allow stdin on Windows

[djdv]

This may be all that has to be done in cmds for now, the rest may be in go-ipfs itself.
Ref: ipfs/kubo#4808

[codecov]

Codecov Report

Merging #74 into master will increase coverage by 0.02%.
The diff coverage is 33.33%.

@@            Coverage Diff             @@
##           master      #74      +/-   ##
==========================================
+ Coverage   43.44%   43.47%   +0.02%     
==========================================
  Files          23       23              
  Lines        1908     1909       +1     
==========================================
+ Hits          829      830       +1     
  Misses        942      942              
  Partials      137      137

Impacted Files	Coverage Δ
cli/parse.go	65.85% <33.33%> (+0.11%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 13acd1a...29d5b1e. Read the comment docs.

[djdv]

FWIW I plan on adding something similar to this in go-ipfs/shareness

  test_expect_success "stdin reading message shows up" '
  case $(uname) in
  MSYS*)
    test_expect_code 1 grep "ipfs: Reading from" add_help_err &&
    test_expect_code 1 grep "send Ctrl-z to stop." add_help_err
    ;;
  *)
    test_expect_code 1 grep "ipfs: Reading from" add_help_err &&
    test_expect_code 1 grep "send Ctrl-d to stop." add_help_err
    ;;
  esac
  '

which should address what codecov is upset over.

[Stebalien]

This may be all that has to be done in cmds for now, the rest may be in go-ipfs itself.

Does this actually fix the issue in question or does something else need to be done to make stdin work on windows?

Do you have any idea why this check was added? I assume there was a reason.

[djdv]

Does this actually fix the issue in question or does something else need to be done to make stdin work on windows?

Partially, see ipfs/kubo#4808 (comment)
stdin+stdout appear to be working as long as the daemon is running, I'll have to investigate go-ipfs to see if something can be done about the lock while the daemon is down. I can't see why it's not released before the next command is issued.

I'm going to bounce back and forth trying to get tests around stdio working on Windows.

Do you have any idea why this check was added? I assume there was a reason.

It's a little unclear unfortunately.
ipfs/kubo#3266

[whyrusleeping]

@djdv any update here? Should we just merge this and push it into go-ipfs?

[djdv]

@whyrusleeping
I would say we merge this now and add handling in go-ipfs for the 1 case that doesn't work. (ipfs|ipfs with daemon down)

Edit: for reference, that issue is not Windows specific.

[Stebalien]

@whyrusleeping @djdv

Let's merge it. I still don't know why we disabled this command on Windows in the past (and not Linux) but we might as well try it.