Is there a capability for or example of creating a Sparql HTTP endpoint with rdflib? We would want it to follow the spec and be able to return json and/or csv formats. This would mostly be for POC usage. It would also be possible to use Javascript/Node.
Thanks!
You might try https://github.com/rdflib/pyLDAPI. It's been touched much more recently than https://github.com/RDFLib/rdflib-web and there are some public examples of it to follow, e.g. https://geofabricld.net. Also, the SKOS-specific tool VocPrez uses it under the hood.
As of earlier this year, pyLDAPI implements the W3C's Content Negotiation by Profile specification which is, I suppose, the latest an greatest Linked Data API-relevant specification, although it's not just for Linked Data APIs.
Feel free to contact me directly if you need more of a hand with this.
Related
I have an web API application that will serve many clients at different times of release and now i need to implement a versioning. Because the API code will be constantly updated and API users will not be able to instantly change their API. Well, the standard situation is when you need to introduce versioning in general. I'm finding a way to organize it inside my API. It's clear that it will not be different folders with an application on the server, conditionally called app_v1, app_v2, app_v2.1, etc., cause this is duplication, redundancy and bad practise.
It's look like will be one application, and in the controllers at the code level there will be a division of the logic already, like If(client_version==1) do function1() else if(client_version==2) do function2(), etc. It seems that git supports tags, this is something similar to versioning, but because all supported versions of the application need be on the server at the same time, this is not about that. how i can realize an architecture in this case?
There are many well-known ways to use API versioning to make code work with older versions. (backward compatibility). The general purpose of API versioning is a way to make sure that different clients can use different versions of an API at the same time. I've seen several ways to do API versioning, such as:
URL Path Versioning: In this method, the number of the version is part of the API endpoint's URL path. For instance:
https://api.example.com/v1/assets
https://api.example.com/v2/assets
URL Query String Parameter: In this method, the version number is added to the API endpoint's URL as a query string parameter. For instance:
https://api.example.com/assets?version=1
https://api.example.com/assets?version=2
HTTP Header: In this method, the version number is put in an HTTP header, like the Accept-Version header. For instance:
Accept-Version: 1
Accept-Version: 2
If you are using dotnet for you project I would like recommend to standard library for that recommend to check this out. Or you can find solid materials in term of WebApi Versioning following link by #Steve Smith.
There is another answer.
I am creating API and trying to figure out is planned approach any good.
That API is not public and it will be used by SPA and mobile app that I build.So I am thinking of GraphQL-like design but without posting json and with regular HTTP methods.
Something like this for GET methods:
Example 1 - get users with specific fields(_join indicates sql table join), ordering and limit:api.com?table=users&displayFields=id,name,email,address,tel,country_join&orderBy=asc&orderColumn=name&offset=0&limit=10
Example 2 - get users based on search parameters with all fields, ordering and limit:api.com?table=users&search=John&searchFields=name,email&orderBy=asc&orderColumn=name&offset=0&limit=10
I assume this is bad since REST is standard, otherwise I would see much more examples of this approach.
But why is this bad? For me it seems easier to develop and more flexible to use.
Is proper REST API for examples I provided easier to implement, safer, easier to use or cache?
The main difference I see between putting the variables in the url vs the request body are:
the length of the data as the url length is limited while the request body is not
special characters to be escaped in the url which can lead to long and unclear url
These are 2 pros in favor of data in request body, but I agree that data in url is much simpler to test and use as tou don't need an http client tool like curl or postman to validate your endpoints.
REST however has stricter conventions if you want to fully implement it:
use the right http requests (get, post, patch, delete and put) to implement crud operations on one single endpoint
return the right http code as a result
use standard data format for input and output (json or XML)
For better interoperability between systems it's advised to comply with REST and microservices design patterns.
For small applications we can follow some shortcuts and not comply fully. I have integrated several services so far and each time I can tell you no one of them implements standard REST :-)
By going through this API documentation page, is it possible to tell which database is being used in the backend?
Zomato API
MySQL would require a php file on the server to handle the requests, make queries, pack data in JSON format then send it back to the device. But in this case parameters are passed to .json files. Please advice
There is no way to "see through" to what the backend service actually used to provide you with the information you may query for. Are you sure you want to continue using this product? The site notes that Zomato will no longer be available to individuals, and that your API key will be disabled if you don't use it monthly.
I haven't read the specs for that particular API. But in general, is it possible to tell what database is being used on the back end by studying an API? No. That's the whole point of an API: It's supposed to shield the API-user from implementation details.
It's probably true that in many cases you could make reasonable guesses about what tools are being used on the back end. Like if you see that the API gives you a syntax for doing comparisons that looks exactly like the proprietary compare function used in Foobar SQL and not found in any other database product, that would be a strong clue. But even something like that wouldn't be proof. Maybe originally they were using Foobar SQL, then they switched to another database, but to maintain compatibility they wrote code to translate the Foobar SQL compare to standard SQL syntax.
I am working on a content platform that should provide semantic features such as querying with SPARQL and providing rdf documents for the contained content.
I would be very thankful for some
clarification on the following
questions:
Did I get that right, that an entity
hub can connect several semantic
stores to a single point of access?
And if not, what is the difference
between a semantic store and an
entity hub?
What frameworks would you use to
store content documents as well as
their semantic annotation?
It is important for the solution to be able to later on retrieve the document (html page / docs such as pdf, doc,...) and their annotated version.
Thanks in advance,
Chris
The only Entityhub term that I know is belong to Apache Stanbol project. And here is a paragraph from the original documentation explaining what Entityhub does:
The Entityhub provides two main services. The Entityhub provides the
connection to external linked open data sites as well as using indexes
of them locally. Its services allow to manage a network of sites to
consume entity information and to manage entities locally.
Entityhub documentation:
http://incubator.apache.org/stanbol/docs/trunk/entityhub.html
Enhancer component of Apache Stanbol provides extracting external entities related with the submitted content using the linked open data sites managed by Entityhub. These enhancements of contents are formed as RDF data. Then, it is also possible to store those content items in Apache Stanbol and run SPARQL queries on top of RDF enhancements. Contenthub component of Apache Stanbol also provides faceted search functionality over the submitted content items.
Documentation of Apache Stanbol:
http://incubator.apache.org/stanbol/docs/trunk/
Access to running demos:
http://dev.iks-project.eu/
You can also ask your further questions to stanbol-dev AT incubator.apache.org.
Alternative suggestion...
Drupal 7 has in-built RDFa support for annotation and is more of a general purpose CMS than Semantic MediaWiki
In more detail...
I'm not really sure what you mean by entity hub, where are you getting that definition from or what do you mean by it?
Yes one can easily write a system that connects to multiple semantic stores, given the context of your question I assume you are referring to RDF Triple Stores?
Any decent CMS should be assigning documents some form of unique/persistent ID to documents so even if the system you go with does not support semantic annotation natively you could build your own extension for this. The extension would simply store annotations against the documents ID in whatever storage layer you chose (I'd assume a Triple Store would be appropriate) and then you can build appropriate query and presentation layers for querying and viewing this data as required.
http://semantic-mediawiki.org/wiki/Semantic_MediaWiki
Apache Stanbol
Do you want to implement a traditional CMS extended with some Semantic capabilities, or do you want to build a Semantic CMS? It could look the same, but actually both a two completely opposite approaches.
It is important for the solution to be able to later on retrieve the document (html page / docs such as pdf, doc,...) and their annotated version.
You can integrate Apache Stanbol with a JCR/CMIS compliant CMS like Alfresco. To get custom annotations, I suggest creating your own custom enhancement engine (maven archetype) based on your domain and adding it to the enhancement engine chain.
https://stanbol.apache.org/docs/trunk/components/enhancer/
One this is done, you can use the REST API endpoints provided by Stanbol to retrieve the results in RDF/Turtle format.
I'm building my first Rest API which serialize data to JSON and XML formats.
I would like to provide an index page to API clients, where they would be able to choose implemented endpoints.
What information do I need to include to make my API most useful, and how should I organize it?
That's a very complex question for a simple answer.
You may want to take a look at existing API frameworks, like Swagger Specification (OpenAPI), and services like apiary.io and apiblueprint.org.
Also, here's an example of the same REST API described, organized and even styled in three different ways. It may be a good start for you to learn from existing common ways.
https://api.coinsecure.in/v1
https://api.coinsecure.in/v1/originalUI
https://api.coinsecure.in/v1/slateUI#!/Blockchain_Tools/v1_bitcoin_search_txid
At the very top level I think quality REST API docs require at least the following:
a list of all your API endpoints (base/relative URLs)
corresponding HTTP GET/POST/... method type for each endpoint
request/response MIME-type (how to encode params and parse replies)
a sample request/response, including HTTP headers
type and format specified for all params, including those in the URL, body and headers
a brief text description and important notes
a short code snippet showing the use of the endpoint in popular web programming languages
Also there are a lot of JSON/XML-based doc frameworks which can parse your API definition or schema and generate a convenient set of docs for you. But the choice for a doc generation system depends on your project, language, development environment and many other things.