Bug Description When running kafka/producer/stream_producer.py, the neighbors property on each node always stays 0. Sometimes the node doesn't even contain the property.

To Reproduce Just run property initialized Memgraph + any backend + stream_producer.py.

Expected behavior As new nodes/edges come in, the neighbors should be incrementally updated.

Desktop: OS: Ubuntu 20.04

If memgraph/memgraph:latest already exists on the machine, memgraph/run.sh memgraph won't actually use the latest Memgraph image.

Probably the best would be to hardcode Memgraph version where everything is working as expected, e.g. v1.6.0.

Bumps tmpl from 1.0.4 to 1.0.5.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

• [x] Create kafka.transform module
• [x] Fix or document problem with neighbors
• [x] Polish Python backend
• [x] Remove redundant code from all backends
• [x] Update README page

closes #21

I added the --pull always flag to the docker run command. It seems weird that docker verifies the hash multiple times before running:

 memgraph fix-memgraph-version ✗ bash run.sh memgraph
19a3f4efbe827fb1bdb5d67241f424d5dec1820627b265b25e5eeda13e65c213
Starting memgraph...
latest: Pulling from memgraph/memgraph
Digest: sha256:9a45e6ad18f28edcc1e8317eb225041a3c017b7841b43d994d90155c97efd905
Status: Image is up to date for memgraph/memgraph:latest
latest: Pulling from memgraph/memgraph
Digest: sha256:9a45e6ad18f28edcc1e8317eb225041a3c017b7841b43d994d90155c97efd905
Status: Image is up to date for memgraph/memgraph:latest
latest: Pulling from memgraph/memgraph
Digest: sha256:9a45e6ad18f28edcc1e8317eb225041a3c017b7841b43d994d90155c97efd905
Status: Image is up to date for memgraph/memgraph:latest
latest: Pulling from memgraph/memgraph
Digest: sha256:9a45e6ad18f28edcc1e8317eb225041a3c017b7841b43d994d90155c97efd905
Status: Image is up to date for memgraph/memgraph:latest

• [x] fix stream producer
• [x] print number of neighbours in all backends
• [x] add a script that runs memgraph queries (why are there so many queries?)
• ~~[ ] add documentation that explains all memgraph queries~~
• [x] add create trigger on nodes that sets neighbors to 0
• [x] fix update trigger not to do bfs (only immediate neighbours are needed)
• [x] update drawing in readme (bigger arrows)

• ~~[ ] add code docs~~
• [x] decouple parts of code into functions/classes
• [x] simplify pom.xml
• ~~[ ] remove test class or add tests~~
• ~~[ ] store kafka message split parts to a Map instead of using String.format()~~
• [x] improve print messages
• [x] add kafka consumer to try arguments along with the session if possible
• ~~[ ] extract magic strings and numbers to global variables~~
• ~~[ ] keep accepting messages even if error occurs after parsing a faulty message~~
• ~~[ ] use csv reader instead of splitting~~
• ~~[ ] print number of neighbours~~
• ~~[ ] use gradle~~

• [x] Init Express and Jest
• [x] Init linter
• [x] Fetch data from Kafka
• [x] Store data to Memgraph
• [x] Calculate something by using BFS
• [x] Log to console and/or return data to the user
• [x] Report bugs in Memgraph
  • Trigger recovery + containing query module call.
  • Trigger --> comment syntax.
  • Property map feature, e.g., (n:Node $map) (Memgraph does not support that yet).
• [x] Improve count correctness (transitive is not working yet)
• [x] Improve README installation / setup details

Combine Memgraph, Kafka, Docker, Docker Compose, and Ansible to showcase simple deployment for each backend. Similar, just more polished, to https://github.com/memgraph/orbicon.