headers: now also accepts process_input - dir input; infile-list;…

… multiple input; auto decompress snappy

src/cmd/headers.rs

@@ -13,6 +13,15 @@ Usage:
     qsv headers [options] [<input>...]
     qsv headers --help
 
+headers arguments:
+    <input>...             The CSV file(s) to read. Use '-' for standard input.
+                           If input is a directory, all files in the directory will
+                           be read as input.
+                           If the input is a file with a '.infile-list' extension,
+                           the file will be read as a list of input files.
+                           If the input are snappy-compressed files(s), it will be
+                           decompressed automatically.
+
 headers options:
     -j, --just-names       Only show the header names (hide column index).
                            This is automatically enabled if more than one
@@ -27,7 +36,7 @@ Common options:
                            Must be a single character. (default: ,)
 "#;
 
-use std::io;
+use std::{io, path::PathBuf};
 
 use serde::Deserialize;
 use tabwriter::TabWriter;
@@ -36,15 +45,17 @@ use crate::{config::Delimiter, util, CliResult};
 
 #[derive(Deserialize)]
 struct Args {
-    arg_input:       Vec<String>,
+    arg_input:       Vec<PathBuf>,
     flag_just_names: bool,
     flag_intersect:  bool,
     flag_trim:       bool,
     flag_delimiter:  Option<Delimiter>,
 }
 
 pub fn run(argv: &[&str]) -> CliResult<()> {
-    let args: Args = util::get_args(USAGE, argv)?;
+    let mut args: Args = util::get_args(USAGE, argv)?;
+    let tmpdir = tempfile::tempdir()?;
+    args.arg_input = util::process_input(args.arg_input, &tmpdir, "")?;
     let configs = util::many_configs(&args.arg_input, args.flag_delimiter, true, false)?;
 
     let num_inputs = configs.len();

tests/test_headers.rs

@@ -171,3 +171,58 @@ h2
 h3";
     assert_eq!(got, expected.to_string());
 }
+
+#[test]
+fn headers_infile() {
+    let wrk = Workdir::new("headers_infile");
+    wrk.create("in1.csv", vec![svec!["a", "b", "c"], svec!["1", "2", "3"]]);
+
+    wrk.create("in2.csv", vec![svec!["c", "d", "e"], svec!["3", "1", "2"]]);
+
+    wrk.create(
+        "in3.csv",
+        vec![svec!["a", "b", "f", "g"], svec!["1", "2", "4", "3"]],
+    );
+
+    wrk.create_from_string("testdata.infile-list", "in1.csv\nin2.csv\nin3.csv\n");
+
+    let mut cmd = wrk.command("headers");
+    cmd.arg("testdata.infile-list");
+
+    let got: Vec<Vec<String>> = wrk.read_stdout(&mut cmd);
+    let expected = vec![
+        ["a"],
+        ["b"],
+        ["c"],
+        ["c"],
+        ["d"],
+        ["e"],
+        ["a"],
+        ["b"],
+        ["f"],
+        ["g"],
+    ];
+    assert_eq!(got, expected);
+}
+
+#[test]
+fn headers_intersect_infile() {
+    let wrk = Workdir::new("headers_intersect_infile");
+    wrk.create("in1.csv", vec![svec!["a", "b", "c"], svec!["1", "2", "3"]]);
+
+    wrk.create("in2.csv", vec![svec!["c", "d", "e"], svec!["3", "1", "2"]]);
+
+    wrk.create(
+        "in3.csv",
+        vec![svec!["a", "b", "f", "g"], svec!["1", "2", "4", "3"]],
+    );
+
+    wrk.create_from_string("testdata.infile-list", "in1.csv\nin2.csv\nin3.csv\n");
+
+    let mut cmd = wrk.command("headers");
+    cmd.arg("--intersect").arg("testdata.infile-list");
+
+    let got: Vec<Vec<String>> = wrk.read_stdout(&mut cmd);
+    let expected = vec![["a"], ["b"], ["c"], ["d"], ["e"], ["f"], ["g"]];
+    assert_eq!(got, expected);
+}