zipit | Backup Utility for Cloud Sites | Continuous Backup library

The IAM role associated to the function is not allowed to upload to S3.

The solution is to create an IAM policy allowing S3 access to your bucket (say read/write), which would look something like:

You need to provide the entire dependency tree, scrapy also has a set of dependencies (and they may also have dependencies).

The easiest way to download all the required dependencies is to use pip

You can use subprocess calling 'zip' and passing the paths as arguments

You read the buffer before you've closed the zip.Writer, so any final data is needs to flush to the buffer has gotten lost. You should remove the defer archive.Close() and instead close the archive before getting the bytes out of the buffer. e.g.

Try this:

First of all it's impossible for the files to be identical and the interpreter program to give different results. You're not writing the whole history in your description.

1. How can I check if the file is corrupted ?

The answer is in the internal tool. You should improve its debug message: "Error while retrieving archive entry". What error ? Debug your old tool. And it's suggesting the problem is in the zip file, not the json file.

1. Should I try to use other API to check it? if yes, which ?

Yes, unzip both files by hand and diff the output.

1. Any other direction ?

Read the code for your old java tool.

It looks like you need to skip adding directories to the zip file header, every time you see a directory in the filepath.Walk callback just skip it:

Assuming you are calling your function as follows:

Names of files and directories in zip archives come from the zip.FileHeader.

Your code already initializes the header from the os.FileInfo. This is important as it populates metadata such as timestamps and uncompressed size.

Furthermore, your code seems to be doing the following:

If invoked with: