Refactor parseTreeEntries, speed up tree list (#21368)

Close #20315 (fix the panic when parsing invalid input), Speed up #20231 (use ls-tree without size field)

Introduce ListEntriesRecursiveFast (ls-tree without size) and ListEntriesRecursiveWithSize (ls-tree with size)
This commit is contained in:
wxiaoguang 2022-10-08 01:20:53 +08:00 committed by GitHub
parent 69fc510d6d
commit c08e42c47e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
7 changed files with 109 additions and 53 deletions

View file

@ -22,70 +22,72 @@ func ParseTreeEntries(data []byte) ([]*TreeEntry, error) {
return parseTreeEntries(data, nil)
}
var sepSpace = []byte{' '}
func parseTreeEntries(data []byte, ptree *Tree) ([]*TreeEntry, error) {
entries := make([]*TreeEntry, 0, 10)
var err error
entries := make([]*TreeEntry, 0, bytes.Count(data, []byte{'\n'})+1)
for pos := 0; pos < len(data); {
// expect line to be of the form "<mode> <type> <sha> <space-padded-size>\t<filename>"
// expect line to be of the form:
// <mode> <type> <sha> <space-padded-size>\t<filename>
// <mode> <type> <sha>\t<filename>
posEnd := bytes.IndexByte(data[pos:], '\n')
if posEnd == -1 {
posEnd = len(data)
} else {
posEnd += pos
}
line := data[pos:posEnd]
posTab := bytes.IndexByte(line, '\t')
if posTab == -1 {
return nil, fmt.Errorf("invalid ls-tree output (no tab): %q", line)
}
entry := new(TreeEntry)
entry.ptree = ptree
if pos+6 > len(data) {
return nil, fmt.Errorf("Invalid ls-tree output: %s", string(data))
entryAttrs := line[:posTab]
entryName := line[posTab+1:]
entryMode, entryAttrs, _ := bytes.Cut(entryAttrs, sepSpace)
_ /* entryType */, entryAttrs, _ = bytes.Cut(entryAttrs, sepSpace) // the type is not used, the mode is enough to determine the type
entryObjectID, entryAttrs, _ := bytes.Cut(entryAttrs, sepSpace)
if len(entryAttrs) > 0 {
entrySize := entryAttrs // the last field is the space-padded-size
entry.size, _ = strconv.ParseInt(strings.TrimSpace(string(entrySize)), 10, 64)
entry.sized = true
}
switch string(data[pos : pos+6]) {
switch string(entryMode) {
case "100644":
entry.entryMode = EntryModeBlob
pos += 12 // skip over "100644 blob "
case "100755":
entry.entryMode = EntryModeExec
pos += 12 // skip over "100755 blob "
case "120000":
entry.entryMode = EntryModeSymlink
pos += 12 // skip over "120000 blob "
case "160000":
entry.entryMode = EntryModeCommit
pos += 14 // skip over "160000 object "
case "040000", "040755": // git uses 040000 for tree object, but some users may get 040755 for unknown reasons
entry.entryMode = EntryModeTree
pos += 12 // skip over "040000 tree "
default:
return nil, fmt.Errorf("unknown type: %v", string(data[pos:pos+6]))
return nil, fmt.Errorf("unknown type: %v", string(entryMode))
}
if pos+40 > len(data) {
return nil, fmt.Errorf("Invalid ls-tree output: %s", string(data))
}
id, err := NewIDFromString(string(data[pos : pos+40]))
entry.ID, err = NewIDFromString(string(entryObjectID))
if err != nil {
return nil, fmt.Errorf("Invalid ls-tree output: %v", err)
}
entry.ID = id
pos += 41 // skip over sha and trailing space
end := pos + bytes.IndexByte(data[pos:], '\t')
if end < pos {
return nil, fmt.Errorf("Invalid ls-tree -l output: %s", string(data))
}
entry.size, _ = strconv.ParseInt(strings.TrimSpace(string(data[pos:end])), 10, 64)
entry.sized = true
pos = end + 1
end = pos + bytes.IndexByte(data[pos:], '\n')
if end < pos {
return nil, fmt.Errorf("Invalid ls-tree output: %s", string(data))
return nil, fmt.Errorf("invalid ls-tree output (invalid object id): %q, err: %w", line, err)
}
// In case entry name is surrounded by double quotes(it happens only in git-shell).
if data[pos] == '"' {
entry.name, err = strconv.Unquote(string(data[pos:end]))
if len(entryName) > 0 && entryName[0] == '"' {
entry.name, err = strconv.Unquote(string(entryName))
if err != nil {
return nil, fmt.Errorf("Invalid ls-tree output: %v", err)
return nil, fmt.Errorf("invalid ls-tree output (invalid name): %q, err: %w", line, err)
}
} else {
entry.name = string(data[pos:end])
entry.name = string(entryName)
}
pos = end + 1
pos = posEnd + 1
entries = append(entries, entry)
}
return entries, nil