cache strings while encoding/decoding to compiler artifacts

3 files changed, 54 insertions, 0 deletions

diff --git a/compiler/rustc_metadata/src/rmeta/decoder.rs b/compiler/rustc_metadata/src/rmeta/decoder.rs
index 40dc4fb052d..b0d9e4a63f5 100644
--- a/compiler/rustc_metadata/src/rmeta/decoder.rs
+++ b/compiler/rustc_metadata/src/rmeta/decoder.rs
@@ -637,6 +637,35 @@ impl<'a, 'tcx> Decodable<DecodeContext<'a, 'tcx>> for Span {
     }
 }
 
+impl<'a, 'tcx> Decodable<DecodeContext<'a, 'tcx>> for Symbol {
+    fn decode(d: &mut DecodeContext<'a, 'tcx>) -> Self {
+        let tag = d.read_u8();
+
+        match tag {
+            SYMBOL_STR => {
+                let s = d.read_str();
+                Symbol::intern(s)
+            }
+            SYMBOL_OFFSET => {
+                // read str offset
+                let pos = d.read_usize();
+                let old_pos = d.opaque.position();
+
+                // move to str ofset and read
+                d.opaque.set_position(pos);
+                let s = d.read_str();
+                let sym = Symbol::intern(s);
+
+                // restore position
+                d.opaque.set_position(old_pos);
+
+                sym
+            }
+            _ => unreachable!(),
+        }
+    }
+}
+
 impl<'a, 'tcx> Decodable<DecodeContext<'a, 'tcx>> for &'tcx [ty::abstract_const::Node<'tcx>] {
     fn decode(d: &mut DecodeContext<'a, 'tcx>) -> Self {
         ty::codec::RefDecodable::decode(d)
diff --git a/compiler/rustc_metadata/src/rmeta/encoder.rs b/compiler/rustc_metadata/src/rmeta/encoder.rs
index f68005c0526..13436af09b7 100644
--- a/compiler/rustc_metadata/src/rmeta/encoder.rs
+++ b/compiler/rustc_metadata/src/rmeta/encoder.rs
@@ -39,6 +39,7 @@ use rustc_span::{
 };
 use rustc_target::abi::VariantIdx;
 use std::borrow::Borrow;
+use std::collections::hash_map::Entry;
 use std::hash::Hash;
 use std::io::{Read, Seek, Write};
 use std::iter;
@@ -75,6 +76,7 @@ pub(super) struct EncodeContext<'a, 'tcx> {
     required_source_files: Option<GrowableBitSet<usize>>,
     is_proc_macro: bool,
     hygiene_ctxt: &'a HygieneEncodeContext,
+    symbol_table: FxHashMap<Symbol, usize>,
 }
 
 /// If the current crate is a proc-macro, returns early with `Lazy:empty()`.
@@ -307,6 +309,24 @@ impl<'a, 'tcx> Encodable<EncodeContext<'a, 'tcx>> for Span {
     }
 }
 
+impl<'a, 'tcx> Encodable<EncodeContext<'a, 'tcx>> for Symbol {
+    fn encode(&self, s: &mut EncodeContext<'a, 'tcx>) {
+        match s.symbol_table.entry(*self) {
+            Entry::Vacant(o) => {
+                s.opaque.emit_u8(SYMBOL_STR);
+                let pos = s.opaque.position();
+                o.insert(pos);
+                s.emit_str(self.as_str());
+            }
+            Entry::Occupied(o) => {
+                let x = o.get().clone();
+                s.emit_u8(SYMBOL_OFFSET);
+                s.emit_usize(x);
+            }
+        }
+    }
+}
+
 impl<'a, 'tcx> TyEncoder for EncodeContext<'a, 'tcx> {
     const CLEAR_CROSS_CRATE: bool = true;
 
@@ -2259,6 +2279,7 @@ fn encode_metadata_impl(tcx: TyCtxt<'_>, path: &Path) {
         required_source_files,
         is_proc_macro: tcx.sess.crate_types().contains(&CrateType::ProcMacro),
         hygiene_ctxt: &hygiene_ctxt,
+        symbol_table: Default::default(),
     };
 
     // Encode the rustc version string in a predictable location.
diff --git a/compiler/rustc_metadata/src/rmeta/mod.rs b/compiler/rustc_metadata/src/rmeta/mod.rs
index 8efe5051b01..2bdb1feb677 100644
--- a/compiler/rustc_metadata/src/rmeta/mod.rs
+++ b/compiler/rustc_metadata/src/rmeta/mod.rs
@@ -445,6 +445,10 @@ const TAG_VALID_SPAN_LOCAL: u8 = 0;
 const TAG_VALID_SPAN_FOREIGN: u8 = 1;
 const TAG_PARTIAL_SPAN: u8 = 2;
 
+// Tags for encoding Symbol's
+const SYMBOL_STR: u8 = 0;
+const SYMBOL_OFFSET: u8 = 1;
+
 pub fn provide(providers: &mut Providers) {
     encoder::provide(providers);
     decoder::provide(providers);