For example, turn this Star Wars SWAPI.co data
{
"name": "Luke Skywalker",
"height": "172",
"mass": "77",
"hair_color": "blond",
"skin_color": "fair",
"eye_color": "blue",
"birth_year": "19BBY",
"gender": "male",
"homeworld": "https://swapi.co/api/planets/1/",
"films": [
"https://swapi.co/api/films/2/",
"https://swapi.co/api/films/6/",
"https://swapi.co/api/films/3/",
"https://swapi.co/api/films/1/",
"https://swapi.co/api/films/7/"
],
"species": [
"https://swapi.co/api/species/1/"
],
"vehicles": [
"https://swapi.co/api/vehicles/14/",
"https://swapi.co/api/vehicles/30/"
],
"starships": [
"https://swapi.co/api/starships/12/",
"https://swapi.co/api/starships/22/"
],
"created": "2014-12-09T13:50:51.644000Z",
"edited": "2014-12-20T21:17:56.891000Z",
"url": "https://swapi.co/api/people/1/"
}
into this:
type people {
name: String
height: String
mass: Date
hair_color: String
skin_color: String
eye_color: String
birth_year: String
gender: String
homeworld: Id
films: [Id]
species: [Id]
vehicles: [Id]
starships: [Id]
created: Date
edited: Date
url: Id
}
The mongo-graphql-starter project "Creates a fully functioning, performant GraphQL endpoint from an existing MongoDB". But it needs some setup work to define your tables.
You can do most of the gruntwork setting up the projectSetupA.js file using this:
>j2gt --uri=mongodb://user:password@host:port/database?options > someOutputFile
You'll then need to
- add table: "abc..." and explicit fields: { ... }
- do some editing and find and replace for some of the values.
> npm install -g json-to-graphql-typesystem
Creates a global CLI app named j2gt (see package.json). Optionally, for more control
> git clone https://github.com/MorganConrad/json-to-graphql-typesystem.git
> cd json-to-graphql-typesystem
> npm install
(showing a mix of j2gt and node cli.js)
> curl https://swapi.co/api/people/1 | j2gt --id=people -
results shown above
> node cli.js --uri=mongodb://user:password@host:port/database?options
converts all collections of that database
> node cli.js file1.json file2.json (try ./tests/data/nested.json)
converts the given json files
[--outdir=dir] put results into dir, which **must exist**
[--outext=.ext] put results into files ending with .ext
(if neither, results go to stdout)
[--clean] try to cleanup input files (takes first {...} from an array of results)
[--suffix=!] add a ! after every non-null field
[--nullData=xxx] the "type" to use if the example data is null. default = "TBD"
[--id=typename] use this for the root "type". Otherwise, filename or collection name will be used
// two input possibilities: files, or a JSON API url
file1.json file2.json... json files to parse and generate schemas
'-' means read from stdin
[--uri=mongodb://...] mongo DB (reads first document of collections)
coll1 coll2... (optional) names of collections to use (if none provided, will parse all)
The following options only apply if the files were exported by something like by Mongo DB Compass. See tests/data/akc.json for an example.
These files contain additional typing details such as {"$date":{"$numberLong":"1397952000000"}}
.
- --BSON use the standard conversions (see json-to-graphql-typesystem.BSON_CONVERSIONS)
- --BSONFile=name use conversions read from this JSON file
- --clean Since the exports are not in proper JSON format, you must correct them
Very Quick and Dirty example code in browser_example.html
Module that does the actual conversion of an object into a string graphql type representation.
- use this if the command line app doesn't meet your needs
- userOptions described below
- userBSON additional BSON types
- data: Javascript object (non-null)
- rootType the name of the Type for this root of this object (e.g. used by the --id=xxx option)
Returns a nicely formatted String, possibly preceeded by nested types, e.g.
type rootType_belongs_to_collection {
id: Int
name: String
poster_path: String
backdrop_path: String
}
type rootType_genres {
id: Int
name: String
}
type rootType {
adult: Boolean
backdrop_path: String
belongs_to_collection: rootType_belongs_to_collection
budget: Int
genres: [rootType_genres]
}
- BSON use standard BSON types
- bson_prefix additional prefix for any BSON types, default = 'BSON_'
- eol default '\n',
- nestedDelimiter how to delimit nested classes, default '_' (see example results above)
- nullData "type" to return if data is null, default = 'TBD'
- suffix default '', add this after every non-null type (usually '!')
- If the example data is null, this program (obviously) cannot determine a type. By default, the type will be set to TBD and hand-cleanup is required.
- v0.1.1 added the --header option.
- v0.1.2
- added nullData test and option
- added --suffix
- v0.2.0 reorg, added main_helper for easier testing
- v0.3.0
- remove URL option, but support reading from stdin instead
- main.js -> cli.js, add bin entry to package.json
- Try converting data from the GitHub issue API
curl https://api.github.com/repos/MorganConrad/json-to-graphql-typesystem/issues?state=closed | j2gt --id=issues --clean -
The result should look like ./tests/data/github_issues.graphql