Embed a single vault entry using the provided embedding function. Chunks the content by heading/paragraph, computes embeddings via `embed_fn`, and stores them in redb. Returns the number of chunks stored. Staleness detection: skips re-embedding if the content hash matches the existing stored hash.
(
&self,
path: &str,
embed_fn: &F,
config: &EmbedConfig,
)
| 49 | /// Staleness detection: skips re-embedding if the content hash matches |
| 50 | /// the existing stored hash. |
| 51 | pub fn vault_embed<F>( |
| 52 | &self, |
| 53 | path: &str, |
| 54 | embed_fn: &F, |
| 55 | config: &EmbedConfig, |
| 56 | ) -> Result<usize, RepositoryError> |
| 57 | where |
| 58 | F: Fn(&str) -> Vec<f32>, |
| 59 | { |
| 60 | let entry = self |
| 61 | .vault_retrieve(path)? |
| 62 | .ok_or_else(|| RepositoryError::FileNotFound { |
| 63 | path: std::path::PathBuf::from(path), |
| 64 | })?; |
| 65 | |
| 66 | let content = String::from_utf8_lossy(&entry.content_bytes); |
| 67 | let content_hash = entry.content_hash; |
| 68 | |
| 69 | // Check staleness — if first chunk has same hash, skip |
| 70 | { |
| 71 | let txn = self |
| 72 | .pristine |
| 73 | .read_txn() |
| 74 | .map_err(|e| RepositoryError::Database(e.to_string()))?; |
| 75 | let still_current = txn |
| 76 | .get_vault_entry(path) |
| 77 | .map_err(|e| RepositoryError::Database(e.to_string()))? |
| 78 | .is_some_and(|current| { |
| 79 | current.entry_type == entry.entry_type |
| 80 | && current.content_hash == content_hash |
| 81 | && current.introduced_by == entry.introduced_by |
| 82 | }); |
| 83 | if !still_current { |
| 84 | return Ok(0); |
| 85 | } |
| 86 | if let Ok(Some(existing)) = txn.get_embedding(path, 0) { |
| 87 | if existing.content_hash == content_hash { |
| 88 | return Ok(0); // Content unchanged, skip |
| 89 | } |
| 90 | } |
| 91 | } |
| 92 | |
| 93 | // Chunk the content |
| 94 | let chunks = chunk_by_heading(&content, config.max_chunk_tokens); |
| 95 | |
| 96 | // Compute vectors before opening the write transaction. This keeps the |
| 97 | // database lock short and gives us one final source validation before |
| 98 | // atomically replacing all derived chunks. |
| 99 | let mut records = Vec::new(); |
| 100 | for (idx, chunk) in chunks.iter().enumerate() { |
| 101 | if chunk.text.trim().is_empty() { |
| 102 | continue; |
| 103 | } |
| 104 | |
| 105 | let record = EmbeddingRecord { |
| 106 | vector: embed_fn(&chunk.text), |
| 107 | content_hash, |
| 108 | introduced_by: entry.introduced_by, |