Merge branch 'main' of https://github.com/barak1412/polars into expos…

…e_regex_escape

crates/polars-plan/src/dsl/function_expr/strings.rs

@@ -9,7 +9,7 @@ use polars_core::utils::handle_casting_failures;
 #[cfg(feature = "dtype-struct")]
 use polars_utils::format_pl_smallstr;
 #[cfg(feature = "regex")]
-use regex::{escape, Regex};
+use regex::{escape, NoExpand, Regex};
 #[cfg(feature = "serde")]
 use serde::{Deserialize, Serialize};
 
@@ -896,15 +896,25 @@ fn replace_all<'a>(
                 "replacement value length ({}) does not match string column length ({})",
                 len_val, ca.len(),
             );
-            let literal = literal || is_literal_pat(&pat);
 
-            if literal {
+            let literal_pat = literal || is_literal_pat(&pat);
+
+            if literal_pat {
                 pat = escape(&pat)
             }
 
             let reg = Regex::new(&pat)?;
 
-            let f = |s: &'a str, val: &'a str| reg.replace_all(s, val);
+            let f = |s: &'a str, val: &'a str| {
+                // According to the docs for replace_all
+                // when literal = True then capture groups are ignored.
+                if literal {
+                    reg.replace_all(s, NoExpand(val))
+                } else {
+                    reg.replace_all(s, val)
+                }
+            };
+
             Ok(iter_and_replace(ca, val, f))
         },
         _ => polars_bail!(

py-polars/tests/unit/operations/namespaces/string/test_string.py

@@ -1006,6 +1006,34 @@ def test_replace_all() -> None:
     )
 
 
+def test_replace_literal_no_caputures() -> None:
+    # When using literal = True, capture groups should be disabled
+
+    # Single row code path in Rust
+    df = pl.DataFrame({"text": ["I found <amt> yesterday."], "amt": ["$1"]})
+    df = df.with_columns(
+        pl.col("text")
+        .str.replace_all("<amt>", pl.col("amt"), literal=True)
+        .alias("text2")
+    )
+    assert df.get_column("text2")[0] == "I found $1 yesterday."
+
+    # Multi-row code path in Rust
+    df2 = pl.DataFrame(
+        {
+            "text": ["I found <amt> yesterday.", "I lost <amt> yesterday."],
+            "amt": ["$1", "$2"],
+        }
+    )
+    df2 = df2.with_columns(
+        pl.col("text")
+        .str.replace_all("<amt>", pl.col("amt"), literal=True)
+        .alias("text2")
+    )
+    assert df2.get_column("text2")[0] == "I found $1 yesterday."
+    assert df2.get_column("text2")[1] == "I lost $2 yesterday."
+
+
 def test_replace_expressions() -> None:
     df = pl.DataFrame({"foo": ["123 bla 45 asd", "xyz 678 910t"], "value": ["A", "B"]})
     out = df.select([pl.col("foo").str.replace(pl.col("foo").first(), pl.col("value"))])