Add more multiline pipeline forms (#4740)

crates/nu-parser/src/lex.rs

@@ -243,7 +243,6 @@ pub fn lex(
         let c = *c;
         if c == b'|' {
             // If the next character is `|`, it's either `|` or `||`.
-
             let idx = curr_offset;
             let prev_idx = idx;
             curr_offset += 1;
@@ -262,10 +261,31 @@ pub fn lex(
             }
 
             // Otherwise, it's just a regular `|` token.
-            output.push(Token::new(
-                TokenContents::Pipe,
-                Span::new(span_offset + idx, span_offset + idx + 1),
-            ));
+
+            // Before we push, check to see if the previous character was a newline.
+            // If so, then this is a continuation of the previous line
+            if let Some(prev) = output.last_mut() {
+                match prev.contents {
+                    TokenContents::Eol => {
+                        *prev = Token::new(
+                            TokenContents::Pipe,
+                            Span::new(span_offset + idx, span_offset + idx + 1),
+                        )
+                    }
+                    _ => {
+                        output.push(Token::new(
+                            TokenContents::Pipe,
+                            Span::new(span_offset + idx, span_offset + idx + 1),
+                        ));
+                    }
+                }
+            } else {
+                output.push(Token::new(
+                    TokenContents::Pipe,
+                    Span::new(span_offset + idx, span_offset + idx + 1),
+                ));
+            }
+
             is_complete = false;
         } else if c == b';' {
             // If the next character is a `;`, we're looking at a semicolon token.
@@ -282,9 +302,11 @@ pub fn lex(
                 TokenContents::Semicolon,
                 Span::new(span_offset + idx, span_offset + idx + 1),
             ));
-        } else if c == b'\n' || c == b'\r' {
+        } else if c == b'\r' {
+            // Ignore a stand-alone carriage return
+            curr_offset += 1;
+        } else if c == b'\n' {
             // If the next character is a newline, we're looking at an EOL (end of line) token.
-
             let idx = curr_offset;
             curr_offset += 1;
             if !additional_whitespace.contains(&c) {

tests/main.rs

@@ -1,5 +1,6 @@
 extern crate nu_test_support;
 
+mod parsing;
 mod path;
 mod plugins;
 mod shell;

tests/parsing/mod.rs

@@ -0,0 +1,46 @@
+use nu_test_support::nu;
+
+#[test]
+fn run_nu_script_single_line() {
+    let actual = nu!(cwd: "tests/parsing/samples", r#"
+        nu single_line.nu
+        "#);
+
+    assert_eq!(actual.out, "5");
+}
+
+#[test]
+fn run_nu_script_multiline_start_pipe() {
+    let actual = nu!(cwd: "tests/parsing/samples", r#"
+        nu multiline_start_pipe.nu
+        "#);
+
+    assert_eq!(actual.out, "4");
+}
+
+#[test]
+fn run_nu_script_multiline_start_pipe_win() {
+    let actual = nu!(cwd: "tests/parsing/samples", r#"
+        nu multiline_start_pipe_win.nu
+        "#);
+
+    assert_eq!(actual.out, "3");
+}
+
+#[test]
+fn run_nu_script_multiline_end_pipe() {
+    let actual = nu!(cwd: "tests/parsing/samples", r#"
+        nu multiline_end_pipe.nu
+        "#);
+
+    assert_eq!(actual.out, "2");
+}
+
+#[test]
+fn run_nu_script_multiline_end_pipe_win() {
+    let actual = nu!(cwd: "tests/parsing/samples", r#"
+        nu multiline_end_pipe_win.nu
+        "#);
+
+    assert_eq!(actual.out, "3");
+}

tests/parsing/samples/multiline_end_pipe.nu

@@ -0,0 +1,2 @@
+echo "hi" |
+str length

tests/parsing/samples/multiline_end_pipe_win.nu

@@ -0,0 +1,2 @@
+echo "how" |
+str length

tests/parsing/samples/multiline_start_pipe.nu

@@ -0,0 +1,2 @@
+echo "four"
+| str length

tests/parsing/samples/multiline_start_pipe_win.nu

@@ -0,0 +1,2 @@
+echo "one"
+| str length

tests/parsing/samples/single_line.nu

@@ -0,0 +1 @@
+echo "hello" | str length