A command line replacement of NBIA data retriever. Forked from ygidtu/NBIA_data_retriever_CLI.
Since the scripts are written in go, some dependencies need to be installed before running build.sh
:
sudo apt install golang-go
go get -v github.com/voxelbrain/goptions
go get -v github.com/rs/zerolog/log
After running build.sh
, three files (nbia_cli_linux_amd64
, nbia_cli_darwin
and nbia_cli_win64.exe
) are built in the directory. Assuming one wants to run the NBIA command line data retriever on linux, after the previous step one can go on and simply run nbia_cli_linux_amd64
with the syntax that follows (see the "Command Line Usage" section for details):
./nbia_cli_linux_amd64 --input ../data/path_to_manifest/NBIA-manifest-xxxxxxxxxxx.tcia --output ../data/path_to_output/
Usage: nbia_cli_xxxx [global options]
Global options:
-i, --input Path to tcia file
-o, --output Output directory, or output file when --meta enabled (default: downloads)
-x, --proxy Proxy
-t, --timeout Due to limitation of target server, please set this time out value as big as possible (default: 1200000)
-p, --process Start how many download at same time (default: 1)
-m, --meta Get Meta info of all files
-v, --version Show version
--debug Show debug info
--help Show this help
The aforementioned TCIA (NBIA-manifest-xxxxxxxxx.tcia
) file can be obtained directly from the dataset page on The TCIA website.
Known issues:
- The
public.cancerimagingarchive.net/nbia-download/servlet
usePOST
to transfer data from server to local, the connection may be terminated even before the download is complete. Therefore, PLEASE set timeout as huge as possible - progress bar is a mess when using multiple process
- I do not have a account of NBIA, therefore this program could not handle the restricted data for now.