A simple Node-js library to download Sheet from Google Drive and parse it to structured JSON.
Google developers console is at this address:
https://console.developers.google.com/apis/
- From your Google developer account, go to the
Library
menu (left side) and find the Google Sheet API. - Activate the API by clicking the
ACTIVATE
button.
The library needs one of the following access keys to download the spreadsheet document:
- Service Account Key: a JSON file.
- API Key: Simplier to create and use, but less secure.
- From the same Google developers console, go to the
Credentials
menu. - Clic the
Create credentials
dropdown button and selectService account key
- select
default
as Service account, andJSON
as Key type - Create and download the key
You get a .json
file that contain a private key to access documents in your Google drive.
(Be carrefull with this secret key, don't commit it in your code!)
- From the developer console, go to the
Credentials
menu. - Clic the
Create credentials
dropdown button and selectAPI key
Copy and paste the provided string as the apiKey
parameter. See the "Usage" section below
In the Google sheet you want to export, clic on the SHARE
button (top-right corner) and check that it's shared
as Anyone with the link can view
.
The file id is a 44 characters hash string inside the shareable link.
For instance, consider the spreadsheet at the following address:
https://docs.google.com/spreadsheets/d/1tDyLD2f_P2n9etVESzx-_CJdiH1VgXL8B9VYmEVA6pQ/edit?usp=sharing
its file id is: 1tDyLD2f_P2n9etVESzx-_CJdiH1VgXL8B9VYmEVA6pQ
npm install sync-google-spreadsheet -save
You need to specify two things:
fileId
the file id of the Google sheet you want to download- An access key:
- either:
clientSecretPath
the path to the json file that contain your Google secret key. - or:
apiKey
the API key string
- either:
var syncSpreadsheet = require('sync-google-sheet');
var params = {
fileId: '1tDyLD2f_P2n9etVESzx-_CJdiH1VgXL8B9VYmEVA6pQ',
clientSecretPath: './keys/google-secret.json',
// apiKey: 'XXxxXxXXX_XXxxxXXXxxxXXXXxxXXxxxXXXX',
metaTableName: 'meta' // optional
};
syncSpreadsheet(params, function onResult(error, result) {
if (error) return console.error(error);
console.log(result);
});
The workbook should contain one meta table. The meta table define the sheets to export and how these sheets should be exported. The format of this meta table is as follow:
name | format | key | headerLine |
---|---|---|---|
SheetName | dictionary | id | line number |
available formats are:
array
dictionary
dictionary*
mappedList
mappedList*
keyvalue
arrayvalue
mappedvalue
The key
column is only used for dictionary
and mappedlist
tables, to specify which attribute
should be used for key. If not specified, id
is used as key name.
Keys can be chained (colon separated) to obtain a recursively structured object.
headerLine
is optional, set it if your title line is not the first one (e.g. 2
).
Column defines one attribute of the objects. The two first row are used to define the attribute name and type. The name can be a dot formatted path used to structure the object.
name |
<- property name |
---|---|
type | <- type of the values |
value |
<- value for item 1 |
value |
<- value for item 2 |
value |
<- value for item 3 |
For instance the following table:
id | stats.atk | stats.def | isBoss |
---|---|---|---|
string |
int |
int |
bool |
bouli | 280 | 430 | TRUE |
spiky | 310 | 240 | FALSE |
mekka | 120 | 510 |
Will result in the following JSON array:
[
{ id: 'bouli', stats: { atk: 280, def: 430 }, isBoss: true },
{ id: 'spiky', stats: { atk: 310, def: 240 }, isBoss: false },
{ id: 'mekka', stats: { atk: 120, def: 510 } }
]
or the following JSON dictionary with id
used as key:
{
bouli: { id: 'bouli', stats: { atk: 280, def: 430 }, isBoss: true },
spiky: { id: 'spiky', stats: { atk: 310, def: 240 }, isBoss: false },
mekka: { id: 'mekka', stats: { atk: 120, def: 510 } }
}
If you want the key to not be included in the dictionary entries,
set the sheet type to dictionary*
(with an asterisk *
) in the meta table.
On the same example, the result would become:
{
bouli: { stats: { atk: 280, def: 430 }, isBoss: true },
spiky: { stats: { atk: 310, def: 240 }, isBoss: false },
mekka: { stats: { atk: 120, def: 510 } }
}
A mapped list is a dictionary that maps to an array of items.
Consider the following table where id
is used as key.
id | x | y | width | height |
---|---|---|---|---|
int |
int |
int |
int |
int |
1 | 0 | 0 | 15 | 20 |
1 | 15 | 2 | 12 | 14 |
1 | 17 | 1 | 8 | 6 |
2 | 2 | 13 | 10 | 14 |
3 | 0 | 0 | 11 | 7 |
3 | 5 | 3 | 9 | 5 |
3 | 4 | 8 | 4 | 4 |
3 | 8 | 5 | 7 | 7 |
4 | 0 | 0 | 16 | 11 |
4 | 7 | 2 | 20 | 22 |
The items that share the same key will be grouped in an array:
{
"1": [
{ x: 0, y: 0, width: 15, height: 20 },
{ x: 15, y: 2, width: 12, height: 14 },
{ x: 17, y: 1, width: 8, height: 6 }
],
"2": [
{ x: 2, y: 13, width: 10, height: 14 }
],
"3": [
{ x: 0, y: 0, width: 11, height: 7 },
{ x: 5, y: 3, width: 9, height: 5 },
{ x: 4, y: 8, width: 4, height: 4 },
{ x: 8, y: 5, width: 7, height: 7 }
],
"4": [
{ x: 0, y: 0, width: 16, height: 11 },
{ x: 7, y: 2, width: 20, height: 22 }
]
}
The keyvalue table should define 3 columns:
- key
- type
- value
For instance, the following table:
key | type | value |
---|---|---|
player.speed | float |
1.4 |
player.power | int |
120 |
player.strength | int |
209 |
area | string |
"tutorial" |
entries | array.int |
[3, 5, 6] |
will produce the following object:
{
player: {
speed: 1.4,
power: 120,
strength: 209,
},
area: 'tutorial',
entries: [3, 5, 6]
}
You can define several values column with an identifier. Each value column will be extracted as a separate object.
key | type | value:en | value:fr | value:ja |
---|---|---|---|---|
cookie | string | Cookie | Biscuit | クッキー |
key | string | Key | Clé | 鍵 |
gem | string | Gem | Diamant | 宝石 |
Will produce the following dictionary:
{
en: { cookie: "Cookie", key: "Key", gem: "Gem" },
fr: { cookie: "Biscuit", key: "Clé", gem: "Diamant" },
ja: { cookie: "クッキー", key: "鍵", gem: "宝石" }
}
To structure dictionary
or mappedlist
to more that one depth level, you can chain
the keys in the meta table by appending them together with a colon character :
For instance, given the following table:
world | map | id | name |
---|---|---|---|
string |
string |
int |
string |
Clearwater Harbor | The Flour Tower | 1 | Leoril |
Clearwater Harbor | Valenstrong Mansion | 1 | Viccoril |
Clearwater Harbor | Shipwreck Edge | 1 | Panneak |
Clearwater Harbor | Shipwreck Edge | 2 | Jamcoril |
Clearwater Harbor | The Smelt Belt | 1 | Yenqarim |
Clearwater Harbor | The Smelt Belt | 2 | Dorxiron |
Clearwater Harbor | The Smelt Belt | 3 | Alydove |
Crewth Piers | Valor Brewery | 1 | Wolgwynn |
Crewth Piers | Valor Brewery | 2 | Oldove |
Crewth Piers | Valor Brewery | 3 | Safrila |
Crewth Piers | Riddlecloud | 1 | Colynn |
Crewth Piers | Riddlecloud | 2 | Kyslynn |
Crewth Piers | Narrow Bridge | 1 | Deltheris |
Crewth Piers | Dawnton Castle | 1 | Aroborin |
Crewth Piers | Dawnton Castle | 2 | Sarovar |
Rainbow Ridge Orchard | Ert Grotto | 1 | Dorfinas |
Rainbow Ridge Orchard | Rivershire Rampart | 1 | Brennoa |
Rainbow Ridge Orchard | Rivershire Rampart | 3 | Alyyra |
By chaining the keys world:map:id
in the meta table, we obtain the following structured
JSON object:
{
"Clearwater Harbor": {
"The Flour Tower": {
"1": { name: "Leoril"}
},
"Valenstrong Mansion": {
"1": { name: "Viccoril" }
},
"Shipwreck Edge": {
"1": { name: "Panneak" },
"2": { name: "Jamcoril" }
},
"The Smelt Belt": {
"1": { name: "Yenqarim" },
"2": { name: "Dorxiron" },
"3": { name: "Alydove" }
}
},
"Crewth Piers": {
"Valor Brewery": {
"1": { name: "Wolgwynn" },
"2": { name: "Oldove" },
"3": { name: "Safrila" }
},
"Riddlecloud": {
"1": { name: "Colynn" },
"2": { name: "Kyslynn" }
},
"Narrow Bridge": {
"1": { name: "Deltheris"}
},
"Dawnton Castle": {
"1": { name: "Aroborin" },
"2": { name: "Sarovar" }
}
},
"Rainbow Ridge Orchard": {
"Ert Grotto": {
"1": { name: "Dorfinas" }
},
"Rivershire Rampart": {
"1": { name: "Brennoa" },
"3": { name: "Alyyra" }
}
}
}
The arrayvalue
format is for exporting a single column to an array.
value |
---|
string |
sword |
shield |
ring |
["sword", "shield", "ring"]
Mapped value let you extract arrays of values, grouped by a key. The value column needs to be named value
.
For instance, extrating the following table:
id | value |
---|---|
string |
int |
star | 9 |
star | 12 |
star | 20 |
luna | 3 |
luna | 7 |
Produces the following JSON object:
{
"star": [9, 12, 20],
"luna": [3, 7]
}
The following types can be defined to tell the script how to parse the values in the spreadsheet.
If no type is specified, the column (or row when spreadsheet is defined as keyvalue
) is ignored.
If a value doesn't match with the type defined, an error is returned in the callback.
int
Base 10 integer. Default to0
.float
Floating point number. Default to0
.string
Text string. Default to the empty string.bool
Boolean value. Note that these values are displayed asTRUE
orFALSE
in Google Sheet. If the value is not defined (i.e. empty cell) the attribute is optimized out and removed completely.
array
JSON encoded array. default to an empty array.array.int
JSON encoded array of integer.array.float
JSON encoded array of number.array.string
JSON encoded array of string.array.bool
JSON encoded array of boolean.
For some of the basic types (int
, float
and string
), the following wildcard symbols are available:
?
if the cell is empty, don't add the attribute (no default value)+
allow the value to be an array of this type.*
allow the value to be empty or an array.
json
string that can be parsed as a valid JSON object. If the value is not defined, the attribute gets optimized out.
Reference type let you point data from another sheet that has been extracted (i.e. the sheet needs to be defined in the meta table before where it is referenced). You define which sheet (and optionaly wich attribute) to point in the type field itself. Empty cell are optimized out and attribute removed completely.
ref:<sheet.path>
single referencearray.ref:<sheet.path>
array of references.
Consider the following table, in which one field is a reference to the zone
sheet
we previously extracted:
name | zones | bgm |
---|---|---|
string |
ref:zone |
string |
tuto | wild | |
plain | 1 | wild |
cave | 2 | underground |
mountain | 3 | wild |
volcano | 4 | fire |
Produce the following JSON object:
{
tuto: { name: "tuto", bgm: "wild" },
plain: {
name: "plain",
zones: [
{ x: 0, y: 0, width: 15, height: 20 },
{ x: 15, y: 2, width: 12, height: 14 },
{ x: 17, y: 1, width: 8, height: 6 }
],
bgm: "wild"
},
cave: {
name: "cave",
zones: [
{ x: 2, y: 13, width: 10, height: 14 }
],
bgm: "underground"
},
mountain: {
name: "mountain",
zones: [
{ x: 0, y: 0, width: 11, height: 7 },
{ x: 5, y: 3, width: 9, height: 5 },
{ x: 4, y: 8, width: 4, height: 4 },
{ x: 8, y: 5, width: 7, height: 7 }
],
bgm: "wild"
},
volcano: {
name: "volcano",
zones: [
{ x: 0, y: 0, width: 16, height: 11 },
{ x: 7, y: 2, width: 20, height: 22 }
],
bgm: "fire"
}
}
This tool is based on grille-downloader and grille-xlsx from Daigo Sato.