Converting csv to json with jq

I'm currently practicing converting plain text to csv, and them csv to json. I need to do the conversion only with jq and basic Linux commands.

I didn't have any major problems converting plain text to csv, but further conversion is a bit tricky. Primarily because the first line and the last one are out of standard.

Here is my current csv:

Asserts Samples
1;expecting command finishes successfully (bash exe);7ms
1;expecting command finishes successfully (loading bin, execute);27ms
0;expecting command fails (bash exe, execute);23ms
...
0;expecting command prints some message (bash exe);12ms
0;expecting command prints some message (loading bin, execute);26ms
5;2;71.43;136ms

I need to convert this csv to the following format:

{
 "testName": "Asserts Samples",
 "tests": [
   {
     "name": "expecting command finishes successfully (bash exe)",
     "status": false,
     "duration": "7ms"
   },
...
   {
     "name": "expecting command prints some message (loading bin, execute)",
     "status": true,
     "duration": "26ms"
   }
 ],
 "summary": {
   "success": 5,
   "failed": 2,
   "rating": 71.43,
   "duration": "136ms"
 }
}

So far I have figured out:

How to add names to values. map({"name": .[1],...
How to replace string with boolean. select(.status == "0")).status |= true
How to make split with one separator. (split(";")

Long attempts have failed to make a json structure with a corresponding arrays and objects structure. I would be glad for advice or guidance.

Solution

You can use jq with raw input option (-R) and -s to slurp the entire file as a single string. Then you split that into array of lines. You access test lines by $lines[1:-1] (we skip 1-st row) and for each test line, split it by ";" and map the fields to the desired keys. Since in your data "0" represents a successful test (status true), you can convert the status using (.[0] == "0"). Then deal with summary line (use -1 index to get it regardles of line count) and you should be good.

Combining the above should give you this monster:

jq -R -s '
  (split("\n") | map(select(length > 0))) as $lines |
  {
    testName: $lines[0],
    tests: ($lines[1:-1] | map(
      (split(";") | {
        name: .[1],
        status: (.[0] == "0"),
        duration: .[2]
      })
    )),
    summary: ($lines[-1] | split(";") | {
      success: (.[0] | tonumber),
      failed: (.[1] | tonumber),
      rating: (.[2] | tonumber),
      duration: .[3]
    })
  }
' input.csv

Once run on your input data output should be

{
  "testName": "Asserts Samples",
  "tests": [
    {
      "name": "expecting command finishes successfully (bash exe)",
      "status": false,
      "duration": "7ms"
    },
    {
      "name": "expecting command finishes successfully (loading bin, execute)",
      "status": false,
      "duration": "27ms"
    },
    {
      "name": "expecting command fails (bash exe, execute)",
      "status": true,
      "duration": "23ms"
    },
    {
      "name": null,
      "status": false,
      "duration": null
    },
    {
      "name": "expecting command prints some message (bash exe)",
      "status": true,
      "duration": "12ms"
    },
    {
      "name": "expecting command prints some message (loading bin, execute)",
      "status": true,
      "duration": "26ms"
    }
  ],
  "summary": {
    "success": 5,
    "failed": 2,
    "rating": 71.43,
    "duration": "136ms"
  }
}