From 83c1e564c44b996cb882f0f92c10f2f15b82cbba Mon Sep 17 00:00:00 2001
From: Catherine <whitequark@whitequark.org>
Date: Tue, 21 Oct 2025 00:49:27 +0000
Subject: [PATCH] Add `stored_size` (size after deduplication) to manifest.

---
 src/manifest.go  | 24 ++++++++++++++++--------
 src/schema.pb.go | 20 +++++++++++++++-----
 src/schema.proto |  3 ++-
 3 files changed, 33 insertions(+), 14 deletions(-)

diff --git a/src/manifest.go b/src/manifest.go
index c387b1c..839865e 100644
--- a/src/manifest.go
+++ b/src/manifest.go
@@ -180,29 +180,37 @@ func StoreManifest(ctx context.Context, name string, manifest *Manifest) (*Manif
 
 	// Replace inline files over certain size with references to external data.
 	extManifest := Manifest{
-		RepoUrl:   manifest.RepoUrl,
-		Branch:    manifest.Branch,
-		Commit:    manifest.Commit,
-		Contents:  make(map[string]*Entry),
-		Redirects: manifest.Redirects,
-		Problems:  manifest.Problems,
-		TotalSize: proto.Int64(0),
+		RepoUrl:    manifest.RepoUrl,
+		Branch:     manifest.Branch,
+		Commit:     manifest.Commit,
+		Contents:   make(map[string]*Entry),
+		Redirects:  manifest.Redirects,
+		Problems:   manifest.Problems,
+		TotalSize:  proto.Int64(0),
+		StoredSize: proto.Int64(0),
 	}
+	extObjectMap := make(map[string]int64)
 	for name, entry := range manifest.Contents {
 		cannotBeInlined := entry.GetType() == Type_InlineFile &&
 			entry.GetSize() > int64(config.Limits.MaxInlineFileSize.Bytes())
 		if cannotBeInlined {
+			dataHash := sha256.Sum256(entry.Data)
 			extManifest.Contents[name] = &Entry{
 				Type: Type_ExternalFile.Enum(),
 				Size: entry.Size,
-				Data: fmt.Appendf(nil, "sha256-%x", sha256.Sum256(entry.Data)),
+				Data: fmt.Appendf(nil, "sha256-%x", dataHash),
 				Xfrm: entry.Xfrm,
 			}
+			extObjectMap[string(dataHash[:])] = *entry.Size
 		} else {
 			extManifest.Contents[name] = entry
 		}
 		*extManifest.TotalSize += entry.GetSize()
 	}
+	// `extObjectMap` stores size once per object, deduplicating it
+	for _, storedSize := range extObjectMap {
+		*extManifest.StoredSize += storedSize
+	}
 
 	// Upload the resulting manifest and the blob it references.
 	extManifestData := EncodeManifest(&extManifest)
diff --git a/src/schema.pb.go b/src/schema.pb.go
index b040af6..677c223 100644
--- a/src/schema.pb.go
+++ b/src/schema.pb.go
@@ -1,6 +1,6 @@
 // Code generated by protoc-gen-go. DO NOT EDIT.
 // versions:
-// 	protoc-gen-go v1.36.9
+// 	protoc-gen-go v1.36.10
 // 	protoc        v6.30.2
 // source: schema.proto
 
@@ -336,8 +336,9 @@ type Manifest struct {
 	Branch  *string `protobuf:"bytes,2,opt,name=branch" json:"branch,omitempty"`
 	Commit  *string `protobuf:"bytes,3,opt,name=commit" json:"commit,omitempty"`
 	// Contents
-	Contents  map[string]*Entry `protobuf:"bytes,4,rep,name=contents" json:"contents,omitempty" protobuf_key:"bytes,1,opt,name=key" protobuf_val:"bytes,2,opt,name=value"`
-	TotalSize *int64            `protobuf:"varint,5,opt,name=total_size,json=totalSize" json:"total_size,omitempty"`
+	Contents   map[string]*Entry `protobuf:"bytes,4,rep,name=contents" json:"contents,omitempty" protobuf_key:"bytes,1,opt,name=key" protobuf_val:"bytes,2,opt,name=value"`
+	TotalSize  *int64            `protobuf:"varint,5,opt,name=total_size,json=totalSize" json:"total_size,omitempty"`
+	StoredSize *int64            `protobuf:"varint,8,opt,name=stored_size,json=storedSize" json:"stored_size,omitempty"` // after deduplication
 	// Netlify-style `_redirects`
 	Redirects []*Redirect `protobuf:"bytes,6,rep,name=redirects" json:"redirects,omitempty"`
 	// Diagnostics for non-fatal errors
@@ -411,6 +412,13 @@ func (x *Manifest) GetTotalSize() int64 {
 	return 0
 }
 
+func (x *Manifest) GetStoredSize() int64 {
+	if x != nil && x.StoredSize != nil {
+		return *x.StoredSize
+	}
+	return 0
+}
+
 func (x *Manifest) GetRedirects() []*Redirect {
 	if x != nil {
 		return x.Redirects
@@ -443,14 +451,16 @@ const file_schema_proto_rawDesc = "" +
 	"\x05force\x18\x04 \x01(\bR\x05force\"3\n" +
 	"\aProblem\x12\x12\n" +
 	"\x04path\x18\x01 \x01(\tR\x04path\x12\x14\n" +
-	"\x05cause\x18\x02 \x01(\tR\x05cause\"\xbd\x02\n" +
+	"\x05cause\x18\x02 \x01(\tR\x05cause\"\xde\x02\n" +
 	"\bManifest\x12\x19\n" +
 	"\brepo_url\x18\x01 \x01(\tR\arepoUrl\x12\x16\n" +
 	"\x06branch\x18\x02 \x01(\tR\x06branch\x12\x16\n" +
 	"\x06commit\x18\x03 \x01(\tR\x06commit\x123\n" +
 	"\bcontents\x18\x04 \x03(\v2\x17.Manifest.ContentsEntryR\bcontents\x12\x1d\n" +
 	"\n" +
-	"total_size\x18\x05 \x01(\x03R\ttotalSize\x12'\n" +
+	"total_size\x18\x05 \x01(\x03R\ttotalSize\x12\x1f\n" +
+	"\vstored_size\x18\b \x01(\x03R\n" +
+	"storedSize\x12'\n" +
 	"\tredirects\x18\x06 \x03(\v2\t.RedirectR\tredirects\x12$\n" +
 	"\bproblems\x18\a \x03(\v2\b.ProblemR\bproblems\x1aC\n" +
 	"\rContentsEntry\x12\x10\n" +
diff --git a/src/schema.proto b/src/schema.proto
index b97ca95..c31d480 100644
--- a/src/schema.proto
+++ b/src/schema.proto
@@ -61,7 +61,8 @@ message Manifest {
 
 	// Contents
 	map<string, Entry> contents = 4;
-	int64 total_size = 5;
+	int64 total_size = 5; // simple sum of each `entry.size`
+	int64 stored_size = 8; // external objects, after deduplication
 
 	// Netlify-style `_redirects`
 	repeated Redirect redirects = 6;