etseidl commented on code in PR #8237:
URL: https://github.com/apache/arrow-rs/pull/8237#discussion_r2337672664
##########
parquet/src/parquet_thrift.rs:
##########
@@ -539,3 +545,394 @@ where
Ok(res)
}
}
+
+/////////////////////////
+// thrift compact output
+
+pub(crate) struct ThriftCompactOutputProtocol<W: Write> {
+ writer: W,
+}
+
+impl<W: Write> ThriftCompactOutputProtocol<W> {
+ pub(crate) fn new(writer: W) -> Self {
+ Self { writer }
+ }
+
+ pub(crate) fn inner(&self) -> &W {
+ &self.writer
+ }
+
+ fn write_byte(&mut self, b: u8) -> Result<()> {
+ self.writer.write_all(&[b])?;
+ Ok(())
+ }
+
+ fn write_vlq(&mut self, val: u64) -> Result<()> {
+ let mut v = val;
+ while v > 0x7f {
+ self.write_byte(v as u8 | 0x80)?;
+ v >>= 7;
+ }
+ self.write_byte(v as u8)
+ }
+
+ fn write_zig_zag(&mut self, val: i64) -> Result<()> {
+ let s = (val < 0) as i64;
+ self.write_vlq((((val ^ -s) << 1) + s) as u64)
+ }
+
+ pub(crate) fn write_field_begin(
+ &mut self,
+ field_type: FieldType,
+ field_id: i16,
+ last_field_id: i16,
+ ) -> Result<()> {
+ let delta = field_id.wrapping_sub(last_field_id);
+ if delta > 0 && delta <= 0xf {
+ self.write_byte((delta as u8) << 4 | field_type as u8)
+ } else {
+ self.write_byte(field_type as u8)?;
+ self.write_i16(field_id)
+ }
+ }
+
+ pub(crate) fn write_list_begin(&mut self, element_type: ElementType, len:
usize) -> Result<()> {
+ if len < 15 {
+ self.write_byte((len as u8) << 4 | element_type as u8)
+ } else {
+ self.write_byte(0xf0u8 | element_type as u8)?;
+ self.write_vlq(len as _)
+ }
+ }
+
+ pub(crate) fn write_struct_end(&mut self) -> Result<()> {
+ self.write_byte(0)
+ }
+
+ pub(crate) fn write_bytes(&mut self, val: &[u8]) -> Result<()> {
+ self.write_vlq(val.len() as u64)?;
+ self.writer.write_all(val)?;
+ Ok(())
+ }
+
+ pub(crate) fn write_empty_struct(&mut self, field_id: i16, last_field_id:
i16) -> Result<i16> {
+ self.write_field_begin(FieldType::Struct, field_id, last_field_id)?;
+ self.write_struct_end()?;
+ Ok(last_field_id)
+ }
+
+ pub(crate) fn write_bool(&mut self, val: bool) -> Result<()> {
+ match val {
+ true => self.write_byte(1),
+ false => self.write_byte(2),
+ }
+ }
+
+ pub(crate) fn write_i8(&mut self, val: i8) -> Result<()> {
+ self.write_byte(val as u8)
+ }
+
+ pub(crate) fn write_i16(&mut self, val: i16) -> Result<()> {
+ self.write_zig_zag(val as _)
+ }
+
+ pub(crate) fn write_i32(&mut self, val: i32) -> Result<()> {
+ self.write_zig_zag(val as _)
+ }
+
+ pub(crate) fn write_i64(&mut self, val: i64) -> Result<()> {
+ self.write_zig_zag(val as _)
+ }
+
+ pub(crate) fn write_double(&mut self, val: f64) -> Result<()> {
+ self.writer.write_all(&val.to_le_bytes())?;
+ Ok(())
+ }
+}
+
+pub(crate) trait WriteThrift {
Review Comment:
I don't think there are many more changes to this particular API so I'll
take a stab at documentation here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
