@lix-js/sdk 0.6.0-preview.1 → 0.6.0-preview.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/SKILL.md +305 -320
- package/dist/engine-wasm/wasm/lix_engine.d.ts +5 -0
- package/dist/engine-wasm/wasm/lix_engine.js +9 -13
- package/dist/engine-wasm/wasm/lix_engine.wasm +0 -0
- package/dist/engine-wasm/wasm/lix_engine.wasm.d.ts +1 -0
- package/dist/open-lix.d.ts +103 -14
- package/dist/open-lix.js +3 -0
- package/dist/sqlite/index.js +99 -22
- package/dist-engine-src/README.md +18 -0
- package/dist-engine-src/src/backend/kv.rs +358 -0
- package/dist-engine-src/src/backend/mod.rs +12 -0
- package/dist-engine-src/src/backend/testing.rs +658 -0
- package/dist-engine-src/src/backend/types.rs +96 -0
- package/dist-engine-src/src/binary_cas/chunking.rs +31 -0
- package/dist-engine-src/src/binary_cas/codec.rs +346 -0
- package/dist-engine-src/src/binary_cas/context.rs +139 -0
- package/dist-engine-src/src/binary_cas/kv.rs +1063 -0
- package/dist-engine-src/src/binary_cas/mod.rs +11 -0
- package/dist-engine-src/src/binary_cas/types.rs +127 -0
- package/dist-engine-src/src/cel/context.rs +86 -0
- package/dist-engine-src/src/cel/error.rs +19 -0
- package/dist-engine-src/src/cel/mod.rs +8 -0
- package/dist-engine-src/src/cel/provider.rs +9 -0
- package/dist-engine-src/src/cel/runtime.rs +167 -0
- package/dist-engine-src/src/cel/value.rs +50 -0
- package/dist-engine-src/src/changelog/codec.rs +321 -0
- package/dist-engine-src/src/changelog/context.rs +92 -0
- package/dist-engine-src/src/changelog/materialization.rs +121 -0
- package/dist-engine-src/src/changelog/mod.rs +13 -0
- package/dist-engine-src/src/changelog/reader.rs +20 -0
- package/dist-engine-src/src/changelog/storage.rs +220 -0
- package/dist-engine-src/src/changelog/types.rs +38 -0
- package/dist-engine-src/src/commit_graph/context.rs +1588 -0
- package/dist-engine-src/src/commit_graph/mod.rs +12 -0
- package/dist-engine-src/src/commit_graph/types.rs +145 -0
- package/dist-engine-src/src/commit_graph/walker.rs +780 -0
- package/dist-engine-src/src/common/error.rs +313 -0
- package/dist-engine-src/src/common/fingerprint.rs +3 -0
- package/dist-engine-src/src/common/fs_path.rs +1336 -0
- package/dist-engine-src/src/common/identity.rs +135 -0
- package/dist-engine-src/src/common/metadata.rs +35 -0
- package/dist-engine-src/src/common/mod.rs +23 -0
- package/dist-engine-src/src/common/types.rs +105 -0
- package/dist-engine-src/src/common/wire.rs +222 -0
- package/dist-engine-src/src/engine.rs +239 -0
- package/dist-engine-src/src/entity_identity.rs +285 -0
- package/dist-engine-src/src/functions/context.rs +327 -0
- package/dist-engine-src/src/functions/deterministic.rs +113 -0
- package/dist-engine-src/src/functions/mod.rs +18 -0
- package/dist-engine-src/src/functions/provider.rs +130 -0
- package/dist-engine-src/src/functions/state.rs +363 -0
- package/dist-engine-src/src/functions/types.rs +37 -0
- package/dist-engine-src/src/init.rs +505 -0
- package/dist-engine-src/src/json_store/compression.rs +77 -0
- package/dist-engine-src/src/json_store/context.rs +129 -0
- package/dist-engine-src/src/json_store/encoded.rs +15 -0
- package/dist-engine-src/src/json_store/mod.rs +9 -0
- package/dist-engine-src/src/json_store/store.rs +236 -0
- package/dist-engine-src/src/json_store/types.rs +52 -0
- package/dist-engine-src/src/lib.rs +61 -0
- package/dist-engine-src/src/live_state/context.rs +2241 -0
- package/dist-engine-src/src/live_state/mod.rs +15 -0
- package/dist-engine-src/src/live_state/overlay.rs +75 -0
- package/dist-engine-src/src/live_state/reader.rs +23 -0
- package/dist-engine-src/src/live_state/types.rs +239 -0
- package/dist-engine-src/src/live_state/visibility.rs +218 -0
- package/dist-engine-src/src/plugin/archive.rs +441 -0
- package/dist-engine-src/src/plugin/component.rs +183 -0
- package/dist-engine-src/src/plugin/install.rs +637 -0
- package/dist-engine-src/src/plugin/manifest.rs +516 -0
- package/dist-engine-src/src/plugin/materializer.rs +477 -0
- package/dist-engine-src/src/plugin/mod.rs +33 -0
- package/dist-engine-src/src/plugin/plugin_manifest.json +119 -0
- package/dist-engine-src/src/plugin/storage.rs +74 -0
- package/dist-engine-src/src/schema/annotations/defaults.rs +280 -0
- package/dist-engine-src/src/schema/annotations/mod.rs +1 -0
- package/dist-engine-src/src/schema/builtin/lix_account.json +22 -0
- package/dist-engine-src/src/schema/builtin/lix_active_account.json +30 -0
- package/dist-engine-src/src/schema/builtin/lix_binary_blob_ref.json +30 -0
- package/dist-engine-src/src/schema/builtin/lix_change.json +62 -0
- package/dist-engine-src/src/schema/builtin/lix_change_author.json +46 -0
- package/dist-engine-src/src/schema/builtin/lix_change_set.json +18 -0
- package/dist-engine-src/src/schema/builtin/lix_change_set_element.json +75 -0
- package/dist-engine-src/src/schema/builtin/lix_commit.json +62 -0
- package/dist-engine-src/src/schema/builtin/lix_commit_edge.json +46 -0
- package/dist-engine-src/src/schema/builtin/lix_directory_descriptor.json +53 -0
- package/dist-engine-src/src/schema/builtin/lix_entity_label.json +63 -0
- package/dist-engine-src/src/schema/builtin/lix_file_descriptor.json +53 -0
- package/dist-engine-src/src/schema/builtin/lix_key_value.json +41 -0
- package/dist-engine-src/src/schema/builtin/lix_label.json +22 -0
- package/dist-engine-src/src/schema/builtin/lix_registered_schema.json +31 -0
- package/dist-engine-src/src/schema/builtin/lix_version_descriptor.json +35 -0
- package/dist-engine-src/src/schema/builtin/lix_version_ref.json +49 -0
- package/dist-engine-src/src/schema/builtin/mod.rs +271 -0
- package/dist-engine-src/src/schema/definition.json +157 -0
- package/dist-engine-src/src/schema/definition.rs +636 -0
- package/dist-engine-src/src/schema/key.rs +206 -0
- package/dist-engine-src/src/schema/mod.rs +20 -0
- package/dist-engine-src/src/schema/seed.rs +14 -0
- package/dist-engine-src/src/schema/tests.rs +739 -0
- package/dist-engine-src/src/schema_registry.rs +294 -0
- package/dist-engine-src/src/session/context.rs +366 -0
- package/dist-engine-src/src/session/create_version.rs +80 -0
- package/dist-engine-src/src/session/execute.rs +447 -0
- package/dist-engine-src/src/session/merge/analysis.rs +102 -0
- package/dist-engine-src/src/session/merge/apply.rs +23 -0
- package/dist-engine-src/src/session/merge/conflicts.rs +62 -0
- package/dist-engine-src/src/session/merge/mod.rs +11 -0
- package/dist-engine-src/src/session/merge/stats.rs +65 -0
- package/dist-engine-src/src/session/merge/version.rs +437 -0
- package/dist-engine-src/src/session/mod.rs +25 -0
- package/dist-engine-src/src/session/switch_version.rs +121 -0
- package/dist-engine-src/src/sql2/change_provider.rs +337 -0
- package/dist-engine-src/src/sql2/classify.rs +147 -0
- package/dist-engine-src/src/sql2/commit_derived_provider.rs +591 -0
- package/dist-engine-src/src/sql2/context.rs +307 -0
- package/dist-engine-src/src/sql2/directory_history_provider.rs +623 -0
- package/dist-engine-src/src/sql2/directory_provider.rs +2405 -0
- package/dist-engine-src/src/sql2/dml.rs +148 -0
- package/dist-engine-src/src/sql2/entity_history_provider.rs +444 -0
- package/dist-engine-src/src/sql2/entity_provider.rs +2700 -0
- package/dist-engine-src/src/sql2/error.rs +196 -0
- package/dist-engine-src/src/sql2/execute.rs +3379 -0
- package/dist-engine-src/src/sql2/file_history_provider.rs +902 -0
- package/dist-engine-src/src/sql2/file_provider.rs +3254 -0
- package/dist-engine-src/src/sql2/filesystem_planner.rs +1526 -0
- package/dist-engine-src/src/sql2/filesystem_predicates.rs +159 -0
- package/dist-engine-src/src/sql2/filesystem_visibility.rs +369 -0
- package/dist-engine-src/src/sql2/history_projection.rs +80 -0
- package/dist-engine-src/src/sql2/history_provider.rs +418 -0
- package/dist-engine-src/src/sql2/history_route.rs +643 -0
- package/dist-engine-src/src/sql2/lix_state_provider.rs +2430 -0
- package/dist-engine-src/src/sql2/mod.rs +43 -0
- package/dist-engine-src/src/sql2/read_only.rs +65 -0
- package/dist-engine-src/src/sql2/record_batch.rs +17 -0
- package/dist-engine-src/src/sql2/result_metadata.rs +29 -0
- package/dist-engine-src/src/sql2/runtime.rs +60 -0
- package/dist-engine-src/src/sql2/session.rs +135 -0
- package/dist-engine-src/src/sql2/udfs/common.rs +295 -0
- package/dist-engine-src/src/sql2/udfs/lix_active_version_commit_id.rs +53 -0
- package/dist-engine-src/src/sql2/udfs/lix_empty_blob.rs +47 -0
- package/dist-engine-src/src/sql2/udfs/lix_json.rs +100 -0
- package/dist-engine-src/src/sql2/udfs/lix_json_get.rs +99 -0
- package/dist-engine-src/src/sql2/udfs/lix_json_get_text.rs +99 -0
- package/dist-engine-src/src/sql2/udfs/lix_text_decode.rs +82 -0
- package/dist-engine-src/src/sql2/udfs/lix_text_encode.rs +85 -0
- package/dist-engine-src/src/sql2/udfs/lix_uuid_v7.rs +76 -0
- package/dist-engine-src/src/sql2/udfs/mod.rs +82 -0
- package/dist-engine-src/src/sql2/version_provider.rs +1187 -0
- package/dist-engine-src/src/sql2/version_scope.rs +394 -0
- package/dist-engine-src/src/sql2/write_normalization.rs +345 -0
- package/dist-engine-src/src/storage/context.rs +356 -0
- package/dist-engine-src/src/storage/mod.rs +14 -0
- package/dist-engine-src/src/storage/read_scope.rs +88 -0
- package/dist-engine-src/src/storage/types.rs +501 -0
- package/dist-engine-src/src/storage_bench.rs +3406 -0
- package/dist-engine-src/src/test_support.rs +81 -0
- package/dist-engine-src/src/tracked_state/by_file_index.rs +102 -0
- package/dist-engine-src/src/tracked_state/codec.rs +747 -0
- package/dist-engine-src/src/tracked_state/context.rs +983 -0
- package/dist-engine-src/src/tracked_state/diff.rs +494 -0
- package/dist-engine-src/src/tracked_state/materialization.rs +141 -0
- package/dist-engine-src/src/tracked_state/merge.rs +474 -0
- package/dist-engine-src/src/tracked_state/mod.rs +31 -0
- package/dist-engine-src/src/tracked_state/rebuild.rs +771 -0
- package/dist-engine-src/src/tracked_state/storage.rs +243 -0
- package/dist-engine-src/src/tracked_state/tree.rs +2744 -0
- package/dist-engine-src/src/tracked_state/tree_types.rs +176 -0
- package/dist-engine-src/src/tracked_state/types.rs +61 -0
- package/dist-engine-src/src/transaction/commit.rs +1224 -0
- package/dist-engine-src/src/transaction/context.rs +1307 -0
- package/dist-engine-src/src/transaction/live_state_overlay.rs +34 -0
- package/dist-engine-src/src/transaction/mod.rs +11 -0
- package/dist-engine-src/src/transaction/normalization.rs +1026 -0
- package/dist-engine-src/src/transaction/schema_resolver.rs +127 -0
- package/dist-engine-src/src/transaction/staging.rs +1436 -0
- package/dist-engine-src/src/transaction/types.rs +351 -0
- package/dist-engine-src/src/transaction/validation.rs +4811 -0
- package/dist-engine-src/src/untracked_state/codec.rs +363 -0
- package/dist-engine-src/src/untracked_state/context.rs +82 -0
- package/dist-engine-src/src/untracked_state/materialization.rs +157 -0
- package/dist-engine-src/src/untracked_state/mod.rs +17 -0
- package/dist-engine-src/src/untracked_state/storage.rs +348 -0
- package/dist-engine-src/src/untracked_state/types.rs +96 -0
- package/dist-engine-src/src/version/context.rs +52 -0
- package/dist-engine-src/src/version/mod.rs +12 -0
- package/dist-engine-src/src/version/refs.rs +421 -0
- package/dist-engine-src/src/version/stage_rows.rs +71 -0
- package/dist-engine-src/src/version/types.rs +21 -0
- package/dist-engine-src/src/wasm/mod.rs +60 -0
- package/package.json +68 -64
|
@@ -0,0 +1,591 @@
|
|
|
1
|
+
use std::any::Any;
|
|
2
|
+
use std::collections::BTreeSet;
|
|
3
|
+
use std::sync::Arc;
|
|
4
|
+
|
|
5
|
+
use async_trait::async_trait;
|
|
6
|
+
use datafusion::arrow::array::{ArrayRef, BooleanArray, StringArray};
|
|
7
|
+
use datafusion::arrow::datatypes::{DataType, Field, Schema, SchemaRef};
|
|
8
|
+
use datafusion::arrow::record_batch::RecordBatch;
|
|
9
|
+
use datafusion::catalog::{Session, TableProvider};
|
|
10
|
+
use datafusion::common::{DataFusionError, Result};
|
|
11
|
+
use datafusion::datasource::TableType;
|
|
12
|
+
use datafusion::execution::TaskContext;
|
|
13
|
+
use datafusion::logical_expr::{Expr, TableProviderFilterPushDown};
|
|
14
|
+
use datafusion::physical_expr::EquivalenceProperties;
|
|
15
|
+
use datafusion::physical_plan::execution_plan::{Boundedness, EmissionType, PlanProperties};
|
|
16
|
+
use datafusion::physical_plan::stream::RecordBatchStreamAdapter;
|
|
17
|
+
use datafusion::physical_plan::{
|
|
18
|
+
DisplayAs, DisplayFormatType, ExecutionPlan, Partitioning, SendableRecordBatchStream,
|
|
19
|
+
};
|
|
20
|
+
use futures_util::stream;
|
|
21
|
+
use tokio::sync::Mutex;
|
|
22
|
+
|
|
23
|
+
use crate::commit_graph::CommitGraphReader;
|
|
24
|
+
use crate::sql2::version_scope::{resolve_provider_version_ids, VersionBinding};
|
|
25
|
+
use crate::version::VersionRefReader;
|
|
26
|
+
use crate::LixError;
|
|
27
|
+
use crate::GLOBAL_VERSION_ID;
|
|
28
|
+
|
|
29
|
+
use super::record_batch::record_batch_with_row_count;
|
|
30
|
+
|
|
31
|
+
pub(crate) async fn register_commit_derived_providers(
|
|
32
|
+
session: &datafusion::prelude::SessionContext,
|
|
33
|
+
commit_graph: Box<dyn CommitGraphReader>,
|
|
34
|
+
version_ref: Arc<dyn VersionRefReader>,
|
|
35
|
+
) -> Result<(), LixError> {
|
|
36
|
+
let commit_graph = Arc::new(Mutex::new(commit_graph));
|
|
37
|
+
for surface in CommitSurface::all() {
|
|
38
|
+
let provider = Arc::new(CommitSurfaceProvider::new(
|
|
39
|
+
surface,
|
|
40
|
+
Arc::clone(&commit_graph),
|
|
41
|
+
Arc::clone(&version_ref),
|
|
42
|
+
));
|
|
43
|
+
session
|
|
44
|
+
.register_table(surface.table_name(), provider)
|
|
45
|
+
.map_err(datafusion_error_to_lix_error)?;
|
|
46
|
+
}
|
|
47
|
+
Ok(())
|
|
48
|
+
}
|
|
49
|
+
|
|
50
|
+
#[derive(Debug, Clone, Copy)]
|
|
51
|
+
enum CommitSurface {
|
|
52
|
+
CommitEdge,
|
|
53
|
+
CommitEdgeByVersion,
|
|
54
|
+
ChangeSet,
|
|
55
|
+
ChangeSetByVersion,
|
|
56
|
+
ChangeSetElement,
|
|
57
|
+
ChangeSetElementByVersion,
|
|
58
|
+
}
|
|
59
|
+
|
|
60
|
+
impl CommitSurface {
|
|
61
|
+
fn all() -> [Self; 6] {
|
|
62
|
+
[
|
|
63
|
+
Self::CommitEdge,
|
|
64
|
+
Self::CommitEdgeByVersion,
|
|
65
|
+
Self::ChangeSet,
|
|
66
|
+
Self::ChangeSetByVersion,
|
|
67
|
+
Self::ChangeSetElement,
|
|
68
|
+
Self::ChangeSetElementByVersion,
|
|
69
|
+
]
|
|
70
|
+
}
|
|
71
|
+
|
|
72
|
+
fn table_name(self) -> &'static str {
|
|
73
|
+
match self {
|
|
74
|
+
Self::CommitEdge => "lix_commit_edge",
|
|
75
|
+
Self::CommitEdgeByVersion => "lix_commit_edge_by_version",
|
|
76
|
+
Self::ChangeSet => "lix_change_set",
|
|
77
|
+
Self::ChangeSetByVersion => "lix_change_set_by_version",
|
|
78
|
+
Self::ChangeSetElement => "lix_change_set_element",
|
|
79
|
+
Self::ChangeSetElementByVersion => "lix_change_set_element_by_version",
|
|
80
|
+
}
|
|
81
|
+
}
|
|
82
|
+
|
|
83
|
+
fn schema(self) -> SchemaRef {
|
|
84
|
+
match self {
|
|
85
|
+
Self::CommitEdge => commit_edge_schema(false),
|
|
86
|
+
Self::CommitEdgeByVersion => commit_edge_schema(true),
|
|
87
|
+
Self::ChangeSet => change_set_schema(false),
|
|
88
|
+
Self::ChangeSetByVersion => change_set_schema(true),
|
|
89
|
+
Self::ChangeSetElement => change_set_element_schema(false),
|
|
90
|
+
Self::ChangeSetElementByVersion => change_set_element_schema(true),
|
|
91
|
+
}
|
|
92
|
+
}
|
|
93
|
+
|
|
94
|
+
fn by_version(self) -> bool {
|
|
95
|
+
matches!(
|
|
96
|
+
self,
|
|
97
|
+
Self::CommitEdgeByVersion | Self::ChangeSetByVersion | Self::ChangeSetElementByVersion
|
|
98
|
+
)
|
|
99
|
+
}
|
|
100
|
+
}
|
|
101
|
+
|
|
102
|
+
struct CommitSurfaceProvider {
|
|
103
|
+
surface: CommitSurface,
|
|
104
|
+
schema: SchemaRef,
|
|
105
|
+
commit_graph: Arc<Mutex<Box<dyn CommitGraphReader>>>,
|
|
106
|
+
version_ref: Arc<dyn VersionRefReader>,
|
|
107
|
+
}
|
|
108
|
+
|
|
109
|
+
impl std::fmt::Debug for CommitSurfaceProvider {
|
|
110
|
+
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
|
111
|
+
f.debug_struct("CommitSurfaceProvider")
|
|
112
|
+
.field("surface", &self.surface)
|
|
113
|
+
.finish()
|
|
114
|
+
}
|
|
115
|
+
}
|
|
116
|
+
|
|
117
|
+
impl CommitSurfaceProvider {
|
|
118
|
+
fn new(
|
|
119
|
+
surface: CommitSurface,
|
|
120
|
+
commit_graph: Arc<Mutex<Box<dyn CommitGraphReader>>>,
|
|
121
|
+
version_ref: Arc<dyn VersionRefReader>,
|
|
122
|
+
) -> Self {
|
|
123
|
+
Self {
|
|
124
|
+
surface,
|
|
125
|
+
schema: surface.schema(),
|
|
126
|
+
commit_graph,
|
|
127
|
+
version_ref,
|
|
128
|
+
}
|
|
129
|
+
}
|
|
130
|
+
}
|
|
131
|
+
|
|
132
|
+
#[async_trait]
|
|
133
|
+
impl TableProvider for CommitSurfaceProvider {
|
|
134
|
+
fn as_any(&self) -> &dyn Any {
|
|
135
|
+
self
|
|
136
|
+
}
|
|
137
|
+
|
|
138
|
+
fn schema(&self) -> SchemaRef {
|
|
139
|
+
Arc::clone(&self.schema)
|
|
140
|
+
}
|
|
141
|
+
|
|
142
|
+
fn table_type(&self) -> TableType {
|
|
143
|
+
TableType::View
|
|
144
|
+
}
|
|
145
|
+
|
|
146
|
+
fn supports_filters_pushdown(
|
|
147
|
+
&self,
|
|
148
|
+
filters: &[&Expr],
|
|
149
|
+
) -> Result<Vec<TableProviderFilterPushDown>> {
|
|
150
|
+
Ok(filters
|
|
151
|
+
.iter()
|
|
152
|
+
.map(|_| TableProviderFilterPushDown::Unsupported)
|
|
153
|
+
.collect())
|
|
154
|
+
}
|
|
155
|
+
|
|
156
|
+
async fn scan(
|
|
157
|
+
&self,
|
|
158
|
+
_state: &dyn Session,
|
|
159
|
+
projection: Option<&Vec<usize>>,
|
|
160
|
+
_filters: &[Expr],
|
|
161
|
+
limit: Option<usize>,
|
|
162
|
+
) -> Result<Arc<dyn ExecutionPlan>> {
|
|
163
|
+
Ok(Arc::new(CommitSurfaceScanExec::new(
|
|
164
|
+
self.surface,
|
|
165
|
+
Arc::clone(&self.commit_graph),
|
|
166
|
+
Arc::clone(&self.version_ref),
|
|
167
|
+
projected_schema(&self.schema, projection),
|
|
168
|
+
projection.cloned(),
|
|
169
|
+
limit,
|
|
170
|
+
)))
|
|
171
|
+
}
|
|
172
|
+
}
|
|
173
|
+
|
|
174
|
+
struct CommitSurfaceScanExec {
|
|
175
|
+
surface: CommitSurface,
|
|
176
|
+
commit_graph: Arc<Mutex<Box<dyn CommitGraphReader>>>,
|
|
177
|
+
version_ref: Arc<dyn VersionRefReader>,
|
|
178
|
+
schema: SchemaRef,
|
|
179
|
+
projection: Option<Vec<usize>>,
|
|
180
|
+
limit: Option<usize>,
|
|
181
|
+
properties: Arc<PlanProperties>,
|
|
182
|
+
}
|
|
183
|
+
|
|
184
|
+
impl std::fmt::Debug for CommitSurfaceScanExec {
|
|
185
|
+
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
|
186
|
+
f.debug_struct("CommitSurfaceScanExec")
|
|
187
|
+
.field("surface", &self.surface)
|
|
188
|
+
.finish()
|
|
189
|
+
}
|
|
190
|
+
}
|
|
191
|
+
|
|
192
|
+
impl CommitSurfaceScanExec {
|
|
193
|
+
fn new(
|
|
194
|
+
surface: CommitSurface,
|
|
195
|
+
commit_graph: Arc<Mutex<Box<dyn CommitGraphReader>>>,
|
|
196
|
+
version_ref: Arc<dyn VersionRefReader>,
|
|
197
|
+
schema: SchemaRef,
|
|
198
|
+
projection: Option<Vec<usize>>,
|
|
199
|
+
limit: Option<usize>,
|
|
200
|
+
) -> Self {
|
|
201
|
+
let properties = PlanProperties::new(
|
|
202
|
+
EquivalenceProperties::new(schema.clone()),
|
|
203
|
+
Partitioning::UnknownPartitioning(1),
|
|
204
|
+
EmissionType::Incremental,
|
|
205
|
+
Boundedness::Bounded,
|
|
206
|
+
);
|
|
207
|
+
Self {
|
|
208
|
+
surface,
|
|
209
|
+
commit_graph,
|
|
210
|
+
version_ref,
|
|
211
|
+
schema,
|
|
212
|
+
projection,
|
|
213
|
+
limit,
|
|
214
|
+
properties: Arc::new(properties),
|
|
215
|
+
}
|
|
216
|
+
}
|
|
217
|
+
}
|
|
218
|
+
|
|
219
|
+
impl DisplayAs for CommitSurfaceScanExec {
|
|
220
|
+
fn fmt_as(&self, t: DisplayFormatType, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
|
221
|
+
match t {
|
|
222
|
+
DisplayFormatType::Default | DisplayFormatType::Verbose => {
|
|
223
|
+
write!(f, "CommitSurfaceScanExec")
|
|
224
|
+
}
|
|
225
|
+
DisplayFormatType::TreeRender => write!(f, "CommitSurfaceScanExec"),
|
|
226
|
+
}
|
|
227
|
+
}
|
|
228
|
+
}
|
|
229
|
+
|
|
230
|
+
impl ExecutionPlan for CommitSurfaceScanExec {
|
|
231
|
+
fn name(&self) -> &str {
|
|
232
|
+
"CommitSurfaceScanExec"
|
|
233
|
+
}
|
|
234
|
+
|
|
235
|
+
fn as_any(&self) -> &dyn Any {
|
|
236
|
+
self
|
|
237
|
+
}
|
|
238
|
+
|
|
239
|
+
fn properties(&self) -> &Arc<PlanProperties> {
|
|
240
|
+
&self.properties
|
|
241
|
+
}
|
|
242
|
+
|
|
243
|
+
fn children(&self) -> Vec<&Arc<dyn ExecutionPlan>> {
|
|
244
|
+
Vec::new()
|
|
245
|
+
}
|
|
246
|
+
|
|
247
|
+
fn with_new_children(
|
|
248
|
+
self: Arc<Self>,
|
|
249
|
+
children: Vec<Arc<dyn ExecutionPlan>>,
|
|
250
|
+
) -> Result<Arc<dyn ExecutionPlan>> {
|
|
251
|
+
if !children.is_empty() {
|
|
252
|
+
return Err(DataFusionError::Execution(
|
|
253
|
+
"CommitSurfaceScanExec does not accept children".to_string(),
|
|
254
|
+
));
|
|
255
|
+
}
|
|
256
|
+
Ok(self)
|
|
257
|
+
}
|
|
258
|
+
|
|
259
|
+
fn execute(
|
|
260
|
+
&self,
|
|
261
|
+
partition: usize,
|
|
262
|
+
_context: Arc<TaskContext>,
|
|
263
|
+
) -> Result<SendableRecordBatchStream> {
|
|
264
|
+
if partition != 0 {
|
|
265
|
+
return Err(DataFusionError::Execution(format!(
|
|
266
|
+
"CommitSurfaceScanExec only exposes one partition, got {partition}"
|
|
267
|
+
)));
|
|
268
|
+
}
|
|
269
|
+
|
|
270
|
+
let surface = self.surface;
|
|
271
|
+
let commit_graph = Arc::clone(&self.commit_graph);
|
|
272
|
+
let version_ref = Arc::clone(&self.version_ref);
|
|
273
|
+
let projection = self.projection.clone();
|
|
274
|
+
let limit = self.limit;
|
|
275
|
+
let schema = Arc::clone(&self.schema);
|
|
276
|
+
let stream = stream::once(async move {
|
|
277
|
+
let version_ids = if surface.by_version() {
|
|
278
|
+
resolve_provider_version_ids(
|
|
279
|
+
version_ref.as_ref(),
|
|
280
|
+
&VersionBinding::explicit(),
|
|
281
|
+
Vec::new(),
|
|
282
|
+
)
|
|
283
|
+
.await
|
|
284
|
+
.map_err(lix_error_to_datafusion_error)?
|
|
285
|
+
} else {
|
|
286
|
+
vec![GLOBAL_VERSION_ID.to_string()]
|
|
287
|
+
};
|
|
288
|
+
let rows = rows_for_surface(surface, &version_ids, Arc::clone(&commit_graph))
|
|
289
|
+
.await
|
|
290
|
+
.map_err(lix_error_to_datafusion_error)?;
|
|
291
|
+
let rows = match limit {
|
|
292
|
+
Some(limit) => rows.into_iter().take(limit).collect::<Vec<_>>(),
|
|
293
|
+
None => rows,
|
|
294
|
+
};
|
|
295
|
+
surface_record_batch(surface, projection.as_ref(), &rows)
|
|
296
|
+
});
|
|
297
|
+
Ok(Box::pin(RecordBatchStreamAdapter::new(schema, stream)))
|
|
298
|
+
}
|
|
299
|
+
}
|
|
300
|
+
|
|
301
|
+
#[derive(Debug, Clone)]
|
|
302
|
+
enum SurfaceRow {
|
|
303
|
+
CommitEdge {
|
|
304
|
+
version_id: Option<String>,
|
|
305
|
+
parent_id: String,
|
|
306
|
+
child_id: String,
|
|
307
|
+
},
|
|
308
|
+
ChangeSet {
|
|
309
|
+
version_id: Option<String>,
|
|
310
|
+
id: String,
|
|
311
|
+
},
|
|
312
|
+
ChangeSetElement {
|
|
313
|
+
version_id: Option<String>,
|
|
314
|
+
change_set_id: String,
|
|
315
|
+
change_id: String,
|
|
316
|
+
entity_id: String,
|
|
317
|
+
schema_key: String,
|
|
318
|
+
file_id: Option<String>,
|
|
319
|
+
},
|
|
320
|
+
}
|
|
321
|
+
|
|
322
|
+
async fn rows_for_surface(
|
|
323
|
+
surface: CommitSurface,
|
|
324
|
+
version_ids: &[String],
|
|
325
|
+
commit_graph: Arc<Mutex<Box<dyn CommitGraphReader>>>,
|
|
326
|
+
) -> Result<Vec<SurfaceRow>, LixError> {
|
|
327
|
+
let mut rows = Vec::new();
|
|
328
|
+
let mut seen = BTreeSet::<String>::new();
|
|
329
|
+
let mut graph = commit_graph.lock().await;
|
|
330
|
+
let commits = graph.all_commits().await?;
|
|
331
|
+
|
|
332
|
+
for version_id in version_ids {
|
|
333
|
+
match surface {
|
|
334
|
+
CommitSurface::CommitEdge | CommitSurface::CommitEdgeByVersion => {
|
|
335
|
+
for edge in graph.commit_edges(&commits) {
|
|
336
|
+
let key = format!(
|
|
337
|
+
"{version_id}\0edge\0{}\0{}",
|
|
338
|
+
edge.parent_commit_id, edge.child_commit_id
|
|
339
|
+
);
|
|
340
|
+
if seen.insert(key) {
|
|
341
|
+
rows.push(SurfaceRow::CommitEdge {
|
|
342
|
+
version_id: surface.by_version().then(|| version_id.clone()),
|
|
343
|
+
parent_id: edge.parent_commit_id,
|
|
344
|
+
child_id: edge.child_commit_id,
|
|
345
|
+
});
|
|
346
|
+
}
|
|
347
|
+
}
|
|
348
|
+
}
|
|
349
|
+
CommitSurface::ChangeSet | CommitSurface::ChangeSetByVersion => {
|
|
350
|
+
for change_set in graph.change_sets(&commits) {
|
|
351
|
+
let key = format!("{version_id}\0change_set\0{}", change_set.id);
|
|
352
|
+
if seen.insert(key) {
|
|
353
|
+
rows.push(SurfaceRow::ChangeSet {
|
|
354
|
+
version_id: surface.by_version().then(|| version_id.clone()),
|
|
355
|
+
id: change_set.id,
|
|
356
|
+
});
|
|
357
|
+
}
|
|
358
|
+
}
|
|
359
|
+
}
|
|
360
|
+
CommitSurface::ChangeSetElement | CommitSurface::ChangeSetElementByVersion => {
|
|
361
|
+
for element in graph.change_set_elements(&commits).await? {
|
|
362
|
+
let key = format!(
|
|
363
|
+
"{version_id}\0change_set_element\0{}\0{}",
|
|
364
|
+
element.change_set_id, element.change.id
|
|
365
|
+
);
|
|
366
|
+
if seen.insert(key) {
|
|
367
|
+
rows.push(SurfaceRow::ChangeSetElement {
|
|
368
|
+
version_id: surface.by_version().then(|| version_id.clone()),
|
|
369
|
+
change_set_id: element.change_set_id,
|
|
370
|
+
entity_id: element.change.entity_id.as_string()?,
|
|
371
|
+
change_id: element.change.id,
|
|
372
|
+
schema_key: element.change.schema_key,
|
|
373
|
+
file_id: element.change.file_id,
|
|
374
|
+
});
|
|
375
|
+
}
|
|
376
|
+
}
|
|
377
|
+
}
|
|
378
|
+
}
|
|
379
|
+
}
|
|
380
|
+
Ok(rows)
|
|
381
|
+
}
|
|
382
|
+
|
|
383
|
+
fn surface_record_batch(
|
|
384
|
+
surface: CommitSurface,
|
|
385
|
+
projection: Option<&Vec<usize>>,
|
|
386
|
+
rows: &[SurfaceRow],
|
|
387
|
+
) -> Result<RecordBatch> {
|
|
388
|
+
let columns = surface_columns(surface, projection);
|
|
389
|
+
let arrays = columns
|
|
390
|
+
.iter()
|
|
391
|
+
.map(|column| column.array(rows))
|
|
392
|
+
.collect::<Vec<_>>();
|
|
393
|
+
record_batch_with_row_count(surface_schema(&columns), arrays, rows.len()).map_err(|error| {
|
|
394
|
+
DataFusionError::Execution(format!(
|
|
395
|
+
"failed to build {} batch: {error}",
|
|
396
|
+
surface.table_name()
|
|
397
|
+
))
|
|
398
|
+
})
|
|
399
|
+
}
|
|
400
|
+
|
|
401
|
+
#[derive(Debug, Clone, Copy)]
|
|
402
|
+
enum SurfaceColumn {
|
|
403
|
+
Id,
|
|
404
|
+
ChangeSetId,
|
|
405
|
+
ParentId,
|
|
406
|
+
ChildId,
|
|
407
|
+
ChangeId,
|
|
408
|
+
EntityId,
|
|
409
|
+
SchemaKey,
|
|
410
|
+
FileId,
|
|
411
|
+
VersionId,
|
|
412
|
+
Global,
|
|
413
|
+
Untracked,
|
|
414
|
+
}
|
|
415
|
+
|
|
416
|
+
impl SurfaceColumn {
|
|
417
|
+
fn field(self) -> Field {
|
|
418
|
+
match self {
|
|
419
|
+
Self::Id => Field::new("id", DataType::Utf8, false),
|
|
420
|
+
Self::ChangeSetId => Field::new("change_set_id", DataType::Utf8, false),
|
|
421
|
+
Self::ParentId => Field::new("parent_id", DataType::Utf8, false),
|
|
422
|
+
Self::ChildId => Field::new("child_id", DataType::Utf8, false),
|
|
423
|
+
Self::ChangeId => Field::new("change_id", DataType::Utf8, false),
|
|
424
|
+
Self::EntityId => Field::new("entity_id", DataType::Utf8, false),
|
|
425
|
+
Self::SchemaKey => Field::new("schema_key", DataType::Utf8, false),
|
|
426
|
+
Self::FileId => Field::new("file_id", DataType::Utf8, true),
|
|
427
|
+
Self::VersionId => Field::new("lixcol_version_id", DataType::Utf8, false),
|
|
428
|
+
Self::Global => Field::new("lixcol_global", DataType::Boolean, false),
|
|
429
|
+
Self::Untracked => Field::new("lixcol_untracked", DataType::Boolean, false),
|
|
430
|
+
}
|
|
431
|
+
}
|
|
432
|
+
|
|
433
|
+
fn array(self, rows: &[SurfaceRow]) -> ArrayRef {
|
|
434
|
+
match self {
|
|
435
|
+
Self::Id => string_array(rows.iter().map(|row| match row {
|
|
436
|
+
SurfaceRow::ChangeSet { id, .. } => Some(id.as_str()),
|
|
437
|
+
_ => None,
|
|
438
|
+
})),
|
|
439
|
+
Self::ChangeSetId => string_array(rows.iter().map(|row| match row {
|
|
440
|
+
SurfaceRow::ChangeSetElement { change_set_id, .. } => Some(change_set_id.as_str()),
|
|
441
|
+
_ => None,
|
|
442
|
+
})),
|
|
443
|
+
Self::ParentId => string_array(rows.iter().map(|row| match row {
|
|
444
|
+
SurfaceRow::CommitEdge { parent_id, .. } => Some(parent_id.as_str()),
|
|
445
|
+
_ => None,
|
|
446
|
+
})),
|
|
447
|
+
Self::ChildId => string_array(rows.iter().map(|row| match row {
|
|
448
|
+
SurfaceRow::CommitEdge { child_id, .. } => Some(child_id.as_str()),
|
|
449
|
+
_ => None,
|
|
450
|
+
})),
|
|
451
|
+
Self::ChangeId => string_array(rows.iter().map(|row| match row {
|
|
452
|
+
SurfaceRow::ChangeSetElement { change_id, .. } => Some(change_id.as_str()),
|
|
453
|
+
_ => None,
|
|
454
|
+
})),
|
|
455
|
+
Self::EntityId => string_array(rows.iter().map(|row| match row {
|
|
456
|
+
SurfaceRow::ChangeSetElement { entity_id, .. } => Some(entity_id.as_str()),
|
|
457
|
+
_ => None,
|
|
458
|
+
})),
|
|
459
|
+
Self::SchemaKey => string_array(rows.iter().map(|row| match row {
|
|
460
|
+
SurfaceRow::ChangeSetElement { schema_key, .. } => Some(schema_key.as_str()),
|
|
461
|
+
_ => None,
|
|
462
|
+
})),
|
|
463
|
+
Self::FileId => string_array(rows.iter().map(|row| match row {
|
|
464
|
+
SurfaceRow::ChangeSetElement { file_id, .. } => file_id.as_deref(),
|
|
465
|
+
_ => None,
|
|
466
|
+
})),
|
|
467
|
+
Self::VersionId => string_array(rows.iter().map(|row| match row {
|
|
468
|
+
SurfaceRow::CommitEdge { version_id, .. }
|
|
469
|
+
| SurfaceRow::ChangeSet { version_id, .. }
|
|
470
|
+
| SurfaceRow::ChangeSetElement { version_id, .. } => version_id.as_deref(),
|
|
471
|
+
})),
|
|
472
|
+
Self::Global => Arc::new(BooleanArray::from(vec![true; rows.len()])) as ArrayRef,
|
|
473
|
+
Self::Untracked => Arc::new(BooleanArray::from(vec![false; rows.len()])) as ArrayRef,
|
|
474
|
+
}
|
|
475
|
+
}
|
|
476
|
+
}
|
|
477
|
+
|
|
478
|
+
fn surface_columns(surface: CommitSurface, projection: Option<&Vec<usize>>) -> Vec<SurfaceColumn> {
|
|
479
|
+
let all_columns = match surface {
|
|
480
|
+
CommitSurface::CommitEdge => vec![
|
|
481
|
+
SurfaceColumn::ParentId,
|
|
482
|
+
SurfaceColumn::ChildId,
|
|
483
|
+
SurfaceColumn::Global,
|
|
484
|
+
SurfaceColumn::Untracked,
|
|
485
|
+
],
|
|
486
|
+
CommitSurface::CommitEdgeByVersion => vec![
|
|
487
|
+
SurfaceColumn::ParentId,
|
|
488
|
+
SurfaceColumn::ChildId,
|
|
489
|
+
SurfaceColumn::VersionId,
|
|
490
|
+
SurfaceColumn::Global,
|
|
491
|
+
SurfaceColumn::Untracked,
|
|
492
|
+
],
|
|
493
|
+
CommitSurface::ChangeSet => vec![
|
|
494
|
+
SurfaceColumn::Id,
|
|
495
|
+
SurfaceColumn::Global,
|
|
496
|
+
SurfaceColumn::Untracked,
|
|
497
|
+
],
|
|
498
|
+
CommitSurface::ChangeSetByVersion => vec![
|
|
499
|
+
SurfaceColumn::Id,
|
|
500
|
+
SurfaceColumn::VersionId,
|
|
501
|
+
SurfaceColumn::Global,
|
|
502
|
+
SurfaceColumn::Untracked,
|
|
503
|
+
],
|
|
504
|
+
CommitSurface::ChangeSetElement => vec![
|
|
505
|
+
SurfaceColumn::ChangeSetId,
|
|
506
|
+
SurfaceColumn::ChangeId,
|
|
507
|
+
SurfaceColumn::EntityId,
|
|
508
|
+
SurfaceColumn::SchemaKey,
|
|
509
|
+
SurfaceColumn::FileId,
|
|
510
|
+
SurfaceColumn::Global,
|
|
511
|
+
SurfaceColumn::Untracked,
|
|
512
|
+
],
|
|
513
|
+
CommitSurface::ChangeSetElementByVersion => vec![
|
|
514
|
+
SurfaceColumn::ChangeSetId,
|
|
515
|
+
SurfaceColumn::ChangeId,
|
|
516
|
+
SurfaceColumn::EntityId,
|
|
517
|
+
SurfaceColumn::SchemaKey,
|
|
518
|
+
SurfaceColumn::FileId,
|
|
519
|
+
SurfaceColumn::VersionId,
|
|
520
|
+
SurfaceColumn::Global,
|
|
521
|
+
SurfaceColumn::Untracked,
|
|
522
|
+
],
|
|
523
|
+
};
|
|
524
|
+
projection.map_or(all_columns.clone(), |indices| {
|
|
525
|
+
indices
|
|
526
|
+
.iter()
|
|
527
|
+
.filter_map(|index| all_columns.get(*index).copied())
|
|
528
|
+
.collect()
|
|
529
|
+
})
|
|
530
|
+
}
|
|
531
|
+
|
|
532
|
+
fn surface_schema(columns: &[SurfaceColumn]) -> SchemaRef {
|
|
533
|
+
Arc::new(Schema::new(
|
|
534
|
+
columns
|
|
535
|
+
.iter()
|
|
536
|
+
.map(|column| column.field())
|
|
537
|
+
.collect::<Vec<_>>(),
|
|
538
|
+
))
|
|
539
|
+
}
|
|
540
|
+
|
|
541
|
+
fn commit_edge_schema(by_version: bool) -> SchemaRef {
|
|
542
|
+
surface_schema(&surface_columns(
|
|
543
|
+
if by_version {
|
|
544
|
+
CommitSurface::CommitEdgeByVersion
|
|
545
|
+
} else {
|
|
546
|
+
CommitSurface::CommitEdge
|
|
547
|
+
},
|
|
548
|
+
None,
|
|
549
|
+
))
|
|
550
|
+
}
|
|
551
|
+
|
|
552
|
+
fn change_set_schema(by_version: bool) -> SchemaRef {
|
|
553
|
+
surface_schema(&surface_columns(
|
|
554
|
+
if by_version {
|
|
555
|
+
CommitSurface::ChangeSetByVersion
|
|
556
|
+
} else {
|
|
557
|
+
CommitSurface::ChangeSet
|
|
558
|
+
},
|
|
559
|
+
None,
|
|
560
|
+
))
|
|
561
|
+
}
|
|
562
|
+
|
|
563
|
+
fn change_set_element_schema(by_version: bool) -> SchemaRef {
|
|
564
|
+
surface_schema(&surface_columns(
|
|
565
|
+
if by_version {
|
|
566
|
+
CommitSurface::ChangeSetElementByVersion
|
|
567
|
+
} else {
|
|
568
|
+
CommitSurface::ChangeSetElement
|
|
569
|
+
},
|
|
570
|
+
None,
|
|
571
|
+
))
|
|
572
|
+
}
|
|
573
|
+
|
|
574
|
+
fn projected_schema(schema: &SchemaRef, projection: Option<&Vec<usize>>) -> SchemaRef {
|
|
575
|
+
match projection {
|
|
576
|
+
Some(projection) => Arc::new(schema.project(projection).expect("projection is valid")),
|
|
577
|
+
None => Arc::clone(schema),
|
|
578
|
+
}
|
|
579
|
+
}
|
|
580
|
+
|
|
581
|
+
fn string_array<'a>(values: impl Iterator<Item = Option<&'a str>>) -> ArrayRef {
|
|
582
|
+
Arc::new(StringArray::from(values.collect::<Vec<_>>())) as ArrayRef
|
|
583
|
+
}
|
|
584
|
+
|
|
585
|
+
fn datafusion_error_to_lix_error(error: DataFusionError) -> LixError {
|
|
586
|
+
super::error::datafusion_error_to_lix_error(error)
|
|
587
|
+
}
|
|
588
|
+
|
|
589
|
+
fn lix_error_to_datafusion_error(error: LixError) -> DataFusionError {
|
|
590
|
+
super::error::lix_error_to_datafusion_error(error)
|
|
591
|
+
}
|