WIP: 1486/first row as headers (#1530)

* headers plugin

* Remove plugin

* Add non-functioning headers command

* Add ability to extract headers from first row

* Refactor header extraction

* Rebuild indexmap with proper headers

* Rebuild result properly

* Compiling, probably wrapped too much?

* Refactoring

* Deal with case of empty header cell

* Deal with case of empty header cell

* Fix formatting

* Fix linting, attempt 2.

* Move whole_stream_command(Headers) to more appropriate section

* ... more linting

* Return Err(ShellError...) instead of panic, yield each row instead of entire table

* Insert Column[index] if no header info is found.

* Update error description

* Add initial test

* Add tests for headers command

* Lint test cases in headers

* Change ShellError for headers, Add sample_headers file to utils.rs

* Add empty sheet to test file

* Revert "Add empty sheet to test file"

This reverts commit a4bf38a31d.

* Show error message when given empty table

crates/nu-cli/src/cli.rs

@@ -311,6 +311,7 @@ pub fn create_default_context(
             whole_stream_command(Shuffle),
             whole_stream_command(Wrap),
             whole_stream_command(Pivot),
+            whole_stream_command(Headers),
             // Data processing
             whole_stream_command(Histogram),
             whole_stream_command(Sum),

crates/nu-cli/src/commands.rs

@@ -45,6 +45,7 @@ pub(crate) mod from_xml;
 pub(crate) mod from_yaml;
 pub(crate) mod get;
 pub(crate) mod group_by;
+pub(crate) mod headers;
 pub(crate) mod help;
 pub(crate) mod histogram;
 pub(crate) mod history;
@@ -155,6 +156,7 @@ pub(crate) use from_yaml::FromYAML;
 pub(crate) use from_yaml::FromYML;
 pub(crate) use get::Get;
 pub(crate) use group_by::GroupBy;
+pub(crate) use headers::Headers;
 pub(crate) use help::Help;
 pub(crate) use histogram::Histogram;
 pub(crate) use history::History;

crates/nu-cli/src/commands/headers.rs

@@ -0,0 +1,80 @@
+use crate::commands::WholeStreamCommand;
+use crate::context::CommandRegistry;
+use crate::prelude::*;
+use futures::stream::StreamExt;
+use indexmap::IndexMap;
+use nu_errors::ShellError;
+use nu_protocol::Dictionary;
+use nu_protocol::{ReturnSuccess, Signature, UntaggedValue, Value};
+
+pub struct Headers;
+#[derive(Deserialize)]
+pub struct HeadersArgs {}
+
+impl WholeStreamCommand for Headers {
+    fn name(&self) -> &str {
+        "headers"
+    }
+    fn signature(&self) -> Signature {
+        Signature::build("headers")
+    }
+    fn usage(&self) -> &str {
+        "Use the first row of the table as column names"
+    }
+    fn run(
+        &self,
+        args: CommandArgs,
+        registry: &CommandRegistry,
+    ) -> Result<OutputStream, ShellError> {
+        args.process(registry, headers)?.run()
+    }
+}
+
+pub fn headers(
+    HeadersArgs {}: HeadersArgs,
+    RunnableContext { input, .. }: RunnableContext,
+) -> Result<OutputStream, ShellError> {
+    let stream = async_stream! {
+        let rows: Vec<Value> = input.values.collect().await;
+
+        if rows.len() < 1 {
+            yield Err(ShellError::untagged_runtime_error("Couldn't find headers, was the input a properly formatted, non-empty table?"));
+        }
+
+        //the headers are the first row in the table
+        let headers: Vec<String> = match &rows[0].value {
+            UntaggedValue::Row(d) => {
+                Ok(d.entries.iter().map(|(k, v)| {
+                    match v.as_string() {
+                        Ok(s) => s,
+                        Err(_) => { //If a cell that should contain a header name is empty, we name the column Column[index]
+                            match d.entries.get_full(k) {
+                                Some((index, _, _)) => format!("Column{}", index),
+                                None => "unknownColumn".to_string()
+                            }
+                        }
+                    }
+                }).collect())
+            }
+            _ => Err(ShellError::unexpected_eof("Could not get headers, is the table empty?", rows[0].tag.span))
+        }?;
+
+        //Each row is a dictionary with the headers as keys
+        for r in rows.iter().skip(1) {
+            match &r.value {
+                UntaggedValue::Row(d) => {
+                    let mut i = 0;
+                    let mut entries = IndexMap::new();
+                    for (_, v) in d.entries.iter() {
+                        entries.insert(headers[i].clone(), v.clone());
+                        i += 1;
+                    }
+                    yield Ok(ReturnSuccess::Value(UntaggedValue::Row(Dictionary{entries}).into_value(r.tag.clone())))
+                }
+                _ => yield Err(ShellError::unexpected_eof("Couldn't iterate through rows, was the input a properly formatted table?", r.tag.span))
+            }
+        }
+    };
+
+    Ok(stream.to_output_stream())
+}

crates/nu-cli/src/utils.rs

@@ -317,6 +317,10 @@ mod tests {
                     loc: fixtures().join("sample_data.xlsx"),
                     at: 0
                 },
+                Res {
+                    loc: fixtures().join("sample_headers.xlsx"),
+                    at: 0
+                },
                 Res {
                     loc: fixtures().join("script.nu"),
                     at: 0

crates/nu-cli/tests/commands/headers.rs

@@ -0,0 +1,31 @@
+use nu_test_support::{nu, pipeline};
+
+#[test]
+fn headers_uses_first_row_as_header() {
+    let actual = nu!(
+    cwd: "tests/fixtures/formats", pipeline(
+        r#"
+            open sample_headers.xlsx
+            | get Sheet1
+            | headers
+            | get header0
+            | from-json"#
+    ));
+
+    assert_eq!(actual, "r1c0r2c0")
+}
+
+#[test]
+fn headers_adds_missing_column_name() {
+    let actual = nu!(
+    cwd: "tests/fixtures/formats", pipeline(
+        r#"
+            open sample_headers.xlsx
+            | get Sheet1
+            | headers
+            | get Column1
+            | from-json"#
+    ));
+
+    assert_eq!(actual, "r1c1r2c1")
+}

crates/nu-cli/tests/commands/mod.rs

@@ -10,6 +10,7 @@ mod first;
 mod format;
 mod get;
 mod group_by;
+mod headers;
 mod histogram;
 mod insert;
 mod last;