Initialize Butchunker project structure and policy system

3 files changed, 62 insertions, 0 deletions

diff --git a/policy/butck/butck_fixed_size/Cargo.lock b/policy/butck/butck_fixed_size/Cargo.lock
new file mode 100644
index 0000000..c1e1873
--- /dev/null
+++ b/policy/butck/butck_fixed_size/Cargo.lock
@@ -0,0 +1,7 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 4
+
+[[package]]
+name = "butck_fixed_size"
+version = "0.1.0"
diff --git a/policy/butck/butck_fixed_size/Cargo.toml b/policy/butck/butck_fixed_size/Cargo.toml
new file mode 100644
index 0000000..1550cb9
--- /dev/null
+++ b/policy/butck/butck_fixed_size/Cargo.toml
@@ -0,0 +1,7 @@
+[package]
+name = "butck_fixed_size"
+authors = ["Butchunker"]
+version = "0.1.0"
+edition = "2024"
+
+[dependencies]
diff --git a/policy/butck/butck_fixed_size/src/lib.rs b/policy/butck/butck_fixed_size/src/lib.rs
new file mode 100644
index 0000000..28cabff
--- /dev/null
+++ b/policy/butck/butck_fixed_size/src/lib.rs
@@ -0,0 +1,48 @@
+use std::collections::HashMap;
+
+const DEFAULT_CHUNK_SIZE: usize = 1024 * 1024; // 1MB
+
+fn get_chunk_size(params: &HashMap<&str, &str>) -> usize {
+    params
+        .get("size")
+        .and_then(|s| s.parse().ok())
+        .unwrap_or(DEFAULT_CHUNK_SIZE)
+}
+
+pub async fn chunk_fixed_size(raw_data: &[u8], params: &HashMap<&str, &str>) -> Vec<u32> {
+    let chunk_size = get_chunk_size(params);
+    (chunk_size..raw_data.len())
+        .step_by(chunk_size)
+        .map(|pos| pos as u32)
+        .collect()
+}
+
+#[derive(Default)]
+pub struct FixedSizeStream {
+    processed_bytes: usize,
+}
+
+pub async fn chunk_fixed_size_stream(
+    _current_data: &[u8],
+    len: u32,
+    stream: &mut FixedSizeStream,
+    params: &HashMap<&str, &str>,
+) -> Option<u32> {
+    let chunk_size = get_chunk_size(params);
+    let valid_len = len as usize;
+
+    let prev_chunk = stream.processed_bytes / chunk_size;
+    let new_processed = stream.processed_bytes + valid_len;
+    let new_chunk = new_processed / chunk_size;
+
+    if prev_chunk != new_chunk {
+        // Find chunk boundary in current data, update processed bytes, return position
+        let boundary_in_chunk = chunk_size - (stream.processed_bytes % chunk_size);
+        stream.processed_bytes += boundary_in_chunk;
+        Some(boundary_in_chunk.min(valid_len) as u32)
+    } else {
+        // Update bytes processed
+        stream.processed_bytes = new_processed;
+        None
+    }
+}