github.com/pelicanplatform/pelican@v1.0.5/client/unique_hash_darwin.go (about)

     1  //go:build darwin
     2  // +build darwin
     3  
     4  /***************************************************************
     5   *
     6   * Copyright (C) 2023, Pelican Project, Morgridge Institute for Research
     7   *
     8   * Licensed under the Apache License, Version 2.0 (the "License"); you
     9   * may not use this file except in compliance with the License.  You may
    10   * obtain a copy of the License at
    11   *
    12   *    http://www.apache.org/licenses/LICENSE-2.0
    13   *
    14   * Unless required by applicable law or agreed to in writing, software
    15   * distributed under the License is distributed on an "AS IS" BASIS,
    16   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    17   * See the License for the specific language governing permissions and
    18   * limitations under the License.
    19   *
    20   ***************************************************************/
    21  
    22  package client
    23  
    24  import (
    25  	"bytes"
    26  	"crypto/sha256"
    27  	"encoding/binary"
    28  	"fmt"
    29  	"syscall"
    30  	"time"
    31  
    32  	log "github.com/sirupsen/logrus"
    33  )
    34  
    35  // Given a local filename, create a unique identifier from filesystem
    36  // metadata; anytime the file contents change, the unique identifier
    37  // should also change.  The hash changes once every 24 hours to allow
    38  // the shadow origin to do garbage collection.
    39  //
    40  // The current algorithm to generate the unique identifier is to
    41  // take the concatenation of the following byte buffers:
    42  //
    43  // - basename of the path.
    44  // - inode # as a 64-bit integer serialized to a byte buffer, represented in network order.
    45  // - ctime # as a double serialized to a buffer.
    46  // - Current unix epoch time, as integer, divided by 86400
    47  //
    48  // and then running it through a SHA256 sum to produce a digest.
    49  //
    50  // The hex digest of the SHA256 sum is returned along with the file size
    51  func unique_hash(filePath string) (string, uint64, error) {
    52  	log.Debugf("Creating a unique hash for filename %s", filePath)
    53  
    54  	var st syscall.Stat_t
    55  	if err := syscall.Stat(filePath, &st); err != nil {
    56  		log.Debugln("Error while stat'ing file for metadata:", err)
    57  		return "", 0, err
    58  	}
    59  
    60  	buf := new(bytes.Buffer)
    61  	if err := binary.Write(buf, binary.BigEndian, st.Ino); err != nil {
    62  		log.Debugln("Error while writing inode to buffer:", err)
    63  		return "", 0, err
    64  	}
    65  	if err := binary.Write(buf, binary.BigEndian, st.Ctimespec.Sec); err != nil {
    66  		log.Debugln("Error while writing ctime to buffer:", err)
    67  		return "", 0, err
    68  	}
    69  	if err := binary.Write(buf, binary.BigEndian, st.Ctimespec.Nsec); err != nil {
    70  		log.Debugln("Error while writing ctime nanoseconds to buffer:", err)
    71  		return "", 0, err
    72  	}
    73  	if err := binary.Write(buf, binary.BigEndian, time.Now().Unix()/86400); err != nil {
    74  		log.Debugln("Error while writing current Unix time to buffer:", err)
    75  		return "", 0, err
    76  	}
    77  
    78  	digest := sha256.New()
    79  	digest.Write([]byte(filePath))
    80  	digest.Write(buf.Bytes())
    81  
    82  	return fmt.Sprintf("%x", digest.Sum(nil)), uint64(st.Size), nil
    83  }