ARROW-11656: [Rust][DataFusion] Remaining Postgres String functions

rust/datafusion/Cargo.toml

@@ -40,10 +40,12 @@ name = "datafusion-cli"
 path = "src/bin/main.rs"
 
 [features]
-default = ["cli", "crypto_expressions"]
+default = ["cli", "crypto_expressions", "regex_expressions", "unicode_expressions"]
 cli = ["rustyline"]
 simd = ["arrow/simd"]
 crypto_expressions = ["md-5", "sha2"]
+regex_expressions = ["regex", "lazy_static"]
+unicode_expressions = ["unicode-segmentation"]
 
 [dependencies]
 ahash = "0.7"
@@ -65,7 +67,9 @@ log = "^0.4"
 md-5 = { version = "^0.9.1", optional = true }
 sha2 = { version = "^0.9.1", optional = true }
 ordered-float = "2.0"
-unicode-segmentation = "^1.7.1"
+unicode-segmentation = { version = "^1.7.1", optional = true }
+regex = { version = "^1.4.3", optional = true }
+lazy_static = { version = "^1.4.0", optional = true }
 
 [dev-dependencies]
 rand = "0.8"

rust/datafusion/README.md

@@ -71,15 +71,20 @@ DataFusion includes a simple command-line interactive SQL utility. See the [CLI
   - [x] lpad
   - [x] ltrim
   - [x] octet_length
+  - [x] regexp_replace
   - [x] repeat
+  - [x] replace
   - [x] reverse
   - [x] right
   - [x] rpad
   - [x] rtrim
+  - [x] split_part
+  - [x] starts_with
+  - [x] strpos
   - [x] substr
   - [x] to_hex
+  - [x] translate
   - [x] trim
-
 - Miscellaneous/Boolean functions
   - [x] nullif
 - Common date/time functions

rust/datafusion/src/lib.rs

@@ -170,3 +170,7 @@ pub mod variable;
 
 #[cfg(test)]
 pub mod test;
+
+#[macro_use]
+#[cfg(feature = "regex_expressions")]
+extern crate lazy_static;

rust/datafusion/src/logical_plan/expr.rs

@@ -1090,6 +1090,8 @@ unary_scalar_expr!(Lpad, lpad);
 unary_scalar_expr!(Ltrim, ltrim);
 unary_scalar_expr!(MD5, md5);
 unary_scalar_expr!(OctetLength, octet_length);
+unary_scalar_expr!(RegexpReplace, regexp_replace);
+unary_scalar_expr!(Replace, replace);
 unary_scalar_expr!(Repeat, repeat);
 unary_scalar_expr!(Reverse, reverse);
 unary_scalar_expr!(Right, right);
@@ -1099,8 +1101,12 @@ unary_scalar_expr!(SHA224, sha224);
 unary_scalar_expr!(SHA256, sha256);
 unary_scalar_expr!(SHA384, sha384);
 unary_scalar_expr!(SHA512, sha512);
+unary_scalar_expr!(SplitPart, split_part);
+unary_scalar_expr!(StartsWith, starts_with);
+unary_scalar_expr!(Strpos, strpos);
 unary_scalar_expr!(Substr, substr);
 unary_scalar_expr!(ToHex, to_hex);
+unary_scalar_expr!(Translate, translate);
 unary_scalar_expr!(Trim, trim);
 unary_scalar_expr!(Upper, upper);
 

rust/datafusion/src/logical_plan/mod.rs

@@ -37,8 +37,9 @@ pub use expr::{
     ceil, character_length, chr, col, combine_filters, concat, concat_ws, cos, count,
     count_distinct, create_udaf, create_udf, exp, exprlist_to_fields, floor, in_list,
     initcap, left, length, lit, ln, log10, log2, lower, lpad, ltrim, max, md5, min,
-    octet_length, or, repeat, reverse, right, round, rpad, rtrim, sha224, sha256, sha384,
-    sha512, signum, sin, sqrt, substr, sum, tan, to_hex, trim, trunc, upper, when, Expr,
+    octet_length, or, regexp_replace, repeat, replace, reverse, right, round, rpad,
+    rtrim, sha224, sha256, sha384, sha512, signum, sin, split_part, sqrt, starts_with,
+    strpos, substr, sum, tan, to_hex, translate, trim, trunc, upper, when, Expr,
     ExprRewriter, ExpressionVisitor, Literal, Recursion,
 };
 pub use extension::UserDefinedLogicalNode;