devstark-com / open-crud-parser

OpenCRUD Parser

Library for parsing GraphQL queries builded by OpenCRUD standard. Standard may be found here: https://www.opencrud.org/

Installation

npm i git+https://github.com/devstark-com/open-crud-parser.git

Usage

Assume you have similar where and orderBy queries:

queryArgs = {
  where1: {
    AND: [{ name: "parrot" }, { id_in: [1, 2] }]
    slug_in: ["slug-b", "slug-c", "slug-d", "slug-e"]
    name_not: "No such name"
    name_gte: "A"
    birds_every: {
      slug_not_in: ["slug-a"]
      name_gte: "A"
    },
    insect_is_null: true
  },
  where2: {
    OR: [
      { type_not_in: ["mammal", "fish"] },
      { AND: [{ id: 1 }, { name: "pigeon" }] }
    ]
    AND: [{ name: "parrot" }, { id_in: [1, 2] }]
    slug_in: ["slug-b", "slug-c", "slug-d", "slug-e"]
    name_not: "No such name"
    name_gte: "A"
    birds_every: {
      slug_not_in: ["slug-a"]
      name_gte: "A"
    }
  },
  orderBy: name_desc
}

an query will be parsed to:

const openCrudParser = require('open-crud-parser')(['insects', 'birds'])

const formattedQuery = formatQuery(query.where1)

formattedQuery === [{
  slug: [{
    value: ['slug-b', 'slug-c', 'slug-d', 'slug-e'],
    negation: false,
    modifier: 'in'
  }],
  name: [{
      value: 'No such name',
      negation: true,
      modifier: 'equals'
    },
    {
      value: 'A',
      negation: false,
      modifier: 'gte'
    },
    {
      value: 'parrot',
      negation: false,
      modifier: 'equals'
    }
  ],
  id: [{
    value: [1, 2],
    negation: false,
    modifier: 'in'
  }],
  related: {
    birds: {
      value: [{
        'birds.slug': [{
          value: ['slug-a'],
          negation: true,
          modifier: 'in'
        }],
        'birds.name': [{
          value: 'A',
          negation: false,
          modifier: 'gte'
        }]
      }],
      modifier: 'every'
    },
    insect: {
      value: [{}],
      modifier: 'is_null'
    }
  }
}]

const formattedQuery = formatQuery(query.where2)

formattedQuery === [{
    slug: [{
      value: ['slug-b', 'slug-c', 'slug-d', 'slug-e'],
      negation: false,
      modifier: 'in'
    }],
    name: [{
        value: 'No such name',
        negation: true,
        modifier: 'equals'
      },
      {
        value: 'A',
        negation: false,
        modifier: 'gte'
      },
      {
        value: 'parrot',
        negation: false,
        modifier: 'equals'
      }
    ],
    type: [{
      value: ['mammal', 'fish'],
      negation: true,
      modifier: 'in'
    }],
    id: [{
      value: [1, 2],
      negation: false,
      modifier: 'in'
    }],
    related: {
      birds: {
        value: [{
          'birds.slug': [{
            value: ['slug-a'],
            negation: true,
            modifier: 'in'
          }],
          'birds.name': [{
            value: 'A',
            negation: false,
            modifier: 'gte'
          }]
        }],
        modifier: 'every'
      }
    }
  },
  {
    slug: [{
      value: ['slug-b', 'slug-c', 'slug-d', 'slug-e'],
      negation: false,
      modifier: 'in'
    }],
    name: [{
        value: 'No such name',
        negation: true,
        modifier: 'equals'
      },
      {
        value: 'A',
        negation: false,
        modifier: 'gte'
      },
      {
        value: 'pigeon',
        negation: false,
        modifier: 'equals'
      }
    ],
    id: [{
      value: 1,
      negation: false,
      modifier: 'equals'
    }],
    related: {
      birds: {
        value: [{
          'birds.slug': [{
            value: ['slug-a'],
            negation: true,
            modifier: 'in'
          }],
          'birds.name': [{
            value: 'A',
            negation: false,
            modifier: 'gte'
          }]
        }],
        modifier: 'every'
      }
    }
  }
]



const formattedOrderBy = formatOrderBy(query.orderBy)

formattedOrderBy === [{ 
  column: 'name',
  order: 'desc'
}]

How it works

formatWhere creates new array element based on top level query arguments plus each element contained in OR statement(so in the result will as much elements as in OR statement). If there is no OR statement in result array will be only one element

This library parses your schema field names by following scheme:

# formatWhere
name_not_gte: String

[name-itself]_[negation?]_[modifier]

# formatOrderBy
name_desc

[column]_[order]