Struct DataflowSession 

pub struct DataflowSession {
    pub build_id: Option<BuildId>,
    pub session_id: SessionId,
    pub git_sources: BTreeMap<NodeId, GitSource>,
    pub local_build: Option<BuildInfo>,
    pub build_fingerprint: Option<String>,
    pub resolved_dataflow: Option<Descriptor>,
    pub source_fingerprint: Option<String>,
}

Fields

build_id: Option<BuildId>

session_id: SessionId

git_sources: BTreeMap<NodeId, GitSource>

local_build: Option<BuildInfo>

build_fingerprint: Option<String>

FNV-1a 64-bit hex digest of the resolved descriptor’s build-inputs. None on sessions written before this field was introduced (treated as “unknown” — invalidate_if_build_inputs_changed clears and rewrites).

resolved_dataflow: Option<Descriptor>

The dataflow descriptor with hub: references desugared into concrete git nodes, as built. dora start / dora run re-read the YAML from disk, which still contains unresolved hub: fields — they use this resolved form instead. None when the dataflow has no hub nodes.

source_fingerprint: Option<String>

FNV-1a digest of the expanded (pre-desugar) descriptor as built. For a hub dataflow the on-disk YAML can’t be re-fingerprinted directly (its hub: references are unresolved, so kind() rejects them), so dora start / dora daemon --run-dataflow compare this instead to detect any on-disk edit since the build. None for non-hub dataflows.

Implementations

impl DataflowSession

pub fn read_session(dataflow_path: &Path) -> Result<Self>

pub fn write_out_for_dataflow(&self, dataflow_path: &Path) -> Result<()>

pub fn fingerprint_source(descriptor: &Descriptor) -> Option<String>

Digest of the expanded (pre-desugar) descriptor. Used to detect any on-disk edit to a hub dataflow since its build, since the on-disk YAML can’t be re-fingerprinted via fingerprint_build_inputs (its hub: references are unresolved). Returns None if serialization fails.

pub fn fingerprint_build_inputs( resolved_nodes: &BTreeMap<NodeId, ResolvedNode>, ) -> String

Compute the build-inputs fingerprint over the resolved descriptor.

Inputs (anything that affects what dora build produces):

• per-node build command (Custom-kind nodes)
• per-node source (Local vs GitBranch with repo + branch/tag/rev)
• per-node env (sorted map, Display’d values — covers global env after resolve_aliases_and_set_defaults merges it)
• per-node deploy.working_dir (changes shift the cwd cargo/python runs in)
• per-operator build command and source-kind tag (Runtime-kind nodes — distinguishes Python vs SharedLibrary vs Wasm because switching kinds is a build-system change)

Deliberately ignored (don’t affect the build):

• path — where the artifact is read from at start time, not built to
• operator python: / shared-library: / wasm: path values, and conda_env — runtime artifact pointers, symmetric to path on Custom nodes
• deploy.machine — affects placement at start time, not build inputs

Discussed in #1444 + extended in #1947 review.

pub fn invalidate_if_build_inputs_changed( &mut self, resolved_nodes: &BTreeMap<NodeId, ResolvedNode>, ) -> bool

Compare current descriptor’s build-inputs fingerprint against the stored one. On mismatch (including “never recorded”), clear stale build metadata (build_id, local_build, git_sources) and store the new fingerprint. Returns true if invalidation occurred.

Call this on dora start / dora run / dora daemon --run-dataflow before consuming build_id, and on dora build after resolving the descriptor (so the fingerprint reflects the build that just ran).

Trait Implementations

impl Clone for DataflowSession

fn clone(&self) -> DataflowSession

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DataflowSession

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for DataflowSession

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for DataflowSession

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for DataflowSession

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.