feat! rename metadata field "sha256" to "label"

The field still contains a cleartext SHa256 of the cleartext chunk
data, but this makes it clearer that it may contain other data.

This is a breaking change: the server API won't work with an old
client, and the new client won't work with an old server. To avoid the
breakage would require more effort than is warranted at this time,
given the very small number of users of Obnam. Sorry.

Sponsored-by: author

6 files changed, 36 insertions, 31 deletions

diff --git a/src/bin/obnam-server.rs b/src/bin/obnam-server.rs
index f06b7b5..0b80854 100644
--- a/src/bin/obnam-server.rs
+++ b/src/bin/obnam-server.rs
@@ -155,8 +155,8 @@ pub async fn search_chunks(
         }
         if key == "generation" && value == "true" {
             store.find_generations().expect("SQL lookup failed")
-        } else if key == "sha256" {
-            store.find_by_sha256(value).expect("SQL lookup failed")
+        } else if key == "label" {
+            store.find_by_label(value).expect("SQL lookup failed")
         } else {
             error!("unknown search key {:?}", key);
             return Ok(ChunkResult::BadRequest);
diff --git a/src/chunk.rs b/src/chunk.rs
index 15e3288..a37aa57 100644
--- a/src/chunk.rs
+++ b/src/chunk.rs
@@ -97,8 +97,8 @@ impl GenerationChunk {
         let json: String =
             serde_json::to_string(self).map_err(GenerationChunkError::JsonGenerate)?;
         let bytes = json.as_bytes().to_vec();
-        let sha = Checksum::sha256(&bytes);
-        let meta = ChunkMeta::new_generation(&sha, ended);
+        let checksum = Checksum::sha256(&bytes);
+        let meta = ChunkMeta::new_generation(&checksum, ended);
         Ok(DataChunk::new(bytes, meta))
     }
 }
diff --git a/src/chunkmeta.rs b/src/chunkmeta.rs
index 06a187b..9a435fe 100644
--- a/src/chunkmeta.rs
+++ b/src/chunkmeta.rs
@@ -10,7 +10,8 @@ use std::str::FromStr;
 /// We manage three bits of metadata about chunks, in addition to its
 /// identifier:
 ///
-/// * for all chunks, a [SHA256][] checksum of the chunk content
+/// * for all chunks, a [SHA256][] checksum of the chunk content; we
+///   expose this to the server as the chunk "label"
 ///
 /// * for generation chunks, an indication that it is a generation
 ///   chunk, and a timestamp for when making the generation snapshot
@@ -23,7 +24,7 @@ use std::str::FromStr;
 ///
 /// ~~~json
 /// {
-///     "sha256": "09ca7e4eaa6e8ae9c7d261167129184883644d07dfba7cbfbc4c8a2e08360d5b",
+///     "label": "09ca7e4eaa6e8ae9c7d261167129184883644d07dfba7cbfbc4c8a2e08360d5b",
 ///     "generation": true,
 ///     "ended": "2020-09-17T08:17:13+03:00"
 /// }
@@ -40,7 +41,7 @@ use std::str::FromStr;
 /// [SHA256]: https://en.wikipedia.org/wiki/SHA-2
 #[derive(Clone, Debug, Default, PartialEq, Serialize, Deserialize)]
 pub struct ChunkMeta {
-    sha256: String,
+    label: String,
     // The remaining fields are Options so that JSON parsing doesn't
     // insist on them being there in the textual representation.
     generation: Option<bool>,
@@ -51,18 +52,18 @@ impl ChunkMeta {
     /// Create a new data chunk.
     ///
     /// Data chunks are not for generations.
-    pub fn new(sha256: &Checksum) -> Self {
+    pub fn new(checksum: &Checksum) -> Self {
         ChunkMeta {
-            sha256: sha256.to_string(),
+            label: checksum.to_string(),
             generation: None,
             ended: None,
         }
     }
 
     /// Create a new generation chunk.
-    pub fn new_generation(sha256: &Checksum, ended: &str) -> Self {
+    pub fn new_generation(checksum: &Checksum, ended: &str) -> Self {
         ChunkMeta {
-            sha256: sha256.to_string(),
+            label: checksum.to_string(),
             generation: Some(true),
             ended: Some(ended.to_string()),
         }
@@ -78,9 +79,13 @@ impl ChunkMeta {
         self.ended.as_deref()
     }
 
-    /// SHA256 checksum of the content of the chunk.
-    pub fn sha256(&self) -> &str {
-        &self.sha256
+    /// The label of the content of the chunk.
+    ///
+    /// The caller should not interpret the label in any way. It
+    /// happens to be a SHA256 of the cleartext contents of the
+    /// checksum for now, but that _will_ change in the future.
+    pub fn label(&self) -> &str {
+        &self.label
     }
 
     /// Serialize from a textual JSON representation.
@@ -118,7 +123,7 @@ mod test {
         let meta = ChunkMeta::new(&sum);
         assert!(!meta.is_generation());
         assert_eq!(meta.ended(), None);
-        assert_eq!(meta.sha256(), "abcdef");
+        assert_eq!(meta.label(), "abcdef");
     }
 
     #[test]
@@ -127,26 +132,26 @@ mod test {
         let meta = ChunkMeta::new_generation(&sum, "2020-09-17T08:17:13+03:00");
         assert!(meta.is_generation());
         assert_eq!(meta.ended(), Some("2020-09-17T08:17:13+03:00"));
-        assert_eq!(meta.sha256(), "abcdef");
+        assert_eq!(meta.label(), "abcdef");
     }
 
     #[test]
     fn data_chunk_from_json() {
-        let meta: ChunkMeta = r#"{"sha256": "abcdef"}"#.parse().unwrap();
+        let meta: ChunkMeta = r#"{"label": "abcdef"}"#.parse().unwrap();
         assert!(!meta.is_generation());
         assert_eq!(meta.ended(), None);
-        assert_eq!(meta.sha256(), "abcdef");
+        assert_eq!(meta.label(), "abcdef");
     }
 
     #[test]
     fn generation_chunk_from_json() {
         let meta: ChunkMeta =
-            r#"{"sha256": "abcdef", "generation": true, "ended": "2020-09-17T08:17:13+03:00"}"#
+            r#"{"label": "abcdef", "generation": true, "ended": "2020-09-17T08:17:13+03:00"}"#
                 .parse()
                 .unwrap();
         assert!(meta.is_generation());
         assert_eq!(meta.ended(), Some("2020-09-17T08:17:13+03:00"));
-        assert_eq!(meta.sha256(), "abcdef");
+        assert_eq!(meta.label(), "abcdef");
     }
 
     #[test]
diff --git a/src/client.rs b/src/client.rs
index bcc31b4..b58f89c 100644
--- a/src/client.rs
+++ b/src/client.rs
@@ -130,7 +130,7 @@ impl BackupClient {
 
     /// Does the server have a chunk?
     pub async fn has_chunk(&self, meta: &ChunkMeta) -> Result<Option<ChunkId>, ClientError> {
-        let body = match self.get("", &[("sha256", meta.sha256())]).await {
+        let body = match self.get("", &[("label", meta.label())]).await {
             Ok((_, body)) => body,
             Err(err) => return Err(err),
         };
diff --git a/src/index.rs b/src/index.rs
index b9d29a2..4a1b9c9 100644
--- a/src/index.rs
+++ b/src/index.rs
@@ -61,8 +61,8 @@ impl Index {
         sql::remove(&self.conn, id)
     }
 
-    /// Find chunks with a given checksum.
-    pub fn find_by_sha256(&self, sha256: &str) -> Result<Vec<ChunkId>, IndexError> {
+    /// Find chunks with a client-assigned label.
+    pub fn find_by_label(&self, sha256: &str) -> Result<Vec<ChunkId>, IndexError> {
         sql::find_by_256(&self.conn, sha256)
     }
 
@@ -98,7 +98,7 @@ mod test {
         let mut idx = new_index(dir.path());
         idx.insert_meta(id.clone(), meta.clone()).unwrap();
         assert_eq!(idx.get_meta(&id).unwrap(), meta);
-        let ids = idx.find_by_sha256("abc").unwrap();
+        let ids = idx.find_by_label("abc").unwrap();
         assert_eq!(ids, vec![id]);
     }
 
@@ -110,7 +110,7 @@ mod test {
         let dir = tempdir().unwrap();
         let mut idx = new_index(dir.path());
         idx.insert_meta(id, meta).unwrap();
-        assert_eq!(idx.find_by_sha256("def").unwrap().len(), 0)
+        assert_eq!(idx.find_by_label("def").unwrap().len(), 0)
     }
 
     #[test]
@@ -122,7 +122,7 @@ mod test {
         let mut idx = new_index(dir.path());
         idx.insert_meta(id.clone(), meta).unwrap();
         idx.remove_meta(&id).unwrap();
-        let ids: Vec<ChunkId> = idx.find_by_sha256("abc").unwrap();
+        let ids: Vec<ChunkId> = idx.find_by_label("abc").unwrap();
         assert_eq!(ids, vec![]);
     }
 
@@ -193,12 +193,12 @@ mod sql {
     /// Insert a new chunk's metadata into database.
     pub fn insert(t: &Transaction, chunkid: &ChunkId, meta: &ChunkMeta) -> Result<(), IndexError> {
         let chunkid = format!("{}", chunkid);
-        let sha256 = meta.sha256();
+        let label = meta.label();
         let generation = if meta.is_generation() { 1 } else { 0 };
         let ended = meta.ended();
         t.execute(
             "INSERT INTO chunks (id, sha256, generation, ended) VALUES (?1, ?2, ?3, ?4)",
-            params![chunkid, sha256, generation, ended],
+            params![chunkid, label, generation, ended],
         )?;
         Ok(())
     }
diff --git a/src/indexedstore.rs b/src/indexedstore.rs
index 49953ee..46f9e72 100644
--- a/src/indexedstore.rs
+++ b/src/indexedstore.rs
@@ -63,9 +63,9 @@ impl IndexedStore {
         Ok(self.index.get_meta(id)?)
     }
 
-    /// Find chunks with a given checksum.
-    pub fn find_by_sha256(&self, sha256: &str) -> Result<Vec<ChunkId>, IndexedError> {
-        Ok(self.index.find_by_sha256(sha256)?)
+    /// Find chunks with a client-assigned label.
+    pub fn find_by_label(&self, label: &str) -> Result<Vec<ChunkId>, IndexedError> {
+        Ok(self.index.find_by_label(label)?)
     }
 
     /// Find all generations.