Switch to a command-line API closer to TaskWarrior

* Use a parser (rather than clap) to process the command line
* Outline some generic support for filtering, reporting, modifying, etc.
* Break argument parsing strictly from invocation, to allow independent testing

cli/src/invocation/cmd/list.rs

@@ -0,0 +1,45 @@
+use crate::argparse::Report;
+use crate::invocation::filtered_tasks;
+use crate::table;
+use failure::Fallible;
+use prettytable::{cell, row, Table};
+use taskchampion::Replica;
+
+pub(crate) fn execute(replica: &mut Replica, report: Report) -> Fallible<()> {
+    let mut t = Table::new();
+    t.set_format(table::format());
+    t.set_titles(row![b->"id", b->"act", b->"description"]);
+    for task in filtered_tasks(replica, &report.filter)? {
+        let uuid = task.get_uuid();
+        let mut id = uuid.to_string();
+        if let Some(i) = replica.get_working_set_index(&uuid)? {
+            id = i.to_string();
+        }
+        let active = match task.is_active() {
+            true => "*",
+            false => "",
+        };
+        t.add_row(row![id, active, task.get_description()]);
+    }
+    t.printstd();
+    Ok(())
+}
+
+#[cfg(test)]
+mod test {
+    use super::*;
+    use crate::argparse::Filter;
+    use crate::invocation::cmd::test::test_replica;
+
+    #[test]
+    fn test_list() {
+        let mut replica = test_replica();
+        let report = Report {
+            filter: Filter {
+                ..Default::default()
+            },
+        };
+        execute(&mut replica, report).unwrap();
+        // output is to stdout, so this is as much as we can check
+    }
+}