commit d0d4ccbd55c3ecbe34ac32a39957194bd882ce4b
Author: Ethan Simmons <eesimmons9105@gmail.com>
Date:   Wed Apr 17 02:01:14 2024 -0500

    Initial Commit
    
    Will write more useful commit messages in the, the start of this project
    was extremely rushed.

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..ea8c4bf
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+/target
diff --git a/Cargo.lock b/Cargo.lock
new file mode 100644
index 0000000..047c59c
--- /dev/null
+++ b/Cargo.lock
@@ -0,0 +1,39 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 3
+
+[[package]]
+name = "anyhow"
+version = "1.0.82"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f538837af36e6f6a9be0faa67f9a314f8119e4e4b5867c6ab40ed60360142519"
+
+[[package]]
+name = "memchr"
+version = "2.7.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c8640c5d730cb13ebd907d8d04b52f55ac9a2eec55b440c8892f40d56c76c1d"
+
+[[package]]
+name = "minimal-lexical"
+version = "0.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "68354c5c6bd36d73ff3feceb05efa59b6acb7626617f4962be322a825e61f79a"
+
+[[package]]
+name = "nom"
+version = "7.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d273983c5a657a70a3e8f2a01329822f3b8c8172b73826411a55751e404a0a4a"
+dependencies = [
+ "memchr",
+ "minimal-lexical",
+]
+
+[[package]]
+name = "vim_undo_extractor"
+version = "0.1.0"
+dependencies = [
+ "anyhow",
+ "nom",
+]
diff --git a/Cargo.toml b/Cargo.toml
new file mode 100644
index 0000000..45feca0
--- /dev/null
+++ b/Cargo.toml
@@ -0,0 +1,10 @@
+[package]
+name = "vim_undo_extractor"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+anyhow = "1.0.82"
+nom = "7.1.3"
diff --git a/layout.txt b/layout.txt
new file mode 100644
index 0000000..eb1eb21
--- /dev/null
+++ b/layout.txt
@@ -0,0 +1,202 @@
+START_MAGIC: 56 69 6D 9F 55 6E 44 6F E5   "Vim.UnDoF"
+
+Version: 00 03
+
+SHA-256: A5 B4 88 3B 3A AB FB 50 30 CF 8E 57 58 30 AE 44 38 25 A8 A3 DF B2 BD DF D6 FB E6 F2 EB E8 14 45
+
+Line Count: 00 00 00 7C 
+
+Line Length : 00 00 00 16
+Line: 20 20 20 20 73 74 64 3A 3A 73 74 72 69 6E 67 20 69 6E 70 75 74 3B    "    std::string input;"
+Line Number: 00 00 00 2B 
+Column Number: 00 00 00 15 
+
+Old Head Sequence or 0 if null: 00 00 00 01
+New Head Sequence or 0 if null: 00 00 00 83
+Current Head Sequence or 0 if null: 00 00 00 83
+
+Numhead: 00 00 00 83 
+Sequence Last: 00 00 00 83
+Sequence Current: 00 00 00 82
+
+
+Time: 00 00 00 00 66 10 CE 39
+
+Optional fields
+    04
+    01 
+    00 00 00 40
+
+End Marker: 00
+
+Header Magic: 5F D0
+
+Next Pointer: 00 00 00 00
+Previous Pointer: 00 00 00 02
+Alt Next Pointer: 00 00 00 00 
+Alt Previous Pointer: 00 00 00 00
+
+Sequence: 00 00 00 01
+
+Position
+Line Number: 00 00 00 54
+Column: 00 00 00 00
+Coladd: 00 00 00 00 
+
+Cursor VCol: FF FF FF FF 
+Flags: 00 01
+00 00 
+
+Marks
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+00 00 00 00
+00 00 00 00
+00 00 00 00
+
+
+Visual Info: 
+    Start Position:
+        00 00 00 00
+        00 00 00 00
+        00 00 00 00
+
+    End Position:
+        00 00 00 00
+        00 00 00 00
+        00 00 00 00
+
+    Vi Mode: 00 00 00 00
+    Vi_curswant: 00 00 00 00
+
+Time: 00 00 00 00 66 10 BC DC
+
+Optional Fields:
+    04
+    01
+    00 00 00 00
+
+End Marker: 00
+
+Entry Magic: F5 18
+
+Entry Type: 00 00 00 00
+Entry Data: 54 00 00 00 00 00 00 00 00 05 00 00 00 00 00 00 00
+
+Entry Magic: F5 18
+
+Entry Type: 00 00 00 00
+Entry Data: 55 00 00 00 04 00 00 00 00 00 00 00 05
+00 00 00 00 00 00 00 05 00 00 00 71 06 00 00 00
+00 00 00 05 00 00 00 00 00 00 00 05 00 00 00 00
+00 00 00
+
+Entry End Magic: 35 81
+
+Header Magic: 5F D0
+00 00 00 01 00 00 00 03 00
+
+
+
+
+
+
+
diff --git a/src/main.rs b/src/main.rs
new file mode 100644
index 0000000..0a3fa64
--- /dev/null
+++ b/src/main.rs
@@ -0,0 +1,28 @@
+use anyhow::Result;
+
+
+mod undo;
+mod parse;
+
+use undo::UndoFile;
+
+
+fn main() -> Result<()> {
+
+    let undo_file = UndoFile::from_path("./undo_file")?;
+
+    for header in undo_file.headers {
+        for entry in header.entries {
+            let section: String = entry.section
+                .iter()
+                .map(|b| char::from_u32(*b as u32).unwrap_or(' '))
+                .collect();
+
+            println!("{}", section);
+                
+        }
+    }
+
+    Ok(())
+}
+
diff --git a/src/parse.rs b/src/parse.rs
new file mode 100644
index 0000000..e6be60e
--- /dev/null
+++ b/src/parse.rs
@@ -0,0 +1,42 @@
+use nom::{
+    bytes::complete::take,
+    combinator::map_res,
+    IResult,
+    error::Error,
+};
+
+pub(crate) mod start_header;
+pub(crate) mod header;
+pub(crate) mod entry;
+
+pub fn bytes_u32(input: &[u8]) -> IResult<&[u8], u32> {
+    map_res(
+        take(4usize),
+        |b: &[u8]| Ok::<u32, Error<&[u8]>>(u32::from_be_bytes(b.try_into().unwrap()))
+    )(input)
+}
+
+fn time(input: &[u8]) -> IResult<&[u8], u64> {
+    map_res(
+        take(8usize),
+        |b: &[u8]| Ok::<u64, Error<&[u8]>>(u64::from_be_bytes(b.try_into().unwrap()))
+    )(input)
+}
+
+#[derive(Debug)]
+struct OptionalFields((u8, u8, u32));
+
+fn optional_fields(input: &[u8]) -> IResult<&[u8], OptionalFields> {
+    map_res(
+        take(6usize),
+        |b: &[u8]| {
+            Ok::<OptionalFields, Error<&[u8]>>(
+                OptionalFields ((
+                    b[0],
+                    b[1],
+                    u32::from_be_bytes(b[2..6].try_into().unwrap())
+                ))
+            )
+        }
+    )(input)
+}
diff --git a/src/parse/entry.rs b/src/parse/entry.rs
new file mode 100644
index 0000000..145d19e
--- /dev/null
+++ b/src/parse/entry.rs
@@ -0,0 +1,97 @@
+use nom::{
+    IResult,
+    Parser,
+    Or,
+    error::Error,
+    combinator::map_res,
+    sequence::{tuple, pair},
+    multi::many0,
+    bytes::complete::{tag, take, take_until},
+};
+
+use super::bytes_u32;
+
+#[derive(Debug, PartialEq, Eq)]
+pub(crate) struct Entry {
+    pub(crate) entry_type: EntryType,
+    pub(crate) section: Vec<u8>,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+enum EntryType {
+    Unknown(u32),
+}
+
+fn magic(input: &[u8]) -> IResult<&[u8], &[u8]> {
+    tag(b"\xf5\x18")(input)
+}
+
+fn entry(input: &[u8]) -> IResult<&[u8], Entry> {
+    let (input, section_type) = take(4usize)(input)?;
+    let (input, section) = take_until(b"\xf5\x18".as_ref())(input)?;
+    let (input, _) = tag(b"\xf5\x18")(input)?;
+
+    let entry_type = match section_type {
+        num => EntryType::Unknown(u32::from_be_bytes(num.try_into().unwrap())),
+    };
+
+    println!("{:#?}", section);
+
+    Ok((
+        input,
+        Entry {
+            entry_type,
+            section: section.to_vec(), 
+        }
+    ))
+
+}
+
+fn sections(input: &[u8]) -> IResult<&[u8], Vec<Entry>> {
+    let (input, sections) = take_until(b"\x35\x81".as_ref())(input)?;
+    let (last_entry, mut entries) = many0(entry)(sections)?;
+    entries.push(
+        Entry {
+            entry_type: match last_entry[0..4] {
+                _ => EntryType::Unknown(u32::from_be_bytes(last_entry[0..4].try_into().unwrap())),
+            },
+            section: last_entry[4..].to_vec(),
+        });
+    Ok((input, entries))
+}
+
+pub(crate) fn parse(input: &[u8]) -> IResult<&[u8], Vec<Entry>> {
+    let (input, (
+                _,
+                entries,
+                _,
+            )) = tuple((
+                        magic,
+                        sections,
+                        tag(b"\x35\x81".as_ref())
+                    ))(input).unwrap();
+
+    Ok((
+        input,
+        entries,
+      ))
+
+}
+
+#[cfg(test)]
+mod tests {
+
+    use super::{sections, Entry, EntryType};
+
+    #[test]
+    fn test_sections() {
+        let test_str = b"\x00\x00\x00\x00\xaa\xaa\xf5\x18\x00\x00\x00\x00\xaa\xaa\xf5\x18\x00\x00\x00\x00\xaa\xaa\x35\x81";
+        
+        assert_eq!(sections(test_str).unwrap().1, vec![
+                   Entry {
+                       entry_type: EntryType::Unknown,
+                       section: vec![b'\xaa', b'\xaa'],
+                   },
+        ])
+    }
+}
diff --git a/src/parse/header.rs b/src/parse/header.rs
new file mode 100644
index 0000000..bf591b2
--- /dev/null
+++ b/src/parse/header.rs
@@ -0,0 +1,175 @@
+use nom::{
+    IResult,
+    error::Error,
+    combinator::map_res,
+    sequence::tuple,
+    bytes::complete::{tag, take},
+};
+
+use super::{bytes_u32, time, optional_fields, OptionalFields};
+
+use crate::parse::entry::Entry;
+
+#[derive(Debug)]
+pub(crate) struct Header {
+    pub(crate) next: u32,
+    pub(crate) previous: u32,
+    pub(crate) alt_next: u32,
+    pub(crate) alt_previous: u32,
+    pub(crate) sequence: u32,
+    pub(crate) position: Position,
+    pub(crate) cursor_vcol: u32,
+    pub(crate) flags: Vec<Flag>,
+    pub(crate) marks: Vec<u8>,
+    pub(crate) visual_info: VisualInfo,
+    pub(crate) time: u64,
+    pub(crate) optional_fields: OptionalFields,
+    pub(crate) entries: Vec<Entry>
+}
+
+#[derive(Debug)]
+struct Position {
+    line_number: u32,
+    column_number: u32,
+    coladd: u32,
+}
+
+#[derive(Debug)]
+enum Flag {
+    Unknown,
+}
+
+#[derive(Debug)]
+struct VisualInfo {
+    start: Position,
+    end: Position,
+    mode: u32,
+    curswant: u32,
+}
+
+fn magic(input: &[u8]) -> IResult<&[u8], &[u8]> {
+    tag(b"\x5f\xd0".as_ref())(input)
+}
+
+fn position(input: &[u8]) -> IResult<&[u8], Position> {
+    map_res(
+        tuple((
+                bytes_u32,
+                bytes_u32,
+                bytes_u32,
+              )),
+        |(line_number, column_number, coladd)| {
+            Ok::<Position, Error<&[u8]>>(Position {
+                line_number,
+                column_number,
+                coladd,
+            })
+        }
+    )(input)
+}
+
+fn visual_info(input: &[u8]) -> IResult<&[u8], VisualInfo> {
+    map_res(
+        tuple((
+                position,
+                position,
+                bytes_u32,
+                bytes_u32,
+              )),
+        |(start, end, mode, curswant)| {
+            Ok::<VisualInfo, Error<&[u8]>>(
+                VisualInfo {
+                    start,
+                    end,
+                    mode,
+                    curswant,
+                }
+            )
+        } 
+    )(input)
+}
+
+fn flags(input: &[u8]) -> IResult<&[u8], Vec<Flag>> {
+    map_res(
+        take(4usize),
+        |flags: &[u8]| {
+            Ok::<Vec<Flag>, Error<&[u8]>>(
+                flags.into_iter().map(|b| {
+                    match b {
+                        _ => Flag::Unknown
+                    }
+                }).collect()
+            )
+        }
+    )(input)
+}
+
+fn marks(input: &[u8]) ->IResult<&[u8], &[u8]> {
+    take(310usize)(input)
+}
+
+fn end_marker(input: &[u8]) -> IResult<&[u8], &[u8]> {
+    tag(b"\x00")(input)
+}
+
+pub(crate) fn parse(input: &[u8]) -> IResult<&[u8], Header> {
+    let next = bytes_u32;
+    let previous = bytes_u32;
+    let alt_next = bytes_u32;
+    let alt_previous = bytes_u32;
+    let sequence = bytes_u32;
+    let cursor_vcol = bytes_u32;
+    let time = time;
+
+    let (input, (
+            _,
+            next,
+            previous,
+            alt_next,
+            alt_previous,
+            sequence,
+            position,
+            cursor_vcol,
+            flags,
+            marks,
+            visual_info,
+            time,
+            optional_fields,
+            _,
+            )) = tuple((
+                    magic,
+                    next,
+                    previous,
+                    alt_next,
+                    alt_previous,
+                    sequence,
+                    position,
+                    cursor_vcol,
+                    flags,
+                    marks,
+                    visual_info,
+                    time,
+                    optional_fields,
+                    end_marker,
+                    ))(input).unwrap();
+
+    Ok((
+
+        input,
+        Header {
+            next,
+            previous,
+            alt_next,
+            alt_previous,
+            sequence,
+            position,
+            visual_info,
+            cursor_vcol,
+            flags,
+            marks: marks.to_vec(),
+            time,
+            optional_fields,
+            entries: Vec::new(),
+        }
+    ))
+}
diff --git a/src/parse/start_header.rs b/src/parse/start_header.rs
new file mode 100644
index 0000000..63782c5
--- /dev/null
+++ b/src/parse/start_header.rs
@@ -0,0 +1,154 @@
+use nom::{
+    IResult,
+    error::Error,
+    combinator::map_res,
+    sequence::tuple,
+    bytes::complete::{tag, take},
+};
+
+use super::{bytes_u32, time, optional_fields, OptionalFields};
+
+
+fn magic(input: &[u8]) -> IResult<&[u8], &[u8]> {
+    tag(b"\x56\x69\x6D\x9F\x55\x6E\x44\x6F\xE5")(input)
+}
+
+fn version(input: &[u8]) -> IResult<&[u8], u16> {
+    map_res(
+        take(2usize),
+        |b: &[u8]| Ok::<u16, Error<&[u8]>>(u16::from_be_bytes(b.try_into().unwrap()))
+    )(input)
+}
+
+fn hash(input: &[u8]) -> IResult<&[u8], &[u8; 32]> {
+    map_res(
+        take(32usize),
+        |b: &[u8]| Ok::<&[u8; 32], Error<&[u8]>>(b.try_into().unwrap())
+    )(input)
+}
+
+fn line_with_length(input: &[u8]) -> IResult<&[u8], String> {
+    let (input, line_length) = bytes_u32(input)?;
+    map_res(
+        take(line_length),
+        |b: &[u8]| Ok::<String, Error<&[u8]>>(String::from_utf8(b.to_vec()).expect("Invalid UTF-8"))
+    )(input)
+}
+
+
+
+fn end_marker(input: &[u8]) -> IResult<&[u8], &[u8]> {
+    tag(b"\x00".as_ref())(input)
+}
+
+#[derive(Debug)]
+pub(crate) struct StartHeader {
+    version: u16,
+    hash: Vec<u8>,
+    line_count: u32,
+    line: String,
+    line_number: u32,
+    column_number: u32,
+    old_head_sequence: u32,
+    new_head_sequence: u32,
+    current_head_sequence: u32,
+    numhead: u32,
+    last_sequence: u32,
+    current_sequence: u32,
+    time: u64,
+    optional_fields: OptionalFields,
+}
+
+pub(crate) fn parse(input: &[u8]) -> IResult<&[u8], StartHeader> {
+    let line_count = bytes_u32;
+    let line_number = bytes_u32;
+    let column_number = bytes_u32;
+    let old_head_sequence = bytes_u32;
+    let new_head_sequence = bytes_u32;
+    let current_head_sequence = bytes_u32;
+    let numhead = bytes_u32;
+    let last_sequence = bytes_u32;
+    let current_sequence = bytes_u32;
+
+
+    let (input, (
+                _,
+                version,
+                hash,
+                line_count,
+                line,
+                line_number,
+                column_number,
+                old_head_sequence,
+                new_head_sequence,
+                current_head_sequence,
+                numhead,
+                last_sequence,
+                current_sequence,
+                time,
+                optional_fields,
+                _,
+            )) = tuple ((
+                     magic,
+                     version,
+                     hash,
+                     line_count,
+                     line_with_length,
+                     line_number,
+                     column_number,
+                     old_head_sequence,
+                     new_head_sequence,
+                     current_head_sequence,
+                     numhead,
+                     last_sequence,
+                     current_sequence,
+                     time,
+                     optional_fields,
+                     end_marker,
+                 ))(&input).unwrap();
+
+    Ok(
+        (
+            input,
+            StartHeader {
+                version,
+                hash: hash.to_vec(),
+                line_count,
+                line,
+                line_number,
+                column_number,
+                old_head_sequence,
+                new_head_sequence,
+                current_head_sequence,
+                numhead,
+                last_sequence,
+                current_sequence,
+                time,
+                optional_fields,
+            }
+        )
+    )
+}
+
+#[cfg(test)]
+mod tests {
+    use super::{magic, version};
+
+    #[test]
+    fn test_magic() {
+        let test_str = b"\x56\x69\x6d\x9f\x55\x6e\x44\x6f\xe5\x00\x03";
+
+        let (_, magic) = magic(test_str).unwrap();
+
+        assert_eq!(magic, b"\x56\x69\x6d\x9f\x55\x6e\x44\x6f\xe5");
+    }
+
+    #[test]
+    fn test_version() {
+        let test_str = b"\x00\x03\xa5\xb4";
+
+        let (_, version) = version(test_str).unwrap();
+
+        assert_eq!(version, 3);
+    }
+}
diff --git a/src/undo.rs b/src/undo.rs
new file mode 100644
index 0000000..5e466e4
--- /dev/null
+++ b/src/undo.rs
@@ -0,0 +1,50 @@
+use std::fs::File;
+use std::io::Read;
+
+use anyhow::{Result, anyhow};
+
+use crate::parse::start_header::{self, StartHeader};
+use crate::parse::header::{self, Header};
+use crate::parse::entry::{self, Entry};
+
+use nom::{
+    sequence::tuple,
+    multi::many0,
+    bytes::complete::{take_until, tag},
+};
+
+#[derive(Debug)]
+pub(crate) struct UndoFile {
+    pub(crate) start_header: StartHeader,
+    pub(crate) headers: Vec<Header>,
+}
+
+impl UndoFile {
+    pub(crate) fn from_path(path: &str) -> Result<Self> {
+        let mut buffer: Vec<u8> = Vec::new();
+        let mut file = File::open(path)?;
+
+        let _ = match file.read_to_end(&mut buffer) {
+            Ok(_) => Ok(()),
+            Err(e) => Err(anyhow!(e)),
+        };
+
+        let (input, start_header) = start_header::parse(&buffer).unwrap();
+        let (_, out) = many0(tuple((header::parse, entry::parse, take_until(b"\x5f\xd0".as_ref()))))(input).unwrap();
+        
+        let headers = out
+            .into_iter()
+            .map(|(mut header, entries, _)| {
+                entries.into_iter().for_each(|entry| header.entries.push(entry));
+                header
+            }).collect();
+
+        Ok(
+            Self {
+                start_header,
+                headers,
+            }
+        )
+    }
+
+}
diff --git a/undo_file b/undo_file
new file mode 100644
index 0000000..73370a5
Binary files /dev/null and b/undo_file differ