Merge branch 'main' into feat/structlog-parseable

.github/workflows/porter_stack_cdp-front.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+    - main
+name: Deploy to cdp-front
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+    - name: Set Github tag
+      id: vars
+      run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+    - name: Setup porter
+      uses: porter-dev/[email protected]
+    - name: Deploy stack
+      timeout-minutes: 30
+      run: exec porter apply
+      env:
+        PORTER_CLUSTER: "3877"
+        PORTER_DEPLOYMENT_TARGET_ID: cd21d246-86df-49e0-ba0a-78e2802572e7
+        PORTER_HOST: https://dashboard.porter.run
+        PORTER_PR_NUMBER: ${{ github.event.number }}
+        PORTER_PROJECT: "10983"
+        PORTER_REPO_NAME: ${{ github.event.repository.name }}
+        PORTER_STACK_NAME: cdp-front
+        PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+        PORTER_TOKEN: ${{ secrets.PORTER_STACK_10983_3877 }}

.github/workflows/porter_stack_cdp.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+    - main
+name: Deploy to cdp
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+    - name: Set Github tag
+      id: vars
+      run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+    - name: Setup porter
+      uses: porter-dev/[email protected]
+    - name: Deploy stack
+      timeout-minutes: 30
+      run: exec porter apply
+      env:
+        PORTER_CLUSTER: "3877"
+        PORTER_DEPLOYMENT_TARGET_ID: cd21d246-86df-49e0-ba0a-78e2802572e7
+        PORTER_HOST: https://dashboard.porter.run
+        PORTER_PR_NUMBER: ${{ github.event.number }}
+        PORTER_PROJECT: "10983"
+        PORTER_REPO_NAME: ${{ github.event.repository.name }}
+        PORTER_STACK_NAME: cdp
+        PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+        PORTER_TOKEN: ${{ secrets.PORTER_STACK_10983_3877 }}

backend/Dockerfile

@@ -24,6 +24,7 @@ RUN apt-get clean && apt-get update && apt-get install -y \
     libreoffice \
     libpq-dev \
     gcc \
+    libhdf5-serial-dev \
     pandoc && \
     rm -rf /var/lib/apt/lists/* && apt-get clean
 
@@ -46,6 +47,8 @@ COPY core/pyproject.toml core/README.md ./core/
 COPY core/quivr_core/__init__.py ./core/quivr_core/__init__.py
 COPY worker/pyproject.toml worker/README.md ./worker/
 COPY worker/quivr_worker/__init__.py ./worker/quivr_worker/__init__.py
+COPY worker/diff-assistant/pyproject.toml worker/diff-assistant/README.md ./worker/diff-assistant/
+COPY worker/diff-assistant/quivr_diff_assistant/__init__.py ./worker/diff-assistant/quivr_diff_assistant/__init__.py
 COPY core/MegaParse/pyproject.toml core/MegaParse/README.md ./core/MegaParse/
 COPY core/MegaParse/megaparse/__init__.py ./core/MegaParse/megaparse/__init__.py
 

backend/Dockerfile.dev

@@ -23,6 +23,7 @@ RUN apt-get clean && apt-get update && apt-get install -y \
     libreoffice \
     libpq-dev \
     gcc \
+    libhdf5-serial-dev \
     pandoc && \
     rm -rf /var/lib/apt/lists/* && apt-get clean
 
@@ -33,6 +34,8 @@ COPY core/pyproject.toml core/README.md ./core/
 COPY core/quivr_core/__init__.py ./core/quivr_core/__init__.py
 COPY worker/pyproject.toml worker/README.md ./worker/
 COPY worker/quivr_worker/__init__.py ./worker/quivr_worker/__init__.py
+COPY worker/diff-assistant/pyproject.toml worker/diff-assistant/README.md ./worker/diff-assistant/
+COPY worker/diff-assistant/quivr_diff_assistant/__init__.py ./worker/diff-assistant/quivr_diff_assistant/__init__.py
 COPY core/MegaParse/pyproject.toml core/MegaParse/README.md ./core/MegaParse/
 COPY core/MegaParse/megaparse/__init__.py ./core/MegaParse/megaparse/__init__.py
 

backend/api/quivr_api/modules/assistant/controller/assistant_routes.py

@@ -2,7 +2,7 @@
 from typing import Annotated, List
 from uuid import uuid4
 
-from fastapi import APIRouter, Depends, HTTPException, Request, UploadFile
+from fastapi import APIRouter, Depends, File, HTTPException, Request, UploadFile
 
 from quivr_api.celery_config import celery
 from quivr_api.logger import get_logger
@@ -16,6 +16,7 @@
 from quivr_api.modules.assistant.entity.assistant_entity import (
     AssistantSettings,
 )
+from quivr_api.modules.assistant.entity.task_entity import TaskMetadata
 from quivr_api.modules.assistant.services.tasks_service import TasksService
 from quivr_api.modules.dependencies import get_service
 from quivr_api.modules.upload.service.upload_file import (
@@ -64,12 +65,15 @@ async def create_task(
     current_user: UserIdentityDep,
     tasks_service: TasksServiceDep,
     request: Request,
-    input: InputAssistant,
+    input: str = File(...),
     files: List[UploadFile] = None,
 ):
+    input = InputAssistant.model_validate_json(input)
+
     assistant = next(
         (assistant for assistant in assistants if assistant.id == input.id), None
     )
+
     if assistant is None:
         raise HTTPException(status_code=404, detail="Assistant not found")
 
@@ -80,7 +84,7 @@ async def create_task(
             raise HTTPException(status_code=400, detail=error)
     else:
         print("Assistant input is valid.")
-    notification_uuid = uuid4()
+    notification_uuid = f"{assistant.name}-{str(uuid4())[:8]}"
 
     # Process files dynamically
     for upload_file in files:
@@ -96,8 +100,14 @@ async def create_task(
 
     task = CreateTask(
         assistant_id=input.id,
-        pretty_id=str(notification_uuid),
+        assistant_name=assistant.name,
+        pretty_id=notification_uuid,
         settings=input.model_dump(mode="json"),
+        task_metadata=TaskMetadata(
+            input_files=[file.filename for file in files]
+        ).model_dump(mode="json")
+        if files
+        else None,  # type: ignore
     )
 
     task_created = await tasks_service.create_task(task, current_user.id)

backend/api/quivr_api/modules/assistant/controller/assistants_definition.py

@@ -1,8 +1,11 @@
 from quivr_api.modules.assistant.dto.inputs import InputAssistant
 from quivr_api.modules.assistant.dto.outputs import (
     AssistantOutput,
+    ConditionalInput,
+    InputBoolean,
     InputFile,
     Inputs,
+    InputSelectText,
     Pricing,
 )
 
@@ -166,10 +169,10 @@ def validate_assistant_input(
 
 assistant1 = AssistantOutput(
     id=1,
-    name="Assistant 1",
-    description="Assistant 1 description",
+    name="Compliance Check",
+    description="Allows analyzing the compliance of the information contained in documents against charter or regulatory requirements.",
     pricing=Pricing(),
-    tags=["tag1", "tag2"],
+    tags=["Disabled"],
     input_description="Input description",
     output_description="Output description",
     inputs=Inputs(
@@ -183,19 +186,66 @@ def validate_assistant_input(
 
 assistant2 = AssistantOutput(
     id=2,
-    name="Assistant 2",
-    description="Assistant 2 description",
+    name="Consistency Check",
+    description="Ensures that the information in one document is replicated identically in another document.",
     pricing=Pricing(),
-    tags=["tag1", "tag2"],
+    tags=[],
     input_description="Input description",
     output_description="Output description",
     icon_url="https://example.com/icon.png",
     inputs=Inputs(
         files=[
-            InputFile(key="file_1", description="File description"),
-            InputFile(key="file_2", description="File description"),
+            InputFile(key="Document 1", description="File description"),
+            InputFile(key="Document 2", description="File description"),
+        ],
+        select_texts=[
+            InputSelectText(
+                key="DocumentsType",
+                description="Select Documents Type",
+                options=[
+                    "Etiquettes VS Cahier des charges",
+                    "Fiche Dev VS Cahier des charges",
+                ],
+            ),
+        ],
+    ),
+)
+
+assistant3 = AssistantOutput(
+    id=3,
+    name="Difference Detection",
+    description="Highlights differences between one document and another after modifications.",
+    pricing=Pricing(),
+    tags=[],
+    input_description="Input description",
+    output_description="Output description",
+    icon_url="https://example.com/icon.png",
+    inputs=Inputs(
+        files=[
+            InputFile(key="Document 1", description="File description"),
+            InputFile(key="Document 2", description="File description"),
+        ],
+        booleans=[
+            InputBoolean(
+                key="Hard-to-Read Document?", description="Boolean description"
+            ),
+        ],
+        select_texts=[
+            InputSelectText(
+                key="DocumentsType",
+                description="Select Documents Type",
+                options=["Etiquettes", "Cahier des charges"],
+            ),
+        ],
+        conditional_inputs=[
+            ConditionalInput(
+                key="DocumentsType",
+                conditional_key="Hard-to-Read Document?",
+                condition="equals",
+                value="Etiquettes",
+            ),
         ],
     ),
 )
 
-assistants = [assistant1, assistant2]
+assistants = [assistant1, assistant2, assistant3]

backend/api/quivr_api/modules/assistant/dto/inputs.py

@@ -1,4 +1,4 @@
-from typing import List, Optional
+from typing import Dict, List, Optional
 from uuid import UUID
 
 from pydantic import BaseModel, root_validator
@@ -7,7 +7,9 @@
 class CreateTask(BaseModel):
     pretty_id: str
     assistant_id: int
+    assistant_name: str
     settings: dict
+    task_metadata: Dict | None = None
 
 
 class BrainInput(BaseModel):

backend/api/quivr_api/modules/assistant/dto/outputs.py

@@ -61,6 +61,21 @@ class InputSelectNumber(BaseModel):
     default: Optional[int] = None
 
 
+class ConditionalInput(BaseModel):
+    """
+    Conditional input is a list of inputs that are conditional to the value of another input.
+    key: The key of the input that is conditional.
+    conditional_key: The key that determines if the input is shown.
+    """
+
+    key: str
+    conditional_key: str
+    condition: Optional[str] = (
+        None  # e.g. "equals", "contains", "starts_with", "ends_with", "regex", "in", "not_in", "is_empty", "is_not_empty"
+    )
+    value: Optional[str] = None
+
+
 class Inputs(BaseModel):
     files: Optional[List[InputFile]] = None
     urls: Optional[List[InputUrl]] = None
@@ -70,6 +85,7 @@ class Inputs(BaseModel):
     select_texts: Optional[List[InputSelectText]] = None
     select_numbers: Optional[List[InputSelectNumber]] = None
     brain: Optional[BrainInput] = None
+    conditional_inputs: Optional[List[ConditionalInput]] = None
 
 
 class Pricing(BaseModel):

backend/api/quivr_api/modules/assistant/entity/task_entity.py

@@ -1,10 +1,15 @@
 from datetime import datetime
-from typing import Dict
+from typing import Dict, List, Optional
 from uuid import UUID
 
+from pydantic import BaseModel
 from sqlmodel import JSON, TIMESTAMP, BigInteger, Column, Field, SQLModel, text
 
 
+class TaskMetadata(BaseModel):
+    input_files: Optional[List[str]] = None
+
+
 class Task(SQLModel, table=True):
     __tablename__ = "tasks"  # type: ignore
 
@@ -17,6 +22,7 @@ class Task(SQLModel, table=True):
         ),
     )
     assistant_id: int
+    assistant_name: str
     pretty_id: str
     user_id: UUID
     status: str = Field(default="pending")
@@ -29,6 +35,4 @@ class Task(SQLModel, table=True):
     )
     settings: Dict = Field(default_factory=dict, sa_column=Column(JSON))
     answer: str | None = Field(default=None)
-
-    class Config:
-        arbitrary_types_allowed = True
+    task_metadata: Dict | None = Field(default_factory=dict, sa_column=Column(JSON))

backend/api/quivr_api/modules/assistant/repository/tasks.py

@@ -3,7 +3,7 @@
 
 from sqlalchemy import exc
 from sqlalchemy.ext.asyncio import AsyncSession
-from sqlmodel import select
+from sqlmodel import col, select
 
 from quivr_api.modules.assistant.dto.inputs import CreateTask
 from quivr_api.modules.assistant.entity.task_entity import Task
@@ -21,9 +21,11 @@ async def create_task(self, task: CreateTask, user_id: UUID) -> Task:
         try:
             task_to_create = Task(
                 assistant_id=task.assistant_id,
+                assistant_name=task.assistant_name,
                 pretty_id=task.pretty_id,
                 user_id=user_id,
                 settings=task.settings,
+                task_metadata=task.task_metadata,  # type: ignore
             )
             self.session.add(task_to_create)
             await self.session.commit()
@@ -40,7 +42,9 @@ async def get_task_by_id(self, task_id: UUID, user_id: UUID) -> Task:
         return response.one()
 
     async def get_tasks_by_user_id(self, user_id: UUID) -> Sequence[Task]:
-        query = select(Task).where(Task.user_id == user_id)
+        query = (
+            select(Task).where(Task.user_id == user_id).order_by(col(Task.id).desc())
+        )
         response = await self.session.exec(query)
         return response.all()
 

backend/api/quivr_api/modules/brain/service/brain_service.py

@@ -1,4 +1,4 @@
-from typing import Optional, Tuple, Dict
+from typing import Dict, Optional, Tuple
 from uuid import UUID
 
 from fastapi import HTTPException

backend/api/quivr_api/modules/chat/controller/chat/utils.py

@@ -1,5 +1,5 @@
-import time
 import os
+import time
 from enum import Enum
 
 from fastapi import HTTPException

backend/api/quivr_api/modules/chat/controller/chat_routes.py

@@ -1,9 +1,10 @@
+import os
 from typing import Annotated, List, Optional
 from uuid import UUID
-import os
 
 from fastapi import APIRouter, BackgroundTasks, Depends, HTTPException, Query, Request
 from fastapi.responses import StreamingResponse
+from quivr_core.config import RetrievalConfig
 
 from quivr_api.logger import get_logger
 from quivr_api.middlewares.auth import AuthBearer, get_current_user
@@ -36,7 +37,6 @@
 from quivr_api.modules.vector.service.vector_service import VectorService
 from quivr_api.utils.telemetry import maybe_send_telemetry
 from quivr_api.utils.uuid_generator import generate_uuid_from_string
-from quivr_core.config import RetrievalConfig
 
 logger = get_logger(__name__)