Massive refactor

WIP

Replace array of (string, int) pairs with tree of nodes.

A tree of nodes more accurately represents the underlying file structure
and hence is a better fit for the problem space.

Regression: Reverse doesn't work in this commit.

I suspect more methods can be simplifed and reduced.

src/display.rs

@@ -2,16 +2,13 @@ extern crate ansi_term;
 
 use self::ansi_term::Colour::Fixed;
 use self::ansi_term::Style;
-use std::cmp::max;
-use std::collections::HashSet;
-use utils::{ensure_end_slash, strip_end_slash_including_root};
+use utils::Node;
 
 static UNITS: [char; 4] = ['T', 'G', 'M', 'K'];
 
 pub struct DisplayData {
     pub short_paths: bool,
     pub is_reversed: bool,
-    pub to_display: Vec<(String, u64)>,
 }
 
 impl DisplayData {
@@ -63,128 +60,54 @@ impl DisplayData {
             num_siblings == max_siblings - 1
         }
     }
-
-    fn get_size(&self, node_to_print: &str) -> Option<u64> {
-        for &(ref k, ref v) in self.to_display.iter() {
-            if *k == *node_to_print {
-                return Some(*v);
-            }
-        }
-        None
-    }
-
-    fn count_siblings(&self, num_slashes: usize, ntp: &str) -> u64 {
-        self.to_display.iter().fold(0, |a, b| {
-            if b.0.starts_with(ntp) && b.0.as_str().matches('/').count() == num_slashes + 1 {
-                a + 1
-            } else {
-                a
-            }
-        })
-    }
-
-    fn has_children(&self, new_depth: Option<u64>, ntp: &str, num_slashes: usize) -> bool {
-        // this shouldn't be needed we should have already stripped
-        if new_depth.is_none() || new_depth.unwrap() != 1 {
-            for &(ref k2, _) in self.to_display.iter() {
-                let ntp_with_slash = String::from(ntp.to_owned() + "/");
-                if k2.starts_with(ntp_with_slash.as_str())
-                    && k2.matches('/').count() == num_slashes + 1
-                {
-                    return true;
-                }
-            }
-        }
-        false
-    }
 }
 
 pub fn draw_it(
     permissions: bool,
     depth: Option<u64>,
-    base_dirs: HashSet<String>,
-    display_data: &DisplayData,
+    use_full_path: bool,
+    is_reversed: bool,
+    root_node: Node,
 ) {
     if !permissions {
         eprintln!("Did not have permissions for all directories");
     }
-    let first_tree_chars = display_data.get_first_chars();
-    let mut found = HashSet::new();
+    let display_data = DisplayData {
+        short_paths: !use_full_path,
+        is_reversed,
+    };
 
-    for &(ref k, _) in display_data.to_display.iter() {
-        if base_dirs.contains(k) {
-            display_node(&k, &mut found, true, depth, first_tree_chars, display_data);
-        }
+    for c in root_node.children {
+        let first_tree_chars = display_data.get_first_chars();
+        display_node(*c, true, depth, first_tree_chars, &display_data)
     }
 }
 
 fn display_node(
-    node: &str,
-    nodes_already_found: &mut HashSet<String>,
+    node: Node,
     is_biggest: bool,
     depth: Option<u64>,
     indent: &str,
     display_data: &DisplayData,
 ) {
-    if nodes_already_found.contains(node) {
-        return;
-    }
-    nodes_already_found.insert(node.to_string());
-
     let new_depth = match depth {
         None => None,
         Some(0) => return,
         Some(d) => Some(d - 1),
     };
+    let short = display_data.short_paths;
+    print_this_node(&node, is_biggest, short, indent);
 
-    match display_data.get_size(node) {
-        None => println!("Can not find path: {}", node),
-        Some(size) => {
-            let short_path = display_data.short_paths;
-            // move this inside display_data?
-            if !display_data.is_reversed {
-                print_this_node(node, size, is_biggest, short_path, indent);
-            }
-            fan_out(node, nodes_already_found, new_depth, indent, display_data);
-            if display_data.is_reversed {
-                print_this_node(node, size, is_biggest, short_path, indent);
-            }
-        }
-    }
-}
+    let mut num_siblings = node.children.len() as u64;
+    let max_sibling = num_siblings;
+    let new_indent = clean_indentation_string(indent);
 
-fn fan_out(
-    node_to_print: &str,
-    nodes_already_found: &mut HashSet<String>,
-    new_depth: Option<u64>,
-    indentation_str: &str,
-    display_data: &DisplayData,
-) {
-    let new_indent = clean_indentation_string(indentation_str);
-    let num_slashes = strip_end_slash_including_root(node_to_print)
-        .matches('/')
-        .count();
-
-    let mut num_siblings = display_data.count_siblings(num_slashes, node_to_print);
-    let max_siblings = num_siblings;
-
-    for &(ref k, _) in display_data.to_display.iter() {
-        let temp = String::from(ensure_end_slash(node_to_print));
-        if k.starts_with(temp.as_str()) && k.matches('/').count() == num_slashes + 1 {
-            num_siblings -= 1;
-            let has_children = display_data.has_children(new_depth, k, num_slashes + 1);
-            let new_tree_chars =
-                display_data.get_tree_chars(num_siblings, max_siblings, has_children);
-            let biggest = display_data.biggest(num_siblings, max_siblings);
-            display_node(
-                k,
-                nodes_already_found,
-                biggest,
-                new_depth,
-                &*(new_indent.to_string() + new_tree_chars),
-                display_data,
-            );
-        }
+    for c in node.children {
+        num_siblings -= 1;
+        let chars = display_data.get_tree_chars(num_siblings, max_sibling, c.children.len() > 0);
+        let is_biggest = display_data.biggest(num_siblings, max_sibling);
+        let full_indent = (new_indent.clone() + chars);
+        display_node(*c, is_biggest, new_depth, &*full_indent, display_data)
     }
 }
 
@@ -205,18 +128,12 @@ fn clean_indentation_string(s: &str) -> String {
     is
 }
 
-fn print_this_node(
-    node_name: &str,
-    size: u64,
-    is_biggest: bool,
-    short_paths: bool,
-    indentation: &str,
-) {
-    let pretty_size = format!("{:>5}", human_readable_number(size),);
+fn print_this_node(node: &Node, is_biggest: bool, short_paths: bool, indentation: &str) {
+    let pretty_size = format!("{:>5}", human_readable_number(node.size),);
     println!(
         "{}",
         format_string(
-            node_name,
+            &*node.name,
             is_biggest,
             short_paths,
             pretty_size.as_ref(),
@@ -224,7 +141,7 @@ fn print_this_node(
         )
     )
 }
-
+// idea: squash these 2
 pub fn format_string(
     dir_name: &str,
     is_biggest: bool,

src/main.rs

@@ -4,11 +4,10 @@ extern crate assert_cli;
 extern crate walkdir;
 
 use self::display::draw_it;
-use self::display::DisplayData;
 use clap::{App, AppSettings, Arg};
 use utils::{
     compare_tuple_smallest_first, find_big_ones, get_dir_tree, simplify_dir_names, sort,
-    trim_deep_ones,
+    trim_deep_ones, Node,
 };
 
 mod display;
@@ -104,13 +103,45 @@ fn main() {
     if options.is_present("reverse") {
         biggest_ones.sort_by(compare_tuple_smallest_first);
     }
-    let dd = DisplayData {
-        short_paths: !use_full_path,
-        is_reversed: options.is_present("reverse"),
-        to_display: biggest_ones,
+    let tree = build_tree(&biggest_ones);
+    //println!("{:?}", tree);
+
+    draw_it(
+        permissions,
+        depth,
+        use_full_path,
+        options.is_present("reverse"),
+        tree,
+    );
+}
+
+fn build_tree(biggest_ones: &Vec<(String, u64)>) -> Node {
+    let mut top_parent = Node {
+        name: "".to_string(),
+        size: 0,
+        children: vec![],
     };
 
-    draw_it(permissions, depth, simplified_dirs, &dd);
+    // assume sorted order
+    for b in biggest_ones.clone() {
+        let n = Node {
+            name: b.0,
+            size: b.1,
+            children: vec![],
+        };
+        recursively_build_tree(&mut top_parent, n)
+    }
+    top_parent
+}
+
+fn recursively_build_tree(parent_node: &mut Node, new_node: Node) {
+    for c in parent_node.children.iter_mut() {
+        if new_node.name.starts_with(&c.name) {
+            return recursively_build_tree(&mut *c, new_node);
+        }
+    }
+    let temp = Box::<Node>::new(new_node);
+    parent_node.children.push(temp);
 }
 
 #[cfg(test)]

src/utils/mod.rs

@@ -7,6 +7,13 @@ use walkdir::WalkDir;
 mod platform;
 use self::platform::*;
 
+#[derive(Debug)]
+pub struct Node {
+    pub name: String,
+    pub size: u64,
+    pub children: Vec<Box<Node>>,
+}
+
 pub fn simplify_dir_names(filenames: Vec<&str>) -> HashSet<String> {
     let mut top_level_names: HashSet<String> = HashSet::new();
 
@@ -64,14 +71,6 @@ pub fn strip_end_slash(s: &str) -> String {
     new_name
 }
 
-pub fn strip_end_slash_including_root(s: &str) -> String {
-    let mut new_name = String::from(s);
-    while new_name.ends_with('/') || new_name.ends_with("/.") {
-        new_name.pop();
-    }
-    new_name
-}
-
 fn examine_dir(
     top_dir: &str,
     apparent_size: bool,
@@ -93,6 +92,7 @@ fn examine_dir(
                             inodes.insert(inode_dev_pair);
                         }
                     }
+                    // This path and all its parent paths have their counter incremented
                     let mut e_path = e.path().to_path_buf();
                     loop {
                         let path_name = e_path.to_string_lossy().to_string();
@@ -112,7 +112,8 @@ fn examine_dir(
         }
     }
 }
-pub fn compare_tuple(a: &(String, u64), b: &(String, u64)) -> Ordering {
+
+pub fn sort_by_size_first_name_second(a: &(String, u64), b: &(String, u64)) -> Ordering {
     let result = b.1.cmp(&a.1);
     if result == Ordering::Equal {
         a.0.cmp(&b.0)
@@ -132,7 +133,7 @@ pub fn compare_tuple_smallest_first(a: &(String, u64), b: &(String, u64)) -> Ord
 
 pub fn sort(data: HashMap<String, u64>) -> Vec<(String, u64)> {
     let mut new_l: Vec<(String, u64)> = data.iter().map(|(a, b)| (a.clone(), *b)).collect();
-    new_l.sort_by(|a, b| compare_tuple(&a, &b));
+    new_l.sort_by(|a, b| sort_by_size_first_name_second(&a, &b));
     new_l
 }