Files
gridline/src-tauri/src/models/db_viewer.rs
T
adrianbonpin 1f081a66e0 feat: Schema Visualizer, docs update & competitor comparison
* chore: add @xyflow/react, dagre, @types/dagre (Task 1)

* feat: add SchemaGraph, TableNode, GraphColumn, Relationship types (Task 2)

* feat: add SchemaGraph, TableNode, GraphColumn, Relationship Rust models (Task 3)

* feat: add cardinality color/label helpers and legend data (Task 4)

* feat: add schema_graph command skeleton with validation (Task 5)

* feat: add PostgreSQL schema graph query builder + cardinality inference (Task 6)

* feat: implement get_schema_graph for PostgreSQL + SQLite (Task 7)

* feat: add getSchemaGraph IPC wrapper (Task 8)

* feat: un-stub Schema Visualizer nav + add view branch placeholder (Task 9)

* feat: add SchemaVisualizerNode custom React Flow table card (Task 10)

* feat: add SchemaVisualizerPage with React Flow canvas (Task 11)

* feat: wire SchemaVisualizerPage + error handling + style polish (Tasks 12-14)

* perf: replace information_schema with pg_catalog for schema graph query (500x+ faster on remote PG)

* fix: add DB selector, ghost-style dropdowns, dark controls, minimap styling, attribution

* fix: always show schema selector, move attribution to top-left

* fix: SelectDropdown close on outside click works in React Flow (capture phase)

* style: remove border from attribution badge

* style: restore bg on attribution, no border

* fix: lower attribution z-index so dropdowns render above

* fix: ensure toolbar + dropdowns stack above canvas attribution

* feat: crow's foot markers on edges + collapsible legend

* fix: restore missing tableCount state (was overwritten by legendOpen)

* fix: move crow's foot marker defs inside ReactFlow SVG, remove duplicate external SVGs

* fix: inject crow's foot SVG markers into ReactFlow SVG via DOM ref

* fix: use hidden SVG before ReactFlow for crow's foot markers, remove DOM injection

* feat: custom CrowsFootEdge component with inline crow's foot markers

* feat: custom CrowsFootEdge with zero-or-one/zero-or-many notation + nullability-based cardinality

* fix: strip markers, use clean text labels only on edges

* fix: add SVG marker defs directly inside ReactFlow + url() references for crow's foot

* fix: use custom CrowsFootEdge with BaseEdge + inline SVG symbols (no marker defs needed)

* debug: add red/green circles at edge endpoints to verify custom edge renders

* fix: remove stale duplicate edge data, use clean data.startMarker/endMarker

* fix: use getSmoothStepPath offset points for correct tangent angle at endpoints

* fix: thicker strokeWidth, position at handle coords, use path tangents

* fix: use straight-line angle (not curve tangent) for marker rotation

* fix: compute marker positions directly with raw math, no SVG transforms

* fix: fixed-orientation symbols — | always vertical, crow's foot fans toward node

* fix: dead simple — | vertical line, ← or → horizontal crow's foot based on edge direction

* fix: increase marker gap to 12px so symbols aren't hidden behind handle dots

* fix: correct offset direction (away from card into gap), G=4

* fix: remove duplicate G offset inside Mark (was canceling out the call-site offset)

* fix: crow's foot back to fork shape — three lines converging to a tip

* fix: flip crow's foot direction

* fix: wider crow's foot spread (4→6)

* feat: add crow's foot symbols to relationship legend

* style: cleaner legend — horizontal edge with endpoint symbols + label

* feat: handles on both sides, edge builder picks closest side based on dagre layout

* fix: compute actual handle distances to pick shortest path

* revert: PK always left, FK always right — one handle per column only

* fix: lock edge marker direction via origRight, TB layout for horizontal spread, truncate long types

* fix: semi-transparent minimap mask, border stroke for viewport visibility

* feat: click edge to highlight (amber glow), all others dim to 15% opacity

* feat: legend highlights matching cardinality row when edge is clicked

* fix: crow's foot symbols now read color from edge style (amber when highlighted)

* fix: highlighted edge gets zIndex 1000 to render on top

* fix: folder empty message now checks unfiltered store, shows filter hint when connections exist but filtered out

* feat: flat SVG DB icons from simple-icons (PostgreSQL, MySQL, SQLite, Redis) replacing emoji

* chore: add *.db, *.sqlite, *.sqlite3 to .gitignore

* docs: update README with 3-way competitor comparison + current roadmap; update AGENTS.md schema visualizer status
2026-07-29 22:36:59 +08:00

480 lines
14 KiB
Rust

use serde::{Deserialize, Serialize};
/// A single filter rule sent from the frontend.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct FilterRule {
pub id: String,
pub column: String,
pub operator: String, // "eq" | "neq" | "contains" | "starts" | "ends" | "gt" | "lt" | "null" | "notnull"
pub value: String,
}
/// A single sort rule sent from the frontend.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SortRule {
pub id: String,
pub column: String,
pub order: String, // "asc" | "desc"
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct TableInfo {
pub name: String,
pub schema: String,
pub table_type: String,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ColumnInfo {
pub name: String,
pub data_type: String,
pub is_nullable: bool,
pub is_pk: bool,
pub is_fk: bool,
pub fk_ref: Option<(String, String)>,
pub default_value: Option<String>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct QueryResult {
pub columns: Vec<ColumnInfo>,
pub rows: Vec<Vec<serde_json::Value>>,
pub total_rows: i64,
pub page: i64,
pub page_size: i64,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Pagination {
pub page: i64,
pub page_size: i64,
pub total_rows: i64,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct FunctionInfo {
pub name: String,
pub schema: String,
pub return_type: String,
pub argument_types: Vec<String>,
pub argument_names: Vec<String>,
pub argument_modes: Vec<String>,
pub language: String,
pub source: Option<String>,
pub kind: String, // 'f' = function, 'p' = procedure
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct TriggerInfo {
pub name: String,
pub schema: String,
pub table_schema: String,
pub table_name: String,
pub event_manipulation: String,
pub action_timing: String,
pub action_orientation: String,
pub action_statement: String,
pub enabled: String,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SequenceInfo {
pub name: String,
pub schema: String,
pub start_value: String,
pub min_value: String,
pub max_value: String,
pub increment: String,
pub current_value: String,
pub cycle: bool,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct EnumInfo {
pub name: String,
pub schema: String,
pub labels: Vec<String>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct ExtensionInfo {
pub name: String,
pub schema: String,
pub version: String,
pub comment: Option<String>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "type", rename_all = "snake_case")]
pub enum Change {
Update {
id: String,
schema: String,
table: String,
primary_key: String,
old_data: String,
new_data: String,
},
Insert {
id: String,
schema: String,
table: String,
data: String,
},
Delete {
id: String,
schema: String,
table: String,
primary_key: String,
},
AlterTable {
id: String,
schema: String,
table: String,
sql: String,
rollback_sql: String,
},
}
impl Change {
pub fn id(&self) -> &str {
match self {
Change::Update { id, .. }
| Change::Insert { id, .. }
| Change::Delete { id, .. }
| Change::AlterTable { id, .. } => id,
}
}
}
/// Complete schema graph for the ER diagram visualizer.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SchemaGraph {
pub tables: Vec<TableNode>,
pub relationships: Vec<Relationship>,
}
/// A table node in the schema graph, including all columns.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct TableNode {
pub name: String,
pub schema: String,
pub table_type: String,
pub columns: Vec<GraphColumn>,
}
/// Column metadata for schema graph visualization.
///
/// Includes PK/FK/UNIQUE flags and an optional foreign-key reference
/// (referenced_schema, referenced_table, referenced_column).
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct GraphColumn {
pub name: String,
pub data_type: String,
pub is_pk: bool,
pub is_fk: bool,
pub is_unique: bool,
pub is_nullable: bool,
pub fk_ref: Option<(String, String, String)>,
}
/// A foreign-key relationship between two tables.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Relationship {
pub source_schema: String,
pub source_table: String,
pub source_column: String,
pub target_schema: String,
pub target_table: String,
pub target_column: String,
/// Inferred cardinality: "1:1", "1:N", or "N:M"
pub cardinality: String,
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn table_info_serialization() {
let info = TableInfo {
name: "users".to_string(),
schema: "public".to_string(),
table_type: "TABLE".to_string(),
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("users"));
assert!(json.contains("public"));
assert!(json.contains("TABLE"));
}
#[test]
fn query_result_can_be_empty() {
let result = QueryResult {
columns: vec![],
rows: vec![],
total_rows: 0,
page: 1,
page_size: 100,
};
let json = serde_json::to_string(&result).unwrap();
assert!(json.contains(r#""rows":[]"#));
}
#[test]
fn change_id_method() {
let update = Change::Update {
id: "chg-1".to_string(),
schema: "public".to_string(),
table: "users".to_string(),
primary_key: "{\"id\": 1}".to_string(),
old_data: "{\"name\": \"old\"}".to_string(),
new_data: "{\"name\": \"new\"}".to_string(),
};
assert_eq!(update.id(), "chg-1");
let inserted = Change::Insert {
id: "chg-2".to_string(),
schema: "public".to_string(),
table: "users".to_string(),
data: "{\"name\": \"alice\"}".to_string(),
};
assert_eq!(inserted.id(), "chg-2");
let deleted = Change::Delete {
id: "chg-3".to_string(),
schema: "public".to_string(),
table: "users".to_string(),
primary_key: "{\"id\": 2}".to_string(),
};
assert_eq!(deleted.id(), "chg-3");
let alter = Change::AlterTable {
id: "chg-4".to_string(),
schema: "public".to_string(),
table: "users".to_string(),
sql: "ALTER TABLE users ADD COLUMN age INT".to_string(),
rollback_sql: "ALTER TABLE users DROP COLUMN age".to_string(),
};
assert_eq!(alter.id(), "chg-4");
}
#[test]
fn change_serde_tag() {
let update = Change::Update {
id: "chg-1".to_string(),
schema: "public".to_string(),
table: "users".to_string(),
primary_key: "{\"id\": 1}".to_string(),
old_data: "{\"name\": \"old\"}".to_string(),
new_data: "{\"name\": \"new\"}".to_string(),
};
let json = serde_json::to_string(&update).unwrap();
assert!(
json.contains(r#""type":"update""#),
"serialized Change::Update should use snake_case tag 'update'; got: {}",
json
);
}
#[test]
fn column_info_fk_ref() {
let col = ColumnInfo {
name: "user_id".to_string(),
data_type: "integer".to_string(),
is_nullable: true,
is_pk: false,
is_fk: true,
fk_ref: Some(("users".to_string(), "id".to_string())),
default_value: None,
};
let json = serde_json::to_string(&col).unwrap();
assert!(json.contains("user_id"));
assert!(json.contains("users"));
}
#[test]
fn pagination_serialization() {
let pagination = Pagination {
page: 2,
page_size: 50,
total_rows: 250,
};
let json = serde_json::to_string(&pagination).unwrap();
assert!(json.contains(r#""page":2"#));
assert!(json.contains(r#""page_size":50"#));
assert!(json.contains(r#""total_rows":250"#));
}
#[test]
fn function_info_serialization() {
let info = FunctionInfo {
name: "get_user".into(),
schema: "public".into(),
return_type: "TABLE(id integer, name text)".into(),
argument_types: vec!["integer".into()],
argument_names: vec!["p_id".into()],
argument_modes: vec!["IN".into()],
language: "plpgsql".into(),
source: Some("BEGIN RETURN; END;".into()),
kind: "f".into(),
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("get_user"));
assert!(json.contains("plpgsql"));
}
#[test]
fn trigger_info_serialization() {
let info = TriggerInfo {
name: "trg_audit".into(),
schema: "public".into(),
table_schema: "public".into(),
table_name: "users".into(),
event_manipulation: "INSERT".into(),
action_timing: "AFTER".into(),
action_orientation: "ROW".into(),
action_statement: "EXECUTE FUNCTION audit_log()".into(),
enabled: "O".into(),
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("trg_audit"));
}
#[test]
fn sequence_info_serialization() {
let info = SequenceInfo {
name: "users_id_seq".into(),
schema: "public".into(),
start_value: "1".into(),
min_value: "1".into(),
max_value: "9223372036854775807".into(),
increment: "1".into(),
current_value: "42".into(),
cycle: false,
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("users_id_seq"));
}
#[test]
fn enum_info_serialization() {
let info = EnumInfo {
name: "user_role".into(),
schema: "public".into(),
labels: vec!["admin".into(), "editor".into(), "viewer".into()],
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("admin"));
}
#[test]
fn extension_info_serialization() {
let info = ExtensionInfo {
name: "pg_stat_statements".into(),
schema: "public".into(),
version: "1.10".into(),
comment: Some("track SQL statistics".into()),
};
let json = serde_json::to_string(&info).unwrap();
assert!(json.contains("pg_stat_statements"));
}
#[test]
fn schema_graph_serialization() {
let graph = SchemaGraph {
tables: vec![TableNode {
name: "users".into(),
schema: "public".into(),
table_type: "TABLE".into(),
columns: vec![
GraphColumn {
name: "id".into(),
data_type: "integer".into(),
is_pk: true,
is_fk: false,
is_unique: true,
is_nullable: false,
fk_ref: None,
},
GraphColumn {
name: "email".into(),
data_type: "text".into(),
is_pk: false,
is_fk: false,
is_unique: true,
is_nullable: false,
fk_ref: None,
},
],
}],
relationships: vec![Relationship {
source_schema: "public".into(),
source_table: "orders".into(),
source_column: "user_id".into(),
target_schema: "public".into(),
target_table: "users".into(),
target_column: "id".into(),
cardinality: "1:N".into(),
}],
};
let json = serde_json::to_string(&graph).unwrap();
assert!(json.contains("users"), "should contain table name");
assert!(json.contains("orders"), "should contain relationship source table");
assert!(json.contains("1:N"), "should contain cardinality");
assert!(json.contains("is_pk"), "should contain is_pk field");
assert!(json.contains("is_fk"), "should contain is_fk field");
assert!(json.contains("is_unique"), "should contain is_unique field");
// Round-trip deserialization
let parsed: SchemaGraph = serde_json::from_str(&json).unwrap();
assert_eq!(parsed.tables.len(), 1);
assert_eq!(parsed.tables[0].columns.len(), 2);
assert_eq!(parsed.relationships.len(), 1);
assert_eq!(parsed.relationships[0].cardinality, "1:N");
}
#[test]
fn schema_graph_empty_is_valid() {
let graph = SchemaGraph {
tables: vec![],
relationships: vec![],
};
let json = serde_json::to_string(&graph).unwrap();
let parsed: SchemaGraph = serde_json::from_str(&json).unwrap();
assert!(parsed.tables.is_empty());
assert!(parsed.relationships.is_empty());
}
#[test]
fn graph_column_fk_ref_serialization() {
// fk_ref = None
let col_none = GraphColumn {
name: "name".into(),
data_type: "text".into(),
is_pk: false,
is_fk: false,
is_unique: false,
is_nullable: false,
fk_ref: None,
};
let json = serde_json::to_string(&col_none).unwrap();
assert!(json.contains("null"), "fk_ref=None should serialize as null");
// fk_ref = Some(...)
let col_some = GraphColumn {
name: "user_id".into(),
data_type: "integer".into(),
is_pk: false,
is_fk: true,
is_unique: false,
is_nullable: false,
fk_ref: Some(("public".into(), "users".into(), "id".into())),
};
let json = serde_json::to_string(&col_some).unwrap();
assert!(json.contains("public"), "should contain referenced schema");
assert!(json.contains("users"), "should contain referenced table");
assert!(json.contains("id"), "should contain referenced column");
}
}