Merge branch 'master' into add-pipe-delimter-to-csv-syntax

2025-10-31 15:12:12 +00:00 · 2024-11-02 22:41:52 -07:00
parent 1942d40863 dbaa0a6d9e
commit d725443116
14 changed files with 154 additions and 34 deletions
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,7 @@
 - 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)
+- Add option `--binary=as-text` for printing binary content, see issue #2974 and PR #2976 (@einfachIrgendwer0815)

 ## Bugfixes

@@ -18,6 +19,7 @@
 - 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)

 ## Other

@@ -63,6 +65,7 @@
 - Map `*.mkd` files to `Markdown` syntax, see issue #3060 and PR #3061 (@einfachIrgendwer0815)
 - 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)

 ## Themes

@@ -112,6 +115,7 @@
 - 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

--- a/Cargo.lock
+++ b/Cargo.lock
@@ -279,12 +279,11 @@ checksum = "702fc72eb24e5a1e48ce58027a675bc24edd52096d5397d4aea7c6dd9eca0bd1"

 [[package]]
 name = "clircle"
-version = "0.5.0"
+version = "0.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ec0b92245ea62a7a751db4b0e4a583f8978e508077ef6de24fcc0d0dc5311a8d"
+checksum = "e136d50bd652710f1d86259a8977263d46bef0ab782a8bfc3887e44338517015"
 dependencies = [
 "cfg-if",
- "libc",
 "serde",
 "serde_derive",
 "winapi",
@@ -422,9 +421,9 @@ checksum = "a357d28ed41a50f9c765dbfe56cbc04a64e53e5fc58ba79fbc34c10ef3df831f"

 [[package]]
 name = "encoding_rs"
-version = "0.8.34"
+version = "0.8.35"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b45de904aa0b010bce2ab45264d0631681847fa7b6f2eaa7dab7619943bc4f59"
+checksum = "75030f3c4f45dafd7586dd6780965a8c7e8e285a5ecb86713e63a79c5b2766f3"
 dependencies = [
 "cfg-if",
 ]
@@ -602,9 +601,9 @@ dependencies = [

 [[package]]
 name = "grep-cli"
-version = "0.1.10"
+version = "0.1.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ea40788c059ab8b622c4d074732750bfb3bd2912e2dd58eabc11798a4d5ad725"
+checksum = "47f1288f0e06f279f84926fa4c17e3fcd2a22b357927a82f2777f7be26e4cec0"
 dependencies = [
 "bstr",
 "globset",
@@ -1147,9 +1146,9 @@ dependencies = [

 [[package]]
 name = "serde_spanned"
-version = "0.6.5"
+version = "0.6.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "eb3622f419d1296904700073ea6cc23ad690adbd66f13ea683df73298736f0c1"
+checksum = "87607cb1398ed59d48732e575a4c28a7a8ebf2454b964fe3f224f2afc07909e1"
 dependencies = [
 "serde",
 ]
@@ -1393,9 +1392,9 @@ checksum = "cda74da7e1a664f795bb1f8a87ec406fb89a02522cf6e50620d016add6dbbf5c"

 [[package]]
 name = "toml"
-version = "0.8.9"
+version = "0.8.19"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c6a4b9e8023eb94392d3dca65d717c53abc5dad49c07cb65bb8fcd87115fa325"
+checksum = "a1ed1f98e3fdc28d6d910e6737ae6ab1a93bf1985935a1193e68f93eeb68d24e"
 dependencies = [
 "indexmap",
 "serde",
@@ -1406,18 +1405,18 @@ dependencies = [

 [[package]]
 name = "toml_datetime"
-version = "0.6.5"
+version = "0.6.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3550f4e9685620ac18a50ed434eb3aec30db8ba93b0287467bca5826ea25baf1"
+checksum = "0dd7358ecb8fc2f8d014bf86f6f638ce72ba252a2c3a2572f2a795f1d23efb41"
 dependencies = [
 "serde",
 ]

 [[package]]
 name = "toml_edit"
-version = "0.21.1"
+version = "0.22.22"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6a8534fd7f78b5405e860340ad6575217ce99f38d4d5c8f2442cb5ecb50090e1"
+checksum = "4ae48d6208a266e853d946088ed816055e556cc6028c5e8e2b84d9fa5dd7c7f5"
 dependencies = [
 "indexmap",
 "serde",
@@ -1747,9 +1746,9 @@ checksum = "dff9641d1cd4be8d1a070daf9e3773c5f67e78b4d9d42263020c057706765c04"

 [[package]]
 name = "winnow"
-version = "0.5.18"
+version = "0.6.20"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "176b6138793677221d420fd2f0aeeced263f197688b36484660da767bca2fa32"
+checksum = "36c1fec1a2bb5866f07c25f68c26e565c4c200aebb96d7e55710c19d3e8ac49b"
 dependencies = [
 "memchr",
 ]
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -58,14 +58,14 @@ serde_derive = "1.0"
 serde_yaml = "0.9.28"
 semver = "1.0"
 path_abs = { version = "0.5", default-features = false }
-clircle = "0.5"
+clircle = "0.6"
 bugreport = { version = "0.5.0", optional = true }
 etcetera = { version = "0.8.0", optional = true }
-grep-cli = { version = "0.1.10", optional = true }
+grep-cli = { version = "0.1.11", optional = true }
 regex = { version = "1.10.2", optional = true }
 walkdir = { version = "2.5", optional = true }
 bytesize = { version = "1.3.0" }
-encoding_rs = "0.8.34"
+encoding_rs = "0.8.35"
 os_str_bytes = { version = "~7.0", optional = true }
 run_script = { version = "^0.10.1", optional = true}

@@ -109,7 +109,7 @@ regex = "1.10.2"
 serde = "1.0"
 serde_derive = "1.0"
 serde_with = { version = "3.8.1", default-features = false, features = ["macros"] }
-toml = { version = "0.8.9", features = ["preserve_order"] }
+toml = { version = "0.8.19", features = ["preserve_order"] }
 walkdir = "2.5"

 [build-dependencies.clap]
--- a/doc/long-help.txt
+++ b/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
--- a/doc/short-help.txt
+++ b/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>
--- a/src/bin/bat/app.rs
+++ b/src/bin/bat/app.rs
@@ -9,6 +9,7 @@ use crate::{
    config::{get_args_from_config_file, get_args_from_env_opts_var, get_args_from_env_vars},
 };
 use bat::style::StyleComponentList;
+use bat::BinaryBehavior;
 use bat::StripAnsiMode;
 use clap::ArgMatches;

@@ -97,12 +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) {
@@ -193,6 +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()) {
--- a/src/bin/bat/clap_app.rs
+++ b/src/bin/bat/clap_app.rs
@@ -77,11 +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)
@@ -306,7 +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")
--- a/src/config.rs
+++ b/src/config.rs
@@ -1,5 +1,5 @@
 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;
@@ -44,6 +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,

--- a/src/lib.rs
+++ b/src/lib.rs
@@ -52,7 +52,7 @@ mod terminal;
 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};
--- a/src/nonprintable_notation.rs
+++ b/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,
+}
--- a/src/printer.rs
+++ b/src/printer.rs
@@ -35,6 +35,7 @@ use crate::style::StyleComponent;
 use crate::terminal::{as_terminal_escaped, to_ansi_color};
 use crate::vscreen::{AnsiStyle, EscapeSequence, EscapeSequenceIterator};
 use crate::wrapping::WrappingMode;
+use crate::BinaryBehavior;
 use crate::StripAnsiMode;

 const ANSI_UNDERLINE_ENABLE: EscapeSequence = EscapeSequence::CSI {
@@ -268,7 +269,8 @@ impl<'a> InteractivePrinter<'a> {
            .content_type
            .map_or(false, |c| c.is_binary() && !config.show_nonprintable);

-        let needs_to_match_syntax = !is_printing_binary
+        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 {
@@ -458,7 +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 \
@@ -539,7 +544,10 @@ impl<'a> Printer for InteractivePrinter<'a> {
            })?;

        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, '┴')?;
@@ -551,7 +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 {
@@ -599,7 +609,9 @@ impl<'a> Printer for InteractivePrinter<'a> {
            .into()
        } else {
            let mut line = match self.content_type {
-                Some(ContentType::BINARY) | None => {
+                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,
--- a/src/syntax_mapping/builtins/common/50-diff.toml
+++ b/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"]
--- a/src/syntax_mapping/builtins/linux/50-pacman.toml
+++ b/src/syntax_mapping/builtins/linux/50-pacman.toml
@@ -1,3 +1,8 @@
 [mappings]
-# pacman hooks
-"INI" = ["/usr/share/libalpm/hooks/*.hook", "/etc/pacman.d/hooks/*.hook"]
+"INI" = [
+    # config
+    "/etc/pacman.conf",
+    # hooks
+    "/usr/share/libalpm/hooks/*.hook",
+    "/etc/pacman.d/hooks/*.hook",
+]
--- a/tests/integration_tests.rs
+++ b/tests/integration_tests.rs
@@ -1019,6 +1019,31 @@ fn enable_pager_if_pp_flag_comes_before_paging() {
        .stdout(predicate::eq("pager-output\n").normalize());
 }

+#[test]
+fn paging_does_not_override_simple_plain() {
+    bat()
+        .env("PAGER", "echo pager-output")
+        .arg("--decorations=always")
+        .arg("--plain")
+        .arg("--paging=never")
+        .arg("test.txt")
+        .assert()
+        .success()
+        .stdout(predicate::eq("hello world\n"));
+}
+
+#[test]
+fn simple_plain_does_not_override_paging() {
+    bat()
+        .env("PAGER", "echo pager-output")
+        .arg("--paging=always")
+        .arg("--plain")
+        .arg("test.txt")
+        .assert()
+        .success()
+        .stdout(predicate::eq("pager-output\n"));
+}
+
 #[test]
 fn pager_failed_to_parse() {
    bat()
@@ -1938,6 +1963,16 @@ fn show_all_with_unicode() {
        .stderr("");
 }

+#[test]
+fn binary_as_text() {
+    bat()
+        .arg("--binary=as-text")
+        .arg("control_characters.txt")
+        .assert()
+        .stdout("\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D\x0E\x0F\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C\x1D\x1E\x1F\x7F")
+        .stderr("");
+}
+
 #[test]
 fn no_paging_arg() {
    bat()