/anon-files

Primary LanguageClojureOtherNOASSERTION

WARNING

This project is deprecated. The functionality was moved into the kifshare service, which in turn is mostly not used anymore as WebDAV is set up in most installations to allow anonymous access to iRODS.

anon-files

A Discovery Environment service that serves up files that have been shared with the anonymous user in iRODS.

Build

docker build --rm -t discoenv/anon-files:dev .

Configure

anon-files's configuration file must be in the properties file format. Here's an unconfigured anon-files properties file:

irods-host =
irods-port =
irods-user =
irods-password =
irods-zone =
irods-home =
anon-user = anonymous
port = 60000
log-file =

The Dockerfile is expecting anon-files to be listening on port 60000, so it's not recommended to place something different in the config file.

Here are the command-line options:

A service that serves up files shared with the iRODS anonymous user.

Usage: anon-files [options]

Options:
  -p, --port PORT                                       Port number
  -c, --config PATH      /etc/iplant/de/anon-files.edn  Path to the config file
  -v, --version                                         Print out the version number.
  -h, --help

Run

docker run -P -d --name anon-files -v /path/to/config:/etc/iplant/de/anon-files.properties discoenv/anon-files

Use docker ps to see which random port anon-files is listening on.

Downloading files

anon-files supports downloading files either all at once or by byte ranges. The caller must already know the path to the file inside iRODS; the path in iRODS corresponds to the path in the URI.

anon-files does not currently support request pre-conditions or multiple byte ranges per request.

If the lower bound on a request is greater than the file's size, anon-files will return a 416 status code as defined in http://tools.ietf.org/html/rfc7233.

If the upper bound on a range request is greater than the file size, then the returned data will go up to the end of the file and have a status of 206.

If a range request consists of a single negative value, then it is used as a negative index into the content and the range is interpreted as being from that point to the end of the content.

curl -H "Range: bytes=-10" http://localhost:8080/path/to/file

If a range request consists of a single positive value, then the range header is ignored and the entire content is downloaded. A single value is not enough information to determine what is wanted from the client, but the RFC implies that it shouldn't be treated as an error (unless I'm missing something, which is likely). To download a single byte from the content, perform a request like the following:

curl -H "Range: bytes=10-10" http://localhost:8080/path/to/file

A successful ranged request will return the requested byte range and have a 206 status code. A requested range that is unsatisfiable (for instance, the lower bound is higher than the file size) will have a status code of 416.

Byte range requests will have the following HTTP headers in the response:

License

See LICENSE