repo_name
stringlengths
1
62
dataset
stringclasses
1 value
lang
stringclasses
11 values
pr_id
int64
1
20.1k
owner
stringlengths
2
34
reviewer
stringlengths
2
39
diff_hunk
stringlengths
15
262k
code_review_comment
stringlengths
1
99.6k
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,366 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
switch to `total`
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,366 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
Remove the input, as this will be too large for our API
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,366 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
there is no user_id on span
langfuse-python
github_2023
others
384
langfuse
maxdeichmann
@@ -7,14 +7,15 @@ license = "MIT" readme = "README.md" [tool.poetry.dependencies] -python = ">=3.8.1,<4.0" +python = ">=3.8.1,<3.12" httpx = ">=0.15.4,<0.26.0" pydantic = ">=1.10.7, <3.0" backoff = "^2.2.1" openai = ">=0.27.8" wrapt = "1.14" langchain = { version = ">=0.0.309", optional = true } chevron = "^...
Do you know by any chance ntil when this implementation is stable? It would be great to also support versions which are a couple of months old.
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,369 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
I think it would be great to have variable declaraitons for these at the top at class level to help our IDEs.
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,369 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
Could you add a quick note that the trace_id is not a uuid or similar but rather a task name?
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,369 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
I think we would want to keep the additional_kwargs in case there is something in there.
langfuse-python
github_2023
python
384
langfuse
maxdeichmann
@@ -0,0 +1,369 @@ +from collections import defaultdict +from typing import Any, Dict, List, Optional, Union, Tuple, Callable +from uuid import uuid4 +import logging + +from langfuse.client import ( + StatefulSpanClient, + StatefulTraceClient, + StatefulGenerationClient, +) +from langfuse.decorators.error_loggi...
can you add a short comment on why we do this at the end and the performance implications for index + query?
langfuse-python
github_2023
python
367
langfuse
maxdeichmann
@@ -0,0 +1,94 @@ +from contextlib import asynccontextmanager +from fastapi import FastAPI, Query, BackgroundTasks +import os +import sys +from langfuse import Langfuse +from langfuse.openai import openai +import uvicorn + + +@asynccontextmanager +async def lifespan(app: FastAPI): + # Operation on startup + + yiel...
I think we can remove the function parameters here, right?
langfuse-python
github_2023
python
367
langfuse
maxdeichmann
@@ -0,0 +1,94 @@ +from contextlib import asynccontextmanager +from fastapi import FastAPI, Query, BackgroundTasks +import os +import sys +from langfuse import Langfuse +from langfuse.openai import openai +import uvicorn +
We can also remove all the unused imports here
langfuse-python
github_2023
others
367
langfuse
maxdeichmann
@@ -0,0 +1,24 @@ +# fastapi_example + +This is an example FastAPI application showcasing integration with Langfuse for event tracing and response generation. + +1. **Shutdown Behavior**: The application defines shutdown logic using FastAPI's lifespan feature. On shutdown, it flushes all events to Langfuse, ensuring dat...
I think we can remove openai here, as this is not used, right?
langfuse-python
github_2023
others
367
langfuse
maxdeichmann
@@ -0,0 +1,19 @@ +[tool.poetry] +name = "fastapi-example" +version = "0.1.0" +description = "" +authors = ["ChrisTho23 <christophe.thomassin23@gmail.com>"] +readme = "README.md" + +[tool.poetry.dependencies] +python = "^3.10" +fastapi = "^0.109.2" +uvicorn = "^0.27.1" +langfuse = "2.13.3a0"
can you use latest langfuse here always?
langfuse-python
github_2023
python
367
langfuse
maxdeichmann
@@ -0,0 +1,11 @@ +from django.shortcuts import render +from django.http import JsonResponse, HttpResponseServerError +from myapp.langfuse_integration import get_response_openai, langfuse
i think here are some unused imports
langfuse-python
github_2023
others
367
langfuse
maxdeichmann
@@ -0,0 +1,16 @@ +[tool.poetry] +name = "django-example" +version = "0.1.0" +description = "" +authors = ["ChrisTho23 <christophe.thomassin23@gmail.com>"] +readme = "README.md" + +[tool.poetry.dependencies] +python = "^3.10" +django = "^5.0.2" +langfuse = "^2.13.3"
Can we also use latest here always?
langfuse-python
github_2023
python
367
langfuse
maxdeichmann
@@ -0,0 +1,16 @@ +import os +import signal +import sys +from .langfuse_integration import langfuse_flush + +def shutdwon_handler(*args): + """ + This function handles the shutdown process. + + It calls the langfuse_flush function to flush any pending changes, + and then exits the program with a s...
I think we can also listen to sigterm here as this is used to gracefully terminate a process. https://www.baeldung.com/linux/sigint-and-other-termination-signals
langfuse-python
github_2023
others
343
langfuse
maxdeichmann
@@ -79,3 +79,11 @@ poetry run pre-commit install - Create PyPi API token: https://pypi.org/manage/account/token/ - Setup: `poetry config pypi-token.pypi your-api-token` 9. Create a release on GitHub with the changelog + +### SDK Reference + +The SDK reference is generated via pdoc. To update the reference, run...
can you also add a comment on hot wo install the docs group before running this command?
langfuse-python
github_2023
python
332
langfuse
maxdeichmann
@@ -23,7 +23,7 @@ def __init__( base_url: str, version: str, timeout: int, - session: httpx.Client, + session: httpx.Client = httpx.Client(),
why is this change required?
langfuse-python
github_2023
python
326
langfuse
maxdeichmann
@@ -59,3 +59,13 @@ def __init__(self, prompt: Prompt): def compile(self, **kwargs) -> str: return chevron.render(self.prompt, kwargs) + + def __eq__(self, other):
This here is only used in testing code at the moment, correct?
langfuse-python
github_2023
python
102
langfuse
maxdeichmann
@@ -524,7 +538,9 @@ def on_llm_end( ) self.runs[run_id].state = self.runs[run_id].state.update( - UpdateGeneration(completion=extracted_response, end_time=datetime.now(), usage=llm_usage) + UpdateGeneration( + completion=ex...
Why do you want to put the token metadata on the metadata as well? We are already tracking that in `usage`.
langfuse-python
github_2023
python
150
langfuse
maxdeichmann
@@ -652,6 +652,36 @@ def initialize_huggingface_llm(prompt: PromptTemplate) -> LLMChain: assert observation["output"] is not None assert observation["output"] != "" +def test_callback_huggingface_pipeline():
``` tests/test_langchain.py::test_callback_huggingface_pipeline FAILED ============================================================================================= FAILURES ============================================================================================== ______________________________________________...
langfuse-python
github_2023
python
215
langfuse
marcklingen
@@ -274,7 +294,8 @@ def on_agent_action( if run_id not in self.runs: raise Exception("run not found") - self.runs[run_id] = self.runs[run_id].update(UpdateSpan(endTime=datetime.now(), output=action, version=self.version)) + self.runs[run_id] = self.runs[run_id].upda...
why arent agent actions spans?
langfuse-python
github_2023
python
215
langfuse
marcklingen
@@ -415,82 +597,228 @@ def _add_default_values(self, body: dict): body["start_time"] = datetime.now() return body - def generation(self, body: CreateGeneration): + def generation( + self, + *, + id: typing.Optional[str] = None, + name: typing.Optional[str] = Non...
event() does not exist in `class Langfuse`
langfuse-python
github_2023
python
215
langfuse
marcklingen
@@ -143,16 +158,29 @@ def _get_langfuse_data_from_kwargs(resource: OpenAiDefinition, langfuse: Langfus modelParameters = { "temperature": kwargs.get("temperature", 1), - "maxTokens": kwargs.get("max_tokens", float("inf")), + "maxTokens": kwargs.get("max_tokens", float("inf")), # casing?
max_tokens? added comment to be able to comment on this
langfuse-python
github_2023
python
134
langfuse
maxdeichmann
@@ -466,6 +466,8 @@ def __on_llm_action( ) if kwargs["invocation_params"]["_type"] in ["anthropic-llm", "anthropic-chat"]: model_name = "anthropic" # unfortunately no model info by anthropic provided. + elif kwargs["invocation_params"]["_type"] == "amazon_bedro...
@kobrinartem thanks for the contribution! Did you test whether Langchain somehow provides the model name?
langfuse-python
github_2023
python
144
langfuse
maxdeichmann
@@ -11,7 +11,10 @@ class TraceWithDetails(Trace): observations: typing.List[str] = pydantic.Field(description=("List of observation ids\n")) - scores: typing.List[str] = pydantic.Field(description=("List of score ids\n")) + scores: typing.List[str] = pydantic.Field( + description=("List of scor...
is this wanted?
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,16 @@ +from dotenv import load_dotenv +from langfuse.integrations import openai + +load_dotenv() + + +def test_openai_chat_completion(): + completion = openai.ChatCompletion.create( + model="gpt-3.5-turbo", messages=[{"role": "user", "content": "1 + 1 = "}], temperature=0 + ) + assert len(com...
can you expose the flush function and call it in the tests? Also, did you try running this locally?
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,83 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion + +from langfuse import Langfuse +from langfuse.client import InitialGeneration +from langfuse.api.resources.commons.types.llm_usage ...
Why is this needed here? So far this worked without
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,83 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion + +from langfuse import Langfuse +from langfuse.client import InitialGeneration +from langfuse.api.resources.commons.types.llm_usage ...
I would maybe rename the class to reflect a bit better of how it works.
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,83 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion + +from langfuse import Langfuse +from langfuse.client import InitialGeneration +from langfuse.api.resources.commons.types.llm_usage ...
maybe `replace_openai_funcs`?
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,40 @@ +import os +from dotenv import load_dotenv +from langfuse.integrations import openai +from langfuse.api.client import FintoLangfuse +from langfuse.version import __version__ as version + +from tests.utils import create_uuid + + +load_dotenv() + +api = FintoLangfuse( + environment=os.environ["HOST"],...
I just tested this out locally. openai is shown on my end as type any in the IDE, whereas when using the original openai SDK, i get autocompletions. Is there a way to solve that?
langfuse-python
github_2023
python
103
langfuse
maxdeichmann
@@ -0,0 +1,99 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion + +from langfuse import Langfuse +from langfuse.client import InitialGeneration +from langfuse.api.resources.commons.types.llm_usage ...
Do you think we can instantiate this as a Singleton? I fear our users will import this multiple times and then create multiple Langfuse objects under the hood. I think one instantiation of the Langfuse object would be ideal. Could we also expose the flush function from there? If yes, we can remove it above in the `_l...
langfuse-python
github_2023
python
103
langfuse
fxmb
@@ -0,0 +1,99 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion
@maxdeichmann we should probably not fully import "lesser known" packages directly to provide more context. Can we change this to `from openai import api_resources` and then call like `api_resources.ChatCompletion`? Just a thought
langfuse-python
github_2023
python
103
langfuse
fxmb
@@ -0,0 +1,99 @@ +import os +import functools +from datetime import datetime +from dotenv import load_dotenv + +import openai +from openai.api_resources import ChatCompletion, Completion + +from langfuse import Langfuse +from langfuse.client import InitialGeneration +from langfuse.api.resources.commons.types.llm_usage ...
This will throw an error if key is not present afaik. Is this intended? Rather use `getenv()` and provide fallback logic in case key is not present? It's a personal preference so feel free to ignore.
langfuse-python
github_2023
python
116
langfuse
github-advanced-security[bot]
@@ -1,5 +1,17 @@ +import os from uuid import uuid4 +from langfuse.api.client import FintoLangfuse + def create_uuid(): return str(uuid4()) + + +def get_api(): + print(os.environ.get("LANGFUSE_PUBLIC_KEY"), os.environ.get("LANGFUSE_SECRET_KEY"), os.environ.get("LANGFUSE_HOST"))
## Clear-text logging of sensitive information This expression logs [sensitive data (secret)](1) as clear text. [Show more details](https://github.com/langfuse/langfuse-python/security/code-scanning/1)
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -36,9 +36,6 @@ class Langfuse(object): def __init__( self, - public_key: str, - secret_key: str, - host: Optional[str] = None,
For backwards compatibility, I'd suggest to keep the constructor arguments (public, secret, host) and make them optional. This applies to the SDK and the langchain callback handler. If set, constructor arguments then take precedence over environment variable. Removing the argument leads to unexpected behavior for exi...
langfuse-python
github_2023
python
100
langfuse
maxdeichmann
@@ -34,14 +35,17 @@ def __init__( statefulTraceClient: Optional[StatefulTraceClient] = None, release: Optional[str] = None, ) -> None: + public_key = public_key if public_key else os.environ.get("LF_PK") + secret_key = secret_key if secret_key else os.environ.get("LF_SK") + ...
Should we set the keys here?
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -31,6 +32,9 @@ def __init__( statefulClient: Optional[StatefulTraceClient | StatefulSpanClient] = None, release: Optional[str] = None, ) -> None: + public_key = public_key if public_key else os.environ.get("LF_PK")
Environment variables will be checked in Langfuse client, no need to do this here again. I'd suggest to just pass public_key and secret_key to Langfuse()
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -71,15 +71,19 @@ def __init__( self.task_manager = TaskManager() + public_key = public_key if public_key else os.environ.get("LF_PK") + secret_key = secret_key if secret_key else os.environ.get("LF_SK") + host = host if host else os.environ.get("HOST") + self.base_url = hos...
This suggests that env var is the only option. I'd suggest `"public_key is required, set as parameter or environment variable "LF_PK""`
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -71,15 +71,19 @@ def __init__( self.task_manager = TaskManager() + public_key = public_key if public_key else os.environ.get("LF_PK") + secret_key = secret_key if secret_key else os.environ.get("LF_SK") + host = host if host else os.environ.get("HOST") + self.base_url = hos...
typo, should be secret_key
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -1,10 +1,13 @@ +import os import requests class LangfuseAPI: - def __init__(self, username=None, password=None, base_url="http://localhost:3000"): - self.auth = (username, password) if username and password else None - self.BASE_URL = base_url + def __init__(self, username=None, password=No...
Is this necessary? I'd assume that we never create the LangfuseAPI directly but only via Langfuse()
langfuse-python
github_2023
python
100
langfuse
marcklingen
@@ -67,23 +73,28 @@ def test_flush(): def test_setup_wthout_pk(): # set up the consumer with more requests than a single batch will allow + os.environ.pop("LF_PK") with pytest.raises(ValueError): - Langfuse(public_key=None, secret_key=os.environ.get("LF_SK")) + Langfuse() + os.environ["...
Great, also add separate tests for LF_SK and LF_HOST parameter. Goal is to check that overriding env in constructor works
langfuse-python
github_2023
others
4
langfuse
derino
@@ -0,0 +1,76 @@ +on: + pull_request: + push: + branches: + - main + - master + workflow_dispatch: + +concurrency: + group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }} + cancel-in-progress: false + +jobs: + ci: + runs-on: ubuntu-latest + env: # Or as an environment variable +...
@maxdeichmann The tests don't seem to be run on the python version that is set up: https://github.com/langfuse/langfuse-python/actions/runs/6309393965/job/17129223793?pr=107#step:12:16
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -12,6 +12,7 @@ class UpdateSpanRequest(pydantic.BaseModel): span_id: str = pydantic.Field(alias="spanId") trace_id: typing.Optional[str] = pydantic.Field(alias="traceId") + name: typing.Optional[str]
Good catch! All files in langfuse/api are auto-generated by fern. Added a PR on the core project and will update this PR as well https://github.com/langfuse/langfuse/pull/101 By keeping the openapi docs in core in sync, we can also update the TS SDK accordingly.
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -138,8 +139,7 @@ def create_trace(): def create_span(): try: - new_body = body.copy(update={"id": new_span_id}) - new_body = body.copy(update={"trace_id": new_trace_id}) + new_body = body.copy(update={"id": new_span_id, "trace_i...
good catch!
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -179,8 +179,7 @@ def create_trace(): def create_generation(): try: - new_body = body.copy(update={"id": new_generation_id}) - new_body = body.copy(update={"trace_id": new_trace_id}) + new_body = body.copy(update={"id": new_gener...
same here!
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -190,8 +189,8 @@ def create_generation(): self.log.exception(e) raise e - self.task_manager.add_task(new_generation_id, create_generation) self.task_manager.add_task(new_trace_id, create_trace) + self.task_manager.add_task(new_generation_...
Ordering does not matter, but more logical to have it this way
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -358,7 +357,7 @@ def update(self, body: UpdateGeneration): def task(): try: - new_body = body.copy(update={"generation_id": self.id}) + new_body = body.copy(update={"generation_id": self.id, "trace_id": self.trace_id})
Why this change? It does not destroy anything but did it not work without?
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -424,3 +426,76 @@ def test_create_trace_with_id_and_generation(): span = trace["observations"][0] assert span["name"] == "generation" assert span["traceId"] == trace["id"] + + +def test_end_generation(): + langfuse = Langfuse(os.environ.get("LF_PK"), os.environ.get("LF_SK"), os.environ.get("HOST"))...
Do you think these type of tests work reliably as they base on time? Maybe we should just check that the endtime is not null?
langfuse-python
github_2023
python
35
langfuse
maxdeichmann
@@ -401,6 +416,14 @@ def task(): except Exception as e: self.log.exception(e) + def end(self): + try: + end_time = datetime.now() + self.update(UpdateGeneration(endTime=end_time))
I missed this one here: i think we need to return this function here so that end() returns a stateful client (also above). Otherwise good for me.
laravel-filament-flexible-content-blocks
github_2023
others
25
statikbe
sten
@@ -68,6 +68,7 @@ Here is a brief overview of the choices made: - `spatie/laravel-sluggable`: for slugs - `spatie/laravel-translatable`: for translations as this works together with the first party filament translatable package. - `dereuromark/media-embed`: to support video embeds of [various media services](https:...
There is no real dependency on this library. It is only used in the examples. So I do not like to include it. You can use whatever you want. In the future, I am planning to implement support for additional fields for SEO metadata types, like articles, pages, how-to lists, etc. Maybe other libs are more useful for t...
laravel-filament-flexible-content-blocks
github_2023
others
25
statikbe
sten
@@ -38,6 +38,7 @@ ], "require": { "php": "^8.1", + "artesaos/seotools": "^1.2",
Please, remove. See comment above.
laravel-filament-flexible-content-blocks
github_2023
php
25
statikbe
sten
@@ -0,0 +1,17 @@ +<?php + +namespace App\View\Components; + +use Illuminate\View\Component; +use Illuminate\View\View; + +class baseLayour extends Component
Typo in class name: BaseLayout
laravel-filament-flexible-content-blocks
github_2023
php
6
statikbe
sten
@@ -134,9 +134,9 @@ public function getChildComponents(): array TextInput::make(static::FIELD_BUTTON_LABEL) ->label(trans('filament-flexible-content-blocks::filament-flexible-content-blocks.form_component.content_blocks.call_to_action_button_label')) ...
zet `->inline(false)` op een new line
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -3,32 +3,52 @@ import java.util.IdentityHashMap; import java.util.Map; +import ai.timefold.solver.core.api.domain.solution.PlanningSolution; import ai.timefold.solver.core.impl.bavet.uni.AbstractForEachUniNode; +import ai.timefold.solver.core.impl.bavet.uni.AbstractForEachUniNode.LifecycleOperation; public a...
Cool!
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -3,32 +3,52 @@ import java.util.IdentityHashMap; import java.util.Map; +import ai.timefold.solver.core.api.domain.solution.PlanningSolution; import ai.timefold.solver.core.impl.bavet.uni.AbstractForEachUniNode; +import ai.timefold.solver.core.impl.bavet.uni.AbstractForEachUniNode.LifecycleOperation; public a...
We are passing an annotation class here. I'm unsure how `nodeNetwork.getForEachNodes()` will locate the solution class.
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -83,7 +80,7 @@ public UniDataStream<Solution_, A> addNull() { throw new UnsupportedOperationException(); } - public AbstractDataset<Solution_, UniTuple<A>> createDataset() { + public UniDataset<Solution_, A> createDataset() {
Should we avoid the use of specialized classes here?
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -10,24 +9,23 @@ import ai.timefold.solver.core.api.score.stream.Joiners; import ai.timefold.solver.core.impl.domain.solution.descriptor.SolutionDescriptor; -import ai.timefold.solver.core.impl.move.streams.maybeapi.stream.DataStreamFactory; -import ai.timefold.solver.core.impl.move.streams.maybeapi.stream.Soluti...
Should this class be abstract, or will there be no default implementation?
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -12,48 +11,40 @@ import ai.timefold.solver.core.impl.bavet.uni.ForEachExcludingUnassignedUniNode; import ai.timefold.solver.core.impl.bavet.uni.ForEachFromSolutionUniNode; import ai.timefold.solver.core.impl.bavet.uni.ForEachIncludingUnassignedUniNode; -import ai.timefold.solver.core.impl.bavet.uni.ForEachStaticU...
Why not to add specialized classes, one for the predicate and another for the FromSolutionValueCollectingFunction?
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -0,0 +1,136 @@ +package ai.timefold.solver.core.impl.move.streams; + +import java.util.Iterator; +import java.util.NoSuchElementException; +import java.util.Objects; +import java.util.Set; +import java.util.function.BiPredicate; + +import ai.timefold.solver.core.impl.bavet.common.tuple.UniTuple; +import ai.timefold....
Nice!
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -81,11 +81,11 @@ public abstract class AbstractScoreDirector<Solution_, Score_ extends Score<Scor private int workingInitScore = 0; private String undoMoveText; - // Null when tracking disabled - private final boolean trackingWorkingSolution; private final SolutionTracker<Solution_> solutionTr...
It cannot be null
timefold-solver
github_2023
java
1,463
TimefoldAI
zepfred
@@ -59,6 +59,7 @@ public BavetConstraintStreamScoreDirector( @Override public void setWorkingSolution(Solution_ workingSolution) { session = scoreDirectorFactory.newSession(workingSolution, constraintMatchPolicy, derived); + session.initialize(workingSolution);
Is this necessary at this stage of the implementation?
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,222 @@ +package ai.timefold.solver.core.impl.bavet.common; + +import java.util.ArrayList; +import java.util.Collections; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.TreeMap; +import java.util.function.BiConsumer; +import java.util.functio...
Let's use `String::format`
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,222 @@ +package ai.timefold.solver.core.impl.bavet.common; + +import java.util.ArrayList; +import java.util.Collections; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.TreeMap; +import java.util.function.BiConsumer; +import java.util.functio...
Please apply the previous comment to all similar occurrences
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,222 @@ +package ai.timefold.solver.core.impl.bavet.common; + +import java.util.ArrayList; +import java.util.Collections; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.TreeMap; +import java.util.function.BiConsumer; +import java.util.functio...
Should it be zero instead of Integer.MIN_VALUE?
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -26,22 +26,23 @@ public void insert(A a) { @Override public void update(A a) { - UniTuple<A> tuple = tupleMap.get(a); + var tuple = tupleMap.get(a); if (tuple == null) { // The tuple was never inserted because it did not pass the filter. insert(a); } else if (...
Why we don't need to filter before removing?
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -26,22 +26,23 @@ public void insert(A a) { @Override public void update(A a) { - UniTuple<A> tuple = tupleMap.get(a); + var tuple = tupleMap.get(a); if (tuple == null) { // The tuple was never inserted because it did not pass the filter. insert(a); } else if (...
The comment is inconsistent
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -3,20 +3,21 @@ import ai.timefold.solver.core.impl.bavet.common.tuple.TupleLifecycle; import ai.timefold.solver.core.impl.bavet.common.tuple.UniTuple; -public final class ForEachIncludingUnassignedUniNode<A> extends AbstractForEachUniNode<A> { +public sealed class ForEachUniNode<A>
Since I don't have deep knowledge of the ConstraintStream ecosystem, I'm missing comments on the node classes that briefly describe their roles.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -29,23 +30,27 @@ public int layerCount() { return layeredNodes.length; } - @SuppressWarnings("unchecked") - public AbstractForEachUniNode<Object>[] getApplicableForEachNodes(Class<?> factClass) { + public Stream<AbstractForEachUniNode<?>> getForEachNodes() { + return declaredClassToNo...
Propagate seems to be a common term used in CS. I'm unsure if the method name change is good idea.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -134,4 +166,8 @@ public VariableDescriptorAwareScoreDirector<Solution_> getScoreDirector() { return scoreDirector; } + public void resetWorkingSolution(Solution_ workingSolution) {
Please add a comment about why we are adding an empty method
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,46 @@ +package ai.timefold.solver.core.impl.move.streams.dataset.common.bridge; + +import java.util.Objects; + +import ai.timefold.solver.core.impl.bavet.common.TupleSource; +import ai.timefold.solver.core.impl.move.streams.dataset.AbstractDataStream; +import ai.timefold.solver.core.impl.move.streams.dataset...
What is the meaning of Aft?
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,46 @@ +package ai.timefold.solver.core.impl.move.streams.dataset.common.bridge; + +import java.util.Objects; + +import ai.timefold.solver.core.impl.bavet.common.TupleSource; +import ai.timefold.solver.core.impl.move.streams.dataset.AbstractDataStream; +import ai.timefold.solver.core.impl.move.streams.dataset...
```suggestion var that = (AftBridgeUniDataStream<?, ?>) o; ```
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,76 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.ArrayList; +import java.util.List; +import java.util.Set; + +import ai.timefold.solver.core.impl.bavet.common.BavetStream; +import ai.timefold.solver.core.impl.bavet.common.TupleSource; +import ai.timefold.solver.core.impl....
Let's use `String::format` in all similar occurrences
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,76 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.Collections; +import java.util.IdentityHashMap; +import java.util.Map; +import java.util.Objects; + +import ai.timefold.solver.core.impl.bavet.AbstractSession; +import ai.timefold.solver.core.impl.bavet.NodeNetwork; +import...
Why are we overriding it?
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,64 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.ArrayList; +import java.util.LinkedHashMap; +import java.util.LinkedHashSet; +import java.util.List; +import java.util.Set; +import java.util.function.Consumer; + +import ai.timefold.solver.core.impl.bavet.NodeNetwork; +imp...
I wonder if we should change the logic to return the list of active datasets. This way, we wouldn't need to pass activeDatasetStreamSet as a parameter.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,147 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.Collection; +import java.util.HashMap; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.function.Consumer; +import java.util.stream.Collectors; +import java.util.stream.Stream; + +imp...
Let's use `String::format`
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,52 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.Objects; +import java.util.function.Predicate; + +import ai.timefold.solver.core.impl.bavet.common.tuple.TupleLifecycle; +import ai.timefold.solver.core.impl.bavet.common.tuple.UniTuple; +import ai.timefold.solver.core.impl...
```suggestion this.predicate = Objects.requireNonNull(predicate, "The predicate (null) cannot be null."); ```
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,115 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.Collection; +import java.util.Objects; +import java.util.Set; +import java.util.function.Predicate; + +import ai.timefold.solver.core.impl.bavet.common.TupleSource; +import ai.timefold.solver.core.impl.bavet.common.tuple.T...
We are using `Objects::requireNonNull` in some places and in others not.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,112 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import java.util.Objects; +import java.util.Set; +import java.util.function.BiPredicate; + +import ai.timefold.solver.core.impl.bavet.bi.joiner.DefaultBiJoiner; +import ai.timefold.solver.core.impl.bavet.common.BavetAbstractConstraintStrea...
I would use an if block instead of a ternary operator
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,178 @@ +package ai.timefold.solver.core.impl.move.streams.generic; + +import java.util.Collection; +import java.util.Collections; +import java.util.List; +import java.util.Objects; + +import ai.timefold.solver.core.api.domain.solution.PlanningSolution; +import ai.timefold.solver.core.api.domain.variable.Plan...
The class is very well-documented. I missed these types of comments in the other move classes and data stream classes.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,8 @@ +package ai.timefold.solver.core.impl.score.stream.bavet.common;
I wonder if these stream classes should be part of the package ` ai.timefold.solver.core.impl.bavet.common`
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -71,7 +71,8 @@ public <Stream_ extends BavetAbstractConstraintStream<Solution_>> Stream_ share( * {@link BavetAbstractConstraintStream} implement equals/hashcode ignoring child streams. * <p> * {@link BavetConstraintSessionFactory#buildSession(Object, ConstraintMatchPolicy, boolean, Consumer)} need...
Please adjust the comment
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,107 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import static org.assertj.core.api.Assertions.assertThat; + +import ai.timefold.solver.core.impl.domain.solution.descriptor.DefaultPlanningListVariableMetaModel; +import ai.timefold.solver.core.impl.domain.solution.descriptor.DefaultPlanni...
We might use this class in other places and will need to make it available to the different classes.
timefold-solver
github_2023
java
1,349
TimefoldAI
zepfred
@@ -0,0 +1,107 @@ +package ai.timefold.solver.core.impl.move.streams.dataset; + +import static org.assertj.core.api.Assertions.assertThat; + +import ai.timefold.solver.core.impl.domain.solution.descriptor.DefaultPlanningListVariableMetaModel; +import ai.timefold.solver.core.impl.domain.solution.descriptor.DefaultPlanni...
Nice!
timefold-solver
github_2023
java
1,454
TimefoldAI
winklerm
@@ -500,6 +501,53 @@ void compareWithConstraintMatchesAndMatchAnalysis() { }); } + @Test + void compareMatchCountsWithDifferentFetchPolicies() { + var constraintPackage = "constraintPackage"; + var constraintName1 = "constraint1"; + var constraintName2 = "constraint2"; + ...
Q: Should `constraintAnalysisMap2` use `FETCH_ALL` for all entries? Or is the mixing of policies intentional?
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -27,7 +27,7 @@ Gradle:: -- [source,shell,subs=attributes+] ---- -curl https://timefold.ai/product/upgrade/upgrade-timefold.gradle > upgrade-timefold.gradle ; gradle -Dorg.gradle.jvmargs=-Xmx2G --init-script upgrade-timefold.gradle rewriteRun -DtimefoldSolverVersion={timefold-solver-version} ; rm upgrade-timefold....
This doesn't work for me. example gives 404, because `upgrade-timefold.gradle` doesn't exist? https://raw.githubusercontent.com/TimefoldAI/timefold-solver/refs/tags/v1.19.0/migration/upgrade-timefold.gradle
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -33,7 +33,7 @@ Gradle:: -- [source,shell,subs=attributes+] ---- -curl https://timefold.ai/product/upgrade/upgrade-timefold.gradle > upgrade-timefold.gradle ; gradle -Dorg.gradle.jvmargs=-Xmx2G --init-script upgrade-timefold.gradle rewriteRun -DtimefoldSolverVersion={timefold-solver-version} ; rm upgrade-timefold....
Ditto other comment
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -140,22 +140,51 @@ Some solver configurations use the `Random` instance a lot more than others. For example, Simulated Annealing depends highly on random numbers, while Tabu Search only depends on it to deal with score ties. The environment mode influences the seed of that `Random` instance. -These are the envir...
```suggestion any two runs of the same dataset with the same solver configuration must have the same result at every step. ```
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
```suggestion The `FULL_ASSERT` mode turns on all assertions and will fail-fast on a bug in a Move implementation, a constraint, the engine itself, ... ``` I think the extra explainer is not needed and makes the sentence harder to read.
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
```suggestion It is also intrusive because it calls the method `calculateScore()` more frequently than a non-assert mode, making the `FULL_ASSERT` mode very slow. ```
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
Remove these 2 with the suggestions above.
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
Should this be a TIP to make it stand out?
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
Do we have a list of the assertions somewhere? Because "ALL" (FULL_ASSERT) and "several" (NONE_INTRUSIVE_FULL_ASSERT) isn't very telling. Side note: having to say we only activate "several assertions" for NONE_INTRUSIVE_FULL_ASSERT is a bit weird. (But probably legacy?)
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
Since Seeds aren't really mentioned clearly before, I wouldn't introduce the concept here.
timefold-solver
github_2023
others
1,437
TimefoldAI
TomCools
@@ -180,76 +209,78 @@ Any variable that changed between the "before move" solution and the "after move Any variable that changed between the "after move" solution and "after undo move" solution without either a `beforeVariableChanged` or `afterVariableChanged` would be reported here. -This mode is <<environmentMode...
I am a bit surprised, if it is only "negligibly slower", I would not expect it to disable concurrency optimizations (that sounds bigger than just negligible to the untrained eye).
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -297,6 +297,10 @@ public TerminationConfig withTerminationClass(Class<? extends Termination> termi return this; } + public @NonNull TerminationConfig withDiminishedReturnsConfig() {
My understanding is this method is to use Diminished Returns with the default configuration? Maybe `withDefaultDiminishedReturnsConfig()`? `withDiminishedReturnsConfig()` looks odd to me without an argument.
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -88,18 +92,32 @@ public boolean isAssertShadowVariablesAreNotStaleAfterStep() { @Override public void solvingStarted(SolverScope<Solution_> solverScope) { - phaseTermination.solvingStarted(solverScope); phaseLifecycleSupport.fireSolvingStarted(solverScope); } @Override pu...
Can we fail fast when the solver is being built? This will fail during solving.
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -40,26 +41,30 @@ public abstract class AbstractSolver<Solution_> implements Solver<Solution_> { protected final BestSolutionRecaller<Solution_> bestSolutionRecaller; // Note that the DefaultSolver.basicPlumbingTermination is a component of this termination. // Called "solverTermination" to clearly dis...
Maybe rename field to `universalTermination`? Since there is a class called `SolverTermination` now.
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -3,89 +3,110 @@ import java.util.ArrayList; import java.util.Arrays; import java.util.List; +import java.util.Objects; import ai.timefold.solver.core.impl.phase.scope.AbstractPhaseScope; import ai.timefold.solver.core.impl.phase.scope.AbstractStepScope; import ai.timefold.solver.core.impl.solver.scope.Solver...
The `step` methods do not get called for the solver terminations? How does the solver terminations get updated then?
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -3,89 +3,110 @@ import java.util.ArrayList; import java.util.Arrays; import java.util.List; +import java.util.Objects; import ai.timefold.solver.core.impl.phase.scope.AbstractPhaseScope; import ai.timefold.solver.core.impl.phase.scope.AbstractStepScope; import ai.timefold.solver.core.impl.solver.scope.Solver...
What happens to universal terminations? Do they get put in both lists?
timefold-solver
github_2023
java
1,431
TimefoldAI
Christopher-Chianelli
@@ -0,0 +1,132 @@ +package ai.timefold.solver.core.impl.solver; + +import ai.timefold.solver.core.api.score.buildin.simple.SimpleScore; +import ai.timefold.solver.core.api.solver.SolverFactory; +import ai.timefold.solver.core.config.constructionheuristic.ConstructionHeuristicPhaseConfig; +import ai.timefold.solver.core...
Outdated comment; no step count termination is set.