Files
addrkit/internal/adapters/registrystore/sqlite.go
T
addrkit 3be331bb65 addrkit: deterministic address resolution service (Go)
- domain: address normalization/parsing, matching with explicit statuses
- adapters: SQLite registry store, ArcGIS outages client, optional LLM rewriter, HTTP API
- seed pipeline with duplicate handling and unparsed reporting
- CLI (seed/verify/resolve), systemd unit, OpenAPI contract, tests
2026-10-09 20:05:32 +07:00

170 lines
5.2 KiB
Go

// Package registrystore persists the canonical registry as a single,
// self-contained SQLite file. It is intentionally boring: one file, no server,
// built once by the seed pipeline and opened read-only by the service.
package registrystore
import (
"database/sql"
"fmt"
"os"
_ "modernc.org/sqlite" // registers the "sqlite" driver
"git.totmin.ru/en2zmax/addrkit/internal/domain/address"
)
const schema = `
CREATE TABLE IF NOT EXISTS meta (
key TEXT PRIMARY KEY,
value TEXT NOT NULL
);
CREATE TABLE IF NOT EXISTS address (
id TEXT PRIMARY KEY,
address TEXT NOT NULL,
district TEXT NOT NULL DEFAULT '',
ordinal TEXT NOT NULL DEFAULT '',
name_key TEXT NOT NULL DEFAULT '',
name_display TEXT NOT NULL DEFAULT '',
street_type TEXT NOT NULL DEFAULT '',
street_key TEXT NOT NULL,
house TEXT NOT NULL DEFAULT '',
house_frac TEXT NOT NULL DEFAULT '',
house_letter TEXT NOT NULL DEFAULT '',
building TEXT NOT NULL DEFAULT '',
building_frac TEXT NOT NULL DEFAULT '',
building_letter TEXT NOT NULL DEFAULT '',
structure TEXT NOT NULL DEFAULT ''
);
CREATE INDEX IF NOT EXISTS idx_address_street ON address(street_key);
`
// Save writes entries to path atomically: it builds a temporary file and
// renames it into place, so readers never observe a half-written index.
//
// The build uses the rollback journal (not WAL) on purpose, so that the result
// is a single file with no companion -wal/-shm that a rename would orphan.
func Save(path, version string, entries []address.Entry) error {
tmp := path + ".tmp"
removeDBFiles(tmp)
db, err := sql.Open("sqlite", "file:"+tmp+
"?_pragma=journal_mode(DELETE)"+
"&_pragma=synchronous(FULL)"+
"&_pragma=busy_timeout(5000)")
if err != nil {
return fmt.Errorf("open temp db: %w", err)
}
if _, err := db.Exec(schema); err != nil {
db.Close()
return fmt.Errorf("create schema: %w", err)
}
if err := insertAll(db, version, entries); err != nil {
db.Close()
return err
}
if err := db.Close(); err != nil {
return fmt.Errorf("close temp db: %w", err)
}
if err := os.Rename(tmp, path); err != nil {
return fmt.Errorf("publish db: %w", err)
}
// Drop only the temp file's journal siblings; the published base file must
// survive.
removeSidecars(tmp)
return nil
}
func insertAll(db *sql.DB, version string, entries []address.Entry) error {
tx, err := db.Begin()
if err != nil {
return fmt.Errorf("begin: %w", err)
}
defer tx.Rollback()
if _, err := tx.Exec(`INSERT INTO meta(key, value) VALUES ('version', ?)`, version); err != nil {
return fmt.Errorf("insert meta: %w", err)
}
stmt, err := tx.Prepare(`INSERT INTO address (
id, address, district, ordinal, name_key, name_display, street_type, street_key,
house, house_frac, house_letter, building, building_frac, building_letter, structure
) VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)`)
if err != nil {
return fmt.Errorf("prepare insert: %w", err)
}
defer stmt.Close()
for i, e := range entries {
id := e.ID
if id == "" {
// Guarantee a primary key even for sources without ids.
id = fmt.Sprintf("row-%d", i)
}
if _, err := stmt.Exec(
id, e.Address, e.District, e.Ordinal, e.NameKey, e.NameDisplay, e.StreetType, e.StreetKey,
e.House, e.HouseFrac, e.HouseLetter, e.Building, e.BuildingFrac, e.BuildingLetter, e.Structure,
); err != nil {
return fmt.Errorf("insert %q: %w", e.Address, err)
}
}
if err := tx.Commit(); err != nil {
return fmt.Errorf("commit: %w", err)
}
return nil
}
// Load opens path read-only and returns the data version and all entries.
func Load(path string) (version string, entries []address.Entry, err error) {
if _, err := os.Stat(path); err != nil {
return "", nil, fmt.Errorf("stat db: %w", err)
}
db, err := sql.Open("sqlite", "file:"+path+
"?mode=ro"+
"&_pragma=query_only(1)"+
"&_pragma=busy_timeout(5000)")
if err != nil {
return "", nil, fmt.Errorf("open db: %w", err)
}
defer db.Close()
if err := db.QueryRow(`SELECT value FROM meta WHERE key='version'`).Scan(&version); err != nil {
return "", nil, fmt.Errorf("read version: %w", err)
}
rows, err := db.Query(`SELECT id, address, district, ordinal, name_key, name_display,
street_type, street_key, house, house_frac, house_letter, building, building_frac,
building_letter, structure FROM address`)
if err != nil {
return "", nil, fmt.Errorf("query: %w", err)
}
defer rows.Close()
for rows.Next() {
var e address.Entry
if err := rows.Scan(
&e.ID, &e.Address, &e.District, &e.Ordinal, &e.NameKey, &e.NameDisplay,
&e.StreetType, &e.StreetKey, &e.House, &e.HouseFrac, &e.HouseLetter,
&e.Building, &e.BuildingFrac, &e.BuildingLetter, &e.Structure,
); err != nil {
return "", nil, fmt.Errorf("scan: %w", err)
}
entries = append(entries, e)
}
if err := rows.Err(); err != nil {
return "", nil, fmt.Errorf("rows: %w", err)
}
return version, entries, nil
}
// removeDBFiles deletes a database file and its journal siblings.
func removeDBFiles(base string) {
_ = os.Remove(base)
removeSidecars(base)
}
// removeSidecars deletes only the -wal/-shm companions of a database file.
func removeSidecars(base string) {
_ = os.Remove(base + "-wal")
_ = os.Remove(base + "-shm")
}