Where’s Waldo?

So far, your room mate and you have created a copy of the DataLad-101 dataset on the same file system but a different place by installing it from a path.

You have observed that the -r/--recursive option needs to be given to datalad install in order to install subdatasets right away. Only then is the subdatasets file content availability metadata to explore the file hierarchy available within the subdataset right from the start. Alternatively, a subsequent datalad install in the subdataset or with a path to the subdataset takes care of the missing installation.

And you have mesmerized your room mate by showing him how Git-annex retrieved large file contents from the original dataset.

Let’s now see the git annex whereis command in more detail, and find out how Git-annex knows where file content can be obtained from. Within the original DataLad-101 dataset, you retrieved some of the .mp3 files via datalad get, but not others. How will this influence the output of git annex whereis, you wonder?

Together with your room mate, you decide to find out. You navigate back into the installed copy, and run git annex whereis on a file that you once retrieved file content for, and on a file that you did not yet retrieve file content for. Here is the output for the retrieved file:

# navigate back into the installed copy of DataLad-101
$ cd ../mock_user/DataLad-101
# navigate into the subdirectory
$ cd recordings/longnow
# file content exists in original DataLad-101 for this file
$ git annex whereis Long_Now__Seminars_About_Long_term_Thinking/2003_11_15__Brian_Eno__The_Long_Now.mp3
whereis Long_Now__Seminars_About_Long_term_Thinking/2003_11_15__Brian_Eno__The_Long_Now.mp3 (3 copies) 
  	00000000-0000-0000-0000-000000000001 -- web
	11ec6b34-3c78-4e53-a38b-5286feb14b49 -- me@muninn:~/dl-101/DataLad-101/recordings/longnow [origin]
   	da3bf937-5bd2-43ea-a07b-bcbe71f3b875 -- mih@medusa:/tmp/seminars-on-longterm-thinking

  web: http://podcast.longnow.org/salt/redirect/salt-020031114-eno-podcast.mp3
ok

And here is the output for a file that you did not yet retrieve content for in your original DataLad-101 dataset.

# but not for this:
$ git annex whereis Long_Now__Seminars_About_Long_term_Thinking/2005_01_15__James_Carse__Religious_War_In_Light_of_the_Infinite_Game.mp3
whereis Long_Now__Seminars_About_Long_term_Thinking/2005_01_15__James_Carse__Religious_War_In_Light_of_the_Infinite_Game.mp3 (2 copies) 
  	00000000-0000-0000-0000-000000000001 -- web
   	da3bf937-5bd2-43ea-a07b-bcbe71f3b875 -- mih@medusa:/tmp/seminars-on-longterm-thinking

  web: http://podcast.longnow.org/salt/redirect/salt-020050114-carse-podcast.mp3
ok

As you can see, the file content previously downloaded with a datalad get has a third source, your original dataset on your computer. The file we did not yet retrieve in the original dataset only has only two sources.

Let’s see how this affects a datalad get:

# get the first file
$ datalad get Long_Now__Seminars_About_Long_term_Thinking/2003_11_15__Brian_Eno__The_Long_Now.mp3
get(ok): Long_Now__Seminars_About_Long_term_Thinking/2003_11_15__Brian_Eno__The_Long_Now.mp3 (file) [from origin...]
# get the second file
$ datalad get Long_Now__Seminars_About_Long_term_Thinking/2005_01_15__James_Carse__Religious_War_In_Light_of_the_Infinite_Game.mp3
get(ok): Long_Now__Seminars_About_Long_term_Thinking/2005_01_15__James_Carse__Religious_War_In_Light_of_the_Infinite_Game.mp3 (file) [from web...]

The most important thing to note is: It worked in both cases, regardless of whether the original DataLad-101 dataset contained the file content or not.

We can see that Git-annex used two different sources to retrieve the content from, though, if we look at the very end of the get summary. The first file was retrieved “from origin...”. Origin is Git terminology for “from where the dataset was copied from” – origin therefore is the original DataLad-101 dataset.

The second file was retrieved “from web...”, and thus from a different source. This source is called web because it actually is a URL through which this particular podcast-episode is made available in the first place. You might also have noticed that the download from web took longer than the retrieval from the directory on the same file system. But we will get into the details of this type of content source once we cover the importfeed and add-url functions1.

Let’s for now add a note on the git annex whereis command. Again, do this in the original DataLad-101 directory, and do not forget to save it.

# navigate back:
$ cd ../../../../DataLad-101

# write the note
$ cat << EOT >> notes.txt
The command "git annex whereis PATH" lists the repositories that have
the file content of an annexed file. When using "datalad get" to retrieve
file content, those repositories will be queried.

EOT
$ datalad status
 modified: notes.txt (file)
$ datalad save -m "add note on git annex whereis"
add(ok): notes.txt (file)
save(ok): . (dataset)
action summary:
  add (ok: 1)
  save (ok: 1)

Footnotes

1

Maybe you wonder what the location mih@medusa is. It is a copy of the data on an account belonging to user mih on the host name medusa. Because we do not have the host names’ address, nor log-in credentials for this user, we can not retrieve content from this location. However, somebody else (for example the user mih) could.