geth/crates/geth-db/src/lib.rs

442 lines
14 KiB
Rust
Raw Normal View History

2026-05-15 15:08:20 +02:00
use geth_types::{DbId, ResourceId};
2026-05-17 20:22:50 +02:00
use rusqlite::types::Value;
use rusqlite::{Connection, OpenFlags};
2026-05-15 15:08:20 +02:00
use serde::{Deserialize, Serialize};
use std::path::Path;
2026-05-15 15:08:20 +02:00
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
pub struct DbResource {
pub id: DbId,
pub resource: ResourceId,
2026-05-16 21:13:33 +02:00
pub name: String,
2026-05-15 15:08:20 +02:00
pub path: String,
2026-05-16 21:13:33 +02:00
pub path_exists: bool,
pub size_bytes: Option<u64>,
pub schema_metadata: String,
2026-05-17 20:01:36 +02:00
pub crsqlite_changes: CrSqliteChangeMetadata,
2026-05-15 15:08:20 +02:00
pub sync_status: String,
}
2026-05-17 20:01:36 +02:00
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
pub struct CrSqliteChangeMetadata {
pub available: bool,
pub change_count: Option<u64>,
pub max_db_version: Option<i64>,
pub columns: Vec<String>,
pub error: Option<String>,
}
2026-05-17 20:22:50 +02:00
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(tag = "type", content = "value", rename_all = "kebab-case")]
pub enum SqliteValue {
Null,
Integer(i64),
Real(f64),
Text(String),
Blob(Vec<u8>),
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct CrSqliteChange {
pub table_name: String,
pub pk: SqliteValue,
pub column_id: String,
pub value: SqliteValue,
pub column_version: i64,
pub db_version: i64,
pub site_id: Option<Vec<u8>>,
pub causal_length: Option<i64>,
pub sequence: Option<i64>,
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct CrSqliteChangeBatch {
pub schema_metadata: String,
pub max_db_version: Option<i64>,
pub changes: Vec<CrSqliteChange>,
}
2026-05-17 20:01:36 +02:00
impl CrSqliteChangeMetadata {
#[must_use]
pub fn unavailable() -> Self {
Self {
available: false,
change_count: None,
max_db_version: None,
columns: Vec::new(),
error: None,
}
}
#[must_use]
pub fn error(error: impl Into<String>) -> Self {
Self {
available: false,
change_count: None,
max_db_version: None,
columns: Vec::new(),
error: Some(error.into()),
}
}
}
2026-05-16 21:13:33 +02:00
#[derive(Debug, thiserror::Error)]
pub enum DbError {
#[error("invalid db resource name: {0}")]
InvalidName(String),
#[error("sqlite error: {0}")]
Sqlite(#[from] rusqlite::Error),
2026-05-17 20:22:50 +02:00
#[error("crsql_changes table or view is missing")]
MissingCrSqliteChanges,
#[error("crsql_changes is missing required column: {0}")]
MissingCrSqliteColumn(&'static str),
2026-05-16 21:13:33 +02:00
}
pub fn validate_db_name(name: &str) -> Result<(), DbError> {
if name.is_empty()
|| !name
.bytes()
.all(|byte| byte.is_ascii_alphanumeric() || matches!(byte, b'.' | b'_' | b'-'))
{
return Err(DbError::InvalidName(name.to_owned()));
}
Ok(())
}
pub fn schema_metadata(path: &Path) -> Result<String, DbError> {
let conn = Connection::open_with_flags(path, OpenFlags::SQLITE_OPEN_READ_ONLY)?;
let mut stmt = conn.prepare(
r#"SELECT type, name, COALESCE(sql, '')
FROM sqlite_master
WHERE name NOT LIKE 'sqlite_%'
ORDER BY type, name"#,
)?;
let rows = stmt.query_map([], |row| {
Ok((
row.get::<_, String>(0)?,
row.get::<_, String>(1)?,
row.get::<_, String>(2)?,
))
})?;
let mut canonical = String::new();
let mut table_count = 0_usize;
let mut index_count = 0_usize;
let mut view_count = 0_usize;
let mut trigger_count = 0_usize;
for row in rows {
let (kind, name, sql) = row?;
match kind.as_str() {
"table" => table_count += 1,
"index" => index_count += 1,
"view" => view_count += 1,
"trigger" => trigger_count += 1,
_ => {}
}
canonical.push_str(&kind);
canonical.push('\0');
canonical.push_str(&name);
canonical.push('\0');
canonical.push_str(&sql);
canonical.push('\n');
}
Ok(format!(
"tables={table_count} indexes={index_count} views={view_count} triggers={trigger_count} schema_hash={}",
blake3::hash(canonical.as_bytes())
))
}
2026-05-17 20:01:36 +02:00
pub fn crsqlite_change_metadata(path: &Path) -> Result<CrSqliteChangeMetadata, DbError> {
let conn = Connection::open_with_flags(path, OpenFlags::SQLITE_OPEN_READ_ONLY)?;
let available: bool = conn.query_row(
r#"SELECT EXISTS(
SELECT 1 FROM sqlite_master
WHERE name = 'crsql_changes' AND type IN ('table', 'view')
)"#,
[],
|row| row.get(0),
)?;
if !available {
return Ok(CrSqliteChangeMetadata::unavailable());
}
let mut columns_stmt = conn.prepare("PRAGMA table_info('crsql_changes')")?;
let columns = columns_stmt
.query_map([], |row| row.get::<_, String>(1))?
.collect::<Result<Vec<_>, _>>()?;
let change_count: u64 =
conn.query_row("SELECT COUNT(*) FROM crsql_changes", [], |row| row.get(0))?;
let max_db_version = if columns.iter().any(|column| column == "db_version") {
conn.query_row("SELECT MAX(db_version) FROM crsql_changes", [], |row| {
row.get(0)
})?
} else {
None
};
Ok(CrSqliteChangeMetadata {
available: true,
change_count: Some(change_count),
max_db_version,
columns,
error: None,
})
}
2026-05-17 20:22:50 +02:00
pub fn extract_crsqlite_changes(
path: &Path,
after_db_version: Option<i64>,
limit: u32,
) -> Result<CrSqliteChangeBatch, DbError> {
let schema_metadata = schema_metadata(path)?;
let conn = Connection::open_with_flags(path, OpenFlags::SQLITE_OPEN_READ_ONLY)?;
let columns = crsqlite_change_columns(&conn)?;
require_crsqlite_column(&columns, "table_name")?;
require_crsqlite_column(&columns, "pk")?;
require_crsqlite_column(&columns, "cid")?;
require_crsqlite_column(&columns, "val")?;
require_crsqlite_column(&columns, "col_version")?;
require_crsqlite_column(&columns, "db_version")?;
let site_id_expr = optional_crsqlite_column_expr(&columns, "site_id");
let causal_length_expr = optional_crsqlite_column_expr(&columns, "cl");
let sequence_expr = optional_crsqlite_column_expr(&columns, "seq");
let bounded_limit = limit.max(1);
let sql = format!(
r#"SELECT table_name, pk, cid, val, col_version, db_version, {site_id_expr}, {causal_length_expr}, {sequence_expr}
FROM crsql_changes
WHERE (?1 IS NULL OR db_version > ?1)
ORDER BY db_version, table_name, pk, cid
LIMIT ?2"#
);
let mut stmt = conn.prepare(&sql)?;
let rows = stmt.query_map((after_db_version, i64::from(bounded_limit)), |row| {
Ok(CrSqliteChange {
table_name: row.get(0)?,
pk: sqlite_value(row.get(1)?),
column_id: row.get(2)?,
value: sqlite_value(row.get(3)?),
column_version: row.get(4)?,
db_version: row.get(5)?,
site_id: row.get(6)?,
causal_length: row.get(7)?,
sequence: row.get(8)?,
})
})?;
let changes = rows.collect::<Result<Vec<_>, _>>()?;
let max_db_version = changes.iter().map(|change| change.db_version).max();
Ok(CrSqliteChangeBatch {
schema_metadata,
max_db_version,
changes,
})
}
fn crsqlite_change_columns(conn: &Connection) -> Result<Vec<String>, DbError> {
let available: bool = conn.query_row(
r#"SELECT EXISTS(
SELECT 1 FROM sqlite_master
WHERE name = 'crsql_changes' AND type IN ('table', 'view')
)"#,
[],
|row| row.get(0),
)?;
if !available {
return Err(DbError::MissingCrSqliteChanges);
}
let mut columns_stmt = conn.prepare("PRAGMA table_info('crsql_changes')")?;
columns_stmt
.query_map([], |row| row.get::<_, String>(1))?
.collect::<Result<Vec<_>, _>>()
.map_err(DbError::from)
}
fn require_crsqlite_column(columns: &[String], column: &'static str) -> Result<(), DbError> {
if columns.iter().any(|candidate| candidate == column) {
Ok(())
} else {
Err(DbError::MissingCrSqliteColumn(column))
}
}
fn optional_crsqlite_column_expr(columns: &[String], column: &'static str) -> &'static str {
if columns.iter().any(|candidate| candidate == column) {
column
} else {
"NULL"
}
}
fn sqlite_value(value: Value) -> SqliteValue {
match value {
Value::Null => SqliteValue::Null,
Value::Integer(value) => SqliteValue::Integer(value),
Value::Real(value) => SqliteValue::Real(value),
Value::Text(value) => SqliteValue::Text(value),
Value::Blob(value) => SqliteValue::Blob(value),
}
}
2026-05-15 15:08:20 +02:00
#[must_use]
pub fn crsqlite_sync_roadmap() -> &'static str {
"future db sync reads crsql_changes, exchanges changes over Iroh, and applies through crsql_changes"
}
2026-05-16 21:13:33 +02:00
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn db_name_validation_rejects_paths_and_empty_names() {
assert!(validate_db_name("notes").is_ok());
assert!(validate_db_name("notes.v1").is_ok());
assert!(validate_db_name("").is_err());
assert!(validate_db_name("../notes").is_err());
assert!(validate_db_name("notes/main").is_err());
assert!(validate_db_name("notes main").is_err());
}
#[test]
fn schema_metadata_reports_counts_and_stable_hash() {
let dir = tempfile::tempdir().expect("tempdir");
let path = dir.path().join("notes.sqlite");
let conn = Connection::open(&path).expect("open sqlite");
conn.execute(
"CREATE TABLE notes(id INTEGER PRIMARY KEY, body TEXT NOT NULL)",
[],
)
.expect("create table");
conn.execute("CREATE INDEX notes_body ON notes(body)", [])
.expect("create index");
drop(conn);
let first = schema_metadata(&path).expect("schema metadata");
let second = schema_metadata(&path).expect("schema metadata again");
assert_eq!(first, second);
assert!(first.contains("tables=1"));
assert!(first.contains("indexes=1"));
assert!(first.contains("schema_hash="));
}
2026-05-17 20:01:36 +02:00
#[test]
fn crsqlite_change_metadata_reports_mock_change_table() {
let dir = tempfile::tempdir().expect("tempdir");
let path = dir.path().join("notes.sqlite");
let conn = Connection::open(&path).expect("open sqlite");
conn.execute(
r#"CREATE TABLE crsql_changes(
table_name TEXT NOT NULL,
pk TEXT NOT NULL,
cid TEXT NOT NULL,
val BLOB,
col_version INTEGER NOT NULL,
db_version INTEGER NOT NULL,
site_id BLOB,
cl INTEGER,
seq INTEGER
)"#,
[],
)
.expect("create crsql_changes table");
conn.execute(
"INSERT INTO crsql_changes(table_name, pk, cid, val, col_version, db_version) VALUES ('notes', '1', 'body', 'hello', 1, 7)",
[],
)
.expect("insert change");
drop(conn);
let metadata = crsqlite_change_metadata(&path).expect("change metadata");
assert!(metadata.available);
assert_eq!(metadata.change_count, Some(1));
assert_eq!(metadata.max_db_version, Some(7));
assert!(metadata.columns.contains(&"db_version".to_owned()));
}
2026-05-17 20:22:50 +02:00
#[test]
fn extract_crsqlite_changes_returns_typed_batch_with_schema_metadata() {
let dir = tempfile::tempdir().expect("tempdir");
let path = dir.path().join("notes.sqlite");
let conn = Connection::open(&path).expect("open sqlite");
conn.execute("CREATE TABLE notes(id INTEGER PRIMARY KEY, body TEXT)", [])
.expect("create notes");
conn.execute(
r#"CREATE TABLE crsql_changes(
table_name TEXT NOT NULL,
pk BLOB NOT NULL,
cid TEXT NOT NULL,
val BLOB,
col_version INTEGER NOT NULL,
db_version INTEGER NOT NULL,
site_id BLOB,
cl INTEGER,
seq INTEGER
)"#,
[],
)
.expect("create crsql_changes table");
conn.execute(
"INSERT INTO crsql_changes(table_name, pk, cid, val, col_version, db_version, site_id, cl, seq) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
(
"notes",
vec![1_u8],
"body",
Vec::from("hello".as_bytes()),
1_i64,
1_i64,
vec![9_u8],
10_i64,
11_i64,
),
)
.expect("insert first change");
conn.execute(
"INSERT INTO crsql_changes(table_name, pk, cid, val, col_version, db_version, site_id, cl, seq) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
(
"notes",
vec![2_u8],
"body",
Vec::from("world".as_bytes()),
1_i64,
2_i64,
vec![9_u8],
12_i64,
13_i64,
),
)
.expect("insert second change");
drop(conn);
let batch = extract_crsqlite_changes(&path, Some(1), 10).expect("extract batch");
assert!(batch.schema_metadata.contains("schema_hash="));
assert_eq!(batch.max_db_version, Some(2));
assert_eq!(batch.changes.len(), 1);
assert_eq!(batch.changes[0].table_name, "notes");
assert_eq!(batch.changes[0].pk, SqliteValue::Blob(vec![2]));
assert_eq!(batch.changes[0].column_id, "body");
assert_eq!(
batch.changes[0].value,
SqliteValue::Blob(Vec::from("world".as_bytes()))
);
assert_eq!(batch.changes[0].site_id, Some(vec![9]));
assert_eq!(batch.changes[0].causal_length, Some(12));
assert_eq!(batch.changes[0].sequence, Some(13));
}
#[test]
fn extract_crsqlite_changes_requires_change_table() {
let dir = tempfile::tempdir().expect("tempdir");
let path = dir.path().join("notes.sqlite");
Connection::open(&path).expect("open sqlite");
assert!(matches!(
extract_crsqlite_changes(&path, None, 10),
Err(DbError::MissingCrSqliteChanges)
));
}
2026-05-16 21:13:33 +02:00
}