[pbs-devel] [PATCH v5 pxar 3/28] fix #3174: encoder: calc filename + metadata byte size

[Christian Ebner]

Introduce SeqSink and impl SeqWrite in order to create an encoder
implementation which instead of writing data to a stream, consumes
the encoded stream and returns the consumed bytes for that stream.

Based on this, implement a helper function `byte_len` which returns the
byte size of the filename entry and metadata entry as encoded by the
archive.

Signed-off-by: Christian Ebner <c.ebner at proxmox.com>
---
Changes since v4:
- Refactor and rename `bytes_len` to `encoded_size`

Changes since v3:
- no changes

Changes since v2:
- no changes

Changes since v1:
- Instead of calculating the metadata size based on the known encoding
  sizes, implement an Encoder instance which counts the encoded bytes.

 src/encoder/mod.rs  | 35 +++++++++++++++++++++++++++++++++++
 src/encoder/sync.rs |  5 +++++
 2 files changed, 40 insertions(+)

diff --git a/src/encoder/mod.rs b/src/encoder/mod.rs
index 0d342ec..860c21f 100644
--- a/src/encoder/mod.rs
+++ b/src/encoder/mod.rs
@@ -85,6 +85,24 @@ where
     }
 }
 
+#[derive(Default)]
+/// Sink to consume sequential byte stream
+pub struct SeqSink;
+
+impl SeqWrite for SeqSink {
+    fn poll_seq_write(
+        self: Pin<&mut Self>,
+        _cx: &mut Context,
+        buf: &[u8],
+    ) -> Poll<io::Result<usize>> {
+        Poll::Ready(Ok(buf.len()))
+    }
+
+    fn poll_flush(self: Pin<&mut Self>, _cx: &mut Context) -> Poll<io::Result<()>> {
+        Poll::Ready(Ok(()))
+    }
+}
+
 /// awaitable verison of `poll_seq_write`.
 async fn seq_write<T: SeqWrite + ?Sized>(
     output: &mut T,
@@ -183,6 +201,23 @@ where
     seq_write_pxar_entry(output, htype, buf, position).await
 }
 
+/// Calculate the encoded byte len of filename and metadata struct
+pub async fn encoded_size(filename: &std::ffi::CStr, metadata: &Metadata) -> io::Result<u64> {
+    let mut this = EncoderImpl {
+        output: EncoderOutput::Owned(SeqSink::default()),
+        state: EncoderState::default(),
+        parent: None,
+        finished: false,
+        file_copy_buffer: Arc::new(Mutex::new(unsafe {
+            crate::util::vec_new_uninitialized(1024 * 1024)
+        })),
+    };
+
+    this.start_file_do(Some(metadata), filename.to_bytes())
+        .await?;
+    Ok(this.position())
+}
+
 /// Error conditions caused by wrong usage of this crate.
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
 pub enum EncodeError {
diff --git a/src/encoder/sync.rs b/src/encoder/sync.rs
index 1ec91b8..93c3b2c 100644
--- a/src/encoder/sync.rs
+++ b/src/encoder/sync.rs
@@ -228,3 +228,8 @@ impl<T: io::Write> SeqWrite for StandardWriter<T> {
         Poll::Ready(self.pin_to_inner().and_then(|inner| inner.flush()))
     }
 }
+
+/// Calculate the encoded byte len of filename and metadata struct
+pub fn encoded_size(filename: &std::ffi::CStr, metadata: &Metadata) -> io::Result<u64> {
+    poll_result_once(crate::encoder::encoded_size(filename, metadata))
+}
-- 
2.39.2