yuqei/difflib
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Make SequenceMatcher generic over all slice types

Browse Source
This commit is contained in:
nmlgc
2018-07-03 20:28:11 +02:00
parent de82e46021
commit 6fd59b0c08
7 changed files with 89 additions and 132 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Cargo.toml
+1 -1
View File
@@ -1,6 +1,6 @@
[package] [package]
name = "difflib" name = "difflib"
version = "0.3.1" version = "0.4.0"
authors = ["Dima Kudosh <dimakudosh@gmail.com>"] authors = ["Dima Kudosh <dimakudosh@gmail.com>"]
description = "Port of Python's difflib library to Rust." description = "Port of Python's difflib library to Rust."
documentation = "https://github.com/DimaKudosh/difflib/wiki" documentation = "https://github.com/DimaKudosh/difflib/wiki"
README.md
+3 -3
View File
@@ -8,7 +8,7 @@ Simply add difflib to your dependencies block in Cargo.toml
```rust ```rust
[dependencies] [dependencies]
difflib = "0.3.0" difflib = "0.4.0"
``` ```
## Documentation ## Documentation
@@ -65,7 +65,7 @@ fn main() {
} }
//SequenceMatcher examples //SequenceMatcher examples
let mut matcher = SequenceMatcher::new("one two three four", "zero one tree four"); let mut matcher = SequenceMatcher::new(b"one two three four", b"zero one tree four");
let m = matcher.find_longest_match(0, 18, 0, 18); let m = matcher.find_longest_match(0, 18, 0, 18);
println!("{:?}", m); println!("{:?}", m);
let all_matches = matcher.get_matching_blocks(); let all_matches = matcher.get_matching_blocks();
@@ -76,7 +76,7 @@ fn main() {
println!("{:?}", grouped_opcodes); println!("{:?}", grouped_opcodes);
let ratio = matcher.ratio(); let ratio = matcher.ratio();
println!("{:?}", ratio); println!("{:?}", ratio);
matcher.set_seqs("aaaaa", "aaaab"); matcher.set_seqs(b"aaaaa", b"aaaab");
println!("{:?}", matcher.ratio()); println!("{:?}", matcher.ratio());
} }
``` ```
examples/example.rs
+2 -2
View File
@@ -47,7 +47,7 @@ fn main() {
} }
//SequenceMatcher examples //SequenceMatcher examples
let mut matcher = SequenceMatcher::new("one two three four", "zero one tree four"); let mut matcher = SequenceMatcher::new(b"one two three four", b"zero one tree four");
let m = matcher.find_longest_match(0, 18, 0, 18); let m = matcher.find_longest_match(0, 18, 0, 18);
println!("{:?}", m); println!("{:?}", m);
let all_matches = matcher.get_matching_blocks(); let all_matches = matcher.get_matching_blocks();
@@ -58,6 +58,6 @@ fn main() {
println!("{:?}", grouped_opcodes); println!("{:?}", grouped_opcodes);
let ratio = matcher.ratio(); let ratio = matcher.ratio();
println!("{:?}", ratio); println!("{:?}", ratio);
matcher.set_seqs("aaaaa", "aaaab"); matcher.set_seqs(b"aaaaa", b"aaaab");
println!("{:?}", matcher.ratio()); println!("{:?}", matcher.ratio());
} }
src/differ.rs
+11 -4
View File
@@ -4,8 +4,8 @@ use utils::{count_leading, str_with_similar_chars};
#[derive(Default)] #[derive(Default)]
pub struct Differ { pub struct Differ {
pub line_junk: Option<fn(&str) -> bool>, pub line_junk: Option<fn(&&str) -> bool>,
pub char_junk: Option<fn(&str) -> bool>, pub char_junk: Option<fn(&char) -> bool>,
} }
impl Differ { impl Differ {
@@ -123,7 +123,12 @@ impl Differ {
} }
continue; continue;
} }
let mut cruncher = SequenceMatcher::new(*first_sequence_str, *second_sequence_str); let (first_sequence_chars, second_sequence_chars) = (
first_sequence_str.chars().collect::<Vec<char>>(),
second_sequence_str.chars().collect::<Vec<char>>(),
);
let mut cruncher =
SequenceMatcher::new(&first_sequence_chars, &second_sequence_chars);
cruncher.set_is_junk(self.char_junk); cruncher.set_is_junk(self.char_junk);
if cruncher.ratio() > best_ratio { if cruncher.ratio() > best_ratio {
best_ratio = cruncher.ratio(); best_ratio = cruncher.ratio();
@@ -167,7 +172,9 @@ impl Differ {
let second_element = &second_sequence[best_j]; let second_element = &second_sequence[best_j];
if eqi.is_none() { if eqi.is_none() {
let (mut first_tag, mut second_tag) = (String::new(), String::new()); let (mut first_tag, mut second_tag) = (String::new(), String::new());
let mut cruncher = SequenceMatcher::new(*first_element, second_element); let first_element_chars: Vec<char> = first_element.chars().collect();
let second_element_chars: Vec<char> = second_element.chars().collect();
let mut cruncher = SequenceMatcher::new(&first_element_chars, &second_element_chars);
cruncher.set_is_junk(self.char_junk); cruncher.set_is_junk(self.char_junk);
for opcode in &cruncher.get_opcodes() { for opcode in &cruncher.get_opcodes() {
let (first_length, second_length) = ( let (first_length, second_length) = (
src/lib.rs
+39 -26
View File
@@ -4,6 +4,7 @@ mod utils;
use sequencematcher::{Sequence, SequenceMatcher}; use sequencematcher::{Sequence, SequenceMatcher};
use std::collections::HashMap; use std::collections::HashMap;
use std::fmt::Display;
use utils::{format_range_context, format_range_unified}; use utils::{format_range_context, format_range_unified};
pub fn get_close_matches<'a>( pub fn get_close_matches<'a>(
@@ -16,9 +17,9 @@ pub fn get_close_matches<'a>(
panic!("Cutoff must be greater than 0.0 and lower than 1.0"); panic!("Cutoff must be greater than 0.0 and lower than 1.0");
} }
let mut res: Vec<(f32, &str)> = Vec::new(); let mut res: Vec<(f32, &str)> = Vec::new();
let mut matcher = SequenceMatcher::new("", word); let mut matcher = SequenceMatcher::new(b"", word.as_bytes());
for i in &possibilities { for i in &possibilities {
matcher.set_first_seq(i); matcher.set_first_seq(i.as_bytes());
let ratio = matcher.ratio(); let ratio = matcher.ratio();
if ratio >= cutoff { if ratio >= cutoff {
res.push((ratio, i)); res.push((ratio, i));
@@ -29,9 +30,9 @@ pub fn get_close_matches<'a>(
res.iter().map(|x| x.1).collect() res.iter().map(|x| x.1).collect()
} }
pub fn unified_diff<T: Sequence>( pub fn unified_diff<T: Sequence + Display>(
first_sequence: &T, first_sequence: &[T],
second_sequence: &T, second_sequence: &[T],
from_file: &str, from_file: &str,
to_file: &str, to_file: &str,
from_file_date: &str, from_file_date: &str,
@@ -59,19 +60,31 @@ pub fn unified_diff<T: Sequence>(
)); ));
for code in group { for code in group {
if code.tag == "equal" { if code.tag == "equal" {
for i in code.first_start..code.first_end { for item in first_sequence
res.push(format!(" {}", first_sequence.at_index(i).unwrap())); .iter()
.take(code.first_end)
.skip(code.first_start)
{
res.push(format!(" {}", item));
} }
continue; continue;
} }
if code.tag == "replace" || code.tag == "delete" { if code.tag == "replace" || code.tag == "delete" {
for i in code.first_start..code.first_end { for item in first_sequence
res.push(format!("-{}", first_sequence.at_index(i).unwrap())); .iter()
.take(code.first_end)
.skip(code.first_start)
{
res.push(format!("-{}", item));
} }
} }
if code.tag == "replace" || code.tag == "insert" { if code.tag == "replace" || code.tag == "insert" {
for i in code.second_start..code.second_end { for item in second_sequence
res.push(format!("+{}", second_sequence.at_index(i).unwrap())); .iter()
.take(code.second_end)
.skip(code.second_start)
{
res.push(format!("+{}", item));
} }
} }
} }
@@ -79,9 +92,9 @@ pub fn unified_diff<T: Sequence>(
res res
} }
pub fn context_diff<T: Sequence>( pub fn context_diff<T: Sequence + Display>(
first_sequence: &T, first_sequence: &[T],
second_sequence: &T, second_sequence: &[T],
from_file: &str, from_file: &str,
to_file: &str, to_file: &str,
from_file_date: &str, from_file_date: &str,
@@ -119,12 +132,12 @@ pub fn context_diff<T: Sequence>(
if any { if any {
for opcode in group { for opcode in group {
if opcode.tag != "insert" { if opcode.tag != "insert" {
for i in opcode.first_start..opcode.first_end { for item in first_sequence
res.push(format!( .iter()
"{}{}", .take(opcode.first_end)
&prefix[&opcode.tag], .skip(opcode.first_start)
first_sequence.at_index(i).unwrap() {
)); res.push(format!("{}{}", &prefix[&opcode.tag], item));
} }
} }
} }
@@ -141,12 +154,12 @@ pub fn context_diff<T: Sequence>(
if any { if any {
for opcode in group { for opcode in group {
if opcode.tag != "delete" { if opcode.tag != "delete" {
for i in opcode.second_start..opcode.second_end { for item in second_sequence
res.push(format!( .iter()
"{}{}", .take(opcode.second_end)
&prefix[&opcode.tag], .skip(opcode.second_start)
second_sequence.at_index(i).unwrap() {
)); res.push(format!("{}{}", prefix[&opcode.tag], item));
} }
} }
} }
src/sequencematcher.rs
+27 -90
View File
@@ -1,6 +1,6 @@
use std::cmp::{max, min}; use std::cmp::{max, min};
use std::collections::HashMap; use std::collections::HashMap;
use std::fmt::Debug; use std::hash::Hash;
use utils::calculate_ratio; use utils::calculate_ratio;
#[derive(Debug, Clone, Copy, PartialEq, PartialOrd, Eq, Ord)] #[derive(Debug, Clone, Copy, PartialEq, PartialOrd, Eq, Ord)]
@@ -47,87 +47,20 @@ impl Opcode {
} }
} }
pub trait Sequence: Debug { pub trait Sequence: Eq + Hash {}
fn len(&self) -> usize; impl<T: Eq + Hash> Sequence for T {}
fn at_index(&self, index: usize) -> Option<&str>;
}
impl Sequence for str { pub struct SequenceMatcher<'a, T: 'a + Sequence> {
fn len(&self) -> usize { first_sequence: &'a [T],
self.len() second_sequence: &'a [T],
}
fn at_index(&self, index: usize) -> Option<&str> {
if index > self.len() {
return None;
}
unsafe { Some(self.slice_unchecked(index, index + 1)) }
}
}
impl<'a> Sequence for Vec<&'a str> {
fn len(&self) -> usize {
self.len()
}
fn at_index(&self, index: usize) -> Option<&str> {
if index < self.len() {
return Some(self[index]);
}
None
}
}
impl Sequence for String {
fn len(&self) -> usize {
self.len()
}
fn at_index(&self, index: usize) -> Option<&str> {
if index > self.len() {
return None;
}
unsafe { Some(self.slice_unchecked(index, index + 1)) }
}
}
impl<'a> Sequence for Vec<String> {
fn len(&self) -> usize {
self.len()
}
fn at_index(&self, index: usize) -> Option<&str> {
if index < self.len() {
return Some(&self[index]);
}
None
}
}
impl<'a> Sequence for [&'a str] {
fn len(&self) -> usize {
self.len()
}
fn at_index(&self, index: usize) -> Option<&str> {
if index < self.len() {
return Some(self[index]);
}
None
}
}
pub struct SequenceMatcher<'a, T: 'a + ?Sized + Sequence> {
first_sequence: &'a T,
second_sequence: &'a T,
matching_blocks: Option<Vec<Match>>, matching_blocks: Option<Vec<Match>>,
opcodes: Option<Vec<Opcode>>, opcodes: Option<Vec<Opcode>>,
is_junk: Option<fn(&str) -> bool>, is_junk: Option<fn(&T) -> bool>,
second_sequence_elements: HashMap<&'a str, Vec<usize>>, second_sequence_elements: HashMap<&'a T, Vec<usize>>,
} }
impl<'a, T: ?Sized + Sequence> SequenceMatcher<'a, T> { impl<'a, T: Sequence> SequenceMatcher<'a, T> {
pub fn new(first_sequence: &'a T, second_sequence: &'a T) -> SequenceMatcher<'a, T> { pub fn new(first_sequence: &'a [T], second_sequence: &'a [T]) -> SequenceMatcher<'a, T> {
let mut matcher = SequenceMatcher { let mut matcher = SequenceMatcher {
first_sequence, first_sequence,
second_sequence, second_sequence,
@@ -140,23 +73,23 @@ impl<'a, T: ?Sized + Sequence> SequenceMatcher<'a, T> {
matcher matcher
} }
pub fn set_is_junk(&mut self, is_junk: Option<fn(&str) -> bool>) { pub fn set_is_junk(&mut self, is_junk: Option<fn(&T) -> bool>) {
self.is_junk = is_junk; self.is_junk = is_junk;
self.set_second_seq(self.second_sequence); self.set_second_seq(self.second_sequence);
} }
pub fn set_seqs(&mut self, first_sequence: &'a T, second_sequence: &'a T) { pub fn set_seqs(&mut self, first_sequence: &'a [T], second_sequence: &'a [T]) {
self.set_first_seq(first_sequence); self.set_first_seq(first_sequence);
self.set_second_seq(second_sequence); self.set_second_seq(second_sequence);
} }
pub fn set_first_seq(&mut self, sequence: &'a T) { pub fn set_first_seq(&mut self, sequence: &'a [T]) {
self.first_sequence = sequence; self.first_sequence = sequence;
self.matching_blocks = None; self.matching_blocks = None;
self.opcodes = None; self.opcodes = None;
} }
pub fn set_second_seq(&mut self, sequence: &'a T) { pub fn set_second_seq(&mut self, sequence: &'a [T]) {
self.second_sequence = sequence; self.second_sequence = sequence;
self.matching_blocks = None; self.matching_blocks = None;
self.opcodes = None; self.opcodes = None;
@@ -166,9 +99,9 @@ impl<'a, T: ?Sized + Sequence> SequenceMatcher<'a, T> {
fn chain_second_seq(&mut self) { fn chain_second_seq(&mut self) {
let second_sequence = self.second_sequence; let second_sequence = self.second_sequence;
let mut second_sequence_elements = HashMap::new(); let mut second_sequence_elements = HashMap::new();
for i in 0..second_sequence.len() { for (i, item) in second_sequence.iter().enumerate() {
let mut counter = second_sequence_elements let mut counter = second_sequence_elements
.entry(second_sequence.at_index(i).unwrap()) .entry(item)
.or_insert_with(Vec::new); .or_insert_with(Vec::new);
counter.push(i); counter.push(i);
} }
@@ -202,10 +135,14 @@ impl<'a, T: ?Sized + Sequence> SequenceMatcher<'a, T> {
let second_sequence_elements = &self.second_sequence_elements; let second_sequence_elements = &self.second_sequence_elements;
let (mut best_i, mut best_j, mut best_size) = (first_start, second_start, 0); let (mut best_i, mut best_j, mut best_size) = (first_start, second_start, 0);
let mut j2len: HashMap<usize, usize> = HashMap::new(); let mut j2len: HashMap<usize, usize> = HashMap::new();
for i in first_start..first_end { for (i, item) in first_sequence
let mut new_j2len: HashMap<usize, usize> = HashMap::new(); .iter()
if let Some(indexes) = second_sequence_elements.get(first_sequence.at_index(i).unwrap()) .enumerate()
.take(first_end)
.skip(first_start)
{ {
let mut new_j2len: HashMap<usize, usize> = HashMap::new();
if let Some(indexes) = second_sequence_elements.get(item) {
for j in indexes { for j in indexes {
let j = *j; let j = *j;
if j < second_start { if j < second_start {
@@ -234,15 +171,15 @@ impl<'a, T: ?Sized + Sequence> SequenceMatcher<'a, T> {
for _ in 0..2 { for _ in 0..2 {
while best_i > first_start while best_i > first_start
&& best_j > second_start && best_j > second_start
&& first_sequence.at_index(best_i - 1) == second_sequence.at_index(best_j - 1) && first_sequence.get(best_i - 1) == second_sequence.get(best_j - 1)
{ {
best_i -= 1; best_i -= 1;
best_j -= 1; best_j -= 1;
best_size += 1; best_size += 1;
} }
while best_i + best_size < first_end && best_j + best_size < second_end while best_i + best_size < first_end
&& first_sequence.at_index(best_i + best_size) && best_j + best_size < second_end
== second_sequence.at_index(best_j + best_size) && first_sequence.get(best_i + best_size) == second_sequence.get(best_j + best_size)
{ {
best_size += 1; best_size += 1;
} }
tests/tests.rs
+6 -6
View File
@@ -5,7 +5,7 @@ use difflib::sequencematcher::{Match, Opcode, SequenceMatcher};
#[test] #[test]
fn test_longest_match() { fn test_longest_match() {
let matcher = SequenceMatcher::new(" abcd", "abcd abcd"); let matcher = SequenceMatcher::new(b" abcd", b"abcd abcd");
let m = matcher.find_longest_match(0, 5, 0, 9); let m = matcher.find_longest_match(0, 5, 0, 9);
assert_eq!(m.first_start, 0); assert_eq!(m.first_start, 0);
assert_eq!(m.second_start, 4); assert_eq!(m.second_start, 4);
@@ -14,7 +14,7 @@ fn test_longest_match() {
#[test] #[test]
fn test_all_matches() { fn test_all_matches() {
let mut matcher = SequenceMatcher::new("abxcd", "abcd"); let mut matcher = SequenceMatcher::new(b"abxcd", b"abcd");
let result = matcher.get_matching_blocks(); let result = matcher.get_matching_blocks();
let mut expected_result = Vec::new(); let mut expected_result = Vec::new();
expected_result.push(Match { expected_result.push(Match {
@@ -37,7 +37,7 @@ fn test_all_matches() {
#[test] #[test]
fn test_get_opcodes() { fn test_get_opcodes() {
let mut matcher = SequenceMatcher::new("qabxcd", "abycdf"); let mut matcher = SequenceMatcher::new(b"qabxcd", b"abycdf");
let result = matcher.get_opcodes(); let result = matcher.get_opcodes();
let mut expected_result = Vec::new(); let mut expected_result = Vec::new();
expected_result.push(Opcode { expected_result.push(Opcode {
@@ -80,7 +80,7 @@ fn test_get_opcodes() {
#[test] #[test]
fn test_ratio() { fn test_ratio() {
let mut matcher = SequenceMatcher::new("abcd", "bcde"); let mut matcher = SequenceMatcher::new(b"abcd", b"bcde");
assert_eq!(matcher.ratio(), 0.75); assert_eq!(matcher.ratio(), 0.75);
} }
@@ -103,8 +103,8 @@ fn test_differ_compare() {
); );
} }
fn is_junk_char(ch: &str) -> bool { fn is_junk_char(ch: &char) -> bool {
if ch == " " || ch == "\t" { if *ch == ' ' || *ch == '\t' {
return true; return true;
} }
false false