seq: implement integer sequences

If we notice that we can represent all arguments as BigInts, take a
different code path. Just like GNU seq this means we can print an
infinite amount of numbers in this case.

tests/by-util/test_seq.rs

@@ -1,5 +1,7 @@
 use crate::common::util::*;
 
+// ---- Tests for the big integer based path ----
+
 #[test]
 fn test_count_up() {
     new_ucmd!()
@@ -45,3 +47,62 @@ fn test_seq_wrong_arg() {
 fn test_zero_step() {
     new_ucmd!().args(&["10", "0", "32"]).fails();
 }
+
+#[test]
+fn test_big_numbers() {
+    new_ucmd!()
+        .args(&[
+            "1000000000000000000000000000",
+            "1000000000000000000000000001",
+        ])
+        .succeeds()
+        .stdout_only("1000000000000000000000000000\n1000000000000000000000000001\n");
+}
+
+// ---- Tests for the floating point based path ----
+
+#[test]
+fn test_count_up_floats() {
+    new_ucmd!()
+        .args(&["10.0"])
+        .run()
+        .stdout_is("1\n2\n3\n4\n5\n6\n7\n8\n9\n10\n");
+}
+
+#[test]
+fn test_count_down_floats() {
+    new_ucmd!()
+        .args(&["--", "5", "-1.0", "1"])
+        .run()
+        .stdout_is("5.0\n4.0\n3.0\n2.0\n1.0\n");
+    new_ucmd!()
+        .args(&["5", "-1", "1.0"])
+        .run()
+        .stdout_is("5\n4\n3\n2\n1\n");
+}
+
+#[test]
+fn test_separator_and_terminator_floats() {
+    new_ucmd!()
+        .args(&["-s", ",", "-t", "!", "2.0", "6"])
+        .run()
+        .stdout_is("2.0,3.0,4.0,5.0,6.0!");
+}
+
+#[test]
+fn test_equalize_widths_floats() {
+    new_ucmd!()
+        .args(&["-w", "5", "10.0"])
+        .run()
+        .stdout_is("05\n06\n07\n08\n09\n10\n");
+}
+
+#[test]
+fn test_seq_wrong_arg_floats() {
+    new_ucmd!().args(&["-w", "5", "10.0", "33", "32"]).fails();
+}
+
+#[test]
+fn test_zero_step_floats() {
+    new_ucmd!().args(&["10.0", "0", "32"]).fails();
+}