| // Copyright 2011 The Go Authors. All rights reserved. | |
| // Use of this source code is governed by a BSD-style | |
| // license that can be found in the LICENSE file. | |
| package sql | |
| import ( | |
| "errors" | |
| "fmt" | |
| "io" | |
| "log" | |
| "strconv" | |
| "strings" | |
| "sync" | |
| "exp/sql/driver" | |
| ) | |
| var _ = log.Printf | |
| // fakeDriver is a fake database that implements Go's driver.Driver | |
| // interface, just for testing. | |
| // | |
| // It speaks a query language that's semantically similar to but | |
| // syntantically different and simpler than SQL. The syntax is as | |
| // follows: | |
| // | |
| // WIPE | |
| // CREATE|<tablename>|<col>=<type>,<col>=<type>,... | |
| // where types are: "string", [u]int{8,16,32,64}, "bool" | |
| // INSERT|<tablename>|col=val,col2=val2,col3=? | |
| // SELECT|<tablename>|projectcol1,projectcol2|filtercol=?,filtercol2=? | |
| // | |
| // When opening a a fakeDriver's database, it starts empty with no | |
| // tables. All tables and data are stored in memory only. | |
| type fakeDriver struct { | |
| mu sync.Mutex | |
| openCount int | |
| dbs map[string]*fakeDB | |
| } | |
| type fakeDB struct { | |
| name string | |
| mu sync.Mutex | |
| free []*fakeConn | |
| tables map[string]*table | |
| } | |
| type table struct { | |
| mu sync.Mutex | |
| colname []string | |
| coltype []string | |
| rows []*row | |
| } | |
| func (t *table) columnIndex(name string) int { | |
| for n, nname := range t.colname { | |
| if name == nname { | |
| return n | |
| } | |
| } | |
| return -1 | |
| } | |
| type row struct { | |
| cols []interface{} // must be same size as its table colname + coltype | |
| } | |
| func (r *row) clone() *row { | |
| nrow := &row{cols: make([]interface{}, len(r.cols))} | |
| copy(nrow.cols, r.cols) | |
| return nrow | |
| } | |
| type fakeConn struct { | |
| db *fakeDB // where to return ourselves to | |
| currTx *fakeTx | |
| } | |
| type fakeTx struct { | |
| c *fakeConn | |
| } | |
| type fakeStmt struct { | |
| c *fakeConn | |
| q string // just for debugging | |
| cmd string | |
| table string | |
| colName []string // used by CREATE, INSERT, SELECT (selected columns) | |
| colType []string // used by CREATE | |
| colValue []interface{} // used by INSERT (mix of strings and "?" for bound params) | |
| placeholders int // used by INSERT/SELECT: number of ? params | |
| whereCol []string // used by SELECT (all placeholders) | |
| placeholderConverter []driver.ValueConverter // used by INSERT | |
| } | |
| var fdriver driver.Driver = &fakeDriver{} | |
| func init() { | |
| Register("test", fdriver) | |
| } | |
| // Supports dsn forms: | |
| // <dbname> | |
| // <dbname>;wipe | |
| func (d *fakeDriver) Open(dsn string) (driver.Conn, error) { | |
| d.mu.Lock() | |
| defer d.mu.Unlock() | |
| d.openCount++ | |
| if d.dbs == nil { | |
| d.dbs = make(map[string]*fakeDB) | |
| } | |
| parts := strings.Split(dsn, ";") | |
| if len(parts) < 1 { | |
| return nil, errors.New("fakedb: no database name") | |
| } | |
| name := parts[0] | |
| db, ok := d.dbs[name] | |
| if !ok { | |
| db = &fakeDB{name: name} | |
| d.dbs[name] = db | |
| } | |
| return &fakeConn{db: db}, nil | |
| } | |
| func (db *fakeDB) wipe() { | |
| db.mu.Lock() | |
| defer db.mu.Unlock() | |
| db.tables = nil | |
| } | |
| func (db *fakeDB) createTable(name string, columnNames, columnTypes []string) error { | |
| db.mu.Lock() | |
| defer db.mu.Unlock() | |
| if db.tables == nil { | |
| db.tables = make(map[string]*table) | |
| } | |
| if _, exist := db.tables[name]; exist { | |
| return fmt.Errorf("table %q already exists", name) | |
| } | |
| if len(columnNames) != len(columnTypes) { | |
| return fmt.Errorf("create table of %q len(names) != len(types): %d vs %d", | |
| name, len(columnNames), len(columnTypes)) | |
| } | |
| db.tables[name] = &table{colname: columnNames, coltype: columnTypes} | |
| return nil | |
| } | |
| // must be called with db.mu lock held | |
| func (db *fakeDB) table(table string) (*table, bool) { | |
| if db.tables == nil { | |
| return nil, false | |
| } | |
| t, ok := db.tables[table] | |
| return t, ok | |
| } | |
| func (db *fakeDB) columnType(table, column string) (typ string, ok bool) { | |
| db.mu.Lock() | |
| defer db.mu.Unlock() | |
| t, ok := db.table(table) | |
| if !ok { | |
| return | |
| } | |
| for n, cname := range t.colname { | |
| if cname == column { | |
| return t.coltype[n], true | |
| } | |
| } | |
| return "", false | |
| } | |
| func (c *fakeConn) Begin() (driver.Tx, error) { | |
| if c.currTx != nil { | |
| return nil, errors.New("already in a transaction") | |
| } | |
| c.currTx = &fakeTx{c: c} | |
| return c.currTx, nil | |
| } | |
| func (c *fakeConn) Close() error { | |
| if c.currTx != nil { | |
| return errors.New("can't close; in a Transaction") | |
| } | |
| if c.db == nil { | |
| return errors.New("can't close; already closed") | |
| } | |
| c.db = nil | |
| return nil | |
| } | |
| func errf(msg string, args ...interface{}) error { | |
| return errors.New("fakedb: " + fmt.Sprintf(msg, args...)) | |
| } | |
| // parts are table|selectCol1,selectCol2|whereCol=?,whereCol2=? | |
| // (note that where where columns must always contain ? marks, | |
| // just a limitation for fakedb) | |
| func (c *fakeConn) prepareSelect(stmt *fakeStmt, parts []string) (driver.Stmt, error) { | |
| if len(parts) != 3 { | |
| return nil, errf("invalid SELECT syntax with %d parts; want 3", len(parts)) | |
| } | |
| stmt.table = parts[0] | |
| stmt.colName = strings.Split(parts[1], ",") | |
| for n, colspec := range strings.Split(parts[2], ",") { | |
| nameVal := strings.Split(colspec, "=") | |
| if len(nameVal) != 2 { | |
| return nil, errf("SELECT on table %q has invalid column spec of %q (index %d)", stmt.table, colspec, n) | |
| } | |
| column, value := nameVal[0], nameVal[1] | |
| _, ok := c.db.columnType(stmt.table, column) | |
| if !ok { | |
| return nil, errf("SELECT on table %q references non-existent column %q", stmt.table, column) | |
| } | |
| if value != "?" { | |
| return nil, errf("SELECT on table %q has pre-bound value for where column %q; need a question mark", | |
| stmt.table, column) | |
| } | |
| stmt.whereCol = append(stmt.whereCol, column) | |
| stmt.placeholders++ | |
| } | |
| return stmt, nil | |
| } | |
| // parts are table|col=type,col2=type2 | |
| func (c *fakeConn) prepareCreate(stmt *fakeStmt, parts []string) (driver.Stmt, error) { | |
| if len(parts) != 2 { | |
| return nil, errf("invalid CREATE syntax with %d parts; want 2", len(parts)) | |
| } | |
| stmt.table = parts[0] | |
| for n, colspec := range strings.Split(parts[1], ",") { | |
| nameType := strings.Split(colspec, "=") | |
| if len(nameType) != 2 { | |
| return nil, errf("CREATE table %q has invalid column spec of %q (index %d)", stmt.table, colspec, n) | |
| } | |
| stmt.colName = append(stmt.colName, nameType[0]) | |
| stmt.colType = append(stmt.colType, nameType[1]) | |
| } | |
| return stmt, nil | |
| } | |
| // parts are table|col=?,col2=val | |
| func (c *fakeConn) prepareInsert(stmt *fakeStmt, parts []string) (driver.Stmt, error) { | |
| if len(parts) != 2 { | |
| return nil, errf("invalid INSERT syntax with %d parts; want 2", len(parts)) | |
| } | |
| stmt.table = parts[0] | |
| for n, colspec := range strings.Split(parts[1], ",") { | |
| nameVal := strings.Split(colspec, "=") | |
| if len(nameVal) != 2 { | |
| return nil, errf("INSERT table %q has invalid column spec of %q (index %d)", stmt.table, colspec, n) | |
| } | |
| column, value := nameVal[0], nameVal[1] | |
| ctype, ok := c.db.columnType(stmt.table, column) | |
| if !ok { | |
| return nil, errf("INSERT table %q references non-existent column %q", stmt.table, column) | |
| } | |
| stmt.colName = append(stmt.colName, column) | |
| if value != "?" { | |
| var subsetVal interface{} | |
| // Convert to driver subset type | |
| switch ctype { | |
| case "string": | |
| subsetVal = []byte(value) | |
| case "int32": | |
| i, err := strconv.Atoi(value) | |
| if err != nil { | |
| return nil, errf("invalid conversion to int32 from %q", value) | |
| } | |
| subsetVal = int64(i) // int64 is a subset type, but not int32 | |
| default: | |
| return nil, errf("unsupported conversion for pre-bound parameter %q to type %q", value, ctype) | |
| } | |
| stmt.colValue = append(stmt.colValue, subsetVal) | |
| } else { | |
| stmt.placeholders++ | |
| stmt.placeholderConverter = append(stmt.placeholderConverter, converterForType(ctype)) | |
| stmt.colValue = append(stmt.colValue, "?") | |
| } | |
| } | |
| return stmt, nil | |
| } | |
| func (c *fakeConn) Prepare(query string) (driver.Stmt, error) { | |
| if c.db == nil { | |
| panic("nil c.db; conn = " + fmt.Sprintf("%#v", c)) | |
| } | |
| parts := strings.Split(query, "|") | |
| if len(parts) < 1 { | |
| return nil, errf("empty query") | |
| } | |
| cmd := parts[0] | |
| parts = parts[1:] | |
| stmt := &fakeStmt{q: query, c: c, cmd: cmd} | |
| switch cmd { | |
| case "WIPE": | |
| // Nothing | |
| case "SELECT": | |
| return c.prepareSelect(stmt, parts) | |
| case "CREATE": | |
| return c.prepareCreate(stmt, parts) | |
| case "INSERT": | |
| return c.prepareInsert(stmt, parts) | |
| default: | |
| return nil, errf("unsupported command type %q", cmd) | |
| } | |
| return stmt, nil | |
| } | |
| func (s *fakeStmt) ColumnConverter(idx int) driver.ValueConverter { | |
| return s.placeholderConverter[idx] | |
| } | |
| func (s *fakeStmt) Close() error { | |
| return nil | |
| } | |
| func (s *fakeStmt) Exec(args []interface{}) (driver.Result, error) { | |
| db := s.c.db | |
| switch s.cmd { | |
| case "WIPE": | |
| db.wipe() | |
| return driver.DDLSuccess, nil | |
| case "CREATE": | |
| if err := db.createTable(s.table, s.colName, s.colType); err != nil { | |
| return nil, err | |
| } | |
| return driver.DDLSuccess, nil | |
| case "INSERT": | |
| return s.execInsert(args) | |
| } | |
| fmt.Printf("EXEC statement, cmd=%q: %#v\n", s.cmd, s) | |
| return nil, fmt.Errorf("unimplemented statement Exec command type of %q", s.cmd) | |
| } | |
| func (s *fakeStmt) execInsert(args []interface{}) (driver.Result, error) { | |
| db := s.c.db | |
| if len(args) != s.placeholders { | |
| panic("error in pkg db; should only get here if size is correct") | |
| } | |
| db.mu.Lock() | |
| t, ok := db.table(s.table) | |
| db.mu.Unlock() | |
| if !ok { | |
| return nil, fmt.Errorf("fakedb: table %q doesn't exist", s.table) | |
| } | |
| t.mu.Lock() | |
| defer t.mu.Unlock() | |
| cols := make([]interface{}, len(t.colname)) | |
| argPos := 0 | |
| for n, colname := range s.colName { | |
| colidx := t.columnIndex(colname) | |
| if colidx == -1 { | |
| return nil, fmt.Errorf("fakedb: column %q doesn't exist or dropped since prepared statement was created", colname) | |
| } | |
| var val interface{} | |
| if strvalue, ok := s.colValue[n].(string); ok && strvalue == "?" { | |
| val = args[argPos] | |
| argPos++ | |
| } else { | |
| val = s.colValue[n] | |
| } | |
| cols[colidx] = val | |
| } | |
| t.rows = append(t.rows, &row{cols: cols}) | |
| return driver.RowsAffected(1), nil | |
| } | |
| func (s *fakeStmt) Query(args []interface{}) (driver.Rows, error) { | |
| db := s.c.db | |
| if len(args) != s.placeholders { | |
| panic("error in pkg db; should only get here if size is correct") | |
| } | |
| db.mu.Lock() | |
| t, ok := db.table(s.table) | |
| db.mu.Unlock() | |
| if !ok { | |
| return nil, fmt.Errorf("fakedb: table %q doesn't exist", s.table) | |
| } | |
| t.mu.Lock() | |
| defer t.mu.Unlock() | |
| colIdx := make(map[string]int) // select column name -> column index in table | |
| for _, name := range s.colName { | |
| idx := t.columnIndex(name) | |
| if idx == -1 { | |
| return nil, fmt.Errorf("fakedb: unknown column name %q", name) | |
| } | |
| colIdx[name] = idx | |
| } | |
| mrows := []*row{} | |
| rows: | |
| for _, trow := range t.rows { | |
| // Process the where clause, skipping non-match rows. This is lazy | |
| // and just uses fmt.Sprintf("%v") to test equality. Good enough | |
| // for test code. | |
| for widx, wcol := range s.whereCol { | |
| idx := t.columnIndex(wcol) | |
| if idx == -1 { | |
| return nil, fmt.Errorf("db: invalid where clause column %q", wcol) | |
| } | |
| tcol := trow.cols[idx] | |
| if bs, ok := tcol.([]byte); ok { | |
| // lazy hack to avoid sprintf %v on a []byte | |
| tcol = string(bs) | |
| } | |
| if fmt.Sprintf("%v", tcol) != fmt.Sprintf("%v", args[widx]) { | |
| continue rows | |
| } | |
| } | |
| mrow := &row{cols: make([]interface{}, len(s.colName))} | |
| for seli, name := range s.colName { | |
| mrow.cols[seli] = trow.cols[colIdx[name]] | |
| } | |
| mrows = append(mrows, mrow) | |
| } | |
| cursor := &rowsCursor{ | |
| pos: -1, | |
| rows: mrows, | |
| cols: s.colName, | |
| } | |
| return cursor, nil | |
| } | |
| func (s *fakeStmt) NumInput() int { | |
| return s.placeholders | |
| } | |
| func (tx *fakeTx) Commit() error { | |
| tx.c.currTx = nil | |
| return nil | |
| } | |
| func (tx *fakeTx) Rollback() error { | |
| tx.c.currTx = nil | |
| return nil | |
| } | |
| type rowsCursor struct { | |
| cols []string | |
| pos int | |
| rows []*row | |
| closed bool | |
| } | |
| func (rc *rowsCursor) Close() error { | |
| rc.closed = true | |
| return nil | |
| } | |
| func (rc *rowsCursor) Columns() []string { | |
| return rc.cols | |
| } | |
| func (rc *rowsCursor) Next(dest []interface{}) error { | |
| if rc.closed { | |
| return errors.New("fakedb: cursor is closed") | |
| } | |
| rc.pos++ | |
| if rc.pos >= len(rc.rows) { | |
| return io.EOF // per interface spec | |
| } | |
| for i, v := range rc.rows[rc.pos].cols { | |
| // TODO(bradfitz): convert to subset types? naah, I | |
| // think the subset types should only be input to | |
| // driver, but the sql package should be able to handle | |
| // a wider range of types coming out of drivers. all | |
| // for ease of drivers, and to prevent drivers from | |
| // messing up conversions or doing them differently. | |
| dest[i] = v | |
| } | |
| return nil | |
| } | |
| func converterForType(typ string) driver.ValueConverter { | |
| switch typ { | |
| case "bool": | |
| return driver.Bool | |
| case "int32": | |
| return driver.Int32 | |
| case "string": | |
| return driver.String | |
| } | |
| panic("invalid fakedb column type of " + typ) | |
| } |