garm/vendor/github.com/go-sql-driver/mysql/utils.go
Gabriel Adrian Samfira f25951decb Add extra specs on pools
Extra specs is an opaque valid JSON that can be set on a pool and which
will be passed along to the provider as part of instance bootstrap params.

This field is meant to allow operators to send extra configuration values
to external or built-in providers. The extra specs is not interpreted or
useful in any way to garm itself, but it may be useful to the provider
which interacts with the IaaS.

The extra specs are not meant to be used for secrets. Adding sensitive
information to this field is highly discouraged. This field is meant as a    
means to add fine tuning knobs to the providers, on a per pool basis.

Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
2023-01-30 13:10:21 +00:00

834 lines
20 KiB
Go

// Go MySQL Driver - A MySQL-Driver for Go's database/sql package
//
// Copyright 2012 The Go-MySQL-Driver Authors. All rights reserved.
//
// This Source Code Form is subject to the terms of the Mozilla Public
// License, v. 2.0. If a copy of the MPL was not distributed with this file,
// You can obtain one at http://mozilla.org/MPL/2.0/.
package mysql
import (
"crypto/tls"
"database/sql"
"database/sql/driver"
"encoding/binary"
"errors"
"fmt"
"io"
"strconv"
"strings"
"sync"
"sync/atomic"
"time"
)
// Registry for custom tls.Configs
var (
tlsConfigLock sync.RWMutex
tlsConfigRegistry map[string]*tls.Config
)
// RegisterTLSConfig registers a custom tls.Config to be used with sql.Open.
// Use the key as a value in the DSN where tls=value.
//
// Note: The provided tls.Config is exclusively owned by the driver after
// registering it.
//
// rootCertPool := x509.NewCertPool()
// pem, err := ioutil.ReadFile("/path/ca-cert.pem")
// if err != nil {
// log.Fatal(err)
// }
// if ok := rootCertPool.AppendCertsFromPEM(pem); !ok {
// log.Fatal("Failed to append PEM.")
// }
// clientCert := make([]tls.Certificate, 0, 1)
// certs, err := tls.LoadX509KeyPair("/path/client-cert.pem", "/path/client-key.pem")
// if err != nil {
// log.Fatal(err)
// }
// clientCert = append(clientCert, certs)
// mysql.RegisterTLSConfig("custom", &tls.Config{
// RootCAs: rootCertPool,
// Certificates: clientCert,
// })
// db, err := sql.Open("mysql", "user@tcp(localhost:3306)/test?tls=custom")
func RegisterTLSConfig(key string, config *tls.Config) error {
if _, isBool := readBool(key); isBool || strings.ToLower(key) == "skip-verify" || strings.ToLower(key) == "preferred" {
return fmt.Errorf("key '%s' is reserved", key)
}
tlsConfigLock.Lock()
if tlsConfigRegistry == nil {
tlsConfigRegistry = make(map[string]*tls.Config)
}
tlsConfigRegistry[key] = config
tlsConfigLock.Unlock()
return nil
}
// DeregisterTLSConfig removes the tls.Config associated with key.
func DeregisterTLSConfig(key string) {
tlsConfigLock.Lock()
if tlsConfigRegistry != nil {
delete(tlsConfigRegistry, key)
}
tlsConfigLock.Unlock()
}
func getTLSConfigClone(key string) (config *tls.Config) {
tlsConfigLock.RLock()
if v, ok := tlsConfigRegistry[key]; ok {
config = v.Clone()
}
tlsConfigLock.RUnlock()
return
}
// Returns the bool value of the input.
// The 2nd return value indicates if the input was a valid bool value
func readBool(input string) (value bool, valid bool) {
switch input {
case "1", "true", "TRUE", "True":
return true, true
case "0", "false", "FALSE", "False":
return false, true
}
// Not a valid bool value
return
}
/******************************************************************************
* Time related utils *
******************************************************************************/
func parseDateTime(b []byte, loc *time.Location) (time.Time, error) {
const base = "0000-00-00 00:00:00.000000"
switch len(b) {
case 10, 19, 21, 22, 23, 24, 25, 26: // up to "YYYY-MM-DD HH:MM:SS.MMMMMM"
if string(b) == base[:len(b)] {
return time.Time{}, nil
}
year, err := parseByteYear(b)
if err != nil {
return time.Time{}, err
}
if b[4] != '-' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[4])
}
m, err := parseByte2Digits(b[5], b[6])
if err != nil {
return time.Time{}, err
}
month := time.Month(m)
if b[7] != '-' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[7])
}
day, err := parseByte2Digits(b[8], b[9])
if err != nil {
return time.Time{}, err
}
if len(b) == 10 {
return time.Date(year, month, day, 0, 0, 0, 0, loc), nil
}
if b[10] != ' ' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[10])
}
hour, err := parseByte2Digits(b[11], b[12])
if err != nil {
return time.Time{}, err
}
if b[13] != ':' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[13])
}
min, err := parseByte2Digits(b[14], b[15])
if err != nil {
return time.Time{}, err
}
if b[16] != ':' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[16])
}
sec, err := parseByte2Digits(b[17], b[18])
if err != nil {
return time.Time{}, err
}
if len(b) == 19 {
return time.Date(year, month, day, hour, min, sec, 0, loc), nil
}
if b[19] != '.' {
return time.Time{}, fmt.Errorf("bad value for field: `%c`", b[19])
}
nsec, err := parseByteNanoSec(b[20:])
if err != nil {
return time.Time{}, err
}
return time.Date(year, month, day, hour, min, sec, nsec, loc), nil
default:
return time.Time{}, fmt.Errorf("invalid time bytes: %s", b)
}
}
func parseByteYear(b []byte) (int, error) {
year, n := 0, 1000
for i := 0; i < 4; i++ {
v, err := bToi(b[i])
if err != nil {
return 0, err
}
year += v * n
n /= 10
}
return year, nil
}
func parseByte2Digits(b1, b2 byte) (int, error) {
d1, err := bToi(b1)
if err != nil {
return 0, err
}
d2, err := bToi(b2)
if err != nil {
return 0, err
}
return d1*10 + d2, nil
}
func parseByteNanoSec(b []byte) (int, error) {
ns, digit := 0, 100000 // max is 6-digits
for i := 0; i < len(b); i++ {
v, err := bToi(b[i])
if err != nil {
return 0, err
}
ns += v * digit
digit /= 10
}
// nanoseconds has 10-digits. (needs to scale digits)
// 10 - 6 = 4, so we have to multiple 1000.
return ns * 1000, nil
}
func bToi(b byte) (int, error) {
if b < '0' || b > '9' {
return 0, errors.New("not [0-9]")
}
return int(b - '0'), nil
}
func parseBinaryDateTime(num uint64, data []byte, loc *time.Location) (driver.Value, error) {
switch num {
case 0:
return time.Time{}, nil
case 4:
return time.Date(
int(binary.LittleEndian.Uint16(data[:2])), // year
time.Month(data[2]), // month
int(data[3]), // day
0, 0, 0, 0,
loc,
), nil
case 7:
return time.Date(
int(binary.LittleEndian.Uint16(data[:2])), // year
time.Month(data[2]), // month
int(data[3]), // day
int(data[4]), // hour
int(data[5]), // minutes
int(data[6]), // seconds
0,
loc,
), nil
case 11:
return time.Date(
int(binary.LittleEndian.Uint16(data[:2])), // year
time.Month(data[2]), // month
int(data[3]), // day
int(data[4]), // hour
int(data[5]), // minutes
int(data[6]), // seconds
int(binary.LittleEndian.Uint32(data[7:11]))*1000, // nanoseconds
loc,
), nil
}
return nil, fmt.Errorf("invalid DATETIME packet length %d", num)
}
func appendDateTime(buf []byte, t time.Time) ([]byte, error) {
year, month, day := t.Date()
hour, min, sec := t.Clock()
nsec := t.Nanosecond()
if year < 1 || year > 9999 {
return buf, errors.New("year is not in the range [1, 9999]: " + strconv.Itoa(year)) // use errors.New instead of fmt.Errorf to avoid year escape to heap
}
year100 := year / 100
year1 := year % 100
var localBuf [len("2006-01-02T15:04:05.999999999")]byte // does not escape
localBuf[0], localBuf[1], localBuf[2], localBuf[3] = digits10[year100], digits01[year100], digits10[year1], digits01[year1]
localBuf[4] = '-'
localBuf[5], localBuf[6] = digits10[month], digits01[month]
localBuf[7] = '-'
localBuf[8], localBuf[9] = digits10[day], digits01[day]
if hour == 0 && min == 0 && sec == 0 && nsec == 0 {
return append(buf, localBuf[:10]...), nil
}
localBuf[10] = ' '
localBuf[11], localBuf[12] = digits10[hour], digits01[hour]
localBuf[13] = ':'
localBuf[14], localBuf[15] = digits10[min], digits01[min]
localBuf[16] = ':'
localBuf[17], localBuf[18] = digits10[sec], digits01[sec]
if nsec == 0 {
return append(buf, localBuf[:19]...), nil
}
nsec100000000 := nsec / 100000000
nsec1000000 := (nsec / 1000000) % 100
nsec10000 := (nsec / 10000) % 100
nsec100 := (nsec / 100) % 100
nsec1 := nsec % 100
localBuf[19] = '.'
// milli second
localBuf[20], localBuf[21], localBuf[22] =
digits01[nsec100000000], digits10[nsec1000000], digits01[nsec1000000]
// micro second
localBuf[23], localBuf[24], localBuf[25] =
digits10[nsec10000], digits01[nsec10000], digits10[nsec100]
// nano second
localBuf[26], localBuf[27], localBuf[28] =
digits01[nsec100], digits10[nsec1], digits01[nsec1]
// trim trailing zeros
n := len(localBuf)
for n > 0 && localBuf[n-1] == '0' {
n--
}
return append(buf, localBuf[:n]...), nil
}
// zeroDateTime is used in formatBinaryDateTime to avoid an allocation
// if the DATE or DATETIME has the zero value.
// It must never be changed.
// The current behavior depends on database/sql copying the result.
var zeroDateTime = []byte("0000-00-00 00:00:00.000000")
const digits01 = "0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789"
const digits10 = "0000000000111111111122222222223333333333444444444455555555556666666666777777777788888888889999999999"
func appendMicrosecs(dst, src []byte, decimals int) []byte {
if decimals <= 0 {
return dst
}
if len(src) == 0 {
return append(dst, ".000000"[:decimals+1]...)
}
microsecs := binary.LittleEndian.Uint32(src[:4])
p1 := byte(microsecs / 10000)
microsecs -= 10000 * uint32(p1)
p2 := byte(microsecs / 100)
microsecs -= 100 * uint32(p2)
p3 := byte(microsecs)
switch decimals {
default:
return append(dst, '.',
digits10[p1], digits01[p1],
digits10[p2], digits01[p2],
digits10[p3], digits01[p3],
)
case 1:
return append(dst, '.',
digits10[p1],
)
case 2:
return append(dst, '.',
digits10[p1], digits01[p1],
)
case 3:
return append(dst, '.',
digits10[p1], digits01[p1],
digits10[p2],
)
case 4:
return append(dst, '.',
digits10[p1], digits01[p1],
digits10[p2], digits01[p2],
)
case 5:
return append(dst, '.',
digits10[p1], digits01[p1],
digits10[p2], digits01[p2],
digits10[p3],
)
}
}
func formatBinaryDateTime(src []byte, length uint8) (driver.Value, error) {
// length expects the deterministic length of the zero value,
// negative time and 100+ hours are automatically added if needed
if len(src) == 0 {
return zeroDateTime[:length], nil
}
var dst []byte // return value
var p1, p2, p3 byte // current digit pair
switch length {
case 10, 19, 21, 22, 23, 24, 25, 26:
default:
t := "DATE"
if length > 10 {
t += "TIME"
}
return nil, fmt.Errorf("illegal %s length %d", t, length)
}
switch len(src) {
case 4, 7, 11:
default:
t := "DATE"
if length > 10 {
t += "TIME"
}
return nil, fmt.Errorf("illegal %s packet length %d", t, len(src))
}
dst = make([]byte, 0, length)
// start with the date
year := binary.LittleEndian.Uint16(src[:2])
pt := year / 100
p1 = byte(year - 100*uint16(pt))
p2, p3 = src[2], src[3]
dst = append(dst,
digits10[pt], digits01[pt],
digits10[p1], digits01[p1], '-',
digits10[p2], digits01[p2], '-',
digits10[p3], digits01[p3],
)
if length == 10 {
return dst, nil
}
if len(src) == 4 {
return append(dst, zeroDateTime[10:length]...), nil
}
dst = append(dst, ' ')
p1 = src[4] // hour
src = src[5:]
// p1 is 2-digit hour, src is after hour
p2, p3 = src[0], src[1]
dst = append(dst,
digits10[p1], digits01[p1], ':',
digits10[p2], digits01[p2], ':',
digits10[p3], digits01[p3],
)
return appendMicrosecs(dst, src[2:], int(length)-20), nil
}
func formatBinaryTime(src []byte, length uint8) (driver.Value, error) {
// length expects the deterministic length of the zero value,
// negative time and 100+ hours are automatically added if needed
if len(src) == 0 {
return zeroDateTime[11 : 11+length], nil
}
var dst []byte // return value
switch length {
case
8, // time (can be up to 10 when negative and 100+ hours)
10, 11, 12, 13, 14, 15: // time with fractional seconds
default:
return nil, fmt.Errorf("illegal TIME length %d", length)
}
switch len(src) {
case 8, 12:
default:
return nil, fmt.Errorf("invalid TIME packet length %d", len(src))
}
// +2 to enable negative time and 100+ hours
dst = make([]byte, 0, length+2)
if src[0] == 1 {
dst = append(dst, '-')
}
days := binary.LittleEndian.Uint32(src[1:5])
hours := int64(days)*24 + int64(src[5])
if hours >= 100 {
dst = strconv.AppendInt(dst, hours, 10)
} else {
dst = append(dst, digits10[hours], digits01[hours])
}
min, sec := src[6], src[7]
dst = append(dst, ':',
digits10[min], digits01[min], ':',
digits10[sec], digits01[sec],
)
return appendMicrosecs(dst, src[8:], int(length)-9), nil
}
/******************************************************************************
* Convert from and to bytes *
******************************************************************************/
func uint64ToBytes(n uint64) []byte {
return []byte{
byte(n),
byte(n >> 8),
byte(n >> 16),
byte(n >> 24),
byte(n >> 32),
byte(n >> 40),
byte(n >> 48),
byte(n >> 56),
}
}
func uint64ToString(n uint64) []byte {
var a [20]byte
i := 20
// U+0030 = 0
// ...
// U+0039 = 9
var q uint64
for n >= 10 {
i--
q = n / 10
a[i] = uint8(n-q*10) + 0x30
n = q
}
i--
a[i] = uint8(n) + 0x30
return a[i:]
}
// treats string value as unsigned integer representation
func stringToInt(b []byte) int {
val := 0
for i := range b {
val *= 10
val += int(b[i] - 0x30)
}
return val
}
// returns the string read as a bytes slice, whether the value is NULL,
// the number of bytes read and an error, in case the string is longer than
// the input slice
func readLengthEncodedString(b []byte) ([]byte, bool, int, error) {
// Get length
num, isNull, n := readLengthEncodedInteger(b)
if num < 1 {
return b[n:n], isNull, n, nil
}
n += int(num)
// Check data length
if len(b) >= n {
return b[n-int(num) : n : n], false, n, nil
}
return nil, false, n, io.EOF
}
// returns the number of bytes skipped and an error, in case the string is
// longer than the input slice
func skipLengthEncodedString(b []byte) (int, error) {
// Get length
num, _, n := readLengthEncodedInteger(b)
if num < 1 {
return n, nil
}
n += int(num)
// Check data length
if len(b) >= n {
return n, nil
}
return n, io.EOF
}
// returns the number read, whether the value is NULL and the number of bytes read
func readLengthEncodedInteger(b []byte) (uint64, bool, int) {
// See issue #349
if len(b) == 0 {
return 0, true, 1
}
switch b[0] {
// 251: NULL
case 0xfb:
return 0, true, 1
// 252: value of following 2
case 0xfc:
return uint64(b[1]) | uint64(b[2])<<8, false, 3
// 253: value of following 3
case 0xfd:
return uint64(b[1]) | uint64(b[2])<<8 | uint64(b[3])<<16, false, 4
// 254: value of following 8
case 0xfe:
return uint64(b[1]) | uint64(b[2])<<8 | uint64(b[3])<<16 |
uint64(b[4])<<24 | uint64(b[5])<<32 | uint64(b[6])<<40 |
uint64(b[7])<<48 | uint64(b[8])<<56,
false, 9
}
// 0-250: value of first byte
return uint64(b[0]), false, 1
}
// encodes a uint64 value and appends it to the given bytes slice
func appendLengthEncodedInteger(b []byte, n uint64) []byte {
switch {
case n <= 250:
return append(b, byte(n))
case n <= 0xffff:
return append(b, 0xfc, byte(n), byte(n>>8))
case n <= 0xffffff:
return append(b, 0xfd, byte(n), byte(n>>8), byte(n>>16))
}
return append(b, 0xfe, byte(n), byte(n>>8), byte(n>>16), byte(n>>24),
byte(n>>32), byte(n>>40), byte(n>>48), byte(n>>56))
}
// reserveBuffer checks cap(buf) and expand buffer to len(buf) + appendSize.
// If cap(buf) is not enough, reallocate new buffer.
func reserveBuffer(buf []byte, appendSize int) []byte {
newSize := len(buf) + appendSize
if cap(buf) < newSize {
// Grow buffer exponentially
newBuf := make([]byte, len(buf)*2+appendSize)
copy(newBuf, buf)
buf = newBuf
}
return buf[:newSize]
}
// escapeBytesBackslash escapes []byte with backslashes (\)
// This escapes the contents of a string (provided as []byte) by adding backslashes before special
// characters, and turning others into specific escape sequences, such as
// turning newlines into \n and null bytes into \0.
// https://github.com/mysql/mysql-server/blob/mysql-5.7.5/mysys/charset.c#L823-L932
func escapeBytesBackslash(buf, v []byte) []byte {
pos := len(buf)
buf = reserveBuffer(buf, len(v)*2)
for _, c := range v {
switch c {
case '\x00':
buf[pos+1] = '0'
buf[pos] = '\\'
pos += 2
case '\n':
buf[pos+1] = 'n'
buf[pos] = '\\'
pos += 2
case '\r':
buf[pos+1] = 'r'
buf[pos] = '\\'
pos += 2
case '\x1a':
buf[pos+1] = 'Z'
buf[pos] = '\\'
pos += 2
case '\'':
buf[pos+1] = '\''
buf[pos] = '\\'
pos += 2
case '"':
buf[pos+1] = '"'
buf[pos] = '\\'
pos += 2
case '\\':
buf[pos+1] = '\\'
buf[pos] = '\\'
pos += 2
default:
buf[pos] = c
pos++
}
}
return buf[:pos]
}
// escapeStringBackslash is similar to escapeBytesBackslash but for string.
func escapeStringBackslash(buf []byte, v string) []byte {
pos := len(buf)
buf = reserveBuffer(buf, len(v)*2)
for i := 0; i < len(v); i++ {
c := v[i]
switch c {
case '\x00':
buf[pos+1] = '0'
buf[pos] = '\\'
pos += 2
case '\n':
buf[pos+1] = 'n'
buf[pos] = '\\'
pos += 2
case '\r':
buf[pos+1] = 'r'
buf[pos] = '\\'
pos += 2
case '\x1a':
buf[pos+1] = 'Z'
buf[pos] = '\\'
pos += 2
case '\'':
buf[pos+1] = '\''
buf[pos] = '\\'
pos += 2
case '"':
buf[pos+1] = '"'
buf[pos] = '\\'
pos += 2
case '\\':
buf[pos+1] = '\\'
buf[pos] = '\\'
pos += 2
default:
buf[pos] = c
pos++
}
}
return buf[:pos]
}
// escapeBytesQuotes escapes apostrophes in []byte by doubling them up.
// This escapes the contents of a string by doubling up any apostrophes that
// it contains. This is used when the NO_BACKSLASH_ESCAPES SQL_MODE is in
// effect on the server.
// https://github.com/mysql/mysql-server/blob/mysql-5.7.5/mysys/charset.c#L963-L1038
func escapeBytesQuotes(buf, v []byte) []byte {
pos := len(buf)
buf = reserveBuffer(buf, len(v)*2)
for _, c := range v {
if c == '\'' {
buf[pos+1] = '\''
buf[pos] = '\''
pos += 2
} else {
buf[pos] = c
pos++
}
}
return buf[:pos]
}
// escapeStringQuotes is similar to escapeBytesQuotes but for string.
func escapeStringQuotes(buf []byte, v string) []byte {
pos := len(buf)
buf = reserveBuffer(buf, len(v)*2)
for i := 0; i < len(v); i++ {
c := v[i]
if c == '\'' {
buf[pos+1] = '\''
buf[pos] = '\''
pos += 2
} else {
buf[pos] = c
pos++
}
}
return buf[:pos]
}
/******************************************************************************
* Sync utils *
******************************************************************************/
// noCopy may be embedded into structs which must not be copied
// after the first use.
//
// See https://github.com/golang/go/issues/8005#issuecomment-190753527
// for details.
type noCopy struct{}
// Lock is a no-op used by -copylocks checker from `go vet`.
func (*noCopy) Lock() {}
// Unlock is a no-op used by -copylocks checker from `go vet`.
// noCopy should implement sync.Locker from Go 1.11
// https://github.com/golang/go/commit/c2eba53e7f80df21d51285879d51ab81bcfbf6bc
// https://github.com/golang/go/issues/26165
func (*noCopy) Unlock() {}
// atomicError is a wrapper for atomically accessed error values
type atomicError struct {
_ noCopy
value atomic.Value
}
// Set sets the error value regardless of the previous value.
// The value must not be nil
func (ae *atomicError) Set(value error) {
ae.value.Store(value)
}
// Value returns the current error value
func (ae *atomicError) Value() error {
if v := ae.value.Load(); v != nil {
// this will panic if the value doesn't implement the error interface
return v.(error)
}
return nil
}
func namedValueToValue(named []driver.NamedValue) ([]driver.Value, error) {
dargs := make([]driver.Value, len(named))
for n, param := range named {
if len(param.Name) > 0 {
// TODO: support the use of Named Parameters #561
return nil, errors.New("mysql: driver does not support the use of Named Parameters")
}
dargs[n] = param.Value
}
return dargs, nil
}
func mapIsolationLevel(level driver.IsolationLevel) (string, error) {
switch sql.IsolationLevel(level) {
case sql.LevelRepeatableRead:
return "REPEATABLE READ", nil
case sql.LevelReadCommitted:
return "READ COMMITTED", nil
case sql.LevelReadUncommitted:
return "READ UNCOMMITTED", nil
case sql.LevelSerializable:
return "SERIALIZABLE", nil
default:
return "", fmt.Errorf("mysql: unsupported isolation level: %v", level)
}
}