get(self.fallback_theme.unwrap_or_else(Self::default_theme))
+                    .get(
+                        self.fallback_theme
+                            .unwrap_or_else(|| default_theme(ColorScheme::Dark)),
+                    )
                     .expect("something is very wrong if the default theme is missing")
             }
         }
@@ -380,7 +333,7 @@ fn asset_from_contents<T: serde::de::DeserializeOwned>(
     } else {
         bincode::deserialize_from(contents)
     }
-    .map_err(|_| format!("Could not parse {}", description).into())
+    .map_err(|_| format!("Could not parse {description}").into())
 }
 
 fn asset_from_cache<T: serde::de::DeserializeOwned>(
@@ -396,27 +349,7 @@ fn asset_from_cache<T: serde::de::DeserializeOwned>(
         )
     })?;
     asset_from_contents(&contents[..], description, compressed)
-        .map_err(|_| format!("Could not parse cached {}", description).into())
-}
-
-#[cfg(target_os = "macos")]
-fn macos_dark_mode_active() -> bool {
-    const PREFERENCES_FILE: &str = "Library/Preferences/.GlobalPreferences.plist";
-    const STYLE_KEY: &str = "AppleInterfaceStyle";
-
-    let preferences_file = home::home_dir()
-        .map(|home| home.join(PREFERENCES_FILE))
-        .expect("Could not get home directory");
-
-    match plist::Value::from_file(preferences_file).map(|file| file.into_dictionary()) {
-        Ok(Some(preferences)) => match preferences.get(STYLE_KEY).and_then(|val| val.as_string()) {
-            Some(value) => value == "Dark",
-            // If the key does not exist, then light theme is currently in use.
-            None => false,
-        },
-        // Unreachable, in theory. All macOS users have a home directory and preferences file setup.
-        Ok(None) | Err(_) => true,
-    }
+        .map_err(|_| format!("Could not parse cached {description}").into())
 }
 
 #[cfg(test)]
@@ -441,7 +374,7 @@ mod tests {
         fn new() -> Self {
             SyntaxDetectionTest {
                 assets: HighlightingAssets::from_binary(),
-                syntax_mapping: SyntaxMapping::builtin(),
+                syntax_mapping: SyntaxMapping::new(),
                 temp_dir: TempDir::new().expect("creation of temporary directory"),
             }
         }
@@ -466,7 +399,7 @@ mod tests {
             let file_path = self.temp_dir.path().join(file_name);
             {
                 let mut temp_file = File::create(&file_path).unwrap();
-                writeln!(temp_file, "{}", first_line).unwrap();
+                writeln!(temp_file, "{first_line}").unwrap();
             }
 
             let input = Input::ordinary_file(&file_path);
@@ -514,8 +447,7 @@ mod tests {
 
             if !consistent {
                 eprintln!(
-                    "Inconsistent syntax detection:\nFor File: {}\nFor Reader: {}",
-                    as_file, as_reader
+                    "Inconsistent syntax detection:\nFor File: {as_file}\nFor Reader: {as_reader}"
                 )
             }
 

src/assets/assets_metadata.rs

@@ -3,7 +3,7 @@ use std::path::Path;
 use std::time::SystemTime;
 
 use semver::Version;
-use serde::{Deserialize, Serialize};
+use serde_derive::{Deserialize, Serialize};
 
 use crate::error::*;
 

src/assets/build_assets.rs

@@ -93,7 +93,7 @@ fn print_unlinked_contexts(syntax_set: &SyntaxSet) {
     if !missing_contexts.is_empty() {
         println!("Some referenced contexts could not be found!");
         for context in missing_contexts {
-            println!("- {}", context);
+            println!("- {context}");
         }
     }
 }
@@ -152,7 +152,7 @@ pub(crate) fn asset_to_contents<T: serde::Serialize>(
     } else {
         bincode::serialize_into(&mut contents, asset)
     }
-    .map_err(|_| format!("Could not serialize {}", description))?;
+    .map_err(|_| format!("Could not serialize {description}"))?;
     Ok(contents)
 }
 

src/assets/build_assets/acknowledgements.rs

@@ -80,7 +80,7 @@ fn handle_license(path: &Path) -> Result<Option<String>> {
     } else if license_not_needed_in_acknowledgements(&license_text) {
         Ok(None)
     } else {
-        Err(format!("ERROR: License is of unknown type: {:?}", path).into())
+        Err(format!("ERROR: License is of unknown type: {path:?}").into())
     }
 }
 
@@ -125,7 +125,7 @@ fn append_to_acknowledgements(
     relative_path: &str,
     license_text: &str,
 ) {
-    write!(acknowledgements, "## {}\n\n{}", relative_path, license_text).ok();
+    write!(acknowledgements, "## {relative_path}\n\n{license_text}").ok();
 
     // Make sure the last char is a newline to not mess up formatting later
     if acknowledgements

src/assets/lazy_theme_set.rs

@@ -3,8 +3,7 @@ use super::*;
 use std::collections::BTreeMap;
 use std::convert::TryFrom;
 
-use serde::Deserialize;
-use serde::Serialize;
+use serde_derive::{Deserialize, Serialize};
 
 use once_cell::unsync::OnceCell;
 
@@ -89,7 +88,7 @@ impl TryFrom<ThemeSet> for LazyThemeSet {
             let lazy_theme = LazyTheme {
                 serialized: crate::assets::build_assets::asset_to_contents(
                     &theme,
-                    &format!("theme {}", name),
+                    &format!("theme {name}"),
                     COMPRESS_LAZY_THEMES,
                 )?,
                 deserialized: OnceCell::new(),

src/bin/bat/app.rs

@@ -2,18 +2,22 @@ use std::collections::HashSet;
 use std::env;
 use std::io::IsTerminal;
 use std::path::{Path, PathBuf};
+use std::str::FromStr;
 
 use crate::{
     clap_app,
     config::{get_args_from_config_file, get_args_from_env_opts_var, get_args_from_env_vars},
 };
+use bat::style::StyleComponentList;
+use bat::theme::{theme, ThemeName, ThemeOptions, ThemePreference};
+use bat::BinaryBehavior;
+use bat::StripAnsiMode;
 use clap::ArgMatches;
 
 use console::Term;
 
 use crate::input::{new_file_input, new_stdin_input};
 use bat::{
-    assets::HighlightingAssets,
     bat_warning,
     config::{Config, VisibleLines},
     error::*,
@@ -85,7 +89,6 @@ impl App {
 
             // .. and the rest at the end
             cli_args.for_each(|a| args.push(a));
-
             args
         };
 
@@ -95,12 +98,30 @@ impl App {
     pub fn config(&self, inputs: &[Input]) -> Result<Config> {
         let style_components = self.style_components()?;
 
+        let extra_plain = self.matches.get_count("plain") > 1;
+        let plain_last_index = self
+            .matches
+            .indices_of("plain")
+            .and_then(Iterator::max)
+            .unwrap_or_default();
+        let paging_last_index = self
+            .matches
+            .indices_of("paging")
+            .and_then(Iterator::max)
+            .unwrap_or_default();
+
         let paging_mode = match self.matches.get_one::<String>("paging").map(|s| s.as_str()) {
-            Some("always") => PagingMode::Always,
+            Some("always") => {
+                // Disable paging if the second -p (or -pp) is specified after --paging=always
+                if extra_plain && plain_last_index > paging_last_index {
+                    PagingMode::Never
+                } else {
+                    PagingMode::Always
+                }
+            }
             Some("never") => PagingMode::Never,
             Some("auto") | None => {
                 // If we have -pp as an option when in auto mode, the pager should be disabled.
-                let extra_plain = self.matches.get_count("plain") > 1;
                 if extra_plain || self.matches.get_flag("no-paging") {
                     PagingMode::Never
                 } else if inputs.iter().any(Input::is_stdin) {
@@ -121,7 +142,11 @@ impl App {
             _ => unreachable!("other values for --paging are not allowed"),
         };
 
-        let mut syntax_mapping = SyntaxMapping::builtin();
+        let mut syntax_mapping = SyntaxMapping::new();
+        // start building glob matchers for builtin mappings immediately
+        // this is an appropriate approach because it's statistically likely that
+        // all the custom mappings need to be checked
+        syntax_mapping.start_offload_build_all();
 
         if let Some(values) = self.matches.get_many::<String>("ignored-suffix") {
             for suffix in values {
@@ -130,7 +155,9 @@ impl App {
         }
 
         if let Some(values) = self.matches.get_many::<String>("map-syntax") {
-            for from_to in values {
+            // later args take precedence over earlier ones, hence `.rev()`
+            // see: https://github.com/sharkdp/bat/pull/2755#discussion_r1456416875
+            for from_to in values.rev() {
                 let parts: Vec<_> = from_to.split(':').collect();
 
                 if parts.len() != 2 {
@@ -185,6 +212,11 @@ impl App {
                 Some("caret") => NonprintableNotation::Caret,
                 _ => unreachable!("other values for --nonprintable-notation are not allowed"),
             },
+            binary: match self.matches.get_one::<String>("binary").map(|s| s.as_str()) {
+                Some("as-text") => BinaryBehavior::AsText,
+                Some("no-printing") => BinaryBehavior::NoPrinting,
+                _ => unreachable!("other values for --binary are not allowed"),
+            },
             wrapping_mode: if self.interactive_output || maybe_term_width.is_some() {
                 if !self.matches.get_flag("chop-long-lines") {
                     match self.matches.get_one::<String>("wrap").map(|s| s.as_str()) {
@@ -236,18 +268,17 @@ impl App {
                         4
                     },
                 ),
-            theme: self
+            strip_ansi: match self
                 .matches
-                .get_one::<String>("theme")
-                .map(String::from)
-                .map(|s| {
-                    if s == "default" {
-                        String::from(HighlightingAssets::default_theme())
-                    } else {
-                        s
-                    }
-                })
-                .unwrap_or_else(|| String::from(HighlightingAssets::default_theme())),
+                .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"),
+            },
+            theme: theme(self.theme_options()).to_string(),
             visible_lines: match self.matches.try_contains_id("diff").unwrap_or_default()
                 && self.matches.get_flag("diff")
             {
@@ -287,6 +318,17 @@ impl App {
             use_custom_assets: !self.matches.get_flag("no-custom-assets"),
             #[cfg(feature = "lessopen")]
             use_lessopen: self.matches.get_flag("lessopen"),
+            set_terminal_title: self.matches.get_flag("set-terminal-title"),
+            squeeze_lines: if self.matches.get_flag("squeeze-blank") {
+                Some(
+                    self.matches
+                        .get_one::<usize>("squeeze-limit")
+                        .map(|limit| limit.to_owned())
+                        .unwrap_or(1),
+                )
+            } else {
+                None
+            },
         })
     }
 
@@ -336,34 +378,57 @@ impl App {
         Ok(file_input)
     }
 
+    fn forced_style_components(&self) -> Option<StyleComponents> {
+        // No components if `--decorations=never``.
+        if self
+            .matches
+            .get_one::<String>("decorations")
+            .map(|s| s.as_str())
+            == Some("never")
+        {
+            return Some(StyleComponents(HashSet::new()));
+        }
+
+        // Only line numbers if `--number`.
+        if self.matches.get_flag("number") {
+            return Some(StyleComponents(HashSet::from([
+                StyleComponent::LineNumbers,
+            ])));
+        }
+
+        // Plain if `--plain` is specified at least once.
+        if self.matches.get_count("plain") > 0 {
+            return Some(StyleComponents(HashSet::from([StyleComponent::Plain])));
+        }
+
+        // Default behavior.
+        None
+    }
+
     fn style_components(&self) -> Result<StyleComponents> {
         let matches = &self.matches;
-        let mut styled_components = StyleComponents(
-            if matches.get_one::<String>("decorations").map(|s| s.as_str()) == Some("never") {
-                HashSet::new()
-            } else if matches.get_flag("number") {
-                [StyleComponent::LineNumbers].iter().cloned().collect()
-            } else if 0 < matches.get_count("plain") {
-                [StyleComponent::Plain].iter().cloned().collect()
-            } else {
-                matches
-                    .get_one::<String>("style")
-                    .map(|styles| {
-                        styles
-                            .split(',')
-                            .map(|style| style.parse::<StyleComponent>())
-                            .filter_map(|style| style.ok())
-                            .collect::<Vec<_>>()
-                    })
-                    .unwrap_or_else(|| vec![StyleComponent::Default])
+        let mut styled_components = match self.forced_style_components() {
+            Some(forced_components) => forced_components,
+
+            // Parse the `--style` arguments and merge them.
+            None if matches.contains_id("style") => {
+                let lists = matches
+                    .get_many::<String>("style")
+                    .expect("styles present")
+                    .map(|v| StyleComponentList::from_str(v))
+                    .collect::<Result<Vec<StyleComponentList>>>()?;
+
+                StyleComponentList::to_components(lists, self.interactive_output, true)
+            }
+
+            // Use the default.
+            None => StyleComponents(HashSet::from_iter(
+                StyleComponent::Default
+                    .components(self.interactive_output)
                     .into_iter()
-                    .map(|style| style.components(self.interactive_output))
-                    .fold(HashSet::new(), |mut acc, components| {
-                        acc.extend(components.iter().cloned());
-                        acc
-                    })
-            },
-        );
+                    .cloned(),
+            )),
+        };
 
         // If `grid` is set, remove `rule` as it is a subset of `grid`, and print a warning.
         if styled_components.grid() && styled_components.0.remove(&StyleComponent::Rule) {
@@ -372,4 +437,25 @@ impl App {
 
         Ok(styled_components)
     }
+
+    fn theme_options(&self) -> ThemeOptions {
+        let theme = self
+            .matches
+            .get_one::<String>("theme")
+            .map(|t| ThemePreference::from_str(t).unwrap())
+            .unwrap_or_default();
+        let theme_dark = self
+            .matches
+            .get_one::<String>("theme-dark")
+            .map(|t| ThemeName::from_str(t).unwrap());
+        let theme_light = self
+            .matches
+            .get_one::<String>("theme-light")
+            .map(|t| ThemeName::from_str(t).unwrap());
+        ThemeOptions {
+            theme,
+            theme_dark,
+            theme_light,
+        }
+    }
 }

src/bin/bat/assets.rs

@@ -44,7 +44,7 @@ pub fn assets_from_cache_or_binary(
 }
 
 fn clear_asset(path: PathBuf, description: &str) {
-    print!("Clearing {} ... ", description);
+    print!("Clearing {description} ... ");
     match fs::remove_file(&path) {
         Err(err) if err.kind() == io::ErrorKind::NotFound => {
             println!("skipped (not present)");

src/bin/bat/clap_app.rs

@@ -1,9 +1,11 @@
+use bat::style::StyleComponentList;
 use clap::{
     crate_name, crate_version, value_parser, Arg, ArgAction, ArgGroup, ColorChoice, Command,
 };
 use once_cell::sync::Lazy;
 use std::env;
 use std::path::{Path, PathBuf};
+use std::str::FromStr;
 
 static VERSION: Lazy<String> = Lazy::new(|| {
     #[cfg(feature = "bugreport")]
@@ -75,11 +77,26 @@ pub fn build_app(interactive_output: bool) -> Command {
                     * caret   (^G, ^J, ^@, ..)",
                 ),
         )
+        .arg(
+            Arg::new("binary")
+                .long("binary")
+                .action(ArgAction::Set)
+                .default_value("no-printing")
+                .value_parser(["no-printing", "as-text"])
+                .value_name("behavior")
+                .hide_default_value(true)
+                .help("How to treat binary content. (default: no-printing)")
+                .long_help(
+                    "How to treat binary content. (default: no-printing)\n\n\
+                    Possible values:\n  \
+                    * no-printing: do not print any binary content\n  \
+                    * as-text: treat binary content as normal text",
+                ),
+        )
         .arg(
             Arg::new("plain")
                 .overrides_with("plain")
                 .overrides_with("number")
-                .overrides_with("paging")
                 .short('p')
                 .long("plain")
                 .action(ArgAction::Count)
@@ -304,7 +321,6 @@ pub fn build_app(interactive_output: bool) -> Command {
                 .long("paging")
                 .overrides_with("paging")
                 .overrides_with("no-paging")
-                .overrides_with("plain")
                 .value_name("when")
                 .value_parser(["auto", "never", "always"])
                 .default_value("auto")
@@ -377,9 +393,40 @@ pub fn build_app(interactive_output: bool) -> Command {
                      see all available themes. To set a default theme, add the \
                      '--theme=\"...\"' option to the configuration file or export the \
                      BAT_THEME environment variable (e.g.: export \
-                     BAT_THEME=\"...\").",
+                     BAT_THEME=\"...\").\n\n\
+                     Special values:\n\n  \
+                     * auto: Picks a dark or light theme depending on the terminal's colors (default).\n          \
+                     Use '--theme-light' and '--theme-dark' to customize the selected theme.\n    \
+                     * auto:always: Detect the terminal's colors even when the output is redirected.\n    \
+                     * auto:system: Detect the color scheme from the system-wide preference (macOS only).\n  \
+                     * dark: Use the dark theme specified by '--theme-dark'.\n  \
+                     * light: Use the light theme specified by '--theme-light'.",
                 ),
         )
+        .arg(
+            Arg::new("theme-light")
+                .long("theme-light")
+                .overrides_with("theme-light")
+                .value_name("theme")
+                .help("Sets the color theme for syntax highlighting used for light backgrounds.")
+                .long_help(
+                    "Sets the theme name for syntax highlighting used when the terminal uses a light background. \
+                    Use '--list-themes' to see all available themes. To set a default theme, add the \
+                    '--theme-light=\"...\" option to the configuration file or export the BAT_THEME_LIGHT \
+                    environment variable (e.g. export BAT_THEME_LIGHT=\"...\")."),
+        )
+        .arg(
+            Arg::new("theme-dark")
+                .long("theme-dark")
+                .overrides_with("theme-dark")
+                .value_name("theme")
+                .help("Sets the color theme for syntax highlighting used for dark backgrounds.")
+                .long_help(
+                    "Sets the theme name for syntax highlighting used when the terminal uses a dark background. \
+                    Use '--list-themes' to see all available themes. To set a default theme, add the \
+                    '--theme-dark=\"...\" option to the configuration file or export the BAT_THEME_DARK \
+                    environment variable (e.g. export BAT_THEME_DARK=\"...\")."),
+        )
         .arg(
             Arg::new("list-themes")
                 .long("list-themes")
@@ -387,37 +434,45 @@ pub fn build_app(interactive_output: bool) -> Command {
                 .help("Display all supported highlighting themes.")
                 .long_help("Display a list of supported themes for syntax highlighting."),
         )
+        .arg(
+            Arg::new("squeeze-blank")
+                .long("squeeze-blank")
+                .short('s')
+                .action(ArgAction::SetTrue)
+                .help("Squeeze consecutive empty lines.")
+                .long_help("Squeeze consecutive empty lines into a single empty line.")
+        )
+        .arg(
+            Arg::new("squeeze-limit")
+                .long("squeeze-limit")
+                .value_parser(|s: &str| s.parse::<usize>().map_err(|_| "Requires a non-negative number".to_owned()))
+                .long_help("Set the maximum number of consecutive empty lines to be printed.")
+                .hide_short_help(true)
+        )
+        .arg(
+            Arg::new("strip-ansi")
+                .long("strip-ansi")
+                .overrides_with("strip-ansi")
+                .value_name("when")
+                .value_parser(["auto", "always", "never"])
+                .default_value("never")
+                .hide_default_value(true)
+                .help("Strip colors from the input (auto, always, *never*)")
+                .long_help("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. Possible values: auto, always, *never*.")
+                .hide_short_help(true)
+        )
         .arg(
             Arg::new("style")
                 .long("style")
+                .action(ArgAction::Append)
                 .value_name("components")
-                .overrides_with("style")
-                .overrides_with("plain")
-                .overrides_with("number")
                 // Cannot use claps built in validation because we have to turn off clap's delimiters
                 .value_parser(|val: &str| {
-                    let mut invalid_vals = val.split(',').filter(|style| {
-                        !&[
-                            "auto",
-                            "full",
-                            "default",
-                            "plain",
-                            "header",
-                            "header-filename",
-                            "header-filesize",
-                            "grid",
-                            "rule",
-                            "numbers",
-                            "snip",
-                            #[cfg(feature = "git")]
-                            "changes",
-                        ].contains(style)
-                    });
-
-                    if let Some(invalid) = invalid_vals.next() {
-                        Err(format!("Unknown style, '{}'", invalid))
-                    } else {
-                        Ok(val.to_owned())
+                    match StyleComponentList::from_str(val) {
+                        Err(err) => Err(err),
+                        Ok(_) => Ok(val.to_owned()),
                     }
                 })
                 .help(
@@ -432,6 +487,14 @@ pub fn build_app(interactive_output: bool) -> Command {
                      pre-defined style ('full'). To set a default style, add the \
                      '--style=\"..\"' option to the configuration file or export the \
                      BAT_STYLE environment variable (e.g.: export BAT_STYLE=\"..\").\n\n\
+                     When styles are specified in multiple places, the \"nearest\" set \
+                     of styles take precedence. The command-line arguments are the highest \
+                     priority, followed by the BAT_STYLE environment variable, and then \
+                     the configuration file. If any set of styles consists entirely of \
+                     components prefixed with \"+\" or \"-\", it will modify the \
+                     previous set of styles instead of replacing them.\n\n\
+                     By default, the following components are enabled:\n  \
+                        changes, grid, header-filename, numbers, snip\n\n\
                      Possible values:\n\n  \
                      * default: enables recommended style components (default).\n  \
                      * full: enables all available components.\n  \
@@ -501,6 +564,17 @@ pub fn build_app(interactive_output: bool) -> Command {
                 .help("Do not load custom assets"),
         );
 
+    #[cfg(feature = "application")]
+    {
+        app = app.arg(
+            Arg::new("completion")
+            .long("completion")
+            .value_name("SHELL")
+            .value_parser(["bash", "fish", "ps1", "zsh"])
+            .help("Show shell completion for a certain shell. [possible values: bash, fish, zsh, ps1]"),
+        );
+    }
+
     #[cfg(feature = "lessopen")]
     {
         app = app
@@ -567,6 +641,13 @@ pub fn build_app(interactive_output: bool) -> Command {
                 .action(ArgAction::SetTrue)
                 .hide_short_help(true)
                 .help("Show acknowledgements."),
+        )
+        .arg(
+            Arg::new("set-terminal-title")
+                .long("set-terminal-title")
+                .action(ArgAction::SetTrue)
+                .hide_short_help(true)
+                .help("Sets terminal title to filenames when using a pager."),
         );
 
     // Check if the current directory contains a file name cache. Otherwise,

src/bin/bat/completions.rs

@@ -0,0 +1,6 @@
+use std::env;
+
+pub const BASH_COMPLETION: &str = include_str!(env!("BAT_GENERATED_COMPLETION_BASH"));
+pub const FISH_COMPLETION: &str = include_str!(env!("BAT_GENERATED_COMPLETION_FISH"));
+pub const PS1_COMPLETION: &str = include_str!(env!("BAT_GENERATED_COMPLETION_PS1"));
+pub const ZSH_COMPLETION: &str = include_str!(env!("BAT_GENERATED_COMPLETION_ZSH"));

src/bin/bat/config.rs

@@ -140,14 +140,19 @@ 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"),
-        ("--theme", "BAT_THEME"),
+        ("--theme", bat::theme::env::BAT_THEME),
+        ("--theme-dark", bat::theme::env::BAT_THEME_DARK),
+        ("--theme-light", bat::theme::env::BAT_THEME_LIGHT),
         ("--pager", "BAT_PAGER"),
         ("--paging", "BAT_PAGING"),
         ("--style", "BAT_STYLE"),
     ]
     .iter()
-    .filter_map(|(flag, key)| env::var(key).ok().map(|var| [flag.to_string(), var]))
-    .flatten()
+    .filter_map(|(flag, key)| {
+        env::var(key)
+            .ok()
+            .map(|var| [flag.to_string(), var].join("="))
+    })
     .map(|a| a.into())
     .collect()
 }

src/bin/bat/main.rs

@@ -3,6 +3,8 @@
 mod app;
 mod assets;
 mod clap_app;
+#[cfg(feature = "application")]
+mod completions;
 mod config;
 mod directories;
 mod input;
@@ -14,6 +16,7 @@ use std::io::{BufReader, Write};
 use std::path::Path;
 use std::process;
 
+use bat::theme::DetectColorScheme;
 use nu_ansi_term::Color::Green;
 use nu_ansi_term::Style;
 
@@ -35,6 +38,7 @@ use bat::{
     error::*,
     input::Input,
     style::{StyleComponent, StyleComponents},
+    theme::{color_scheme, default_theme, ColorScheme},
     MappingTarget, PagingMode,
 };
 
@@ -78,9 +82,11 @@ fn run_cache_subcommand(
     Ok(())
 }
 
-fn get_syntax_mapping_to_paths<'a>(
-    mappings: &[(GlobMatcher, MappingTarget<'a>)],
-) -> HashMap<&'a str, Vec<String>> {
+fn get_syntax_mapping_to_paths<'r, 't, I>(mappings: I) -> HashMap<&'t str, Vec<String>>
+where
+    I: IntoIterator<Item = (&'r GlobMatcher, &'r MappingTarget<'t>)>,
+    't: 'r, // target text outlives rule
+{
     let mut map = HashMap::new();
     for mapping in mappings {
         if let (matcher, MappingTarget::MapTo(s)) = mapping {
@@ -123,7 +129,7 @@ pub fn get_languages(config: &Config, cache_dir: &Path) -> Result<String> {
 
     languages.sort_by_key(|lang| lang.name.to_uppercase());
 
-    let configured_languages = get_syntax_mapping_to_paths(config.syntax_mapping.mappings());
+    let configured_languages = get_syntax_mapping_to_paths(config.syntax_mapping.all_mappings());
 
     for lang in &mut languages {
         if let Some(additional_paths) = configured_languages.get(lang.name.as_str()) {
@@ -186,7 +192,12 @@ fn theme_preview_file<'a>() -> Input<'a> {
     Input::from_reader(Box::new(BufReader::new(THEME_PREVIEW_DATA)))
 }
 
-pub fn list_themes(cfg: &Config, config_dir: &Path, cache_dir: &Path) -> Result<()> {
+pub fn list_themes(
+    cfg: &Config,
+    config_dir: &Path,
+    cache_dir: &Path,
+    detect_color_scheme: DetectColorScheme,
+) -> Result<()> {
     let assets = assets_from_cache_or_binary(cfg.use_custom_assets, cache_dir)?;
     let mut config = cfg.clone();
     let mut style = HashSet::new();
@@ -197,19 +208,35 @@ pub fn list_themes(cfg: &Config, config_dir: &Path, cache_dir: &Path) -> Result<
     let stdout = io::stdout();
     let mut stdout = stdout.lock();
 
-    if config.colored_output {
-        for theme in assets.themes() {
+    let default_theme_name = default_theme(color_scheme(detect_color_scheme).unwrap_or_default());
+    for theme in assets.themes() {
+        let default_theme_info = if !config.loop_through && default_theme_name == theme {
+            " (default)"
+        } else if default_theme(ColorScheme::Dark) == theme {
+            " (default dark)"
+        } else if default_theme(ColorScheme::Light) == theme {
+            " (default light)"
+        } else {
+            ""
+        };
+        if config.colored_output {
             writeln!(
                 stdout,
-                "Theme: {}\n",
-                Style::new().bold().paint(theme.to_string())
+                "Theme: {}{}\n",
+                Style::new().bold().paint(theme.to_string()),
+                default_theme_info
             )?;
             config.theme = theme.to_string();
             Controller::new(&config, &assets)
                 .run(vec![theme_preview_file()], None)
                 .ok();
             writeln!(stdout)?;
+        } else {
+            writeln!(stdout, "{theme}{default_theme_info}")?;
         }
+    }
+
+    if config.colored_output {
         writeln!(
             stdout,
             "Further themes can be installed to '{}', \
@@ -218,18 +245,35 @@ pub fn list_themes(cfg: &Config, config_dir: &Path, cache_dir: &Path) -> Result<
             https://github.com/sharkdp/bat#adding-new-themes",
             config_dir.join("themes").to_string_lossy()
         )?;
-    } else {
-        for theme in assets.themes() {
-            writeln!(stdout, "{}", theme)?;
-        }
     }
 
     Ok(())
 }
 
+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}");
+    io::stdout().flush().unwrap();
+}
+
+fn get_new_terminal_title(inputs: &Vec<Input>) -> String {
+    let mut new_terminal_title = "bat: ".to_string();
+    for (index, input) in inputs.iter().enumerate() {
+        new_terminal_title += input.description().title();
+        if index < inputs.len() - 1 {
+            new_terminal_title += ", ";
+        }
+    }
+    new_terminal_title
+}
+
 fn run_controller(inputs: Vec<Input>, config: &Config, cache_dir: &Path) -> Result<bool> {
     let assets = assets_from_cache_or_binary(config.use_custom_assets, cache_dir)?;
     let controller = Controller::new(config, &assets);
+    if config.paging_mode != PagingMode::Never && config.set_terminal_title {
+        set_terminal_title_to(get_new_terminal_title(&inputs));
+    }
     controller.run(inputs, None)
 }
 
@@ -249,24 +293,25 @@ fn invoke_bugreport(app: &App, cache_dir: &Path) {
         .info(OperatingSystem::default())
         .info(CommandLine::default())
         .info(EnvironmentVariables::list(&[
-            "SHELL",
-            "PAGER",
-            "LESS",
-            "LANG",
-            "LC_ALL",
-            "BAT_PAGER",
-            "BAT_PAGING",
             "BAT_CACHE_PATH",
             "BAT_CONFIG_PATH",
             "BAT_OPTS",
+            "BAT_PAGER",
+            "BAT_PAGING",
             "BAT_STYLE",
             "BAT_TABS",
             "BAT_THEME",
-            "XDG_CONFIG_HOME",
-            "XDG_CACHE_HOME",
             "COLORTERM",
-            "NO_COLOR",
+            "LANG",
+            "LC_ALL",
+            "LESS",
             "MANPAGER",
+            "NO_COLOR",
+            "PAGER",
+            "SHELL",
+            "TERM",
+            "XDG_CACHE_HOME",
+            "XDG_CONFIG_HOME",
         ]))
         .info(FileContent::new("System Config file", system_config_file()))
         .info(FileContent::new("Config file", config_file()))
@@ -304,6 +349,18 @@ fn run() -> Result<bool> {
         return Ok(true);
     }
 
+    #[cfg(feature = "application")]
+    if let Some(shell) = app.matches.get_one::<String>("completion") {
+        match shell.as_str() {
+            "bash" => println!("{}", completions::BASH_COMPLETION),
+            "fish" => println!("{}", completions::FISH_COMPLETION),
+            "ps1" => println!("{}", completions::PS1_COMPLETION),
+            "zsh" => println!("{}", completions::ZSH_COMPLETION),
+            _ => unreachable!("No completion for shell '{}' available.", shell),
+        }
+        return Ok(true);
+    }
+
     match app.matches.subcommand() {
         Some(("cache", cache_matches)) => {
             // If there is a file named 'cache' in the current working directory,
@@ -338,7 +395,7 @@ fn run() -> Result<bool> {
                 };
                 run_controller(inputs, &plain_config, cache_dir)
             } else if app.matches.get_flag("list-themes") {
-                list_themes(&config, config_dir, cache_dir)?;
+                list_themes(&config, config_dir, cache_dir, DetectColorScheme::default())?;
                 Ok(true)
             } else if app.matches.get_flag("config-file") {
                 println!("{}", config_file().to_string_lossy());

src/config.rs

@@ -1,10 +1,11 @@
 use crate::line_range::{HighlightedLineRanges, LineRanges};
-use crate::nonprintable_notation::NonprintableNotation;
+use crate::nonprintable_notation::{BinaryBehavior, NonprintableNotation};
 #[cfg(feature = "paging")]
 use crate::paging::PagingMode;
 use crate::style::StyleComponents;
 use crate::syntax_mapping::SyntaxMapping;
 use crate::wrapping::WrappingMode;
+use crate::StripAnsiMode;
 
 #[derive(Debug, Clone)]
 pub enum VisibleLines {
@@ -43,6 +44,9 @@ pub struct Config<'a> {
     /// The configured notation for non-printable characters
     pub nonprintable_notation: NonprintableNotation,
 
+    /// How to treat binary content
+    pub binary: BinaryBehavior,
+
     /// The character width of the terminal
     pub term_width: usize,
 
@@ -94,6 +98,15 @@ pub struct Config<'a> {
     // Whether or not to use $LESSOPEN if set
     #[cfg(feature = "lessopen")]
     pub use_lessopen: bool,
+
+    // Weather or not to set terminal title when using a pager
+    pub set_terminal_title: bool,
+
+    /// The maximum number of consecutive empty lines to display
+    pub squeeze_lines: Option<usize>,
+
+    // Weather or not to set terminal title when using a pager
+    pub strip_ansi: StripAnsiMode,
 }
 
 #[cfg(all(feature = "minimal-application", feature = "paging"))]

src/controller.rs

@@ -47,7 +47,7 @@ impl<'b> Controller<'b> {
         &self,
         inputs: Vec<Input>,
         output_buffer: Option<&mut dyn std::fmt::Write>,
-        handle_error: impl Fn(&Error, &mut dyn Write),
+        mut handle_error: impl FnMut(&Error, &mut dyn Write),
     ) -> Result<bool> {
         let mut output_type;
 

src/decorations.rs

@@ -46,7 +46,7 @@ impl Decoration for LineNumberDecoration {
         _printer: &InteractivePrinter,
     ) -> DecorationText {
         if continuation {
-            if line_number > self.cached_wrap_invalid_at {
+            if line_number >= self.cached_wrap_invalid_at {
                 let new_width = self.cached_wrap.width + 1;
                 return DecorationText {
                     text: self.color.paint(" ".repeat(new_width)).to_string(),
@@ -56,7 +56,7 @@ impl Decoration for LineNumberDecoration {
 
             self.cached_wrap.clone()
         } else {
-            let plain: String = format!("{:4}", line_number);
+            let plain: String = format!("{line_number:4}");
             DecorationText {
                 width: plain.len(),
                 text: self.color.paint(plain).to_string(),

src/input.rs

@@ -197,7 +197,7 @@ impl<'a> Input<'a> {
             InputKind::StdIn => {
                 if let Some(stdout) = stdout_identifier {
                     let input_identifier = Identifier::try_from(clircle::Stdio::Stdin)
-                        .map_err(|e| format!("Stdin: Error identifying file: {}", e))?;
+                        .map_err(|e| format!("Stdin: Error identifying file: {e}"))?;
                     if stdout.surely_conflicts_with(&input_identifier) {
                         return Err("IO circle detected. The input from stdin is also an output. Aborting to avoid infinite loop.".into());
                     }

src/lessopen.rs

@@ -3,13 +3,12 @@
 use std::convert::TryFrom;
 use std::env;
 use std::fs::File;
-use std::io::{BufRead, BufReader, Cursor, Read, Write};
+use std::io::{BufRead, BufReader, Cursor, Read};
 use std::path::PathBuf;
-use std::str;
+use std::process::{ExitStatus, Stdio};
 
 use clircle::{Clircle, Identifier};
-use os_str_bytes::RawOsString;
-use run_script::{IoOptions, ScriptOptions};
+use execute::{shell, Execute};
 
 use crate::error::Result;
 use crate::{
@@ -21,7 +20,6 @@ use crate::{
 pub(crate) struct LessOpenPreprocessor {
     lessopen: String,
     lessclose: Option<String>,
-    command_options: ScriptOptions,
     kind: LessOpenKind,
     /// Whether or not data piped via stdin is to be preprocessed
     preprocess_stdin: bool,
@@ -52,7 +50,7 @@ impl LessOpenPreprocessor {
         // Otherwise, if output is empty and exit code is nonzero, use original file contents
         let (kind, lessopen) = if lessopen.starts_with("||") {
             (LessOpenKind::Piped, lessopen.chars().skip(2).collect())
-        // "|" means pipe, but ignore exit code, always using preprocessor output
+        // "|" means pipe as above, but ignore exit code and always use preprocessor output even if empty
         } else if lessopen.starts_with('|') {
             (
                 LessOpenKind::PipedIgnoreExitCode,
@@ -70,16 +68,9 @@ impl LessOpenPreprocessor {
             (false, lessopen)
         };
 
-        let mut command_options = ScriptOptions::new();
-        command_options.runner = env::var("SHELL").ok();
-        command_options.input_redirection = IoOptions::Pipe;
-
         Ok(Self {
-            lessopen: lessopen.replacen("%s", "$1", 1),
-            lessclose: env::var("LESSCLOSE")
-                .ok()
-                .map(|str| str.replacen("%s", "$1", 1).replacen("%s", "$2", 1)),
-            command_options,
+            lessopen,
+            lessclose: env::var("LESSCLOSE").ok(),
             kind,
             preprocess_stdin: stdin,
         })
@@ -98,21 +89,21 @@ impl LessOpenPreprocessor {
                     None => return input.open(stdin, stdout_identifier),
                 };
 
-                let (exit_code, lessopen_stdout, _) = match run_script::run(
-                    &self.lessopen,
-                    &vec![path_str.to_string()],
-                    &self.command_options,
-                ) {
+                let mut lessopen_command = shell(self.lessopen.replacen("%s", path_str, 1));
+                lessopen_command.stdout(Stdio::piped());
+
+                let lessopen_output = match lessopen_command.execute_output() {
                     Ok(output) => output,
                     Err(_) => return input.open(stdin, stdout_identifier),
                 };
 
-                if self.fall_back_to_original_file(&lessopen_stdout, exit_code) {
+                if self.fall_back_to_original_file(&lessopen_output.stdout, lessopen_output.status)
+                {
                     return input.open(stdin, stdout_identifier);
                 }
 
                 (
-                    RawOsString::from_string(lessopen_stdout),
+                    lessopen_output.stdout,
                     path_str.to_string(),
                     OpenedInputKind::OrdinaryFile(path.to_path_buf()),
                 )
@@ -127,47 +118,31 @@ impl LessOpenPreprocessor {
                         }
                     }
 
-                    // stdin isn't Clone, so copy it to a cloneable buffer
+                    // stdin isn't Clone or AsRef<[u8]>, so move it into a cloneable buffer
+                    // so the data can be used multiple times if necessary
+                    // NOTE: stdin will be empty from this point onwards
                     let mut stdin_buffer = Vec::new();
-                    stdin.read_to_end(&mut stdin_buffer).unwrap();
+                    stdin.read_to_end(&mut stdin_buffer)?;
 
-                    let mut lessopen_handle = match run_script::spawn(
-                        &self.lessopen,
-                        &vec!["-".to_string()],
-                        &self.command_options,
-                    ) {
-                        Ok(handle) => handle,
-                        Err(_) => {
-                            return input.open(stdin, stdout_identifier);
-                        }
-                    };
+                    let mut lessopen_command = shell(self.lessopen.replacen("%s", "-", 1));
+                    lessopen_command.stdout(Stdio::piped());
 
-                    if lessopen_handle
-                        .stdin
-                        .as_mut()
-                        .unwrap()
-                        .write_all(&stdin_buffer.clone())
-                        .is_err()
+                    let lessopen_output = match lessopen_command.execute_input_output(&stdin_buffer)
                     {
-                        return input.open(stdin, stdout_identifier);
-                    }
-
-                    let lessopen_output = match lessopen_handle.wait_with_output() {
                         Ok(output) => output,
                         Err(_) => {
                             return input.open(Cursor::new(stdin_buffer), stdout_identifier);
                         }
                     };
 
-                    if lessopen_output.stdout.is_empty()
-                        && (!lessopen_output.status.success()
-                            || matches!(self.kind, LessOpenKind::PipedIgnoreExitCode))
+                    if self
+                        .fall_back_to_original_file(&lessopen_output.stdout, lessopen_output.status)
                     {
                         return input.open(Cursor::new(stdin_buffer), stdout_identifier);
                     }
 
                     (
-                        RawOsString::assert_from_raw_vec(lessopen_output.stdout),
+                        lessopen_output.stdout,
                         "-".to_string(),
                         OpenedInputKind::StdIn,
                     )
@@ -184,13 +159,17 @@ impl LessOpenPreprocessor {
             kind,
             reader: InputReader::new(BufReader::new(
                 if matches!(self.kind, LessOpenKind::TempFile) {
-                    // Remove newline at end of temporary file path returned by $LESSOPEN
-                    let stdout = match lessopen_stdout.strip_suffix("\n") {
-                        Some(stripped) => stripped.to_owned(),
-                        None => lessopen_stdout,
+                    let lessopen_string = match String::from_utf8(lessopen_stdout) {
+                        Ok(string) => string,
+                        Err(_) => {
+                            return input.open(stdin, stdout_identifier);
+                        }
+                    };
+                    // Remove newline at end of temporary file path returned by $LESSOPEN
+                    let stdout = match lessopen_string.strip_suffix("\n") {
+                        Some(stripped) => stripped.to_owned(),
+                        None => lessopen_string,
                     };
-
-                    let stdout = stdout.into_os_string();
 
                     let file = match File::open(PathBuf::from(&stdout)) {
                         Ok(file) => file,
@@ -201,16 +180,18 @@ impl LessOpenPreprocessor {
 
                     Preprocessed {
                         kind: PreprocessedKind::TempFile(file),
-                        lessclose: self.lessclose.clone(),
-                        command_args: vec![path_str, stdout.to_str().unwrap().to_string()],
-                        command_options: self.command_options.clone(),
+                        lessclose: self
+                            .lessclose
+                            .as_ref()
+                            .map(|s| s.replacen("%s", &path_str, 1).replacen("%s", &stdout, 1)),
                     }
                 } else {
                     Preprocessed {
-                        kind: PreprocessedKind::Piped(Cursor::new(lessopen_stdout.into_raw_vec())),
-                        lessclose: self.lessclose.clone(),
-                        command_args: vec![path_str, "-".to_string()],
-                        command_options: self.command_options.clone(),
+                        kind: PreprocessedKind::Piped(Cursor::new(lessopen_stdout)),
+                        lessclose: self
+                            .lessclose
+                            .as_ref()
+                            .map(|s| s.replacen("%s", &path_str, 1).replacen("%s", "-", 1)),
                     }
                 },
             )),
@@ -219,9 +200,9 @@ impl LessOpenPreprocessor {
         })
     }
 
-    fn fall_back_to_original_file(&self, lessopen_output: &str, exit_code: i32) -> bool {
-        lessopen_output.is_empty()
-            && (exit_code != 0 || matches!(self.kind, LessOpenKind::PipedIgnoreExitCode))
+    fn fall_back_to_original_file(&self, lessopen_stdout: &Vec<u8>, exit_code: ExitStatus) -> bool {
+        lessopen_stdout.is_empty()
+            && (!exit_code.success() || matches!(self.kind, LessOpenKind::PipedIgnoreExitCode))
     }
 
     #[cfg(test)]
@@ -261,8 +242,6 @@ impl Read for PreprocessedKind {
 pub struct Preprocessed {
     kind: PreprocessedKind,
     lessclose: Option<String>,
-    command_args: Vec<String>,
-    command_options: ScriptOptions,
 }
 
 impl Read for Preprocessed {
@@ -273,11 +252,20 @@ impl Read for Preprocessed {
 
 impl Drop for Preprocessed {
     fn drop(&mut self) {
-        if let Some(ref command) = self.lessclose {
-            self.command_options.output_redirection = IoOptions::Inherit;
+        if let Some(lessclose) = self.lessclose.clone() {
+            let mut lessclose_command = shell(lessclose);
 
-            run_script::run(command, &self.command_args, &self.command_options)
-                .expect("failed to run $LESSCLOSE to clean up file");
+            let lessclose_output = match lessclose_command.execute_output() {
+                Ok(output) => output,
+                Err(_) => {
+                    bat_warning!("failed to run $LESSCLOSE to clean up temporary file");
+                    return;
+                }
+            };
+
+            if lessclose_output.status.success() {
+                bat_warning!("$LESSCLOSE exited with nonzero exit code",)
+            };
         }
     }
 }
@@ -301,7 +289,7 @@ mod tests {
     fn test_just_lessopen() -> Result<()> {
         let preprocessor = LessOpenPreprocessor::mock_new(Some("|batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(preprocessor.lessclose.is_none());
 
         reset_env_vars();
@@ -327,8 +315,8 @@ mod tests {
         let preprocessor =
             LessOpenPreprocessor::mock_new(Some("lessopen.sh %s"), Some("lessclose.sh %s %s"))?;
 
-        assert_eq!(preprocessor.lessopen, "lessopen.sh $1");
-        assert_eq!(preprocessor.lessclose.unwrap(), "lessclose.sh $1 $2");
+        assert_eq!(preprocessor.lessopen, "lessopen.sh %s");
+        assert_eq!(preprocessor.lessclose.unwrap(), "lessclose.sh %s %s");
 
         reset_env_vars();
 
@@ -340,13 +328,13 @@ mod tests {
     fn test_lessopen_prefixes() -> Result<()> {
         let preprocessor = LessOpenPreprocessor::mock_new(Some("batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(preprocessor.kind, LessOpenKind::TempFile));
         assert!(!preprocessor.preprocess_stdin);
 
         let preprocessor = LessOpenPreprocessor::mock_new(Some("|batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(
             preprocessor.kind,
             LessOpenKind::PipedIgnoreExitCode
@@ -355,19 +343,19 @@ mod tests {
 
         let preprocessor = LessOpenPreprocessor::mock_new(Some("||batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(preprocessor.kind, LessOpenKind::Piped));
         assert!(!preprocessor.preprocess_stdin);
 
         let preprocessor = LessOpenPreprocessor::mock_new(Some("-batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(preprocessor.kind, LessOpenKind::TempFile));
         assert!(preprocessor.preprocess_stdin);
 
         let preprocessor = LessOpenPreprocessor::mock_new(Some("|-batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(
             preprocessor.kind,
             LessOpenKind::PipedIgnoreExitCode
@@ -376,7 +364,7 @@ mod tests {
 
         let preprocessor = LessOpenPreprocessor::mock_new(Some("||-batpipe %s"), None)?;
 
-        assert_eq!(preprocessor.lessopen, "batpipe $1");
+        assert_eq!(preprocessor.lessopen, "batpipe %s");
         assert!(matches!(preprocessor.kind, LessOpenKind::Piped));
         assert!(preprocessor.preprocess_stdin);
 
@@ -391,8 +379,8 @@ mod tests {
         let preprocessor =
             LessOpenPreprocessor::mock_new(Some("|echo File:%s"), Some("echo File:%s Temp:%s"))?;
 
-        assert_eq!(preprocessor.lessopen, "echo File:$1");
-        assert_eq!(preprocessor.lessclose.unwrap(), "echo File:$1 Temp:$2");
+        assert_eq!(preprocessor.lessopen, "echo File:%s");
+        assert_eq!(preprocessor.lessclose.unwrap(), "echo File:%s Temp:%s");
 
         reset_env_vars();
 

src/lib.rs

@@ -49,10 +49,12 @@ pub(crate) mod printer;
 pub mod style;
 pub(crate) mod syntax_mapping;
 mod terminal;
+pub mod theme;
 mod vscreen;
 pub(crate) mod wrapping;
 
-pub use nonprintable_notation::NonprintableNotation;
+pub use nonprintable_notation::{BinaryBehavior, NonprintableNotation};
+pub use preprocessor::StripAnsiMode;
 pub use pretty_printer::{Input, PrettyPrinter, Syntax};
 pub use syntax_mapping::{MappingTarget, SyntaxMapping};
 pub use wrapping::WrappingMode;

src/nonprintable_notation.rs

@@ -10,3 +10,15 @@ pub enum NonprintableNotation {
     #[default]
     Unicode,
 }
+
+/// How to treat binary content
+#[derive(Debug, Default, Clone, Copy, PartialEq, Eq)]
+#[non_exhaustive]
+pub enum BinaryBehavior {
+    /// Do not print any binary content
+    #[default]
+    NoPrinting,
+
+    /// Treat binary content as normal text
+    AsText,
+}

src/preprocessor.rs

@@ -1,17 +1,18 @@
 use std::fmt::Write;
 
-use console::AnsiCodeIterator;
-
-use crate::nonprintable_notation::NonprintableNotation;
+use crate::{
+    nonprintable_notation::NonprintableNotation,
+    vscreen::{EscapeSequenceOffsets, EscapeSequenceOffsetsIterator},
+};
 
 /// Expand tabs like an ANSI-enabled expand(1).
 pub fn expand_tabs(line: &str, width: usize, cursor: &mut usize) -> String {
     let mut buffer = String::with_capacity(line.len() * 2);
 
-    for chunk in AnsiCodeIterator::new(line) {
-        match chunk {
-            (text, true) => buffer.push_str(text),
-            (mut text, false) => {
+    for seq in EscapeSequenceOffsetsIterator::new(line) {
+        match seq {
+            EscapeSequenceOffsets::Text { .. } => {
+                let mut text = &line[seq.index_of_start()..seq.index_past_end()];
                 while let Some(index) = text.find('\t') {
                     // Add previous text.
                     if index > 0 {
@@ -31,6 +32,10 @@ pub fn expand_tabs(line: &str, width: usize, cursor: &mut usize) -> String {
                 *cursor += text.len();
                 buffer.push_str(text);
             }
+            _ => {
+                // Copy the ANSI escape sequence.
+                buffer.push_str(&line[seq.index_of_start()..seq.index_past_end()])
+            }
         }
     }
 
@@ -131,6 +136,27 @@ pub fn replace_nonprintable(
     output
 }
 
+/// 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
+}
+
+#[derive(Debug, PartialEq, Clone, Copy, Default)]
+pub enum StripAnsiMode {
+    #[default]
+    Never,
+    Always,
+    Auto,
+}
+
 #[test]
 fn test_try_parse_utf8_char() {
     assert_eq!(try_parse_utf8_char(&[0x20]), Some((' ', 1)));
@@ -174,3 +200,14 @@ fn test_try_parse_utf8_char() {
     assert_eq!(try_parse_utf8_char(&[0xef, 0x20]), None);
     assert_eq!(try_parse_utf8_char(&[0xf0, 0xf0]), None);
 }
+
+#[test]
+fn test_strip_ansi() {
+    // The sequence detection is covered by the tests in the vscreen module.
+    assert_eq!(strip_ansi("no ansi"), "no ansi");
+    assert_eq!(strip_ansi("\x1B[33mone"), "one");
+    assert_eq!(
+        strip_ansi("\x1B]1\x07multiple\x1B[J sequences"),
+        "multiple sequences"
+    );
+}

src/pretty_printer.rs

@@ -11,7 +11,7 @@ use crate::{
     input,
     line_range::{HighlightedLineRanges, LineRange, LineRanges},
     style::StyleComponent,
-    SyntaxMapping, WrappingMode,
+    StripAnsiMode, SyntaxMapping, WrappingMode,
 };
 
 #[cfg(feature = "paging")]
@@ -182,6 +182,15 @@ impl<'a> PrettyPrinter<'a> {
         self
     }
 
+    /// Whether to remove ANSI escape sequences from the input (default: never)
+    ///
+    /// If `Auto` is used, escape sequences will only be removed when the input
+    /// is not plain text.
+    pub fn strip_ansi(&mut self, mode: StripAnsiMode) -> &mut Self {
+        self.config.strip_ansi = mode;
+        self
+    }
+
     /// Text wrapping mode (default: do not wrap)
     pub fn wrapping_mode(&mut self, mode: WrappingMode) -> &mut Self {
         self.config.wrapping_mode = mode;
@@ -230,7 +239,15 @@ impl<'a> PrettyPrinter<'a> {
         self
     }
 
-    /// Specify the highlighting theme
+    /// Specify the maximum number of consecutive empty lines to print.
+    pub fn squeeze_empty_lines(&mut self, maximum: Option<usize>) -> &mut Self {
+        self.config.squeeze_lines = maximum;
+        self
+    }
+
+    /// Specify the highlighting theme.
+    /// You can use [`crate::theme::theme`] to pick a theme based on user preferences
+    /// and the terminal's background color.
     pub fn theme(&mut self, theme: impl AsRef<str>) -> &mut Self {
         self.config.theme = theme.as_ref().to_owned();
         self
@@ -264,6 +281,11 @@ impl<'a> PrettyPrinter<'a> {
     /// If you want to call 'print' multiple times, you have to call the appropriate
     /// input_* methods again.
     pub fn print(&mut self) -> Result<bool> {
+        self.print_with_writer(None::<&mut dyn std::fmt::Write>)
+    }
+
+    /// Pretty-print all specified inputs to a specified writer.
+    pub fn print_with_writer<W: std::fmt::Write>(&mut self, writer: Option<W>) -> Result<bool> {
         let highlight_lines = std::mem::take(&mut self.highlighted_lines);
         self.config.highlighted_lines = HighlightedLineRanges(LineRanges::from(highlight_lines));
         self.config.term_width = self
@@ -300,7 +322,13 @@ impl<'a> PrettyPrinter<'a> {
 
         // Run the controller
         let controller = Controller::new(&self.config, &self.assets);
-        controller.run(inputs.into_iter().map(|i| i.into()).collect(), None)
+
+        // If writer is provided, pass it to the controller, otherwise pass None
+        if let Some(mut w) = writer {
+            controller.run(inputs.into_iter().map(|i| i.into()).collect(), Some(&mut w))
+        } else {
+            controller.run(inputs.into_iter().map(|i| i.into()).collect(), None)
+        }
     }
 }
 

src/printer.rs

@@ -7,10 +7,9 @@ use nu_ansi_term::Style;
 
 use bytesize::ByteSize;
 
-use console::AnsiCodeIterator;
-
 use syntect::easy::HighlightLines;
 use syntect::highlighting::Color;
+use syntect::highlighting::FontStyle;
 use syntect::highlighting::Theme;
 use syntect::parsing::SyntaxSet;
 
@@ -30,11 +29,44 @@ use crate::diff::LineChanges;
 use crate::error::*;
 use crate::input::OpenedInput;
 use crate::line_range::RangeCheckResult;
+use crate::preprocessor::strip_ansi;
 use crate::preprocessor::{expand_tabs, replace_nonprintable};
 use crate::style::StyleComponent;
 use crate::terminal::{as_terminal_escaped, to_ansi_color};
-use crate::vscreen::AnsiStyle;
+use crate::vscreen::{AnsiStyle, EscapeSequence, EscapeSequenceIterator};
 use crate::wrapping::WrappingMode;
+use crate::BinaryBehavior;
+use crate::StripAnsiMode;
+
+const ANSI_UNDERLINE_ENABLE: EscapeSequence = EscapeSequence::CSI {
+    raw_sequence: "\x1B[4m",
+    parameters: "4",
+    intermediates: "",
+    final_byte: "m",
+};
+
+const ANSI_UNDERLINE_DISABLE: EscapeSequence = EscapeSequence::CSI {
+    raw_sequence: "\x1B[24m",
+    parameters: "24",
+    intermediates: "",
+    final_byte: "m",
+};
+
+const EMPTY_SYNTECT_STYLE: syntect::highlighting::Style = syntect::highlighting::Style {
+    foreground: Color {
+        r: 127,
+        g: 127,
+        b: 127,
+        a: 255,
+    },
+    background: Color {
+        r: 127,
+        g: 127,
+        b: 127,
+        a: 255,
+    },
+    font_style: FontStyle::empty(),
+};
 
 pub enum OutputHandle<'a> {
     IoWrite(&'a mut dyn io::Write),
@@ -72,11 +104,15 @@ pub(crate) trait Printer {
 
 pub struct SimplePrinter<'a> {
     config: &'a Config<'a>,
+    consecutive_empty_lines: usize,
 }
 
 impl<'a> SimplePrinter<'a> {
     pub fn new(config: &'a Config) -> Self {
-        SimplePrinter { config }
+        SimplePrinter {
+            config,
+            consecutive_empty_lines: 0,
+        }
     }
 }
 
@@ -105,6 +141,21 @@ impl<'a> Printer for SimplePrinter<'a> {
         _line_number: usize,
         line_buffer: &[u8],
     ) -> Result<()> {
+        // Skip squeezed lines.
+        if let Some(squeeze_limit) = self.config.squeeze_lines {
+            if String::from_utf8_lossy(line_buffer)
+                .trim_end_matches(|c| c == '\r' || c == '\n')
+                .is_empty()
+            {
+                self.consecutive_empty_lines += 1;
+                if self.consecutive_empty_lines > squeeze_limit {
+                    return Ok(());
+                }
+            } else {
+                self.consecutive_empty_lines = 0;
+            }
+        }
+
         if !out_of_range {
             if self.config.show_nonprintable {
                 let line = replace_nonprintable(
@@ -112,7 +163,7 @@ impl<'a> Printer for SimplePrinter<'a> {
                     self.config.tab_width,
                     self.config.nonprintable_notation,
                 );
-                write!(handle, "{}", line)?;
+                write!(handle, "{line}")?;
             } else {
                 match handle {
                     OutputHandle::IoWrite(handle) => handle.write_all(line_buffer)?,
@@ -158,6 +209,8 @@ pub(crate) struct InteractivePrinter<'a> {
     pub line_changes: &'a Option<LineChanges>,
     highlighter_from_set: Option<HighlighterFromSet<'a>>,
     background_color_highlight: Option<Color>,
+    consecutive_empty_lines: usize,
+    strip_ansi: bool,
 }
 
 impl<'a> InteractivePrinter<'a> {
@@ -210,24 +263,48 @@ impl<'a> InteractivePrinter<'a> {
             panel_width = 0;
         }
 
-        let highlighter_from_set = if input
+        // Get the highlighter for the output.
+        let is_printing_binary = input
             .reader
             .content_type
-            .map_or(false, |c| c.is_binary() && !config.show_nonprintable)
-        {
-            None
-        } else {
-            // Determine the type of syntax for highlighting
-            let syntax_in_set =
-                match assets.get_syntax(config.language, input, &config.syntax_mapping) {
-                    Ok(syntax_in_set) => syntax_in_set,
-                    Err(Error::UndetectedSyntax(_)) => assets
-                        .find_syntax_by_name("Plain Text")?
-                        .expect("A plain text syntax is available"),
-                    Err(e) => return Err(e),
-                };
+            .map_or(false, |c| c.is_binary() && !config.show_nonprintable);
 
-            Some(HighlighterFromSet::new(syntax_in_set, theme))
+        let needs_to_match_syntax = (!is_printing_binary
+            || matches!(config.binary, BinaryBehavior::AsText))
+            && (config.colored_output || config.strip_ansi == StripAnsiMode::Auto);
+
+        let (is_plain_text, highlighter_from_set) = if needs_to_match_syntax {
+            // Determine the type of syntax for highlighting
+            const PLAIN_TEXT_SYNTAX: &str = "Plain Text";
+            match assets.get_syntax(config.language, input, &config.syntax_mapping) {
+                Ok(syntax_in_set) => (
+                    syntax_in_set.syntax.name == PLAIN_TEXT_SYNTAX,
+                    Some(HighlighterFromSet::new(syntax_in_set, theme)),
+                ),
+
+                Err(Error::UndetectedSyntax(_)) => (
+                    true,
+                    Some(
+                        assets
+                            .find_syntax_by_name(PLAIN_TEXT_SYNTAX)?
+                            .map(|s| HighlighterFromSet::new(s, theme))
+                            .expect("A plain text syntax is available"),
+                    ),
+                ),
+
+                Err(e) => return Err(e),
+            }
+        } else {
+            (false, None)
+        };
+
+        // Determine when to strip ANSI sequences
+        let strip_ansi = match config.strip_ansi {
+            _ if config.show_nonprintable => false,
+            StripAnsiMode::Always => true,
+            StripAnsiMode::Auto if is_plain_text => false, // Plain text may already contain escape sequences.
+            StripAnsiMode::Auto => true,
+            _ => false,
         };
 
         Ok(InteractivePrinter {
@@ -241,6 +318,8 @@ impl<'a> InteractivePrinter<'a> {
             line_changes,
             highlighter_from_set,
             background_color_highlight,
+            consecutive_empty_lines: 0,
+            strip_ansi,
         })
     }
 
@@ -281,12 +360,20 @@ impl<'a> InteractivePrinter<'a> {
             " ".repeat(self.panel_width - 1 - text_truncated.len())
         );
         if self.config.style_components.grid() {
-            format!("{} │ ", text_filled)
+            format!("{text_filled} │ ")
         } else {
             text_filled
         }
     }
 
+    fn get_header_component_indent_length(&self) -> usize {
+        if self.config.style_components.grid() && self.panel_width > 0 {
+            self.panel_width + 2
+        } else {
+            self.panel_width
+        }
+    }
+
     fn print_header_component_indent(&mut self, handle: &mut OutputHandle) -> Result<()> {
         if self.config.style_components.grid() {
             write!(
@@ -302,6 +389,55 @@ impl<'a> InteractivePrinter<'a> {
         }
     }
 
+    fn print_header_component_with_indent(
+        &mut self,
+        handle: &mut OutputHandle,
+        content: &str,
+    ) -> Result<()> {
+        self.print_header_component_indent(handle)?;
+        writeln!(handle, "{content}")
+    }
+
+    fn print_header_multiline_component(
+        &mut self,
+        handle: &mut OutputHandle,
+        content: &str,
+    ) -> Result<()> {
+        let mut content = content;
+        let content_width = self.config.term_width - self.get_header_component_indent_length();
+        while content.len() > content_width {
+            let (content_line, remaining) = content.split_at(content_width);
+            self.print_header_component_with_indent(handle, content_line)?;
+            content = remaining;
+        }
+        self.print_header_component_with_indent(handle, content)
+    }
+
+    fn highlight_regions_for_line<'b>(
+        &mut self,
+        line: &'b str,
+    ) -> Result<Vec<(syntect::highlighting::Style, &'b str)>> {
+        let highlighter_from_set = match self.highlighter_from_set {
+            Some(ref mut highlighter_from_set) => highlighter_from_set,
+            _ => return Ok(vec![(EMPTY_SYNTECT_STYLE, line)]),
+        };
+
+        // skip syntax highlighting on long lines
+        let too_long = line.len() > 1024 * 16;
+
+        let for_highlighting: &str = if too_long { "\n" } else { line };
+
+        let mut highlighted_line = highlighter_from_set
+            .highlighter
+            .highlight_line(for_highlighting, highlighter_from_set.syntax_set)?;
+
+        if too_long {
+            highlighted_line[0].1 = &line;
+        }
+
+        Ok(highlighted_line)
+    }
+
     fn preprocess(&self, text: &str, cursor: &mut usize) -> String {
         if self.config.tab_width > 0 {
             return expand_tabs(text, self.config.tab_width, cursor);
@@ -324,7 +460,10 @@ impl<'a> Printer for InteractivePrinter<'a> {
         }
 
         if !self.config.style_components.header() {
-            if Some(ContentType::BINARY) == self.content_type && !self.config.show_nonprintable {
+            if Some(ContentType::BINARY) == self.content_type
+                && !self.config.show_nonprintable
+                && !matches!(self.config.binary, BinaryBehavior::AsText)
+            {
                 writeln!(
                     handle,
                     "{}: Binary content from {} will not be printed to the terminal \
@@ -377,34 +516,38 @@ impl<'a> Printer for InteractivePrinter<'a> {
             }
         }
 
-        header_components.iter().try_for_each(|component| {
-            self.print_header_component_indent(handle)?;
-
-            match component {
-                StyleComponent::HeaderFilename => writeln!(
-                    handle,
-                    "{}{}{}",
-                    description
-                        .kind()
-                        .map(|kind| format!("{}: ", kind))
-                        .unwrap_or_else(|| "".into()),
-                    self.colors.header_value.paint(description.title()),
-                    mode
-                ),
-
+        header_components
+            .iter()
+            .try_for_each(|component| match component {
+                StyleComponent::HeaderFilename => {
+                    let header_filename = format!(
+                        "{}{}{}",
+                        description
+                            .kind()
+                            .map(|kind| format!("{kind}: "))
+                            .unwrap_or_else(|| "".into()),
+                        self.colors.header_value.paint(description.title()),
+                        mode
+                    );
+                    self.print_header_multiline_component(handle, &header_filename)
+                }
                 StyleComponent::HeaderFilesize => {
                     let bsize = metadata
                         .size
                         .map(|s| format!("{}", ByteSize(s)))
                         .unwrap_or_else(|| "-".into());
-                    writeln!(handle, "Size: {}", self.colors.header_value.paint(bsize))
+                    let header_filesize =
+                        format!("Size: {}", self.colors.header_value.paint(bsize));
+                    self.print_header_multiline_component(handle, &header_filesize)
                 }
                 _ => Ok(()),
-            }
-        })?;
+            })?;
 
         if self.config.style_components.grid() {
-            if self.content_type.map_or(false, |c| c.is_text()) || self.config.show_nonprintable {
+            if self.content_type.map_or(false, |c| c.is_text())
+                || self.config.show_nonprintable
+                || matches!(self.config.binary, BinaryBehavior::AsText)
+            {
                 self.print_horizontal_line(handle, '┼')?;
             } else {
                 self.print_horizontal_line(handle, '┴')?;
@@ -416,7 +559,9 @@ impl<'a> Printer for InteractivePrinter<'a> {
 
     fn print_footer(&mut self, handle: &mut OutputHandle, _input: &OpenedInput) -> Result<()> {
         if self.config.style_components.grid()
-            && (self.content_type.map_or(false, |c| c.is_text()) || self.config.show_nonprintable)
+            && (self.content_type.map_or(false, |c| c.is_text())
+                || self.config.show_nonprintable
+                || matches!(self.config.binary, BinaryBehavior::AsText))
         {
             self.print_horizontal_line(handle, '┴')
         } else {
@@ -442,7 +587,7 @@ impl<'a> Printer for InteractivePrinter<'a> {
             "{}",
             self.colors
                 .grid
-                .paint(format!("{}{}{}{}", panel, snip_left, title, snip_right))
+                .paint(format!("{panel}{snip_left}{title}{snip_right}"))
         )?;
 
         Ok(())
@@ -463,8 +608,10 @@ impl<'a> Printer for InteractivePrinter<'a> {
             )
             .into()
         } else {
-            match self.content_type {
-                Some(ContentType::BINARY) | None => {
+            let mut line = match self.content_type {
+                Some(ContentType::BINARY) | None
+                    if !matches!(self.config.binary, BinaryBehavior::AsText) =>
+                {
                     return Ok(());
                 }
                 Some(ContentType::UTF_16LE) => UTF_16LE.decode_with_bom_removal(line_buffer).0,
@@ -480,37 +627,33 @@ impl<'a> Printer for InteractivePrinter<'a> {
                         line
                     }
                 }
-            }
-        };
-
-        let regions = {
-            let highlighter_from_set = match self.highlighter_from_set {
-                Some(ref mut highlighter_from_set) => highlighter_from_set,
-                _ => {
-                    return Ok(());
-                }
             };
 
-            // skip syntax highlighting on long lines
-            let too_long = line.len() > 1024 * 16;
-
-            let for_highlighting: &str = if too_long { "\n" } else { &line };
-
-            let mut highlighted_line = highlighter_from_set
-                .highlighter
-                .highlight_line(for_highlighting, highlighter_from_set.syntax_set)?;
-
-            if too_long {
-                highlighted_line[0].1 = &line;
+            // If ANSI escape sequences are supposed to be stripped, do it before syntax highlighting.
+            if self.strip_ansi {
+                line = strip_ansi(&line).into()
             }
 
-            highlighted_line
+            line
         };
 
+        let regions = self.highlight_regions_for_line(&line)?;
         if out_of_range {
             return Ok(());
         }
 
+        // Skip squeezed lines.
+        if let Some(squeeze_limit) = self.config.squeeze_lines {
+            if line.trim_end_matches(|c| c == '\r' || c == '\n').is_empty() {
+                self.consecutive_empty_lines += 1;
+                if self.consecutive_empty_lines > squeeze_limit {
+                    return Ok(());
+                }
+            } else {
+                self.consecutive_empty_lines = 0;
+            }
+        }
+
         let mut cursor: usize = 0;
         let mut cursor_max: usize = self.config.term_width;
         let mut cursor_total: usize = 0;
@@ -521,7 +664,7 @@ impl<'a> Printer for InteractivePrinter<'a> {
             self.config.highlighted_lines.0.check(line_number) == RangeCheckResult::InRange;
 
         if highlight_this_line && self.config.theme == "ansi" {
-            self.ansi_style.update("^[4m");
+            self.ansi_style.update(ANSI_UNDERLINE_ENABLE);
         }
 
         let background_color = self
@@ -548,23 +691,17 @@ impl<'a> Printer for InteractivePrinter<'a> {
             let italics = self.config.use_italic_text;
 
             for &(style, region) in &regions {
-                let ansi_iterator = AnsiCodeIterator::new(region);
+                let ansi_iterator = EscapeSequenceIterator::new(region);
                 for chunk in ansi_iterator {
                     match chunk {
-                        // ANSI escape passthrough.
-                        (ansi, true) => {
-                            self.ansi_style.update(ansi);
-                            write!(handle, "{}", ansi)?;
-                        }
-
                         // Regular text.
-                        (text, false) => {
-                            let text = &*self.preprocess(text, &mut cursor_total);
+                        EscapeSequence::Text(text) => {
+                            let text = self.preprocess(text, &mut cursor_total);
                             let text_trimmed = text.trim_end_matches(|c| c == '\r' || c == '\n');
 
                             write!(
                                 handle,
-                                "{}",
+                                "{}{}",
                                 as_terminal_escaped(
                                     style,
                                     &format!("{}{}", self.ansi_style, text_trimmed),
@@ -572,9 +709,11 @@ impl<'a> Printer for InteractivePrinter<'a> {
                                     colored_output,
                                     italics,
                                     background_color
-                                )
+                                ),
+                                self.ansi_style.to_reset_sequence(),
                             )?;
 
+                            // Pad the rest of the line.
                             if text.len() != text_trimmed.len() {
                                 if let Some(background_color) = background_color {
                                     let ansi_style = Style {
@@ -592,6 +731,12 @@ impl<'a> Printer for InteractivePrinter<'a> {
                                 write!(handle, "{}", &text[text_trimmed.len()..])?;
                             }
                         }
+
+                        // ANSI escape passthrough.
+                        _ => {
+                            write!(handle, "{}", chunk.raw())?;
+                            self.ansi_style.update(chunk);
+                        }
                     }
                 }
             }
@@ -601,17 +746,11 @@ impl<'a> Printer for InteractivePrinter<'a> {
             }
         } else {
             for &(style, region) in &regions {
-                let ansi_iterator = AnsiCodeIterator::new(region);
+                let ansi_iterator = EscapeSequenceIterator::new(region);
                 for chunk in ansi_iterator {
                     match chunk {
-                        // ANSI escape passthrough.
-                        (ansi, true) => {
-                            self.ansi_style.update(ansi);
-                            write!(handle, "{}", ansi)?;
-                        }
-
                         // Regular text.
-                        (text, false) => {
+                        EscapeSequence::Text(text) => {
                             let text = self.preprocess(
                                 text.trim_end_matches(|c| c == '\r' || c == '\n'),
                                 &mut cursor_total,
@@ -654,7 +793,7 @@ impl<'a> Printer for InteractivePrinter<'a> {
                                     // It wraps.
                                     write!(
                                         handle,
-                                        "{}\n{}",
+                                        "{}{}\n{}",
                                         as_terminal_escaped(
                                             style,
                                             &format!("{}{}", self.ansi_style, line_buf),
@@ -663,6 +802,7 @@ impl<'a> Printer for InteractivePrinter<'a> {
                                             self.config.use_italic_text,
                                             background_color
                                         ),
+                                        self.ansi_style.to_reset_sequence(),
                                         panel_wrap.clone().unwrap()
                                     )?;
 
@@ -691,6 +831,12 @@ impl<'a> Printer for InteractivePrinter<'a> {
                                 )
                             )?;
                         }
+
+                        // ANSI escape passthrough.
+                        _ => {
+                            write!(handle, "{}", chunk.raw())?;
+                            self.ansi_style.update(chunk);
+                        }
                     }
                 }
             }
@@ -711,8 +857,8 @@ impl<'a> Printer for InteractivePrinter<'a> {
         }
 
         if highlight_this_line && self.config.theme == "ansi" {
-            self.ansi_style.update("^[24m");
-            write!(handle, "\x1B[24m")?;
+            write!(handle, "{}", ANSI_UNDERLINE_DISABLE.raw())?;
+            self.ansi_style.update(ANSI_UNDERLINE_DISABLE);
         }
 
         Ok(())

src/style.rs

@@ -80,7 +80,7 @@ impl FromStr for StyleComponent {
             "full" => Ok(StyleComponent::Full),
             "default" => Ok(StyleComponent::Default),
             "plain" => Ok(StyleComponent::Plain),
-            _ => Err(format!("Unknown style '{}'", s).into()),
+            _ => Err(format!("Unknown style '{s}'").into()),
         }
     }
 }
@@ -138,3 +138,227 @@ impl StyleComponents {
         self.0.clear();
     }
 }
+
+#[derive(Debug, PartialEq)]
+enum ComponentAction {
+    Override,
+    Add,
+    Remove,
+}
+
+impl ComponentAction {
+    fn extract_from_str(string: &str) -> (ComponentAction, &str) {
+        match string.chars().next() {
+            Some('-') => (ComponentAction::Remove, string.strip_prefix('-').unwrap()),
+            Some('+') => (ComponentAction::Add, string.strip_prefix('+').unwrap()),
+            _ => (ComponentAction::Override, string),
+        }
+    }
+}
+
+/// A list of [StyleComponent] that can be parsed from a string.
+pub struct StyleComponentList(Vec<(ComponentAction, StyleComponent)>);
+
+impl StyleComponentList {
+    fn expand_into(&self, components: &mut HashSet<StyleComponent>, interactive_terminal: bool) {
+        for (action, component) in self.0.iter() {
+            let subcomponents = component.components(interactive_terminal);
+
+            use ComponentAction::*;
+            match action {
+                Override | Add => components.extend(subcomponents),
+                Remove => components.retain(|c| !subcomponents.contains(c)),
+            }
+        }
+    }
+
+    /// Returns `true` if any component in the list was not prefixed with `+` or `-`.
+    fn contains_override(&self) -> bool {
+        self.0.iter().any(|(a, _)| *a == ComponentAction::Override)
+    }
+
+    /// Combines multiple [StyleComponentList]s into a single [StyleComponents] set.
+    ///
+    /// ## Precedence
+    /// The most recent list will take precedence and override all previous lists
+    /// unless it only contains components prefixed with `-` or `+`. When this
+    /// happens, the list's components will be merged into the previous list.
+    ///
+    /// ## Example
+    /// ```text
+    /// [numbers,grid] + [header,changes]  -> [header,changes]
+    /// [numbers,grid] + [+header,-grid]   -> [numbers,header]
+    /// ```
+    ///
+    /// ## Parameters
+    ///  - `with_default`: If true, the styles lists will build upon the StyleComponent::Auto style.
+    pub fn to_components(
+        lists: impl IntoIterator<Item = StyleComponentList>,
+        interactive_terminal: bool,
+        with_default: bool,
+    ) -> StyleComponents {
+        let mut components: HashSet<StyleComponent> = HashSet::new();
+        if with_default {
+            components.extend(StyleComponent::Auto.components(interactive_terminal))
+        }
+
+        StyleComponents(lists.into_iter().fold(components, |mut components, list| {
+            if list.contains_override() {
+                components.clear();
+            }
+
+            list.expand_into(&mut components, interactive_terminal);
+            components
+        }))
+    }
+}
+
+impl Default for StyleComponentList {
+    fn default() -> Self {
+        StyleComponentList(vec![(ComponentAction::Override, StyleComponent::Default)])
+    }
+}
+
+impl FromStr for StyleComponentList {
+    type Err = Error;
+
+    fn from_str(s: &str) -> Result<Self> {
+        Ok(StyleComponentList(
+            s.split(",")
+                .map(|s| ComponentAction::extract_from_str(s)) // If the component starts with "-", it's meant to be removed
+                .map(|(a, s)| Ok((a, StyleComponent::from_str(s)?)))
+                .collect::<Result<Vec<(ComponentAction, StyleComponent)>>>()?,
+        ))
+    }
+}
+
+#[cfg(test)]
+mod test {
+    use std::collections::HashSet;
+    use std::str::FromStr;
+
+    use super::ComponentAction::*;
+    use super::StyleComponent;
+    use super::StyleComponent::*;
+    use super::StyleComponentList;
+
+    #[test]
+    pub fn style_component_list_parse() {
+        assert_eq!(
+            StyleComponentList::from_str("grid,+numbers,snip,-snip,header")
+                .expect("no error")
+                .0,
+            vec![
+                (Override, Grid),
+                (Add, LineNumbers),
+                (Override, Snip),
+                (Remove, Snip),
+                (Override, Header),
+            ]
+        );
+
+        assert!(StyleComponentList::from_str("not-a-component").is_err());
+        assert!(StyleComponentList::from_str("grid,not-a-component").is_err());
+        assert!(StyleComponentList::from_str("numbers,-not-a-component").is_err());
+    }
+
+    #[test]
+    pub fn style_component_list_to_components() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![StyleComponentList::from_str("grid,numbers").expect("no error")],
+                false,
+                false
+            )
+            .0,
+            HashSet::from([Grid, LineNumbers])
+        );
+    }
+
+    #[test]
+    pub fn style_component_list_to_components_removes_negated() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![StyleComponentList::from_str("grid,numbers,-grid").expect("no error")],
+                false,
+                false
+            )
+            .0,
+            HashSet::from([LineNumbers])
+        );
+    }
+
+    #[test]
+    pub fn style_component_list_to_components_expands_subcomponents() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![StyleComponentList::from_str("full").expect("no error")],
+                false,
+                false
+            )
+            .0,
+            HashSet::from_iter(Full.components(true).to_owned())
+        );
+    }
+
+    #[test]
+    pub fn style_component_list_expand_negates_subcomponents() {
+        assert!(!StyleComponentList::to_components(
+            vec![StyleComponentList::from_str("full,-numbers").expect("no error")],
+            true,
+            false
+        )
+        .numbers());
+    }
+
+    #[test]
+    pub fn style_component_list_to_components_precedence_overrides_previous_lists() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![
+                    StyleComponentList::from_str("grid").expect("no error"),
+                    StyleComponentList::from_str("numbers").expect("no error"),
+                ],
+                false,
+                false
+            )
+            .0,
+            HashSet::from([LineNumbers])
+        );
+    }
+
+    #[test]
+    pub fn style_component_list_to_components_precedence_merges_previous_lists() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![
+                    StyleComponentList::from_str("grid,header").expect("no error"),
+                    StyleComponentList::from_str("-grid").expect("no error"),
+                    StyleComponentList::from_str("+numbers").expect("no error"),
+                ],
+                false,
+                false
+            )
+            .0,
+            HashSet::from([HeaderFilename, LineNumbers])
+        );
+    }
+
+    #[test]
+    pub fn style_component_list_default_builds_on_auto() {
+        assert_eq!(
+            StyleComponentList::to_components(
+                vec![StyleComponentList::from_str("-numbers").expect("no error"),],
+                true,
+                true
+            )
+            .0,
+            {
+                let mut expected: HashSet<StyleComponent> = HashSet::new();
+                expected.extend(Auto.components(true));
+                expected.remove(&LineNumbers);
+                expected
+            }
+        );
+    }
+}

src/syntax_mapping.rs

@@ -1,12 +1,31 @@
-use std::path::Path;
-
-use crate::error::Result;
-use ignored_suffixes::IgnoredSuffixes;
+use std::{
+    path::Path,
+    sync::{
+        atomic::{AtomicBool, Ordering},
+        Arc,
+    },
+    thread,
+};
 
 use globset::{Candidate, GlobBuilder, GlobMatcher};
+use once_cell::sync::Lazy;
 
+use crate::error::Result;
+use builtin::BUILTIN_MAPPINGS;
+use ignored_suffixes::IgnoredSuffixes;
+
+mod builtin;
 pub mod ignored_suffixes;
 
+fn make_glob_matcher(from: &str) -> Result<GlobMatcher> {
+    let matcher = GlobBuilder::new(from)
+        .case_insensitive(true)
+        .literal_separator(true)
+        .build()?
+        .compile_matcher();
+    Ok(matcher)
+}
+
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
 #[non_exhaustive]
 pub enum MappingTarget<'a> {
@@ -29,204 +48,108 @@ pub enum MappingTarget<'a> {
 
 #[derive(Debug, Clone, Default)]
 pub struct SyntaxMapping<'a> {
-    mappings: Vec<(GlobMatcher, MappingTarget<'a>)>,
+    /// User-defined mappings at run time.
+    ///
+    /// Rules in front have precedence.
+    custom_mappings: Vec<(GlobMatcher, MappingTarget<'a>)>,
+
     pub(crate) ignored_suffixes: IgnoredSuffixes<'a>,
+
+    /// A flag to halt glob matcher building, which is offloaded to another thread.
+    ///
+    /// We have this so that we can signal the thread to halt early when appropriate.
+    halt_glob_build: Arc<AtomicBool>,
+}
+
+impl<'a> Drop for SyntaxMapping<'a> {
+    fn drop(&mut self) {
+        // signal the offload thread to halt early
+        self.halt_glob_build.store(true, Ordering::Relaxed);
+    }
 }
 
 impl<'a> SyntaxMapping<'a> {
-    pub fn empty() -> SyntaxMapping<'a> {
+    pub fn new() -> SyntaxMapping<'a> {
         Default::default()
     }
 
-    pub fn builtin() -> SyntaxMapping<'a> {
-        let mut mapping = Self::empty();
-        mapping.insert("*.h", MappingTarget::MapTo("C++")).unwrap();
-        mapping
-            .insert(".clang-format", MappingTarget::MapTo("YAML"))
-            .unwrap();
-        mapping.insert("*.fs", MappingTarget::MapTo("F#")).unwrap();
-        mapping
-            .insert("build", MappingTarget::MapToUnknown)
-            .unwrap();
-        mapping
-            .insert("**/.ssh/config", MappingTarget::MapTo("SSH Config"))
-            .unwrap();
-        mapping
-            .insert(
-                "**/bat/config",
-                MappingTarget::MapTo("Bourne Again Shell (bash)"),
-            )
-            .unwrap();
-        mapping
-            .insert(
-                "/etc/profile",
-                MappingTarget::MapTo("Bourne Again Shell (bash)"),
-            )
-            .unwrap();
-        mapping
-            .insert(
-                "os-release",
-                MappingTarget::MapTo("Bourne Again Shell (bash)"),
-            )
-            .unwrap();
-        mapping
-            .insert("*.pac", MappingTarget::MapTo("JavaScript (Babel)"))
-            .unwrap();
-        mapping
-            .insert("fish_history", MappingTarget::MapTo("YAML"))
-            .unwrap();
-
-        for glob in ["*.jsonl", "*.sarif"] {
-            mapping.insert(glob, MappingTarget::MapTo("JSON")).unwrap();
-        }
-
-        // See #2151, https://nmap.org/book/nse-language.html
-        mapping
-            .insert("*.nse", MappingTarget::MapTo("Lua"))
-            .unwrap();
-
-        // See #1008
-        mapping
-            .insert("rails", MappingTarget::MapToUnknown)
-            .unwrap();
-
-        mapping
-            .insert("Containerfile", MappingTarget::MapTo("Dockerfile"))
-            .unwrap();
-
-        mapping
-            .insert("*.ksh", MappingTarget::MapTo("Bourne Again Shell (bash)"))
-            .unwrap();
-
-        // Nginx and Apache syntax files both want to style all ".conf" files
-        // see #1131 and #1137
-        mapping
-            .insert("*.conf", MappingTarget::MapExtensionToUnknown)
-            .unwrap();
-
-        for glob in &[
-            "/etc/nginx/**/*.conf",
-            "/etc/nginx/sites-*/**/*",
-            "nginx.conf",
-            "mime.types",
-        ] {
-            mapping.insert(glob, MappingTarget::MapTo("nginx")).unwrap();
-        }
-
-        for glob in &[
-            "/etc/apache2/**/*.conf",
-            "/etc/apache2/sites-*/**/*",
-            "httpd.conf",
-        ] {
-            mapping
-                .insert(glob, MappingTarget::MapTo("Apache Conf"))
-                .unwrap();
-        }
-
-        for glob in &[
-            "**/systemd/**/*.conf",
-            "**/systemd/**/*.example",
-            "*.automount",
-            "*.device",
-            "*.dnssd",
-            "*.link",
-            "*.mount",
-            "*.netdev",
-            "*.network",
-            "*.nspawn",
-            "*.path",
-            "*.service",
-            "*.scope",
-            "*.slice",
-            "*.socket",
-            "*.swap",
-            "*.target",
-            "*.timer",
-        ] {
-            mapping.insert(glob, MappingTarget::MapTo("INI")).unwrap();
-        }
-
-        // unix mail spool
-        for glob in &["/var/spool/mail/*", "/var/mail/*"] {
-            mapping.insert(glob, MappingTarget::MapTo("Email")).unwrap()
-        }
-
-        // pacman hooks
-        mapping
-            .insert("*.hook", MappingTarget::MapTo("INI"))
-            .unwrap();
-
-        mapping
-            .insert("*.ron", MappingTarget::MapTo("Rust"))
-            .unwrap();
-
-        // Global git config files rooted in `$XDG_CONFIG_HOME/git/` or `$HOME/.config/git/`
-        // See e.g. https://git-scm.com/docs/git-config#FILES
-        match (
-            std::env::var_os("XDG_CONFIG_HOME").filter(|val| !val.is_empty()),
-            std::env::var_os("HOME")
-                .filter(|val| !val.is_empty())
-                .map(|home| Path::new(&home).join(".config")),
-        ) {
-            (Some(xdg_config_home), Some(default_config_home))
-                if xdg_config_home == default_config_home => {
-                insert_git_config_global(&mut mapping, &xdg_config_home)
+    /// Start a thread to build the glob matchers for all builtin mappings.
+    ///
+    /// The use of this function while not necessary, is useful to speed up startup
+    /// times by starting this work early in parallel.
+    ///
+    /// The thread halts if/when `halt_glob_build` is set to true.
+    pub fn start_offload_build_all(&self) {
+        let halt = Arc::clone(&self.halt_glob_build);
+        thread::spawn(move || {
+            for (matcher, _) in BUILTIN_MAPPINGS.iter() {
+                if halt.load(Ordering::Relaxed) {
+                    break;
+                }
+                Lazy::force(matcher);
             }
-            (Some(xdg_config_home), Some(default_config_home)) /* else guard */ => {
-                insert_git_config_global(&mut mapping, &xdg_config_home);
-                insert_git_config_global(&mut mapping, &default_config_home)
-            }
-            (Some(config_home), None) => insert_git_config_global(&mut mapping, &config_home),
-            (None, Some(config_home)) => insert_git_config_global(&mut mapping, &config_home),
-            (None, None) => (),
-        };
-
-        fn insert_git_config_global(mapping: &mut SyntaxMapping, config_home: impl AsRef<Path>) {
-            let git_config_path = config_home.as_ref().join("git");
-
-            mapping
-                .insert(
-                    &git_config_path.join("config").to_string_lossy(),
-                    MappingTarget::MapTo("Git Config"),
-                )
-                .ok();
-
-            mapping
-                .insert(
-                    &git_config_path.join("ignore").to_string_lossy(),
-                    MappingTarget::MapTo("Git Ignore"),
-                )
-                .ok();
-
-            mapping
-                .insert(
-                    &git_config_path.join("attributes").to_string_lossy(),
-                    MappingTarget::MapTo("Git Attributes"),
-                )
-                .ok();
-        }
-
-        mapping
+        });
+        // Note that this thread is not joined upon completion because there's
+        // no shared resources that need synchronization to be safely dropped.
+        // If we later add code into this thread that requires interesting
+        // resources (e.g. IO), it would be a good idea to store the handle
+        // and join it on drop.
     }
 
     pub fn insert(&mut self, from: &str, to: MappingTarget<'a>) -> Result<()> {
-        let glob = GlobBuilder::new(from)
-            .case_insensitive(true)
-            .literal_separator(true)
-            .build()?;
-        self.mappings.push((glob.compile_matcher(), to));
+        let matcher = make_glob_matcher(from)?;
+        self.custom_mappings.push((matcher, to));
         Ok(())
     }
 
-    pub fn mappings(&self) -> &[(GlobMatcher, MappingTarget<'a>)] {
-        &self.mappings
+    /// Returns an iterator over all mappings. User-defined mappings are listed
+    /// before builtin mappings; mappings in front have higher precedence.
+    ///
+    /// Builtin mappings' `GlobMatcher`s are lazily compiled.
+    ///
+    /// Note that this function only returns mappings that are valid under the
+    /// current environment. For details see [`Self::builtin_mappings`].
+    pub fn all_mappings(&self) -> impl Iterator<Item = (&GlobMatcher, &MappingTarget<'a>)> {
+        self.custom_mappings()
+            .iter()
+            .map(|(matcher, target)| (matcher, target)) // as_ref
+            .chain(
+                // we need a map with a closure to "do" the lifetime variance
+                // see: https://discord.com/channels/273534239310479360/1120124565591425034/1170543402870382653
+                // also, clippy false positive:
+                // see: https://github.com/rust-lang/rust-clippy/issues/9280
+                #[allow(clippy::map_identity)]
+                self.builtin_mappings().map(|rule| rule),
+            )
     }
 
-    pub(crate) fn get_syntax_for(&self, path: impl AsRef<Path>) -> Option<MappingTarget<'a>> {
+    /// Returns an iterator over all valid builtin mappings. Mappings in front
+    /// have higher precedence.
+    ///
+    /// The `GlabMatcher`s are lazily compiled.
+    ///
+    /// Mappings that are invalid under the current environment (i.e. rule
+    /// requires environment variable(s) that is unset, or the joined string
+    /// after variable(s) replacement is not a valid glob expression) are
+    /// ignored.
+    pub fn builtin_mappings(
+        &self,
+    ) -> impl Iterator<Item = (&'static GlobMatcher, &'static MappingTarget<'static>)> {
+        BUILTIN_MAPPINGS
+            .iter()
+            .filter_map(|(matcher, target)| matcher.as_ref().map(|glob| (glob, target)))
+    }
+
+    /// Returns all user-defined mappings.
+    pub fn custom_mappings(&self) -> &[(GlobMatcher, MappingTarget<'a>)] {
+        &self.custom_mappings
+    }
+
+    pub fn get_syntax_for(&self, path: impl AsRef<Path>) -> Option<MappingTarget<'a>> {
         // Try matching on the file name as-is.
         let candidate = Candidate::new(&path);
         let candidate_filename = path.as_ref().file_name().map(Candidate::new);
-        for (ref glob, ref syntax) in self.mappings.iter().rev() {
+        for (glob, syntax) in self.all_mappings() {
             if glob.is_match_candidate(&candidate)
                 || candidate_filename
                     .as_ref()
@@ -252,9 +175,46 @@ impl<'a> SyntaxMapping<'a> {
 #[cfg(test)]
 mod tests {
     use super::*;
+
     #[test]
-    fn basic() {
-        let mut map = SyntaxMapping::empty();
+    fn builtin_mappings_work() {
+        let map = SyntaxMapping::new();
+
+        assert_eq!(
+            map.get_syntax_for("/path/to/build"),
+            Some(MappingTarget::MapToUnknown)
+        );
+    }
+
+    #[test]
+    fn all_fixed_builtin_mappings_can_compile() {
+        let map = SyntaxMapping::new();
+
+        // collect call evaluates all lazy closures
+        // fixed builtin mappings will panic if they fail to compile
+        let _mappings = map.builtin_mappings().collect::<Vec<_>>();
+    }
+
+    #[test]
+    fn builtin_mappings_matcher_only_compile_once() {
+        let map = SyntaxMapping::new();
+
+        let two_iterations: Vec<_> = (0..2)
+            .map(|_| {
+                // addresses of every matcher
+                map.builtin_mappings()
+                    .map(|(matcher, _)| matcher as *const _ as usize)
+                    .collect::<Vec<_>>()
+            })
+            .collect();
+
+        // if the matchers are only compiled once, their address should remain the same
+        assert_eq!(two_iterations[0], two_iterations[1]);
+    }
+
+    #[test]
+    fn custom_mappings_work() {
+        let mut map = SyntaxMapping::new();
         map.insert("/path/to/Cargo.lock", MappingTarget::MapTo("TOML"))
             .ok();
         map.insert("/path/to/.ignore", MappingTarget::MapTo("Git Ignore"))
@@ -273,52 +233,32 @@ mod tests {
     }
 
     #[test]
-    fn user_can_override_builtin_mappings() {
-        let mut map = SyntaxMapping::builtin();
+    fn custom_mappings_override_builtin() {
+        let mut map = SyntaxMapping::new();
 
         assert_eq!(
-            map.get_syntax_for("/etc/profile"),
-            Some(MappingTarget::MapTo("Bourne Again Shell (bash)"))
+            map.get_syntax_for("/path/to/httpd.conf"),
+            Some(MappingTarget::MapTo("Apache Conf"))
         );
-        map.insert("/etc/profile", MappingTarget::MapTo("My Syntax"))
+        map.insert("httpd.conf", MappingTarget::MapTo("My Syntax"))
             .ok();
         assert_eq!(
-            map.get_syntax_for("/etc/profile"),
+            map.get_syntax_for("/path/to/httpd.conf"),
             Some(MappingTarget::MapTo("My Syntax"))
         );
     }
 
     #[test]
-    fn builtin_mappings() {
-        let map = SyntaxMapping::builtin();
+    fn custom_mappings_precedence() {
+        let mut map = SyntaxMapping::new();
 
+        map.insert("/path/to/foo", MappingTarget::MapTo("alpha"))
+            .ok();
+        map.insert("/path/to/foo", MappingTarget::MapTo("bravo"))
+            .ok();
         assert_eq!(
-            map.get_syntax_for("/path/to/build"),
-            Some(MappingTarget::MapToUnknown)
+            map.get_syntax_for("/path/to/foo"),
+            Some(MappingTarget::MapTo("alpha"))
         );
     }
-
-    #[test]
-    /// verifies that SyntaxMapping::builtin() doesn't repeat `Glob`-based keys
-    fn no_duplicate_builtin_keys() {
-        let mappings = SyntaxMapping::builtin().mappings;
-        for i in 0..mappings.len() {
-            let tail = mappings[i + 1..].into_iter();
-            let (dupl, _): (Vec<_>, Vec<_>) =
-                tail.partition(|item| item.0.glob() == mappings[i].0.glob());
-
-            // emit repeats on failure
-            assert_eq!(
-                dupl.len(),
-                0,
-                "Glob pattern `{}` mapped to multiple: {:?}",
-                mappings[i].0.glob().glob(),
-                {
-                    let (_, mut dupl_targets): (Vec<GlobMatcher>, Vec<MappingTarget>) =
-                        dupl.into_iter().cloned().unzip();
-                    dupl_targets.push(mappings[i].1)
-                },
-            )
-        }
-    }
 }

src/syntax_mapping/builtin.rs

@@ -0,0 +1,91 @@
+use std::env;
+
+use globset::GlobMatcher;
+use once_cell::sync::Lazy;
+
+use crate::syntax_mapping::{make_glob_matcher, MappingTarget};
+
+// Static syntax mappings generated from /src/syntax_mapping/builtins/ by the
+// build script (/build/syntax_mapping.rs).
+include!(concat!(
+    env!("OUT_DIR"),
+    "/codegen_static_syntax_mappings.rs"
+));
+
+// The defined matcher strings are analysed at compile time and converted into
+// lazily-compiled `GlobMatcher`s. This is so that the string searches are moved
+// from run time to compile time, thus improving startup performance.
+//
+// To any future maintainer (including possibly myself) wondering why there is
+// not a `BuiltinMatcher` enum that looks like this:
+//
+// ```
+// enum BuiltinMatcher {
+//     Fixed(&'static str),
+//     Dynamic(Lazy<Option<String>>),
+// }
+// ```
+//
+// Because there was. I tried it and threw it out.
+//
+// Naively looking at the problem from a distance, this may seem like a good
+// design (strongly typed etc. etc.). It would also save on compiled size by
+// extracting out common behaviour into functions. But while actually
+// implementing the lazy matcher compilation logic, I realised that it's most
+// convenient for `BUILTIN_MAPPINGS` to have the following type:
+//
+// `[(Lazy<Option<GlobMatcher>>, MappingTarget); N]`
+//
+// The benefit for this is that operations like listing all builtin mappings
+// would be effectively memoised. The caller would not have to compile another
+// `GlobMatcher` for rules that they have previously visited.
+//
+// Unfortunately, this means we are going to have to store a distinct closure
+// for each rule anyway, which makes a `BuiltinMatcher` enum a pointless layer
+// of indirection.
+//
+// In the current implementation, the closure within each generated rule simply
+// calls either `build_matcher_fixed` or `build_matcher_dynamic`, depending on
+// whether the defined matcher contains dynamic segments or not.
+
+/// Compile a fixed glob string into a glob matcher.
+///
+/// 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")
+}
+
+/// Join a list of matcher segments to create a glob string, replacing all
+/// environment variables, then compile to a glob matcher.
+///
+/// Returns `None` if any replacement fails, or if the joined glob string fails
+/// to compile.
+///
+/// Used internally by `Lazy<Option<GlobMatcher>>`'s lazy evaluation closure.
+fn build_matcher_dynamic(segs: &[MatcherSegment]) -> Option<GlobMatcher> {
+    // join segments
+    let mut buf = String::new();
+    for seg in segs {
+        match seg {
+            MatcherSegment::Text(s) => buf.push_str(s),
+            MatcherSegment::Env(var) => {
+                let replaced = env::var(var).ok()?;
+                buf.push_str(&replaced);
+            }
+        }
+    }
+    // compile glob matcher
+    let matcher = make_glob_matcher(&buf).ok()?;
+    Some(matcher)
+}
+
+/// A segment of a dynamic builtin matcher.
+///
+/// Used internally by `Lazy<Option<GlobMatcher>>`'s lazy evaluation closure.
+#[derive(Clone, Debug)]
+enum MatcherSegment {
+    Text(&'static str),
+    Env(&'static str),
+}

src/syntax_mapping/builtins/README.md

@@ -0,0 +1,116 @@
+# `/src/syntax_mapping/builtins`
+
+The files in this directory define path/name-based syntax mappings, which amend
+and take precedence over the extension/content-based syntax mappings provided by
+[syntect](https://github.com/trishume/syntect).
+
+## File organisation
+
+Each TOML file should describe the syntax mappings of a single application, or
+otherwise a set of logically-related rules.
+
+What defines "a single application" here is deliberately vague, since the
+file-splitting is purely for maintainability reasons. (Technically, we could
+just as well use a single TOML file.) So just use common sense.
+
+TOML files should reside in the corresponding subdirectory of the platform(s)
+that they intend to target. At compile time, the build script will go through
+each subdirectory that is applicable to the compilation target, collect the
+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").
+
+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".
+
+For example, if `foo-application` uses both TOML and YAML configuration files,
+we could write something like this:
+
+```toml
+# 30-foo-application.toml
+[mappings]
+"TOML" = [
+    # rules for TOML syntax go here
+    "/usr/share/foo-application/toml-config/*.conf",
+    "/etc/foo-application/toml-config/*.conf",
+]
+"YAML" = [
+    # rules for YAML syntax go here
+    # ...
+]
+```
+
+### Dynamic environment variable replacement
+
+In additional to the standard glob matcher syntax, rules also support dynamic
+replacement of environment variables at runtime. This allows us to concisely
+handle things like [XDG](https://specifications.freedesktop.org/basedir-spec/latest/).
+
+All environment variables intended to be replaced at runtime must be enclosed in
+`${}`, for example `"/foo/*/${YOUR_ENV}-suffix/*.log"`. Note that this is the
+**only** admissible syntax; other variable substitution syntaxes are not
+supported and will either cause a compile time error, or be treated as plain
+text.
+
+For example, if `foo-application` also supports per-user configuration files, we
+could write something like this:
+
+```toml
+# 30-foo-application.toml
+[mappings]
+"TOML" = [
+    # rules for TOML syntax go here
+    "/usr/share/foo-application/toml-config/*.conf",
+    "/etc/foo-application/toml-config/*.conf",
+    "${XDG_CONFIG_HOME}/foo-application/toml-config/*.conf",
+    "${HOME}/.config/foo-application/toml-config/*.conf",
+]
+"YAML" = [
+    # rules for YAML syntax go here
+    # ...
+]
+```
+
+If any environment variable replacement in a rule fails (for example when a
+variable is unset), or if the glob string after replacements is invalid, the
+entire rule will be ignored.
+
+### Explicitly mapping to unknown
+
+Sometimes it may be necessary to "unset" a particular syntect mapping - perhaps
+a syntax's matching rules are "too greedy", and is claiming files that it should
+not. In this case, there are two special identifiers:
+`MappingTarget::MapToUnknown` and `MappingTarget::MapExtensionToUnknown`
+(corresponding to the two variants of the `syntax_mapping::MappingTarget` enum).
+
+An example of this would be `*.conf` files in general. So we may write something
+like this:
+
+```toml
+# 99-unset-ambiguous-extensions.toml
+[mappings]
+"MappingTarget::MapExtensionToUnknown" = [
+    "*.conf",
+]
+```
+
+## Ordering
+
+At compile time, all TOML files applicable to the target are processed in
+lexicographical filename order. So `00-foo.toml` takes precedence over
+`10-bar.toml`, which takes precedence over `20-baz.toml`, and so on. Note that
+**only** the filenames of the TOML files are taken into account; the
+subdirectories they are placed in have no influence on ordering.
+
+This behaviour can be occasionally useful for creating high/low priority rules,
+such as in the aforementioned example of explicitly mapping `*.conf` files to
+unknown. Generally this should not be much of a concern though, since rules
+should be written as specifically as possible for each application.
+
+Rules within each TOML file are processed (and therefore matched) in the order
+in which they are defined. At runtime, the syntax selection algorithm will
+short-circuit and return the target of the first matching rule.

src/syntax_mapping/builtins/bsd-family/50-os-release.toml

@@ -0,0 +1,2 @@
+[mappings]
+"Bourne Again Shell (bash)" = ["/etc/os-release", "/var/run/os-release"]

src/syntax_mapping/builtins/common/50-apache.toml

@@ -0,0 +1,2 @@
+[mappings]
+"Apache Conf" = ["httpd.conf"]

src/syntax_mapping/builtins/common/50-aws-credentials.toml

@@ -0,0 +1,2 @@
+[mappings]
+"INI" = ["**/.aws/credentials", "**/.aws/config"]

src/syntax_mapping/builtins/common/50-bat.toml

@@ -0,0 +1,2 @@
+[mappings]
+"Bourne Again Shell (bash)" = ["**/bat/config"]

src/syntax_mapping/builtins/common/50-citation.toml

@@ -0,0 +1,2 @@
+[mappings]
+"YAML" = ["CITATION.cff"]

src/syntax_mapping/builtins/common/50-container.toml

@@ -0,0 +1,2 @@
+[mappings]
+"Dockerfile" = ["Containerfile"]

src/syntax_mapping/builtins/common/50-cpp.toml

@@ -0,0 +1,6 @@
+[mappings]
+"C++" = [
+    # probably better than the default Objective C mapping #877
+    "*.h",
+]
+"YAML" = [".clang-format"]

src/syntax_mapping/builtins/common/50-diff.toml

@@ -0,0 +1,3 @@
+# .debdiff is the extension used for diffs in Debian packaging
+[mappings]
+"Diff" = ["*.debdiff"]

src/syntax_mapping/builtins/common/50-f-sharp.toml

@@ -0,0 +1,2 @@
+[mappings]
+"F#" = ["*.fs"]

src/syntax_mapping/builtins/common/50-git.toml

@@ -0,0 +1,10 @@
+# Global git config files rooted in `$XDG_CONFIG_HOME/git/` or `$HOME/.config/git/`
+# See e.g. https://git-scm.com/docs/git-config#FILES
+
+[mappings]
+"Git Config" = ["${XDG_CONFIG_HOME}/git/config", "${HOME}/.config/git/config"]
+"Git Ignore" = ["${XDG_CONFIG_HOME}/git/ignore", "${HOME}/.config/git/ignore"]
+"Git Attributes" = [
+    "${XDG_CONFIG_HOME}/git/attributes",
+    "${HOME}/.config/git/attributes",
+]

src/syntax_mapping/builtins/common/50-json.toml

@@ -0,0 +1,3 @@
+# JSON Lines is a simple variation of JSON #2535
+[mappings]
+"JSON" = ["*.jsonl", "*.jsonc", "*.jsonld", "*.geojson"]

src/syntax_mapping/builtins/common/50-markdown.toml

@@ -0,0 +1,2 @@
+[mappings]
+"Markdown" = ["*.mkd"]

src/syntax_mapping/builtins/common/50-nginx.toml

@@ -0,0 +1,2 @@
+[mappings]
+"nginx" = ["nginx.conf", "mime.types"]

src/syntax_mapping/builtins/common/50-nmap.toml

@@ -0,0 +1,3 @@
+[mappings]
+# See #2151, https://nmap.org/book/nse-language.html
+"Lua" = ["*.nse"]

src/syntax_mapping/builtins/common/50-proxy-auto-config.toml

@@ -0,0 +1,3 @@
+# 1515
+[mappings]
+"JavaScript (Babel)" = ["*.pac"]

src/syntax_mapping/builtins/common/50-ron.toml

@@ -0,0 +1,3 @@
+# Rusty Object Notation #2427
+[mappings]
+"Rust" = ["*.ron"]

src/syntax_mapping/builtins/common/50-sarif.toml

@@ -0,0 +1,3 @@
+# SARIF is a format for reporting static analysis results #2695
+[mappings]
+"JSON" = ["*.sarif"]

src/syntax_mapping/builtins/common/50-ssh.toml

@@ -0,0 +1,2 @@
+[mappings]
+"SSH Config" = ["**/.ssh/config"]

src/syntax_mapping/builtins/common/99-unset-ambiguous-extensions.toml

@@ -0,0 +1,5 @@
+[mappings]
+"MappingTarget::MapExtensionToUnknown" = [
+    # common extension used for all kinds of formats
+    "*.conf",
+]

src/syntax_mapping/builtins/common/99-unset-ambiguous-filenames.toml

@@ -0,0 +1,7 @@
+[mappings]
+"MappingTarget::MapToUnknown" = [
+    # "NAnt Build File" should only match *.build files, not files named "build"
+    "build",
+    # "bin/rails" scripts in a Ruby project misidentified as HTML (Rails) #1008
+    "rails",
+]

src/syntax_mapping/builtins/common/xonsh.toml

@@ -0,0 +1,3 @@
+# Xonsh shell (https://xon.sh/)
+[mappings]
+"Python" = ["*.xsh", "*.xonshrc"]

src/syntax_mapping/builtins/linux/50-containers.toml

@@ -0,0 +1,8 @@
+# see https://github.com/containers/image/tree/main/docs
+[mappings]
+"TOML" = [
+    "/usr/share/containers/**/*.conf",
+    "/etc/containers/**/*.conf",
+    "${HOME}/.config/containers/**/*.conf",
+    "${XDG_CONFIG_HOME}/containers/**/*.conf",
+]