veidenberg/pline

JSON-based web interfaces for command-line programs

pline's Issues

STDERR verbosity of the task window expanded by default

While one of the Pline tasks is running, I can see the last line of the prompt being shown. If I click on the single-line prompt of the running task, the window expands and shows me everything printed in the prompt (STDERR/STDOUT), sometimes going beyond the bottom of the browser window.

Is there a way to set the height of the prompt window for a specific task? E.g. if we want the task window to show the 10 latest lines of the prompt, or 500 pixels, is there a way to set that up? Thanks.

'Drop file here' vs 'Select' behaviour

My "file" type options always show a Drop file here box with a Select button in them.

The drop file behaviour works perfectly, but the Select button doesn't work in my Pline (same in the Pline demo site): the menu of the equivalent 'File explorer' succesfully opens, but after selecting the file and clicking 'Open', the Pline box remains empty.

I've tried with the following combinations:

Chrome / Firefox
Windows 10 / Linux Ubuntu
Safari / OSX

If I try the same with the wasabi website, the Select menu e.g. for Import does work.

But it doesn't in the Pline demo site: http://wasabiapp.org/pline-demo/ . E.g. I can try 'MAFFT' then click Select on the 'Sequences File' and it won't work.

Hopefully the fact that it is reproducible in the Pline demo site means you can have a look on your end,

Thanks in advance

reboot server

Hi,

Another small one that may be very obvious to a more experience javascript user:

Every time I change something in the code, I interrupt the running server, e.g. with Ctrl+c, then restart it. By default, it opens up a Firefox window, which I rather not have opening up (I do a sleep 2 && pkill firefox).

What's the recommended way of rebooting/restarting the server during code development? Is there a way for Firefox not to kick in after every reboot?

Also, is there a command-line to restart a second server on the same machine with a different port for the development server, while the "production" server keeps on for the users?

Thanks in advance

job locking mechanism for certain tools

Is there a flag to make sure that only one concurrent job can be executed for a certain tool in a running Pline server?
It is so successful on our end, some jobs trip over each other when different people kick them off, due to the nature of the job.

generate URLs with parameters?

Hi,

Would it be possible for Pline to run on a URL-generated request? E.g.

http://localhost:8000?param1=foo&param2=bar

Thanks in advance,

how to pass two parameters into bash script

Hi,

I have a plugin with two text parameters, and I want to run them in a bash script like so:

mybash.sh

echo $1 and $2

So far it works with $1, but $2 is ignored. Any ideas? Thanks.

wider Tools box width

Some of the tool names I have are rather long, and I was wondering how to make the Tools box width wider, or dynamic to the longest tool name.

What part of the css or config file should I look into? Thanks in advance

order of the tools

Hi, I've tried pline today and I find it a very useful way to enable wet-lab scientists in using tools with a CLI.

Is there a way of reordering the list of tools based on a specific "prefix" or "category"?

I see the plugin for pagan has prefix "--" and category "aligners", but in my list, Codeml appears first, with a prefix "" and no category.

What's the best way to group and then sort based on the preferred grouping? Thanks in advance.

hidden toogle for tool list but still available inside a pipeline

Hi,

I've got a pipeline composed of 3 tools, 1 of which only makes sense as part of the pipeline, and don't want to have listed in the 'Tools' list.

I've added a hidden: "true" in the top-level json, but upon refresh, I can still see the tool listed. Is it possible to hide a tool from the 'Tools' list but still have it available as part of a pipeline?

Thanks in advance,

{
  program: "p -m list_to_vdjc_aggr -stdout -i ",
  name: "list_to_vdjc_aggr",
  desc: "desc",
  hidden: "true",
  outFiles: "outfilename",
  jobName: "list_to_vdjc_aggr results",
  submit: ["Run tests"],
  prefix: "-",
  category: "generic",
  options: [
        {group: "Options", desc: "Hover option labels for description", options: [
            {file: "", desc: "Input file", required: "in csv format", order: "end"},
        ]}
  ]
}

pipeline description

Hi, I've started experimenting with the pipelines that are generated via 'Add step' and 'Import|Export'.

I managed to create a pipeline with two steps, configure it, give it a name and a description, then 'Export'. It produces a beautiful '.json' file that I can copy in their own folder in 'pipelines' (I presume it doesn't need to have its own folder, as it worked without it, right?).

A small question: where does the pipeline description ({"desc": "description here"}) appear in the menu when loaded? I couldn't see it anywhere obvious or hovering the mouse over the pipeline title. Does it need a special flag somewhere?

See example below:

The description doesn't immediately appear in the .json, but I can add it just above the pipeline key:

{
 "name": "csvtk_cut_freq",
 "pipeline": [
  {
   "plugin": "csvtk_cut",
   "name": "csvtk_cut results",
   "inputs": {
    "f": "foo,bar,xyz"
   }
  },
  {
   "plugin": "csvtk_freq",
   "name": "csvtk_freq results",
   "inputs": {
    "f": "foo,bar",
    "trackName1": "_pipe_",
    "trackName1_sel": "_pipe_"
   }
  }
 ]
}

With the description (I took the read mapping pipeline as an example):

{
 "name": "csvtk_cut_freq",
 "desc":"Does this and that",
"pipeline": [
  {
   "plugin": "csvtk_cut",
   "name": "csvtk_cut results",
   "inputs": {
    "f": "foo,bar,xyz"
   }
  },
  {
   "plugin": "csvtk_freq",
   "name": "csvtk_freq results",
   "inputs": {
    "f": "foo,bar",
    "trackName1": "_pipe_",
    "trackName1_sel": "_pipe_"
   }
  }
 ]
}

Example screenshot of the second json: