Possibility of saving {md5} hash into archive.db to avoid downloading duplicate files?

[guzmelek]

Hello everyone!

It seem like I stuck and cannot make gallery-dl to calculate hashes for the downlaoded files. My current config is below. Unfortunately, I am still getting None instead of hash

  "extractor": {
    "reddit": {
      "directory": ["downloads", "{extractor}"],
      "filename": "{id}_{md5}.{extension}",
      "archive": "/tmp/gallery-dl/gallery-dl-archive.db",
      "archive-format": "{md5}",
      "hash": {
        "event": "file",
        "filename": true,
        "hashes": "md5"
      }
    }
  },
  "path-restrict": {
    "?": "_",
    ":": "_",
    "/": "_",
    "\\": "_",
    "<": "_",
    ">": "_",
    "|": "_",
    "*": "_",
    "\"": "_"
  },
  "archive": "/tmp/gallery-dl/gallery-dl-archive.db"
}

./gallery-dl/downloads/None/u0vzud_None.jpg
./gallery-dl/downloads/None/u0px0y_None.jpg
./gallery-dl/downloads/None/u0pprb_None.jpg

$ sqlite3 /tmp/gallery-dl/gallery-dl-archive.db "SELECT * FROM archive;"
redditNone                                                                                                                                                                                                                                      

[mikf]

      "hash": {
        "event": "file",
        "filename": true,
        "hashes": "md5"
      }

This should be

"postprocessors": {
    "name": "hash",
    "event": "file",
    "filename": true,
    "hashes": "md5"
}

but it would still not work with an archive. The archive check happens before a file download, but the MD5 hash can only be computed after a file was downloaded.