Not so minor adjustments to JSON

[alexdmccabe]

I don't know how to update the .obgjen file reliably - I made the changes by writing a script that recurses over the data and changes the keys. Simple find/replace changes some of the values, too.

Any ideas?

[alexdmccabe]

Also updated the question types to use underscores, and changed the booleans to proper booleans instead of strings.

[hestenet]

Oh right, I guess I did use objen now that you mention it - that's an online JSON formatter here: http://www.objgen.com/json (and I just saved the input format to a file for reference)

[alexdmccabe]

Updated the objgen file to print out the same JSON I created.

[alexdmccabe]

I think the choices could also be simplified to a simple key -> value, for example, with age:

category_age
  id = age
  description = The age bracket of the respondent.
  questions
    how_old
      type = select_one
      self_identify b = false
      prefer_not_to_answer b = true
      does_not_apply b = true
      prompt = How old are you?
      choices
        full_set
          0-15 = 0 - 15
          15-19 = 15 - 19
          20-24 = 20 - 24
          25-29 = 25 - 29
          30-39 = 30 - 39
          40-49 = 40 - 49
          50-49 = 50 - 59
          60-69 = 60 - 69
          70-79 = 70 - 79
          80+ = 80+
          prefer-not-to-answer = Prefer not to answer
        short_set
          0-19 = 0 - 19
          20-39 = 20 - 39
          40-60 = 40 - 60
          61+ = 61+
          prefer-not-to-answer = Prefer not to answer

which produces

{
  "category_age": {
    "id": "age",
    "description": "The age bracket of the respondent.",
    "questions": {
      "how_old": {
        "type": "select_one",
        "self_identify": false,
        "prefer_not_to_answer": true,
        "does_not_apply": true,
        "prompt": "How old are you?",
        "choices": {
          "full_set": {
            "0-15": "0 - 15",
            "15-19": "15 - 19",
            "20-24": "20 - 24",
            "25-29": "25 - 29",
            "30-39": "30 - 39",
            "40-49": "40 - 49",
            "50-49": "50 - 59",
            "60-69": "60 - 69",
            "70-79": "70 - 79",
            "80+": "80+",
            "prefer-not-to-answer": "Prefer not to answer"
          },
          "short_set": {
            "0-19": "0 - 19",
            "20-39": "20 - 39",
            "40-60": "40 - 60",
            "61+": "61+",
            "prefer-not-to-answer": "Prefer not to answer"
          }
        }
      }
    }
  },

instead of this style:

category_age
  id = age
  description = The age bracket of the respondent.
  questions
    how_old
      type = select_one
      self_identify b = false
      prefer_not_to_answer b = true
      does_not_apply b = true
      prompt = How old are you?
      choices 
        full_set [0]
          id = 0-15
          value = 0-15
          label = 0 - 15
        full_set [1]
          id = 15-19
          value = 15-19
          label = 15 - 19
        full_set [2]
           id = 20-24
           value = 20-24
           label = 20 - 24
        full_set [3]
           id = 25-29
           value = 25-29
           label = 25 - 29
        full_set [4]
           id = 30-39
           value = 30-39
           label = 30 - 39
        full_set [5]
           id = 40-49
           value = 40-49
           label = 40 - 49
        full_set [6]
           id = 50-59
           value = 50-59
           label = 50 - 59
        full_set [7]
           id = 60-69
           value = 60-69
           label = 60 - 69
        full_set [8]
           id = 70-79
           value = 70-79
           label = 70 - 79
        full_set [9]
           id = 80+
           value = 80+
           label = 80+
        full_set [10]
           id = prefer-not-to-answer
           value = prefer-not-to-answer
           label = Prefer not to answer
        short_set [0]
          id = 0-19
          value = 0-19
          label = 0 - 19
        short_set [1]
          id = 20-39
          value = 20-39
          label = 20 - 39
        short_set [2]
          id = 40-60
          value = 40-60
          label = 40 - 60
        short_set [3]
          id = 61+
          value = 61+
          label = 61+
        short_set [4]
           id = prefer-not-to-answer
           value = prefer-not-to-answer
           label = Prefer not to answer

which produces

{
  "category_age": {
    "id": "age",
    "description": "The age bracket of the respondent.",
    "questions": {
      "how_old": {
        "type": "select_one",
        "self_identify": false,
        "prefer_not_to_answer": true,
        "does_not_apply": true,
        "prompt": "How old are you?",
        "choices": {
          "full_set": [
            {
              "id": "0-15",
              "value": "0-15",
              "label": "0 - 15"
            },
            {
              "id": "15-19",
              "value": "15-19",
              "label": "15 - 19"
            },
            {
              "id": "20-24",
              "value": "20-24",
              "label": "20 - 24"
            },
            {
              "id": "25-29",
              "value": "25-29",
              "label": "25 - 29"
            },
            {
              "id": "30-39",
              "value": "30-39",
              "label": "30 - 39"
            },
            {
              "id": "40-49",
              "value": "40-49",
              "label": "40 - 49"
            },
            {
              "id": "50-59",
              "value": "50-59",
              "label": "50 - 59"
            },
            {
              "id": "60-69",
              "value": "60-69",
              "label": "60 - 69"
            },
            {
              "id": "70-79",
              "value": "70-79",
              "label": "70 - 79"
            },
            {
              "id": "80+",
              "value": "80+",
              "label": "80+"
            },
            {
              "id": "prefer-not-to-answer",
              "value": "prefer-not-to-answer",
              "label": "Prefer not to answer"
            }
          ],
          "short_set": [
            {
              "id": "0-19",
              "value": "0-19",
              "label": "0 - 19"
            },
            {
              "id": "20-39",
              "value": "20-39",
              "label": "20 - 39"
            },
            {
              "id": "40-60",
              "value": "40-60",
              "label": "40 - 60"
            },
            {
              "id": "61+",
              "value": "61+",
              "label": "61+"
            },
            {
              "id": "prefer-not-to-answer",
              "value": "prefer-not-to-answer",
              "label": "Prefer not to answer"
            }
          ]
        }
      }
    }
  },

[alexdmccabe]

I also removed the "self-identify" choices from the choice lists because that should be coming from the self_identify property of the question, not a choice.

[alexdmccabe]

Initial implementation with these changes in D7: https://cgit.drupalcode.org/open_demographics/tree/

Needs a lot of work still, but it's a strong start, I think.

[alexdmccabe]

I also removed the "self-identify" choices from the choice lists because that should be coming from the self_identify property of the question, not a choice.

We should probably do the same thing with the prefer_not_to_answer property (or do it in reverse and get rid of the property in favor of the question choice). And on that note, what is the does_not_apply property supposed to mean?

[hestenet]

I think does_not_apply may not make sense here - we'll already be able to control whether fields are required are not on the implementation level - so it seems like that should be removed.

Agree with inverting the the prefer_not_to_answer logic in the same way you changed the self-identify property.

[bookworm2000]

Looks good to me.