HTTP JSON API Service¶
The JSON API provides a significantly simpler way to interact with a ledger than the Ledger API by providing basic active contract set functionality:
- creating contracts,
- exercising choices on contracts,
- querying the current active contract set, and
- retrieving all known parties.
The goal of this API is to get your distributed ledger application up and running quickly, so we have deliberately excluded complicating concerns including, but not limited to:
- inspecting transactions,
- asynchronous submit/completion workflows,
- temporal queries (e.g. active contracts as of a certain time), and
For these and other features, use the Ledger API instead.
We welcome feedback about the JSON API on our issue tracker, or on our forum.
Running the JSON API¶
Start a Daml Ledger¶
You can run the JSON API alongside any ledger exposing the gRPC Ledger API you want. If you don’t have an existing ledger, you can start an in-memory sandbox:
daml new my_project --template quickstart-java
cd my_project
daml build
daml sandbox --wall-clock-time --ledgerid MyLedger ./.daml/dist/quickstart-0.0.1.dar
Start the HTTP JSON API Service¶
Basic¶
The most basic way to start the JSON API is with the command:
daml json-api --ledger-host localhost --ledger-port 6865 --http-port 7575
This will start the JSON API on port 7575 and connect it to a ledger running on localhost:6865
.
Note
Your JSON API service should never be exposed to the internet. When running in production the JSON API should be behind a reverse proxy, such as via NGINX.
Standalone JAR¶
The daml json-api
command is great during development since it is
included with the SDK and integrates with daml start
and other
commands. Once you are ready to deploy your application, you can
download the standalone JAR from
Github releases. It is much smaller
than the whole SDK and easier to deploy since it only requires a JVM
but no other dependencies and no installation process. The JAR accepts
exactly the same command line parameters as daml json-api
, so to
start the standalone JAR, you can use the following command:
java -jar http-json-1.5.0.jar --ledger-host localhost --ledger-port 6865 --http-port 7575
Replace the version number 1.5.0
by the version of the SDK you are
using.
With Query Store¶
In production setups, you should configure the JSON API to use a PostgreSQL backend as a cache. The in-memory backend will call the ledger to fetch the entire active contract set for the templates in your query every time so it is generally not recommended to rely on this in production. Note that the PostgreSQL backend acts purely as a cache. It is safe to reinitialize the database at any time.
To enable the PostgreSQL backend you can use the --query-store-jdbc-config
flag, an example of which is below.
Note
When you use the Query Store you’ll want your first run to specify createSchema=true
so that all the necessary tables are created. After the first run make sure createSchema=false
so that it doesn’t attempt to create the tables again.
daml json-api --ledger-host localhost --ledger-port 6865 --http-port 7575 \
--query-store-jdbc-config "driver=org.postgresql.Driver,url=jdbc:postgresql://localhost:5432/test?&ssl=true,user=postgres,password=password,createSchema=false"
Note
The JSON API provides many other useful configuration flags, run daml json-api --help
to see all of them.
Access Tokens¶
The JSON API essentially performs two separate tasks:
- It talks to the Ledger API to get data it needs to operate, for this you need to provide an access token if your Ledger requires authorization. Learn more in the Authorization docs.
- It accepts requests from Parties and passes them on to the Ledger API, for this each party needs to provide an access token with each request it sends to the JSON API.
Note
By default, the Daml Sandbox does not does not require access tokens. In this case, you can omit the token used by the JSON API to request packages. However, you still need to provide a party-specific access token when submitting commands or queries as a party. The token will not be validated in this case but it will be decoded to extract information like the party submitting the command.
Internal Access Token¶
This access token is used exclusively by the JSON API service for maintaining the internal list of known packages and templates that it gets from the Ledger API.
Note
At no point should this access token be provided to an end user, these are for internal use only.
Every access token is different and will depend on your specific ledger operator’s requirements. The JSON API server requires no access to party-specific data, only access to the ledger identity and package services. These services are public meaning that you need a valid token to access them but no party-specific claims nor an admin claim. Please refer to your ledger operator’s documentation to find out how to get these tokens from your ledger operator.
Once you have retrieved your access token, you can provide it to the JSON API by storing it in a file
and starting daml json-api
with the flag --access-token-file /path/to/your/token.file
.
If the token cannot be read from the provided path or the Ledger API reports an authentication error (for example due to token expiration), the JSON API will report the error via logging.
Note
If the token file is updated with a new token it will be picked up at the next attempt to send a request. You can use this to handle cases where an old token expires without restarting your JSON API service.
Party-specific Access Tokens¶
Party-specific requests, i.e., command submissions and queries, require a JWT with some additional restrictions compared to the format described in the Token Payload section here. For command submissions, actAs
must contain at least one party and readAs
can contain 0 or more parties. Queries require at least one party in either actAs
or readAs
(note that before SDK 1.7.0, every request required exactly one party and before SDK 1.8.0 actAs was limited to exactly one party). In addition to that, the application id and ledger id are mandatory. HTTP requests pass the token in a header, while WebSocket requests pass the token in a subprotocol.
Note
While the JSON API receives the token it doesn’t validate it itself. Upon receiving a token it will pass it, and all data contained within the request, on to the Ledger API’s AuthService which will then determine if the token is valid and authorized. However, the JSON API does decode the token to extract the ledger id, application id and party so it requires that you use the JWT format documented below.
For a ledger without authorization, e.g., the default configuration of Daml Sandbox, you can use https://jwt.io (or the JWT library of your choice) to generate your token. You can use an arbitrary secret here. The default “header” is fine. Under “Payload”, fill in:
{
"https://daml.com/ledger-api": {
"ledgerId": "MyLedger",
"applicationId": "foobar",
"actAs": ["Alice"]
}
}
The value of the ledgerId
field has to match the ledgerId
of your underlying Daml Ledger.
For the Sandbox this corresponds to the --ledgerid MyLedger
flag.
Note
The value of applicationId
will be used for commands submitted using that token.
The value for actAs
is specified as a list and you provide it with the party that you want to use.
Such as the example which uses Alice
for a party. Each request can only be for one party.
For example you couldn’t have actAs
defined as ["Alice", "Bob"]
.
The party should reference an already allocated party.
Note
As mentioned above the JSON API does not validate tokens so if your ledger runs without authorization you can use an arbitrary secret.
Then the “Encoded” box should have your token, ready for passing to the service as described in the following sections.
Alternatively, here are two tokens you can use for testing:
{"https://daml.com/ledger-api": {"ledgerId": "MyLedger", "applicationId": "HTTP-JSON-API-Gateway", "actAs": ["Alice"]}}
:
eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJodHRwczovL2RhbWwuY29tL2xlZGdlci1hcGkiOnsibGVkZ2VySWQiOiJNeUxlZGdlciIsImFwcGxpY2F0aW9uSWQiOiJIVFRQLUpTT04tQVBJLUdhdGV3YXkiLCJhY3RBcyI6WyJBbGljZSJdfX0.34zzF_fbWv7p60r5s1kKzwndvGdsJDX-W4Xhm4oVdpk
{"https://daml.com/ledger-api": {"ledgerId": "MyLedger", "applicationId": "HTTP-JSON-API-Gateway", "actAs": ["Bob"]}}
:
eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJodHRwczovL2RhbWwuY29tL2xlZGdlci1hcGkiOnsibGVkZ2VySWQiOiJNeUxlZGdlciIsImFwcGxpY2F0aW9uSWQiOiJIVFRQLUpTT04tQVBJLUdhdGV3YXkiLCJhY3RBcyI6WyJCb2IiXX19.0uPPZtM1AmKvnGixt_Qo53cMDcpnziCjKKiWLvMX2VM
Auth via HTTP¶
Set HTTP header Authorization: Bearer paste-jwt-here
Example:
Authorization: Bearer eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJodHRwczovL2RhbWwuY29tL2xlZGdlci1hcGkiOnsibGVkZ2VySWQiOiJNeUxlZGdlciIsImFwcGxpY2F0aW9uSWQiOiJIVFRQLUpTT04tQVBJLUdhdGV3YXkiLCJhY3RBcyI6WyJBbGljZSJdfX0.34zzF_fbWv7p60r5s1kKzwndvGdsJDX-W4Xhm4oVdpk
Auth via WebSockets¶
WebSocket clients support a “subprotocols” argument (sometimes simply called “protocols”); this is usually in a list form but occasionally in comma-separated form. Check documentation for your WebSocket library of choice for details.
For HTTP JSON requests, you must pass two subprotocols:
daml.ws.auth
jwt.token.paste-jwt-here
Example:
jwt.token.eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJodHRwczovL2RhbWwuY29tL2xlZGdlci1hcGkiOnsibGVkZ2VySWQiOiJNeUxlZGdlciIsImFwcGxpY2F0aW9uSWQiOiJIVFRQLUpTT04tQVBJLUdhdGV3YXkiLCJhY3RBcyI6WyJBbGljZSJdfX0.34zzF_fbWv7p60r5s1kKzwndvGdsJDX-W4Xhm4oVdpk
HTTP Status Codes¶
The JSON API reports errors using standard HTTP status codes. It divides HTTP status codes into 3 groups indicating:
- success (200)
- failure due to a client-side problem (400, 401, 404)
- failure due to a server-side problem (500)
The JSON API can return one of the following HTTP status codes:
- 200 - OK
- 400 - Bad Request (Client Error)
- 401 - Unauthorized, authentication required
- 404 - Not Found
- 500 - Internal Server Error
If a client’s HTTP GET or POST request reaches an API endpoint, the corresponding response will always contain a JSON object with a status
field, either an errors
or result
field and an optional warnings
:
{
"status": <400 | 401 | 404 | 500>,
"errors": <JSON array of strings>, | "result": <JSON object or array>,
["warnings": <JSON object> ]
}
Where:
status
– a JSON number which matches the HTTP response status code returned in the HTTP header,errors
– a JSON array of strings, each string represents one error,result
– a JSON object or JSON array, representing one or many results,warnings
– an optional field with a JSON object, representing one or many warnings.
See the following blog post for more details about error handling best practices: REST API Error Codes 101.
Successful response, HTTP status: 200 OK¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": <JSON object>
}
Successful response with a warning, HTTP status: 200 OK¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": <JSON object>,
"warnings": <JSON object>
}
Failure, HTTP status: 400 | 401 | 404 | 500¶
- Content-Type:
application/json
- Content:
{
"status": <400 | 401 | 404 | 500>,
"errors": <JSON array of strings>
}
Examples¶
Result with JSON Object without Warnings:
{"status": 200, "result": {...}}
Result with JSON Array and Warnings:
{"status": 200, "result": [...], "warnings": {"unknownTemplateIds": ["UnknownModule:UnknownEntity"]}}
Bad Request Error:
{"status": 400, "errors": ["JSON parser error: Unexpected character 'f' at input index 27 (line 1, position 28)"]}
Bad Request Error with Warnings:
{"status":400, "errors":["Cannot resolve any template ID from request"], "warnings":{"unknownTemplateIds":["XXX:YYY","AAA:BBB"]}}
Authentication Error:
{"status": 401, "errors": ["Authentication Required"]}
Not Found Error:
{"status": 404, "errors": ["HttpMethod(POST), uri: http://localhost:7575/v1/query1"]}
Internal Server Error:
{"status": 500, "errors": ["Cannot initialize Ledger API"]}
Create a new Contract¶
To create an Iou
contract from the Quickstart guide:
template Iou
with
issuer : Party
owner : Party
currency : Text
amount : Decimal
observers : [Party]
HTTP Request¶
- URL:
/v1/create
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateId": "Iou:Iou",
"payload": {
"issuer": "Alice",
"owner": "Alice",
"currency": "USD",
"amount": "999.99",
"observers": []
}
}
Where:
templateId
is the contract template identifier, which can be formatted as either:"<package ID>:<module>:<entity>"
or"<module>:<entity>"
if contract template can be uniquely identified by its module and entity name.
payload
field contains contract fields as defined in the Daml template and formatted according to Daml-LF JSON Encoding.
HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": {
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#124:0",
"templateId": "11c8f3ace75868d28136adc5cfc1de265a9ee5ad73fe8f2db97510e3631096a2:Iou:Iou"
}
}
Where:
status
field matches the HTTP response status code returned in the HTTP header,result
field contains created contract details. Keep in mind thattemplateId
in the JSON API response is always fully qualified (always contains package ID).
Creating a Contract with a Command ID¶
When creating a new contract you may specify an optional meta
field. This allows you to control the commandId used when submitting a command to the ledger.
Note
You cannot currently use commandIds
anywhere else in the JSON API, but you can use it for observing the results of its commands outside the JSON API in logs or via the Ledger API’s Command Services
{
"templateId": "Iou:Iou",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"meta": {
"commandId": "a unique ID"
}
}
Where:
commandId
– optional field, a unique string identifying the command.
Exercise by Contract ID¶
The JSON command below, demonstrates how to exercise an Iou_Transfer
choice on an Iou
contract:
controller owner can
Iou_Transfer : ContractId IouTransfer
with
newOwner : Party
do create IouTransfer with iou = this; newOwner
HTTP Request¶
- URL:
/v1/exercise
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateId": "Iou:Iou",
"contractId": "#124:0",
"choice": "Iou_Transfer",
"argument": {
"newOwner": "Alice"
}
}
Where:
templateId
– contract template identifier, same as in create request,contractId
– contract identifier, the value from the create response,choice
– Daml contract choice, that is being exercised,argument
– contract choice argument(s).
HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": {
"exerciseResult": "#201:1",
"events": [
{
"archived": {
"contractId": "#124:0",
"templateId": "11c8f3ace75868d28136adc5cfc1de265a9ee5ad73fe8f2db97510e3631096a2:Iou:Iou"
}
},
{
"created": {
"observers": [],
"agreementText": "",
"payload": {
"iou": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"newOwner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#201:1",
"templateId": "11c8f3ace75868d28136adc5cfc1de265a9ee5ad73fe8f2db97510e3631096a2:Iou:IouTransfer"
}
}
]
}
}
Where:
status
field matches the HTTP response status code returned in the HTTP header,result
field contains contract choice execution details:exerciseResult
field contains the return value of the exercised contract choice,events
contains an array of contracts that were archived and created as part of the choice execution. The array may contain: zero or many{"archived": {...}}
and zero or many{"created": {...}}
elements. The order of the contracts is the same as on the ledger.
Exercise by Contract Key¶
The JSON command below, demonstrates how to exercise the Archive
choice on the Account
contract with a (Party, Text)
contract key defined like this:
template Account with
owner : Party
number : Text
status : AccountStatus
where
signatory owner
key (owner, number) : (Party, Text)
maintainer key._1
HTTP Request¶
- URL:
/v1/exercise
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateId": "Account:Account",
"key": {
"_1": "Alice",
"_2": "abc123"
},
"choice": "Archive",
"argument": {}
}
Where:
templateId
– contract template identifier, same as in create request,key
– contract key, formatted according to the Daml-LF JSON Encoding,choice
– Daml contract choice, that is being exercised,argument
– contract choice argument(s), empty, becauseArchive
does not take any.
HTTP Response¶
Formatted similar to Exercise by Contract ID response.
Create and Exercise in the Same Transaction¶
This command allows creating a contract and exercising a choice on the newly created contract in the same transaction.
HTTP Request¶
- URL:
/v1/create-and-exercise
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateId": "Iou:Iou",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"choice": "Iou_Transfer",
"argument": {
"newOwner": "Bob"
}
}
Where:
templateId
– the initial contract template identifier, in the same format as in the create request,payload
– the initial contract fields as defined in the Daml template and formatted according to Daml-LF JSON Encoding,choice
– Daml contract choice, that is being exercised,argument
– contract choice argument(s).
HTTP Response¶
Please note that the response below is for a consuming choice, so it contains:
created
andarchived
events for the initial contract ("contractId": "#1:0"
), which was created and archived right away when a consuming choice was exercised on it,- a
created
event for the contract that is the result of exercising the choice ("contractId": "#1:2"
). - Content-Type:
application/json
- Content:
{
"result": {
"exerciseResult": "#1:2",
"events": [
{
"created": {
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#1:0",
"templateId": "a3b788b4dc18dc060bfb82366ae6dc055b1e361d646d5cfdb1b729607e344336:Iou:Iou"
}
},
{
"archived": {
"contractId": "#1:0",
"templateId": "a3b788b4dc18dc060bfb82366ae6dc055b1e361d646d5cfdb1b729607e344336:Iou:Iou"
}
},
{
"created": {
"observers": [
"Bob"
],
"agreementText": "",
"payload": {
"iou": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"newOwner": "Bob"
},
"signatories": [
"Alice"
],
"contractId": "#1:2",
"templateId": "a3b788b4dc18dc060bfb82366ae6dc055b1e361d646d5cfdb1b729607e344336:Iou:IouTransfer"
}
}
]
},
"status": 200
}
Fetch Contract by Contract ID¶
HTTP Request¶
- URL:
/v1/fetch
- Method:
POST
- Content-Type:
application/json
- Content:
application/json body:
{
"contractId": "#201:1"
}
Contract Not Found HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": null
}
Contract Found HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": {
"observers": [],
"agreementText": "",
"payload": {
"iou": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"newOwner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#201:1",
"templateId": "11c8f3ace75868d28136adc5cfc1de265a9ee5ad73fe8f2db97510e3631096a2:Iou:IouTransfer"
}
}
Fetch Contract by Key¶
HTTP Request¶
- URL:
/v1/fetch
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateId": "Account:Account",
"key": {
"_1": "Alice",
"_2": "abc123"
}
}
Contract Not Found HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": null
}
Contract Found HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": {
"observers": [],
"agreementText": "",
"payload": {
"owner": "Alice",
"number": "abc123",
"status": {
"tag": "Enabled",
"value": "2020-01-01T00:00:01Z"
}
},
"signatories": [
"Alice"
],
"key": {
"_1": "Alice",
"_2": "abc123"
},
"contractId": "#697:0",
"templateId": "11c8f3ace75868d28136adc5cfc1de265a9ee5ad73fe8f2db97510e3631096a2:Account:Account"
}
}
Get all Active Contracts¶
List all currently active contracts for all known templates.
Note
Retrieved contracts do not get persisted into a query store database. Query store is a search index and can be used to optimize search latency. See Start HTTP service for information on how to start JSON API service with a query store enabled.
Note
You can only query active contracts with the /v1/query
endpoint. Archived contracts (those that were archived or consumed during an exercise operation) will not be shown in the results.
HTTP Request¶
- URL:
/v1/query
- Method:
GET
- Content: <EMPTY>
HTTP Response¶
The response is the same as for the POST method below.
Get all Active Contracts Matching a Given Query¶
List currently active contracts that match a given query.
HTTP Request¶
- URL:
/v1/query
- Method:
POST
- Content-Type:
application/json
- Content:
{
"templateIds": ["Iou:Iou"],
"query": {"amount": 999.99}
}
Where:
templateIds
– an array of contract template identifiers to search through,query
– search criteria to apply to the specifiedtemplateIds
, formatted according to the Query language.
Nonempty HTTP Response¶
- Content-Type:
application/json
- Content:
{
"result": [
{
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#52:0",
"templateId": "b10d22d6c2f2fae41b353315cf893ed66996ecb0abe4424ea6a81576918f658a:Iou:Iou"
}
],
"status": 200
}
Where
result
contains an array of contracts, each contract formatted according to Daml-LF JSON Encoding,status
matches the HTTP status code returned in the HTTP header.
Nonempty HTTP Response with Unknown Template IDs Warning¶
- Content-Type:
application/json
- Content:
{
"warnings": {
"unknownTemplateIds": ["UnknownModule:UnknownEntity"]
},
"result": [
{
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"signatories": [
"Alice"
],
"contractId": "#52:0",
"templateId": "b10d22d6c2f2fae41b353315cf893ed66996ecb0abe4424ea6a81576918f658a:Iou:Iou"
}
],
"status": 200
}
Fetch Parties by Identifiers¶
- URL:
/v1/parties
- Method:
POST
- Content-Type:
application/json
- Content:
["Alice", "Bob", "Dave"]
If an empty JSON array is passed: []
, this endpoint returns BadRequest(400) error:
{
"status": 400,
"errors": [
"JsonReaderError. Cannot read JSON: <[]>. Cause: spray.json.DeserializationException: must be a list with at least 1 element"
]
}
HTTP Response¶
- Content-Type:
application/json
- Content:
{
"status": 200,
"result": [
{
"identifier": "Alice",
"displayName": "Alice & Co. LLC",
"isLocal": true
},
{
"identifier": "Bob",
"displayName": "Bob & Co. LLC",
"isLocal": true
},
{
"identifier": "Dave",
"isLocal": true
}
]
}
Please note that the order of the party objects in the response is not guaranteed to match the order of the passed party identifiers.
Where
identifier
– a stable unique identifier of a Daml party,displayName
– optional human readable name associated with the party. Might not be unique,isLocal
– true if party is hosted by the backing participant.
Response with Unknown Parties Warning¶
- Content-Type:
application/json
- Content:
{
"result": [
{
"identifier": "Alice",
"displayName": "Alice & Co. LLC",
"isLocal": true
}
],
"warnings": {
"unknownParties": ["Erin"]
},
"status": 200
}
The result
might be an empty JSON array if none of the requested parties is known.
Fetch All Known Parties¶
- URL:
/v1/parties
- Method:
GET
- Content: <EMPTY>
HTTP Response¶
The response is the same as for the POST method above.
Allocate a New Party¶
This endpoint is a JSON API proxy for the Ledger API’s AllocatePartyRequest. For more information about party management, please refer to Provisioning Identifiers part of the Ledger API documentation.
HTTP Request¶
- URL:
/v1/parties/allocate
- Method:
POST
- Content-Type:
application/json
- Content:
{
"identifierHint": "Carol",
"displayName": "Carol & Co. LLC"
}
Please refer to AllocateParty documentation for information about the meaning of the fields.
All fields in the request are optional, this means that an empty JSON object is a valid request to allocate a new party:
{}
HTTP Response¶
{
"result": {
"identifier": "Carol",
"displayName": "Carol & Co. LLC",
"isLocal": true
},
"status": 200
}
List All DALF Packages¶
HTTP Request¶
- URL:
/v1/packages
- Method:
GET
- Content: <EMPTY>
HTTP Response¶
{
"result": [
"c1f1f00558799eec139fb4f4c76f95fb52fa1837a5dd29600baa1c8ed1bdccfd",
"733e38d36a2759688a4b2c4cec69d48e7b55ecc8dedc8067b815926c917a182a",
"bfcd37bd6b84768e86e432f5f6c33e25d9e7724a9d42e33875ff74f6348e733f",
"40f452260bef3f29dede136108fc08a88d5a5250310281067087da6f0baddff7",
"8a7806365bbd98d88b4c13832ebfa305f6abaeaf32cfa2b7dd25c4fa489b79fb"
],
"status": 200
}
Where result
is the JSON array containing the package IDs of all loaded DALFs.
Download a DALF Package¶
HTTP Request¶
- URL:
/v1/packages/<package ID>
- Method:
GET
- Content: <EMPTY>
Note that the desired package ID is specified in the URL.
HTTP Response, status: 200 OK¶
- Transfer-Encoding:
chunked
- Content-Type:
application/octet-stream
- Content: <DALF bytes>
The content (body) of the HTTP response contains raw DALF package bytes, without any encoding. Note that the package ID specified in the URL is actually the SHA-256 hash of the downloaded DALF package and can be used to validate the integrity of the downloaded content.
HTTP Response with Error, any status different from 200 OK¶
Any status different from 200 OK
will be in the format specified below.
- Content-Type:
application/json
- Content:
{
"errors": [
"io.grpc.StatusRuntimeException: NOT_FOUND"
],
"status": 500
}
Upload a DAR File¶
HTTP Request¶
- URL:
/v1/packages
- Method:
POST
- Content-Type:
application/octet-stream
- Content: <DAR bytes>
The content (body) of the HTTP request contains raw DAR file bytes, without any encoding.
HTTP Response, status: 200 OK¶
- Content-Type:
application/json
- Content:
{
"result": 1,
"status": 200
}
HTTP Response with Error¶
- Content-Type:
application/json
- Content:
{
"errors": [
"io.grpc.StatusRuntimeException: INVALID_ARGUMENT: Invalid argument: Invalid DAR: package-upload, content: [}]"
],
"status": 500
}
Streaming API¶
Two subprotocols must be passed with every request, as described in Auth via WebSockets.
JavaScript/Node.js example demonstrating how to establish Streaming API connection:
const wsProtocol = "daml.ws.auth";
const tokenPrefix = "jwt.token.";
const jwt =
"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJodHRwczovL2RhbWwuY29tL2xlZGdlci1hcGkiOnsibGVkZ2VySWQiOiJNeUxlZGdlciIsImFwcGxpY2F0aW9uSWQiOiJIVFRQLUpTT04tQVBJLUdhdGV3YXkiLCJhY3RBcyI6WyJBbGljZSJdfX0.34zzF_fbWv7p60r5s1kKzwndvGdsJDX-W4Xhm4oVdp";
const subprotocols = [`${tokenPrefix}${jwt}`, wsProtocol];
const ws = new WebSocket("ws://localhost:7575/v1/stream/query", subprotocols);
ws.addEventListener("open", function open() {
ws.send(JSON.stringify({templateIds: ["Iou:Iou"]}));
});
ws.addEventListener("message", function incoming(data) {
console.log(data);
});
Please note that Streaming API does not allow multiple requests over the same WebSocket connection. The server returns an error and disconnects if second request received over the same WebSocket connection.
Error and Warning Reporting¶
Errors and warnings reported as part of the regular on-message
flow: ws.addEventListener("message", ...)
.
Streaming API error messages formatted the same way as synchronous API errors.
Streaming API reports only one type of warnings – unknown template IDs, which is formatted as:
{"warnings":{"unknownTemplateIds":<JSON Array of template ID strings>>}}
Error and Warning Examples¶
{"warnings": {"unknownTemplateIds": ["UnknownModule:UnknownEntity"]}}
{
"errors":["JsonReaderError. Cannot read JSON: <{\"templateIds\":[]}>. Cause: spray.json.DeserializationException: search requires at least one item in 'templateIds'"],
"status":400
}
{
"errors":["Multiple requests over the same WebSocket connection are not allowed."],
"status":400
}
{
"errors":["Could not resolve any template ID from request."],
"status":400
}
Contracts Query Stream¶
- URL:
/v1/stream/query
- Scheme:
ws
- Protocol:
WebSocket
List currently active contracts that match a given query, with continuous updates.
application/json
body must be sent first, formatted according to the
Query language:
{"templateIds": ["Iou:Iou"]}
Multiple queries may be specified in an array, for overlapping or different sets of template IDs:
[
{"templateIds": ["Iou:Iou"], "query": {"amount": {"%lte": 50}}},
{"templateIds": ["Iou:Iou"], "query": {"amount": {"%gt": 50}}},
{"templateIds": ["Iou:Iou"]}
]
Queries have two ways to specify an offset.
An offset
, a string supplied by an earlier query output message, may
optionally be specified alongside each query itself:
[
{"templateIds": ["Iou:Iou"], "query": {"amount": {"%lte": 50}}},
{"templateIds": ["Iou:Iou"], "query": {"amount": {"%gt": 50}}},
{"templateIds": ["Iou:Iou"], "offset": "5609"}
]
If specified, the stream will include only contract creations and archivals after the response body that included that offset. Queries with no offset will begin with all active contracts for that query, as usual.
If an offset is specified before the queries, as a separate body, it will be used as a default offset for all queries that do not include an offset themselves:
{"offset": "4307"}
For example, if this message preceded the above 3-query example, it
would be as if "4307"
had been specified for the first two queries,
while "5609"
would be used for the third query.
The output is a series of JSON documents, each payload
formatted
according to Daml-LF JSON Encoding:
{
"events": [{
"created": {
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "999.99",
"currency": "USD",
"owner": "Alice"
},
"signatories": ["Alice"],
"contractId": "#1:0",
"templateId": "eb3b150383a979d6765b8570a17dd24ae8d8b63418ee5fd20df20ad2a1c13976:Iou:Iou"
},
"matchedQueries": [1, 2]
}]
}
where matchedQueries
indicates the 0-based indices into the request
list of queries that matched this contract.
Every events
block following the end of contracts that existed when
the request started includes an offset
. The stream is guaranteed to
send an offset immediately at the beginning of this “live” data, which
may or may not contain any events
; if it does not contain events and
no events were emitted before, it may be null
if there was no
transaction on the ledger
or a string representing the current ledger end;
otherwise, it will be a string. For example, you might use it to turn
off an initial “loading” indicator:
{
"events": [],
"offset": "2"
}
Note
Events in the following “live” data may include events
that precede
this offset
if an earlier per-query offset
was specified.
This has been done with the intent of allowing to use per-query offset
s to
efficiently use a single connection to multiplex various requests. To give an
example of how this would work, let’s say that there are two contract templates,
A
and B
. Your application first queries for A
s without specifying
an offset. Then some client-side interaction requires the application to do the
same for B
s. The application can save the latest observed offset for the
previous query, which let’s say is 42
, and issue a new request that queries
for all B
s without specifying an offset and all A
s from 42
. While
this happens on the client, a few more A
s and B
s are created and the
new request is issued once the latest offset is 47
. The response to this
will contain a message with all active B
s, followed by the message reporting
the offset 47
, followed by a stream of live updates that contains new A
s
starting from 42
and new B
s starting from 47
.
To keep the stream alive, you’ll occasionally see messages like this, which can be safely ignored if you do not need to capture the last seen ledger offset:
{"events":[],"offset":"5609"}
where offset
is the last seen ledger offset.
After submitting an Iou_Split
exercise, which creates two contracts
and archives the one above, the same stream will eventually produce:
{
"events": [{
"archived": {
"contractId": "#1:0",
"templateId": "eb3b150383a979d6765b8570a17dd24ae8d8b63418ee5fd20df20ad2a1c13976:Iou:Iou"
}
}, {
"created": {
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "42.42",
"currency": "USD",
"owner": "Alice"
},
"signatories": ["Alice"],
"contractId": "#2:1",
"templateId": "eb3b150383a979d6765b8570a17dd24ae8d8b63418ee5fd20df20ad2a1c13976:Iou:Iou"
},
"matchedQueries": [0, 2]
}, {
"created": {
"observers": [],
"agreementText": "",
"payload": {
"observers": [],
"issuer": "Alice",
"amount": "957.57",
"currency": "USD",
"owner": "Alice"
},
"signatories": ["Alice"],
"contractId": "#2:2",
"templateId": "eb3b150383a979d6765b8570a17dd24ae8d8b63418ee5fd20df20ad2a1c13976:Iou:Iou"
},
"matchedQueries": [1, 2]
}],
"offset": "3"
}
If any template IDs are found not to resolve, the first element of the stream will report them:
{"warnings": {"unknownTemplateIds": ["UnknownModule:UnknownEntity"]}}
and the stream will continue, provided that at least one template ID resolved properly.
Aside from "created"
and "archived"
elements, "error"
elements may appear, which contain a string describing the error. The
stream will continue in these cases, rather than terminating.
Some notes on behavior:
- Each result array means “this is what would have changed if you just
polled
/v1/query
iteratively.” In particular, just as polling search can “miss” contracts (as a create and archive can be paired between polls), such contracts may or may not appear in any result object. - No
archived
ever contains a contract ID occurring within acreated
in the same array. So, for example, supposing you are keeping an internal map of active contracts keyed by contract ID, you can apply thecreated
first or thearchived
first, forwards, backwards, or in random order, and be guaranteed to get the same results. - Within a given array, if an
archived
andcreated
refer to contracts with the same template ID and contract key, thearchived
is guaranteed to occur before thecreated
. - Except in cases of #3, within a single response array, the order of
created
andarchived
is undefined and does not imply that any element occurred “before” or “after” any other one. - You will almost certainly receive contract IDs in
archived
that you never received acreated
for. These are contracts that query filtered out, but for which the server no longer is aware of that. You can safely ignore these. However, such “phantom archives” are guaranteed to represent an actual archival on the ledger, so if you are keeping a more global dataset outside the context of this specific search, you can use that archival information as you wish.
Fetch by Key Contracts Stream¶
- URL:
/v1/stream/fetch
- Scheme:
ws
- Protocol:
WebSocket
List currently active contracts that match one of the given {templateId, key}
pairs, with continuous updates.
application/json
body must be sent first, formatted according to the following rule:
[
{"templateId": "<template ID 1>", "key": <key 1>},
{"templateId": "<template ID 2>", "key": <key 2>},
...
{"templateId": "<template ID N>", "key": <key N>}
]
Where:
templateId
– contract template identifier, same as in create request,key
– contract key, formatted according to the Daml-LF JSON Encoding,
Example:
[
{"templateId": "Account:Account", "key": {"_1": "Alice", "_2": "abc123"}},
{"templateId": "Account:Account", "key": {"_1": "Alice", "_2": "def345"}}
]
The output stream has the same format as the output from the
Contracts Query Stream. We further guarantee that for every archived
event
appearing on the stream there has been a matching created
event
earlier in the stream, except in the case of missing
contractIdAtOffset
fields in the case described below.
You may supply optional offset
s for the stream, exactly as with
query streams. However, you should supply with each {templateId,
key}
pair a contractIdAtOffset
, which is the contract ID currently
associated with that pair at the point of the given offset, or null
if no contract ID was associated with the pair at that offset. For
example, with the above keys, if you had one "abc123"
contract but
no "def345"
contract, you might specify:
[
{"templateId": "Account:Account", "key": {"_1": "Alice", "_2": "abc123"},
"contractIdAtOffset": "#1:0"},
{"templateId": "Account:Account", "key": {"_1": "Alice", "_2": "def345"},
"contractIdAtOffset": null}
]
If every contractIdAtOffset
is specified, as is so in the example
above, you will not receive any archived
events for contracts
created before the offset unless those contracts are identified in a
contractIdAtOffset
. By contrast, if any contractIdAtOffset
is
missing, archived
event filtering will be disabled, and you will
receive “phantom archives” as with query streams.
Healthcheck Endpoints¶
The HTTP JSON API provides two healthcheck endpoints for integration with schedulers like Kubernetes.
Liveness check¶
- URL:
/livez
- Method:
GET
A status code of 200
indicates a successful liveness check.
This is an unauthenticated endpoint intended to be used as a liveness probe.
Readiness check¶
- URL:
/readyz
- Method:
GET
A status code of 200
indicates a successful readiness check.
This is an unauthenticated endpoint intended to be used as a liveness probe. It validates both the ledger connection as well as the database connection.