unicode/norm: lazy load recompMap from string

recompMap gets populated on init, which requires 21kB of memory.
Instead populate it only when we need it.

This will put it in the rodata section of the binary, which means that
the Go compiler can apply better optimizations to it. The string is an
alternating series of 4 byte keys and 4 byte values, which get built
into a map the first time they are required.

Updates golang/go#26752.

Change-Id: I4f9b7d74af6df42b58999cf42cf5ada9646aa98d
Reviewed-on: https://go-review.googlesource.com/127926
Run-TryBot: Brad Fitzpatrick <[email protected]>
TryBot-Result: Gobot Gobot <[email protected]>
Reviewed-by: Brad Fitzpatrick <[email protected]>

unicode/norm/forminfo.go

@@ -4,6 +4,8 @@
 
 package norm
 
+import "encoding/binary"
+
 // This file contains Form-specific logic and wrappers for data in tables.go.
 
 // Rune info is stored in a separate trie per composing form. A composing form
@@ -178,6 +180,17 @@ func (p Properties) TrailCCC() uint8 {
 	return ccc[p.tccc]
 }
 
+func buildRecompMap() {
+	recompMap = make(map[uint32]rune, len(recompMapPacked)/8)
+	var buf [8]byte
+	for i := 0; i < len(recompMapPacked); i += 8 {
+		copy(buf[:], recompMapPacked[i:i+8])
+		key := binary.BigEndian.Uint32(buf[:4])
+		val := binary.BigEndian.Uint32(buf[4:])
+		recompMap[key] = rune(val)
+	}
+}
+
 // Recomposition
 // We use 32-bit keys instead of 64-bit for the two codepoint keys.
 // This clips off the bits of three entries, but we know this will not
@@ -188,6 +201,7 @@ func (p Properties) TrailCCC() uint8 {
 // combine returns the combined rune or 0 if it doesn't exist.
 func combine(a, b rune) rune {
 	key := uint32(uint16(a))<<16 + uint32(uint16(b))
+	recompMapOnce.Do(buildRecompMap)
 	return recompMap[key]
 }
 

unicode/norm/maketables.go

@@ -12,6 +12,7 @@ package main
 
 import (
 	"bytes"
+	"encoding/binary"
 	"flag"
 	"fmt"
 	"io"
@@ -735,6 +736,8 @@ func makeTables() {
 			max = n
 		}
 	}
+	fmt.Fprintln(w, `import "sync"`)
+	fmt.Fprintln(w)
 
 	fmt.Fprintln(w, "const (")
 	fmt.Fprintln(w, "\t// Version is the Unicode edition from which the tables are derived.")
@@ -782,16 +785,23 @@ func makeTables() {
 		sz := nrentries * 8
 		size += sz
 		fmt.Fprintf(w, "// recompMap: %d bytes (entries only)\n", sz)
-		fmt.Fprintln(w, "var recompMap = map[uint32]rune{")
+		fmt.Fprintln(w, "var recompMap map[uint32]rune")
+		fmt.Fprintln(w, "var recompMapOnce sync.Once\n")
+		fmt.Fprintln(w, `const recompMapPacked = "" +`)
+		var buf [8]byte
 		for i, c := range chars {
 			f := c.forms[FCanonical]
 			d := f.decomp
 			if !f.isOneWay && len(d) > 0 {
 				key := uint32(uint16(d[0]))<<16 + uint32(uint16(d[1]))
-				fmt.Fprintf(w, "0x%.8X: 0x%.4X,\n", key, i)
+				binary.BigEndian.PutUint32(buf[:4], key)
+				binary.BigEndian.PutUint32(buf[4:], uint32(i))
+				fmt.Fprintf(w, "\t\t%q + // 0x%.8X: 0x%.8X\n", string(buf[:]), key, uint32(i))
 			}
 		}
-		fmt.Fprintf(w, "}\n\n")
+		// hack so we don't have to special case the trailing plus sign
+		fmt.Fprintf(w, `	""`)
+		fmt.Fprintln(w)
 	}
 
 	fmt.Fprintf(w, "// Total size of tables: %dKB (%d bytes)\n", (size+512)/1024, size)