Merge branch 'master' of https://github.com/OFFIS-DAI/mango into msgs…

…_counting

.github/workflows/publish-mango.yml

@@ -0,0 +1,40 @@
+name: Upload mango
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.8'
+      - name: Install dependencies
+        run: |
+          pip install virtualenv
+          virtualenv venv
+          source venv/bin/activate
+          pip3 install -r requirements.txt
+          pip3 install -e .
+          sudo apt update
+          sudo apt install --assume-yes mosquitto
+          sudo service mosquitto start
+          pip3 install pytest
+          pip3 install coverage
+      - name: Test+Coverage
+        run: |
+          source venv/bin/activate
+          coverage run --source ./mango -m pytest
+          coverage report
+      - name: Build package
+        run: |
+          source venv/bin/activate
+          python setup.py sdist bdist_wheel
+      - name: Publish package
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          password: ${{ secrets.PYPI_API_TOKEN }}

.github/workflows/test-mango.yml

@@ -0,0 +1,34 @@
+name: Test mango
+
+on: [push]
+
+permissions:
+  contents: read
+
+jobs: 
+  build:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: '3.8'
+    - name: Install dependencies
+      run: |
+        pip install virtualenv
+        virtualenv venv
+        source venv/bin/activate
+        pip3 install -r requirements.txt
+        pip3 install -e .
+        sudo apt update
+        sudo apt install --assume-yes mosquitto
+        sudo service mosquitto start
+        pip3 install pytest
+        pip3 install coverage
+    - name: Test+Coverage
+      run: |
+        source venv/bin/activate
+        coverage run --source ./mango -m pytest
+        coverage report

.readthedocs.yaml

@@ -5,6 +5,11 @@
 # Required
 version: 2
 
+build:
+  os: ubuntu-22.04
+  tools:
+    python: "3.7"
+
 # Build documentation in the docs/ directory with Sphinx
 sphinx:
    configuration: docs/source/conf.py
@@ -15,6 +20,5 @@ formats:
 
 # Optionally set the version of Python and requirements required to build your docs
 python:
-   version: "3.7"
    install:
-   - requirements: docs/requirements.txt
+   - requirements: docs/requirements.txt

docs/source/agents-container.rst

@@ -8,7 +8,7 @@ mango container
 
 In mango, agents live in a ``container``. The container is responsible for everything network related of the agent.
 This includes in particular sending and receiving of messages, but also message distribution to the correct agent or
-(de-)serialization of messages.
+serialization and deserialization of messages.
 Container also help to to speed up message exchange between agents that run on the same physical hardware,
 as data that is exchanged between such agents will not have to be sent through the network.
 

docs/source/conf.py

@@ -22,7 +22,7 @@
 # -- Project information -----------------------------------------------------
 
 project = 'mango'
-copyright = '2021, mango team'
+copyright = '2023, mango team'
 author = 'mango team'
 
 # The full version, including alpha/beta/rc tags

docs/source/development.rst

@@ -7,7 +7,7 @@ As we mainly work in a critical domain, we set great value on code quality not o
 Quickstart
 ##########
 
-In mango it is not possible to directly push on the branches *development* or *master*. Both branches are protected and changes can only be merged using a gitlab merge-request. So when you work on a feature, the typical process would be to create a feature-branch. When you are finished you just have to create a merge-request, pass the CI/CD pipeline, make a maintainer review the changes, and you are ready to merge! 
+In mango it is not possible to directly push on the branches *development* or *master*. Both branches are protected and changes can only be merged using a github pull-request. So when you work on a feature, the typical process would be to create a feature-branch. When you are finished you just have to create a merge-request, pass the CI/CD pipeline, make a maintainer review the changes, and you are ready to merge! 
 
 CI/CD
 #####

docs/source/impressum.rst

@@ -34,7 +34,7 @@ DE 811582102
 
 **Verantwortlich im Sinne der Presse**
 
-| Dr. Christoph Mayer (Bereichsleiter)
+| Dr. Ing. Jürgen Meister (Bereichsleiter)
 | OFFIS e.V.
 | Escherweg 2
 | 26121 Oldenburg

docs/source/legals.rst

@@ -35,7 +35,7 @@ DE 811582102
 
 **Responsible in the sense of press law**
 
-| Dr. Christoph Mayer (Director)
+| Dr. Ing. Jürgen Meister (Director)
 | OFFIS e.V.
 | Escherweg 2
 | 26121 Oldenburg

docs/source/scheduling.rst

@@ -20,6 +20,12 @@ The core of this API is the scheduler, which is part of every agent. To schedule
      - Executes a coroutine periodically with a static delay between the cycles
    * - ConditionalScheduledTask
      - Executes the coroutine when a specified condition evaluates to True
+   * - AwaitingTask
+     - Execute a given coroutine after another given coroutine has been awaited
+   * - RecurrentScheduledTask
+     - Will get executed periodically with a specified delay
+
+
 
 Furthermore there are convenience methods to get rid of the class imports when using these types of tasks.
 
@@ -56,6 +62,28 @@ Dispatch Tasks to other Process
 As asyncio does not provide real parallelism to utilize multiple cores and agents may have tasks, which need a lot computational power, the need to dispatch certain tasks to other processes appear. Handling inter process communication manually is quite exhausting and having multiple process pools across different roles or agents leads to inefficient resource allocations. As a result mango offers a way to dispatch tasks, based on coroutine-functions, to other processes, managed by the framework. 
 
 Analogues to the normal API there are two different ways, first you create a ScheduledProcessTask and call ``schedule_process_task``, second you invoke the convnience methods with "process" in the name. These methods exists on any Agent, the RoleContext and the Scheduler.
+In mango the following process tasks are available:
+
+.. list-table:: Available ProcessTasks
+   :widths: 30 70
+   :header-rows: 1
+
+   * - Class
+     - Description
+   * - ScheduledProcessTask
+     - Marks a ScheduledTask as process compatible
+   * - TimestampScheduledProcessTask
+     - Timestamp based one-shot task
+   * - AwaitingProcessTask
+     - Await a coroutine, then execute another
+   * - InstantScheduledProcessTask
+     - One-shot task, which will get executed instantly
+   * - PeriodicScheduledProcessTask
+     - Executes a coroutine periodically with a static delay between the cycles
+   * - RecurrentScheduledProcessTask
+     - Will get executed periodically with a specified delay
+   * - ConditionalProcessTask
+     - Will get executed as soon as the given condition is fulfilled
 
 .. code-block:: python3
 

examples/tutorial/v2_inter_container_messaging_and_basic_functionality.py

@@ -50,7 +50,7 @@ def handle_ask_feed_in(self, sender_addr, sender_id):
         content = reported_feed_in
 
         acl_meta = {
-            "sender_addr": self.context.addr,
+            "sender_addr": self._context.addr,
             "sender_id": self.aid,
             "performative": Performatives.inform,
         }
@@ -68,7 +68,7 @@ def handle_set_feed_in_max(self, max_feed_in, sender_addr, sender_id):
         self.max_feed_in = float(max_feed_in)
         print(f"{self.aid}: Limiting my feed_in to {max_feed_in}")
         self.schedule_instant_task(
-            self.context.send_acl_message(
+            self._context.send_acl_message(
                 content=None,
                 receiver_addr=sender_addr,
                 receiver_id=sender_id,
@@ -125,13 +125,13 @@ async def run(self):
         for addr, aid in self.known_agents:
             content = None
             acl_meta = {
-                "sender_addr": self.context.addr,
+                "sender_addr": self._context.addr,
                 "sender_id": self.aid,
                 "performative": Performatives.request,
             }
             # alternatively we could call send_acl_message here directly and await it
             self.schedule_instant_task(
-                self.context.send_acl_message(
+                self._context.send_acl_message(
                     content=content,
                     receiver_addr=addr,
                     receiver_id=aid,
@@ -149,14 +149,14 @@ async def run(self):
         for addr, aid in self.known_agents:
             content = min_feed_in
             acl_meta = {
-                "sender_addr": self.context.addr,
+                "sender_addr": self._context.addr,
                 "sender_id": self.aid,
                 "performative": Performatives.propose,
             }
 
             # alternatively we could call send_acl_message here directly and await it
             self.schedule_instant_task(
-                self.context.send_acl_message(
+                self._context.send_acl_message(
                     content=content,
                     receiver_addr=addr,
                     receiver_id=aid,

examples/tutorial/v3_codecs_and_typing.py

@@ -69,7 +69,7 @@ def handle_ask_feed_in(self, sender_addr, sender_id):
         msg = FeedInReplyMsg(reported_feed_in)
 
         self.schedule_instant_task(
-            self.context.send_acl_message(
+            self._context.send_acl_message(
                 content=msg,
                 receiver_addr=sender_addr,
                 receiver_id=sender_id,
@@ -82,7 +82,7 @@ def handle_set_feed_in_max(self, max_feed_in, sender_addr, sender_id):
         msg = MaxFeedInAck()
 
         self.schedule_instant_task(
-            self.context.send_acl_message(
+            self._context.send_acl_message(
                 content=msg,
                 receiver_addr=sender_addr,
                 receiver_id=sender_id,
@@ -132,11 +132,11 @@ async def run(self):
         # ask pv agent feed-ins
         for addr, aid in self.known_agents:
             msg = AskFeedInMsg()
-            acl_meta = {"sender_addr": self.context.addr, "sender_id": self.aid}
+            acl_meta = {"sender_addr": self._context.addr, "sender_id": self.aid}
 
             # alternatively we could call send_acl_message here directly and await it
             self.schedule_instant_task(
-                self.context.send_acl_message(
+                self._context.send_acl_message(
                     content=msg,
                     receiver_addr=addr,
                     receiver_id=aid,
@@ -153,11 +153,11 @@ async def run(self):
 
         for addr, aid in self.known_agents:
             msg = SetMaxFeedInMsg(min_feed_in)
-            acl_meta = {"sender_addr": self.context.addr, "sender_id": self.aid}
+            acl_meta = {"sender_addr": self._context.addr, "sender_id": self.aid}
 
             # alternatively we could call send_acl_message here directly and await it
             self.schedule_instant_task(
-                self.context.send_acl_message(
+                self._context.send_acl_message(
                     content=msg,
                     receiver_addr=addr,
                     receiver_id=aid,

mango/agent/core.py

@@ -224,41 +224,6 @@ def schedule_conditional_task(
             src=src,
         )
 
-    def schedule_datetime_process_task(
-        self, coroutine_creator, date_time: datetime, on_stop=None, src=None
-    ):
-        """Schedule a task at specified datetime in another process.
-
-        :param coroutine_creator: coroutine_creator creating couroutine to be scheduled
-        :type coroutine_creator: Coroutine-creator
-        :param date_time: datetime defining when the task should start
-        :type date_time: datetime
-        :param src: creator of the task
-        :type src: Object
-        """
-        return self._scheduler.schedule_datetime_process_task(
-            coroutine_creator=coroutine_creator,
-            date_time=date_time,
-            on_stop=on_stop,
-            src=src,
-        )
-
-    def schedule_datetime_task(
-        self, coroutine, date_time: datetime, on_stop=None, src=None
-    ):
-        """Schedule a task at specified datetime.
-
-        :param coroutine: coroutine to be scheduled
-        :type coroutine: Coroutine
-        :param date_time: datetime defining when the task should start
-        :type date_time: datetime
-        :param src: creator of the task
-        :type src: Object
-        """
-        return self._scheduler.schedule_datetime_task(
-            coroutine=coroutine, date_time=date_time, on_stop=on_stop, src=src
-        )
-
     def schedule_timestamp_task(
         self, coroutine, timestamp: float, on_stop=None, src=None
     ):