From 5cdf9763e417d248ae8748de072808354a3283d6 Mon Sep 17 00:00:00 2001
From: Gabriel de Perthuis <g2p.code@gmail.com>
Date: Thu, 18 Jan 2024 14:32:32 +0100
Subject: [PATCH] Add a flag to anchor patterns in the input

Closes #1476.
---
 src/cli.rs     | 37 +++++++++++++++++++++++++++++++++++--
 src/main.rs    | 35 ++++++++++++++++++++++++++++-------
 tests/tests.rs | 27 +++++++++++++++++++++++++++
 3 files changed, 90 insertions(+), 9 deletions(-)

diff --git a/src/cli.rs b/src/cli.rs
index 6e61636..f802a5a 100644
--- a/src/cli.rs
+++ b/src/cli.rs
@@ -168,8 +168,8 @@ pub struct Opts {
     pub regex: bool,
 
     /// Treat the pattern as a literal string instead of a regular expression. Note
-    /// that this also performs substring comparison. If you want to match on an
-    /// exact filename, consider using '--glob'.
+    /// that the pattern would still match on a substring of the input. If you want
+    /// to match on an exact filename, consider adding '--anchor=input' as well.
     #[arg(
         long,
         short = 'F',
@@ -246,6 +246,20 @@ pub struct Opts {
     )]
     pub full_path: bool,
 
+    /// By default, the search pattern for --regex and --fixed-strings can match any part of the input.
+    /// (See the --full-path option for what constitutes input)
+    ///
+    /// This flag allows anchoring the pattern.
+    ///
+    /// Conflicts with the --glob flag: globs always match the entire input
+    #[arg(
+        long,
+        help = "Where to anchor the pattern",
+        conflicts_with("glob"),
+        long_help
+    )]
+    pub anchor: Option<Anchor>,
+
     /// Separate search results by the null character (instead of newlines).
     /// Useful for piping results to 'xargs'.
     #[arg(
@@ -680,6 +694,17 @@ impl Opts {
         self.rg_alias_hidden_ignore > 0
     }
 
+    pub fn anchor(&self) -> Option<Anchor> {
+        if self.glob {
+            // globset has no way to use an anchor.
+            // Otherwise we'd guard like this:
+            // && !self.no_anchor && self.anchor.is_none()
+            Some(Anchor::Input)
+        } else {
+            self.anchor
+        }
+    }
+
     pub fn max_depth(&self) -> Option<usize> {
         self.max_depth.or(self.exact_depth)
     }
@@ -725,6 +750,14 @@ fn default_num_threads() -> NonZeroUsize {
         .min(limit)
 }
 
+#[derive(Copy, Clone, PartialEq, Eq, ValueEnum)]
+pub enum Anchor {
+    InputStart,
+    InputEnd,
+    Input,
+    Word,
+}
+
 #[derive(Copy, Clone, PartialEq, Eq, ValueEnum)]
 pub enum FileType {
     #[value(alias = "f")]
diff --git a/src/main.rs b/src/main.rs
index bef4120..e0b3e6a 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -162,15 +162,36 @@ fn ensure_search_pattern_is_not_a_path(opts: &Opts) -> Result<()> {
     }
 }
 
+fn apply_anchors(re: String, anchors: Option<cli::Anchor>) -> String {
+    use cli::Anchor;
+    match anchors {
+        None => re,
+        Some(Anchor::InputStart) => "^".to_owned() + &re,
+        Some(Anchor::InputEnd) => re + "$",
+        Some(Anchor::Input) => "^".to_owned() + &re + "$",
+        // https://docs.rs/regex/latest/regex/#empty-matches
+        Some(Anchor::Word) => r"\<".to_owned() + &re + r"\>",
+    }
+}
+
 fn build_pattern_regex(pattern: &str, opts: &Opts) -> Result<String> {
-    Ok(if opts.glob && !pattern.is_empty() {
-        let glob = GlobBuilder::new(pattern).literal_separator(true).build()?;
-        glob.regex().to_owned()
-    } else if opts.fixed_strings {
-        // Treat pattern as literal string if '--fixed-strings' is used
-        regex::escape(pattern)
+    Ok(if opts.glob {
+        if !pattern.is_empty() {
+            let glob = GlobBuilder::new(pattern).literal_separator(true).build()?;
+            glob.regex().to_owned()
+        } else {
+            "".to_owned()
+        }
     } else {
-        String::from(pattern)
+        apply_anchors(
+            if opts.fixed_strings {
+                // Treat pattern as literal string if '--fixed-strings' is used
+                regex::escape(pattern)
+            } else {
+                String::from(pattern)
+            },
+            opts.anchor(),
+        )
     })
 }
 
diff --git a/tests/tests.rs b/tests/tests.rs
index 1edbeae..d8f4a8d 100644
--- a/tests/tests.rs
+++ b/tests/tests.rs
@@ -604,6 +604,33 @@ fn test_full_path() {
     );
 }
 
+/// Anchoring (--anchor)
+#[test]
+fn test_anchors() {
+    let te = TestEnv::new(DEFAULT_DIRS, DEFAULT_FILES);
+
+    te.assert_output(&["--anchor=input", "foo"], "");
+    te.assert_output(&["--anchor=input", "b.foo"], "one/b.foo");
+    te.assert_output(&["--anchor=input-start", "foo"], "");
+    te.assert_output(&["--anchor=input-start", "b."], "one/b.foo");
+    te.assert_output(
+        &["--anchor=input-end", "oo"],
+        "a.foo
+         one/b.foo
+         one/two/c.foo
+         one/two/three/d.foo
+         one/two/three/directory_foo/",
+    );
+    te.assert_output(&["--anchor=word", "oo"], "");
+    te.assert_output(
+        &["--anchor=word", "foo"],
+        "a.foo
+         one/b.foo
+         one/two/c.foo
+         one/two/three/d.foo",
+    );
+}
+
 /// Hidden files (--hidden)
 #[test]
 fn test_hidden() {