1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213
|
package pack
import (
"fmt"
"hash"
"os"
"path/filepath"
"regexp"
"sort"
"strings"
"github.com/git-lfs/gitobj/v2/errors"
)
// Set allows access of objects stored across a set of packfiles.
type Set struct {
// m maps the leading byte of a SHA-1 object name to a set of packfiles
// that might contain that object, in order of which packfile is most
// likely to contain that object.
m map[byte][]*Packfile
// closeFn is a function that is run by Close(), designated to free
// resources held by the *Set, like open packfiles.
closeFn func() error
}
var (
// nameRe is a regular expression that matches the basename of a
// filepath that is a packfile.
//
// It includes one matchgroup, which is the SHA-1 name of the pack.
nameRe = regexp.MustCompile(`^(.*)\.pack$`)
)
// NewSet creates a new *Set of all packfiles found in a given object database's
// root (i.e., "/path/to/repo/.git/objects").
//
// It finds all packfiles in the "pack" subdirectory, and instantiates a *Set
// containing them. If there was an error parsing the packfiles in that
// directory, or the directory was otherwise unable to be observed, NewSet
// returns that error.
func NewSet(db string, algo hash.Hash) (*Set, error) {
pd := filepath.Join(db, "pack")
paths, err := filepath.Glob(filepath.Join(escapeGlobPattern(pd), "*.pack"))
if err != nil {
return nil, err
}
packs := make([]*Packfile, 0, len(paths))
for _, path := range paths {
submatch := nameRe.FindStringSubmatch(filepath.Base(path))
if len(submatch) != 2 {
continue
}
name := submatch[1]
idxf, err := os.Open(filepath.Join(pd, fmt.Sprintf("%s.idx", name)))
if err != nil {
// We have a pack (since it matched the regex), but the
// index is missing or unusable. Skip this pack and
// continue on with the next one, as Git does.
if idxf != nil {
// In the unlikely event that we did open a
// file, close it, but discard any error in
// doing so.
idxf.Close()
}
continue
}
packf, err := os.Open(filepath.Join(pd, fmt.Sprintf("%s.pack", name)))
if err != nil {
return nil, err
}
pack, err := DecodePackfile(packf, algo)
if err != nil {
return nil, err
}
idx, err := DecodeIndex(idxf, algo)
if err != nil {
return nil, err
}
pack.idx = idx
packs = append(packs, pack)
}
return NewSetPacks(packs...), nil
}
// globEscapes uses these escapes because filepath.Glob does not understand
// backslash escapes on Windows.
var globEscapes = map[string]string{
"*": "[*]",
"?": "[?]",
"[": "[[]",
}
func escapeGlobPattern(s string) string {
for char, escape := range globEscapes {
s = strings.Replace(s, char, escape, -1)
}
return s
}
// NewSetPacks creates a new *Set from the given packfiles.
func NewSetPacks(packs ...*Packfile) *Set {
m := make(map[byte][]*Packfile)
for i := 0; i < 256; i++ {
n := byte(i)
for j := 0; j < len(packs); j++ {
pack := packs[j]
var count uint32
if n == 0 {
count = pack.idx.fanout[n]
} else {
count = pack.idx.fanout[n] - pack.idx.fanout[n-1]
}
if count > 0 {
m[n] = append(m[n], pack)
}
}
sort.Slice(m[n], func(i, j int) bool {
ni := m[n][i].idx.fanout[n]
nj := m[n][j].idx.fanout[n]
return ni > nj
})
}
return &Set{
m: m,
closeFn: func() error {
for _, pack := range packs {
if err := pack.Close(); err != nil {
return err
}
}
return nil
},
}
}
// Close closes all open packfiles, returning an error if one was encountered.
func (s *Set) Close() error {
if s.closeFn == nil {
return nil
}
return s.closeFn()
}
// Object opens (but does not unpack, or, apply the delta-base chain) a given
// object in the first packfile that matches it.
//
// Object searches packfiles contained in the set in order of how many objects
// they have that begin with the first by of the given SHA-1 "name", in
// descending order.
//
// If the object was unable to be found in any of the packfiles, (nil,
// ErrNotFound) will be returned.
//
// If there was otherwise an error opening the object for reading from any of
// the packfiles, it will be returned, and no other packfiles will be searched.
//
// Otherwise, the object will be returned without error.
func (s *Set) Object(name []byte) (*Object, error) {
return s.each(name, func(p *Packfile) (*Object, error) {
return p.Object(name)
})
}
// iterFn is a function that takes a given packfile and opens an object from it.
type iterFn func(p *Packfile) (o *Object, err error)
// each executes the given iterFn "fn" on each Packfile that has any objects
// beginning with a prefix of the SHA-1 "name", in order of which packfiles have
// the most objects beginning with that prefix.
//
// If any invocation of "fn" returns a non-nil error, it will either be a)
// returned immediately, if the error is not ErrIsNotFound, or b) continued
// immediately, if the error is ErrNotFound.
//
// If no packfiles match the given file, return errors.NoSuchObject, along with
// no object.
func (s *Set) each(name []byte, fn iterFn) (*Object, error) {
var key byte
if len(name) > 0 {
key = name[0]
}
for _, pack := range s.m[key] {
o, err := fn(pack)
if err != nil {
if IsNotFound(err) {
continue
}
return nil, err
}
return o, nil
}
return nil, errors.NoSuchObject(name)
}
|