Rollup merge of rust-lang#30565 - michaelwoerister:opaque_encoder, r=…

…brson

This PR changes the `emit_opaque` and `read_opaque` methods in the RBML library to use a space-efficient binary encoder that does not emit any tags and uses the LEB128 variable-length integer format for all numbers it emits.

The space savings are nice, albeit a bit underwhelming, especially for dynamic libraries where metadata is already compressed.

| RLIBs        |  NEW   |   OLD     |
|--------------|--------|-----------|
|libstd        | 8.8 MB |  10.5 MB  |
|libcore       |15.6 MB |   19.7 MB |
|libcollections| 3.7 MB |    4.8 MB |
|librustc      |34.0 MB |   37.8 MB |
|libsyntax     |28.3 MB |   32.1 MB |

| SOs           |     NEW   |    OLD |
|---------------|-----------|--------|
| libstd        |  4.8 MB   | 5.1 MB |
| librustc      |  8.6 MB   | 9.2 MB |
| libsyntax     |  7.8 MB   | 8.4 MB |

At least this should make up for the size increase caused recently by also storing MIR in crate metadata.

Can this be a breaking change for anyone?
cc @rust-lang/compiler

src/librbml/leb128.rs

@@ -0,0 +1,162 @@
+// Copyright 2012-2015 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+#[inline]
+pub fn write_to_vec(vec: &mut Vec<u8>, position: &mut usize, byte: u8)
+{
+    if *position == vec.len() {
+        vec.push(byte);
+    } else {
+        vec[*position] = byte;
+    }
+
+    *position += 1;
+}
+
+pub fn write_unsigned_leb128(out: &mut Vec<u8>,
+                             start_position: usize,
+                             mut value: u64)
+                             -> usize {
+    let mut position = start_position;
+    loop
+    {
+        let mut byte = (value & 0x7F) as u8;
+        value >>= 7;
+        if value != 0 {
+            byte |= 0x80;
+        }
+
+        write_to_vec(out, &mut position, byte);
+
+        if value == 0 {
+            break;
+        }
+    }
+
+    return position - start_position;
+}
+
+pub fn read_unsigned_leb128(data: &[u8],
+                            start_position: usize)
+                            -> (u64, usize) {
+    let mut result = 0;
+    let mut shift = 0;
+    let mut position = start_position;
+    loop {
+        let byte = data[position];
+        position += 1;
+        result |= ((byte & 0x7F) as u64) << shift;
+        if (byte & 0x80) == 0 {
+            break;
+        }
+        shift += 7;
+    }
+
+    (result, position - start_position)
+}
+
+
+pub fn write_signed_leb128(out: &mut Vec<u8>,
+                           start_position: usize,
+                           mut value: i64) -> usize {
+    let mut position = start_position;
+
+    loop {
+        let mut byte = (value as u8) & 0x7f;
+        value >>= 7;
+        let more = !((((value == 0 ) && ((byte & 0x40) == 0)) ||
+                      ((value == -1) && ((byte & 0x40) != 0))));
+        if more {
+            byte |= 0x80; // Mark this byte to show that more bytes will follow.
+        }
+
+        write_to_vec(out, &mut position, byte);
+
+        if !more {
+            break;
+        }
+    }
+
+    return position - start_position;
+}
+
+pub fn read_signed_leb128(data: &[u8],
+                          start_position: usize)
+                          -> (i64, usize) {
+    let mut result = 0;
+    let mut shift = 0;
+    let mut position = start_position;
+    let mut byte;
+
+    loop {
+        byte = data[position];
+        position += 1;
+        result |= ((byte & 0x7F) as i64) << shift;
+        shift += 7;
+
+        if (byte & 0x80) == 0 {
+            break;
+        }
+    }
+
+    if (shift < 64) && ((byte & 0x40) != 0) {
+        /* sign extend */
+        result |= -(1i64 << shift);
+    }
+
+    (result, position - start_position)
+}
+
+#[test]
+fn test_unsigned_leb128() {
+    let mut stream = Vec::with_capacity(10000);
+
+    for x in 0..62 {
+        let pos = stream.len();
+        let bytes_written = write_unsigned_leb128(&mut stream, pos, 3 << x);
+        assert_eq!(stream.len(), pos + bytes_written);
+    }
+
+    let mut position = 0;
+    for x in 0..62 {
+        let expected = 3 << x;
+        let (actual, bytes_read) = read_unsigned_leb128(&stream, position);
+        assert_eq!(expected, actual);
+        position += bytes_read;
+    }
+    assert_eq!(stream.len(), position);
+}
+
+#[test]
+fn test_signed_leb128() {
+    let mut values = Vec::new();
+
+    let mut i = -500;
+    while i < 500 {
+        values.push(i * 123457i64);
+        i += 1;
+    }
+
+    let mut stream = Vec::new();
+
+    for &x in &values {
+        let pos = stream.len();
+        let bytes_written = write_signed_leb128(&mut stream, pos, x);
+        assert_eq!(stream.len(), pos + bytes_written);
+    }
+
+    let mut pos = 0;
+    for &x in &values {
+        let (value, bytes_read) = read_signed_leb128(&mut stream, pos);
+        pos += bytes_read;
+        assert_eq!(x, value);
+    }
+    assert_eq!(pos, stream.len());
+}

src/librbml/lib.rs

@@ -128,10 +128,17 @@
 #![cfg_attr(test, feature(test))]
 
 extern crate serialize;
+
+#[cfg(test)]
+extern crate serialize as rustc_serialize; // Used by RustcEncodable
+
 #[macro_use] extern crate log;
 
 #[cfg(test)] extern crate test;
 
+pub mod opaque;
+pub mod leb128;
+
 pub use self::EbmlEncoderTag::*;
 pub use self::Error::*;
 
@@ -241,6 +248,7 @@ pub mod reader {
 
     use serialize;
 
+    use super::opaque;
     use super::{ ApplicationError, EsVec, EsMap, EsEnum, EsSub8, EsSub32,
         EsVecElt, EsMapKey, EsU64, EsU32, EsU16, EsU8, EsI64,
         EsI32, EsI16, EsI8, EsBool, EsF64, EsF32, EsChar, EsStr, EsMapVal,
@@ -621,18 +629,16 @@ pub mod reader {
         }
 
         pub fn read_opaque<R, F>(&mut self, op: F) -> DecodeResult<R> where
-            F: FnOnce(&mut Decoder, Doc) -> DecodeResult<R>,
+            F: FnOnce(&mut opaque::Decoder, Doc) -> DecodeResult<R>,
         {
             let doc = try!(self.next_doc(EsOpaque));
 
-            let (old_parent, old_pos) = (self.parent, self.pos);
-            self.parent = doc;
-            self.pos = doc.start;
-
-            let result = try!(op(self, doc));
+            let result = {
+                let mut opaque_decoder = opaque::Decoder::new(doc.data,
+                                                              doc.start);
+                try!(op(&mut opaque_decoder, doc))
+            };
 
-            self.parent = old_parent;
-            self.pos = old_pos;
             Ok(result)
         }
 
@@ -877,6 +883,7 @@ pub mod writer {
     use std::io::prelude::*;
     use std::io::{self, SeekFrom, Cursor};
 
+    use super::opaque;
     use super::{ EsVec, EsMap, EsEnum, EsSub8, EsSub32, EsVecElt, EsMapKey,
         EsU64, EsU32, EsU16, EsU8, EsI64, EsI32, EsI16, EsI8,
         EsBool, EsF64, EsF32, EsChar, EsStr, EsMapVal,
@@ -1120,10 +1127,16 @@ pub mod writer {
         }
 
         pub fn emit_opaque<F>(&mut self, f: F) -> EncodeResult where
-            F: FnOnce(&mut Encoder) -> EncodeResult,
+            F: FnOnce(&mut opaque::Encoder) -> EncodeResult,
         {
             try!(self.start_tag(EsOpaque as usize));
-            try!(f(self));
+
+            {
+                let mut opaque_encoder = opaque::Encoder::new(self.writer);
+                try!(f(&mut opaque_encoder));
+            }
+
+            self.mark_stable_position();
             self.end_tag()
         }
     }