bloomberg / blpapi-http Goto Github PK

Related to #15 but until #102 is resolved I really don't want to advertise the functionality too much (i.e. let's wait until auth is more usable before worrying about docs).

Provide a complete set of Ruby examples for accessing the blpapi-http server over HTTPS. Examples should cover the following:

• Historical data
• Reference data
• Field lookup and search
• Overrides
• Varied options and parameters
• Using different securities that are not just equities (i.e., stocks)
• Getting the result back in some usable language-specific structure

Related to #15.

Right now, we only have AuthorizationTokenRequest which sends you a token via MSG, which is lame.

http-api-guide.md is falling behind a bit:

1. Subscriptions aren't explained.
2. Auth will need to get added when it lands.
3. All the references (as well as examples across the board) will need to be updated for the URL scheme when #59 is resolved).

This issue is to track the progress of adding support for the //blp/mktvwap service. Please update with progress, findings, and status.

See comment on #73 re: switch and add that to the style guide.

After we have mocks for request/response (and eventually for everything), we should hook up running the examples during CI, to detect when there are errors in the examples. (See #81, for example.)

This is to come to a consensus among the URL scheme and design of the HTTP API. Once we understand how all the services work that is being addressed in #15, #54, #55, #56, #57, #58 we decide upon what the final HTTP API should look like with confidence.

We should provide the same set of examples that the blpapi-node project provides in its examples directory for completeness. curl or wget should be sufficient for request/response examples and some language should be decided upon for the subscription examples.

This issue is to track the progress of adding support for the //blp/instruments service. Please update with progress, findings, and status.

NodeJS v0.12.0 is now stable.

After an identity is authorized, changes in its status need to be surfaced and handled somehow. This will require work in the node project as well.

We need an example for websocket subscription using python

This issue is to serve as discussion thread to determine the initial contribution guidelines for the project. I have initially started this work in my own personal fork at http://github.com/ericvw/blpapi-http under the contributing branch. Once the initial guidelines have been settled upon, we can close this issue with a PR that addresses everything discussed.

The primary topics that would be desirable to initially cover are:

• the development workflow/branching model
• guidelines for awesome issue/feature requests
• determining if and how we want to use milestones to organize the project
• what the testing and quality assurance tools are (this may be premature to define)
• how we want to deal with documentation with respect to the Bloomberg Open API to avoid duplication, but keep the content relevant.

Ready...set...GO!

We should be careful about cleaning up state of the blpapi-wrapper Session when errors occur.

We need to update the documentation regarding subscriptions.

Provide a complete set of curl and/or wget examples for accessing the blpapi-http server over HTTPS. Examples should cover the following:

• Historical data
• Reference data
• Field lookup and search
• Overrides
• Varied options and parameters
• Using different securities that are not just equities (i.e., stocks)
• Getting the result back in some usable language-specific structure

Provide a FAQ that should cover the following and more:

• Specific environment/tooling oddities and solutions
• Links to other externally hosted and in-repository documentation

We need a quickstart guide to standing up the http server across all environments.

Log all requests through API to learn what the most common requests/datasets are to create better documentation and highlight what information is interesting

Right now in our request/response code we use http status code to negotiate the errors with client. This approach won't work for cases where we encounter errors in the middle of the request, since we are sending the data received from underlying session right away(not buffering).

The two possible approaches are:

• Buffer the response until the last bit and then send the data in 1 trunk, so that we could leverage the http status code.
• Always send 200 as http status code and leverage the status field in the response body to indicate the status of the response.

Open to discussion, the goal is to come up with a consistency approach for our server.

Provide a complete set of NodeJS examples for accessing the blpapi-http server over HTTPS. Examples should cover the following:

• Historical data
• Reference data
• Field lookup and search
• Overrides
• Varied options and parameters
• Using different securities that are not just equities (i.e., stocks)
• Getting the result back in some usable language-specific structure

We need this.

• We need to decided upon some testing infrastructure (i.e., mocha?)
• How can we stub the BLP API?
• Can the testing infrastructure work with TravisCI?

Provide a complete set of Python 2.7 and Python 3.x examples for accessing the blpapi-http server over HTTPS. Examples should cover the following:

• Historical data
• Reference data
• Field lookup and search
• Overrides
• Varied options and parameters
• Using different securities that are not just equities (i.e., stocks)
• Getting the result back in some usable language-specific structure

Migrate blpapi-wrapper.ts into the blpapi-node project as a higher-level user-facing API. Some things to consider:

• Should the existing blpapi-node API exist alongside this higher-level API?
• How much of this functionality should be pushed down into the native/binding layer?
• What's the strategy for migrating to depend upon NodeJS 0.12.x features (i.e., generators)?

We need an example for long-polling subscription using python

This is somewhat related to #13, but will probably happen in one PR. However, this issues to converge the management to one underlying BLP API session.

Documentation says you can use http, but the server requires https.

We need an example for long-polling subscription using node.js

See #86 -- we should do the same sort of simplification for subscriptions.

This issues is to understand how the Open API authorization service and system works. Section 6 of the BLP API Developers Guide discusses the //blp/apiauth service, but we need a better understanding how this works in practice.

This investigation encompasses setting up our own BPIPE installation and user management to properly test authorization.

This should address issues not covered by lint checks.

style-guide.md should have a section specifying the format and conventions for object literals.

Provide a complete set of Java examples for accessing the blpapi-http server over HTTPS. Examples should cover the following:

• Historical data
• Reference data
• Field lookup and search
• Overrides
• Varied options and parameters
• Using different securities that are not just equities (i.e., stocks)
• Getting the result back in some usable language-specific structure

This issue is to track the progress of adding support for the //blp/mktbar service. Please update with progress, findings, and status.

Currently, any time the wrapper returns an error, we reject with that error, which gets handled as a 500. We should allow the wrapper to return various error types and surface that to the client.

This issue is to track the progress of adding support for the //blp/pagedata service. Please update with progress, findings, and status.

Right now index.ts is mixing up the command-line setup and spin-up the server. It would be nice to have a separate module to wrap the business logic and let index.ts to be called purely on command line. This will help writing test cases and include servers in other applications as well.

Currently the http server is not supporting CRL. We need to configure the node server to read a CRL. Other consideration including auto-bouncing the server once the CRL change( as part of research in #13 we found out that node server will not update when CRL file changes).

This issue is to track the progress of adding support for the //blp/tasvc service. Please update with progress, findings, and status.

When a socket is connected and I kill the underlying blpSession, I did not received a SessionTerminated events. The event is received if there's no socket connected.

During Treehacks, we were seeing responses returning incomplete data(Only 1 field having data returned) or being invalid json objects. I have a feeling it is because the requests are M x N which are causing the issues.

Sample json request that was causing issues:

{
   "securities": ["FOXA US Equity", "CTSH US Equity", "APC US Equity"],
    "fields": ["PX_LAST", "OPEN", "EPS_ANNUALIZED"],
    "startDate": "20120601",
    "endDate": "20120602",
    "periodicitySelection": "DAILY"
}

Currently #104 only support auth for request/response model. We need to extend it to cover websocket cases too.

This issue is to track the progress of adding nested object/array support to blpapi-node. Specifically, this support will allow for full usage of the FieldSearchRequest. Please feel free to use this issue as notes, updates, and progress.

This issue is to investigate and implement the usage of HTTPS client certificates as a means of authentication and identity management.

Some investigative questions to answer are:

• Do we have access to the underlying certificate information?
  • If so, what part of the certificate can we use to uniquely identify the user?
• How does revocation lists of certificates work with the http server?

Please feel free to add any other investigative questions, answers, and progress.

We need to update the documentation regarding authorization.

It will be useful to have a script used for development that generate server and client-side certificates.