Periodic Reporting for period 2 - BEHAPI (Behavioural Application Program Interfaces)
Période du rapport: 2020-03-01 au 2023-12-31
The primary goal of this project is to tackle two major shortcomings currently present in the API economy:
(i) the lack of rigour in documentation practices of APIs and
(ii) the lack of support for provably correct methodologies in the round-trip engineering of APIs.
We advocate behavioural APIs with precise descriptions of their behaviour to validate component-based applications. Behavioural interfaces will enable more effective Software Development Life-Cycles (SDLCs) for the API economy. Traditional approaches to software development do not cope well with the development of open, scalable communication-centric systems. In particular,
- software is traditionally built in a top-down fashion, from requirements and specifications to code;
- current bottom-up approaches are very limited and not linked to the documentation of software;
- the continuous evolution of APIs makes the adoption of round-trip software engineering a requirement for SDLCs as opposed to more traditional top-down approaches;
- developing team(s) assume context boundaries for subsystems which create duplication of work or impose conflicting requirements.
We envisage behavioural specifications as lightweight yet effective instruments. Such specifications precisely and (semi-)automatically document components and characterise their behaviour. Behavioural interfaces facilitate several software engineering activities such as component upgrades, their dynamic composition, the evaluation of their compatibility, etc. This, in turn, requires support for developing “open” protocols. Bottom-up approaches allow different teams to develop parts of systems independently to make components available to other projects. Moreover, bottom-up approaches allow the handling of legacy code.
BehAPI enables the definition of techniques and tools for designing, testing, and deploying APIs and API-based software while reducing the time-to-market of software production. It extends existing prototypes developed in academia, and integrates them with industrial platforms and languages. The consortium tackles the following challenges:
- How to formally document the behaviour of existing API, extending existing descriptions which only specify parameter type signature.
- How to identify the often “hidden” communication at different layers of abstraction in order to attain scalability.
- How to ensure a smooth integration of third-party APIs with the system/app being developed.
- How to cope with upgrades while minimising disruptions to running (live) systems.
BehAPI supports developers by easing the API documentation process. It considers automata, types, DSLs, and graphs, to enrich interfaces with behavioural information used as verifiable and lightweight documentation. More concretely, it
- provides models and tools for the analysis, construction, and validation of API-based software.
- integrates said models together, provide algorithms and tools based on them, and embed them within existing industrial development frameworks and SDLCs.
- enhances testing for APIs and provide methodologies and tools for supporting traceability among the different artefacts.
- extends industry-strength programming languages with primitives for programming b-APIs.
- defines specification and verification techniques for (non-)functional properties, resilience, and dependability.
Conclusions:
The project reached all of its 10 milestones and produced all of the 17 deliverables. It contributed
- 44 open-access publications,
- 27 open-source tools,
- a project website,
- 2 summer schools and 1 boot camp,
- 2 industrial case studies,
- 23 tutorials and exercises,
- numerous participations at inters-ectorial fora.
(i) the lack of rigour in documentation practices of APIs and
(ii) the lack of support for provably correct methodologies in the round-trip engineering of APIs.
We advocate behavioural APIs with precise descriptions of their behaviour to validate component-based applications. Behavioural interfaces will enable more effective Software Development Life-Cycles (SDLCs) for the API economy. Traditional approaches to software development do not cope well with the development of open, scalable communication-centric systems. In particular,
- software is traditionally built in a top-down fashion, from requirements and specifications to code;
- current bottom-up approaches are very limited and not linked to the documentation of software;
- the continuous evolution of APIs makes the adoption of round-trip software engineering a requirement for SDLCs as opposed to more traditional top-down approaches;
- developing team(s) assume context boundaries for subsystems which create duplication of work or impose conflicting requirements.
We envisage behavioural specifications as lightweight yet effective instruments. Such specifications precisely and (semi-)automatically document components and characterise their behaviour. Behavioural interfaces facilitate several software engineering activities such as component upgrades, their dynamic composition, the evaluation of their compatibility, etc. This, in turn, requires support for developing “open” protocols. Bottom-up approaches allow different teams to develop parts of systems independently to make components available to other projects. Moreover, bottom-up approaches allow the handling of legacy code.
BehAPI enables the definition of techniques and tools for designing, testing, and deploying APIs and API-based software while reducing the time-to-market of software production. It extends existing prototypes developed in academia, and integrates them with industrial platforms and languages. The consortium tackles the following challenges:
- How to formally document the behaviour of existing API, extending existing descriptions which only specify parameter type signature.
- How to identify the often “hidden” communication at different layers of abstraction in order to attain scalability.
- How to ensure a smooth integration of third-party APIs with the system/app being developed.
- How to cope with upgrades while minimising disruptions to running (live) systems.
BehAPI supports developers by easing the API documentation process. It considers automata, types, DSLs, and graphs, to enrich interfaces with behavioural information used as verifiable and lightweight documentation. More concretely, it
- provides models and tools for the analysis, construction, and validation of API-based software.
- integrates said models together, provide algorithms and tools based on them, and embed them within existing industrial development frameworks and SDLCs.
- enhances testing for APIs and provide methodologies and tools for supporting traceability among the different artefacts.
- extends industry-strength programming languages with primitives for programming b-APIs.
- defines specification and verification techniques for (non-)functional properties, resilience, and dependability.
Conclusions:
The project reached all of its 10 milestones and produced all of the 17 deliverables. It contributed
- 44 open-access publications,
- 27 open-source tools,
- a project website,
- 2 summer schools and 1 boot camp,
- 2 industrial case studies,
- 23 tutorials and exercises,
- numerous participations at inters-ectorial fora.
We have surveyed the state-of-the-art and assembled portfolios containing theories, tools and prototypes for API-based software development using behavioural information (WP2, Task T2.1 WP3 Task T3.1 WP4, Task T4.1). We have also engaged in disseminating these results with the public via workshops, exhibitions, boot camps and summer schools (WP5 Tasks T5.2-8). Concretely, the results achieved up to this point are:
1. The following channels for internal and external communication have been established (Task 5.6):
(i) two mailing lists to communicate within the consortium.
(ii) a web portal, www.um.edu.mt/projects/behapi/ hosting repositories and information that can be accessed by the general public.
(iii) two social media channels: www.facebook.com/behapi.rise and www.twitter.com/behapi_rise used for communication and dissemination purposes.
2. Courses for graduate researchers interested in the area of communication-based concurrency supported by behavioural types were developed, used at summer schools and as tutorials at top peer-reviewed conferences (Tasks T5.4).
3. Various workshops and one BehAPI Summer School/Boot Camp were organised offering courses from both academia and industry concerning both the foundations and practice of API-based software development. (Tasks T5.2 T5.3)
4. State-of-the-art reports and portfolios were compiled on the following areas of API software development:
(i) current practices of API provision (Task T2.1)
(ii) current practices of API consumption (Task T3.1)
(iii) tools, prototypes and platforms for API-based software development. (Task T4.1)
5. Ongoing development of languages and models to specify and extract behavioural API information for existing software. (Tasks T2.2 T3.2)
6. Ongoing study of dynamic and static analysis techniques to assist the development of new API-based software and the integration of existing software (Tasks T2.4 T2.5 T3.4 T3.5).
7. Ongoing development of tools and platforms realising the aforementioned models and verification techniques to form part of existing software development life-cycles (Tasks T4.2 T4.3 T4.4).
8. 20 publications were produced by members of the consortium reporting on the above findings.
1. The following channels for internal and external communication have been established (Task 5.6):
(i) two mailing lists to communicate within the consortium.
(ii) a web portal, www.um.edu.mt/projects/behapi/ hosting repositories and information that can be accessed by the general public.
(iii) two social media channels: www.facebook.com/behapi.rise and www.twitter.com/behapi_rise used for communication and dissemination purposes.
2. Courses for graduate researchers interested in the area of communication-based concurrency supported by behavioural types were developed, used at summer schools and as tutorials at top peer-reviewed conferences (Tasks T5.4).
3. Various workshops and one BehAPI Summer School/Boot Camp were organised offering courses from both academia and industry concerning both the foundations and practice of API-based software development. (Tasks T5.2 T5.3)
4. State-of-the-art reports and portfolios were compiled on the following areas of API software development:
(i) current practices of API provision (Task T2.1)
(ii) current practices of API consumption (Task T3.1)
(iii) tools, prototypes and platforms for API-based software development. (Task T4.1)
5. Ongoing development of languages and models to specify and extract behavioural API information for existing software. (Tasks T2.2 T3.2)
6. Ongoing study of dynamic and static analysis techniques to assist the development of new API-based software and the integration of existing software (Tasks T2.4 T2.5 T3.4 T3.5).
7. Ongoing development of tools and platforms realising the aforementioned models and verification techniques to form part of existing software development life-cycles (Tasks T4.2 T4.3 T4.4).
8. 20 publications were produced by members of the consortium reporting on the above findings.
Progress beyond the state of the art:
1. New results regarding the behavioural type subtyping for APIs in an asynchronous setting.
2. New results on incorporating probabilistic information to API behavioural descriptions.
3. New results on the hierarchical classification on the monitorability of linear-time properties.
4. New toolchain development bringing together existing and newly improved tools.
Expected results until the end of the project:
1. The adaptation of existing behavioural type theory to the realities of API provision and consumption.
2. The development of tools manifesting these adaptations.
3. The integration of these tools in an API Software Development Life-cycle process.
Potential impacts:
1. (Short term) Adoption of our tools and methods by the software development community.
2. (Medium term) The qualitative improvement of API-based software in terms of aspects such as reliability and security.
3. (Long term) Further lowering of barriers to entry to the API Economy and fairer markets with increased competition potential
4. (Long term) The proliferation of more API-based software into everyday life, improving the well-being of EU citizens in terms of information access, healthcare, etc
1. New results regarding the behavioural type subtyping for APIs in an asynchronous setting.
2. New results on incorporating probabilistic information to API behavioural descriptions.
3. New results on the hierarchical classification on the monitorability of linear-time properties.
4. New toolchain development bringing together existing and newly improved tools.
Expected results until the end of the project:
1. The adaptation of existing behavioural type theory to the realities of API provision and consumption.
2. The development of tools manifesting these adaptations.
3. The integration of these tools in an API Software Development Life-cycle process.
Potential impacts:
1. (Short term) Adoption of our tools and methods by the software development community.
2. (Medium term) The qualitative improvement of API-based software in terms of aspects such as reliability and security.
3. (Long term) Further lowering of barriers to entry to the API Economy and fairer markets with increased competition potential
4. (Long term) The proliferation of more API-based software into everyday life, improving the well-being of EU citizens in terms of information access, healthcare, etc