Zipspy is a CLI tool to extract files from zip archives in S3 without needing to download the entire archive
Zipspy allows you to search for and download files from remote archives quicker. For example, imagine you have a 10MB file (compressed) within a 10GB zip archive in AWS S3. Instead of downloading the entire 10GB file, extracting your file, and deleting the excess, Zipspy downloads only the 10MB compressed file you care about (plus 1-65K for the central directory).
To install the Zipspy, run the following command:
$ go install github.com/alec-rabold/zipspy
You may check that it's installed correctly by running:
$ zipspy -v
zipspy version db00fe2
Zipspy currently supports reading from two storage locations:
- AWS S3 Bucket (
s3:https://) - Local File on Disk (
file:https://) [note: mainly for development]
The underlying providers for each are determined by the protocol specified in the global, required flag --location.
For example, an S3 location may look like "s3:https://my-bucket/archive.zip" while a local file location would look like file:https://path/to/archive.zip.
For S3, all AWS configuration will be read from your environment through the shared config functionality.
To see all available commands, simply type zipspy:
$ zipspy
____ __ ____ ____ ____ _ _
(__ )( )( _ \/ ___)( _ \( \/ )
/ _/ )( ) __/\___ \ ) __/ ) /
(____)(__)(__) (____/(__) (__/
Zipspy allows you interact with ZIP archives stored in remote locations without
requiring a local copy. For example, you can list the filenames in an S3 ZIP archive,
download a subset of files, search and retrieve files with regular expressions, and more!
Usage:
zipspy [command]
Available Commands:
completion Generate the autocompletion script for the specified shell
extract Extract one or more files from the zip archive.
help Help about any command
list List all file names from a zip archive.
Flags:
--development whether or not to use development settings
-h, --help help for zipspy
--location string (required) protocol and address of your ZIP archive ("file:https://archive.zip", "s3:https://<bucket_name>/archive.zip")
--verbosity string global log level (trace, debug, info, warn, error, fatal, panic) (default "warning")
-v, --version version for zipspy
Use "zipspy [command] --help" for more information about a command.
As indicated, you may also use zipspy [command] --help for more information about a command:
$ zipspy list --help
Prints out the names of all files contained within a zip archive.
Usage:
zipspy list [--include-directory-names] [flags]
Flags:
-h, --help help for list
--include-directory-names (optional) include the leaf names of directories
--no-newlines (optional) omit the newlines appended to file names
-o, --out string (optional) name of a file to write output to
--separator string (optional) separator when combining the output of multiple file names
Global Flags:
--development whether or not to use development settings
--location string (required) protocol and address of your ZIP archive ("file:https://archive.zip", "s3:https://<bucket_name>/archive.zip")
Suppose you have a zip file named archive.zip at the top level of an AWS S3 bucket called my-bucket.
The contents of archive.zip are as follows:
test/
├── important.txt
├── path
│ ├── bin
│ │ └── program
│ └── to
│ ├── file.txt
│ └── movie.mp4
└── test.txt
To list all the file names, run the following command:
$ zipspy list --location "s3:https://my-bucket/archive.zip"
archive/important.txt
archive/path/bin/program
archive/path/to/file.txt
archive/path/to/movie.mp4
archive/test.txt
You may choose to include bare directory names by using the --include-directory-names flag:
$ zipspy list --location "s3:https://my-bucket/archive.zip" --include-directory-names
test/
test/important.txt
test/path/
test/path/bin/
test/path/bin/program
test/path/to/
test/path/to/file.txt
test/path/to/movie.mp4
test/test.txt
To extract a particular file, use the extract command:
$ zipspy extract --location "s3:https://my-bucket/archive.zip" -f "archive/important.txt"
Contents of important document.You extract multiple files simultaneously, separated by a newline by default:
$ zipspy extract --location "s3:https://my-bucket/archive.zip" -f "archive/important.txt" -f "archive/path/to/file.txt"
Contents of important document.
Notes from file.If the number of files provided does not match the number of files found, a warning will be output:
$ zipspy extract --location "s3:https://my-bucket/archive.zip" -f "archive/important.txt" -f "archive/DNE.jpg"
WARN[0000] number of input files does not match number of found files (input: 2) (found: 1)
Contents of important document.You may silence these warning with the --silence-warnings flag.
To use a custom separator, use the --separator flag:
$ zipspy extract --location "s3:https://my-bucket/archive.zip" -f "archive/important.txt" -f "archive/path/to/file.txt" --separator "---"
Contents of important document.
---
Notes from file.Zipspy uses a plugin-based architecture. A plugin must simply satisfy the zipspy.Reader interface:
type Reader interface {
io.ReaderAt
Size() (int64, error)
}For remote locations, it's preferable to use HTTP Range Requests where possible.
While this will likely produce a greater number of requests, the target consumers for zipspy will benefit from substantially greater speed and lower network consumption.