RAGFlow is an agent-native database: a Retrieval-Augmented Generation (RAG) engine that grew into context infrastructure.
Go agent-harnessagent-native-databaseagentic-aiagentic-retrievalagentic-searchaiai-agentscontext-enginecontext-engineeringcontext-managementharness-engineeringknowledge-compilationragretrieval-augmented-generationretrieval-harnesssearch-harness
Stop splitting resume documents into per-page parsing tasks (#18713)
Yash Raj Pandeybfccb5177d8e
diff --git a/api/db/services/task_service.py b/api/db/services/task_service.pyindex 683ab2d244..7c14e3e652 100644--- a/api/db/services/task_service.py+++ b/api/db/services/task_service.py@@ -491,9 +491,13 @@ def queue_tasks(doc: dict, bucket: str, name: str, priority: int):passif is_mineru:logging.info("Document %s selected MinerU unsplit-task mode with page size %s", doc["id"], MAXIMUM_TASK_PAGE_NUMBER)- if doc["parser_id"] in ["one", "knowledge_graph"] or doc["parser_config"].get("toc_extraction", False) or is_mineru:+ if doc["parser_id"] in ["one", "knowledge_graph", "resume"] or doc["parser_config"].get("toc_extraction", False) or is_mineru:page_size = MAXIMUM_TASK_PAGE_NUMBERpage_ranges = doc["parser_config"].get("pages") or [(1, MAXIMUM_PAGE_NUMBER)]+ if doc["parser_id"] == "resume":+ # The resume parser ignores from_page and to_page, so every task parses the whole+ # file. Collapse the configured ranges into one range to keep a resume document at a single task.+ page_ranges = [(1, MAXIMUM_PAGE_NUMBER)]for s, e in page_ranges:s -= 1s = max(0, s)diff --git a/test/unit_test/api/db/services/test_task_service_queue_tasks.py b/test/unit_test/api/db/services/test_task_service_queue_tasks.pynew file mode 100644index 0000000000..ec28bc222e--- /dev/null+++ b/test/unit_test/api/db/services/test_task_service_queue_tasks.py@@ -0,0 +1,64 @@+#+# Copyright 2026 The InfiniFlow Authors. All Rights Reserved.+#+# Licensed under the Apache License, Version 2.0 (the "License");+# you may not use this file except in compliance with the License.+# You may obtain a copy of the License at+#+# http://www.apache.org/licenses/LICENSE-2.0+#+# Unless required by applicable law or agreed to in writing, software+# distributed under the License is distributed on an "AS IS" BASIS,+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.+# See the License for the specific language governing permissions and+# limitations under the License.+#+from types import SimpleNamespace++import pytest++from api.db import FileType+from api.db.services import task_service+++def _queue_tasks(monkeypatch, parser_id, parser_config):+ queued_tasks = []+ monkeypatch.setattr(task_service.settings, "STORAGE_IMPL", SimpleNamespace(get=lambda *_args: b"pdf"))+ monkeypatch.setattr(task_service.PdfParser, "total_page_number", lambda *_args: 20)+ monkeypatch.setattr(task_service.DocumentService, "get_chunking_config", lambda *_args: {})+ monkeypatch.setattr(task_service.TaskService, "get_tasks", lambda *_args: [])+ monkeypatch.setattr(task_service.DocumentService, "update_by_id", lambda *_args: None)+ monkeypatch.setattr(task_service, "bulk_insert_into_db", lambda _model, tasks, _replace: queued_tasks.extend(tasks))+ monkeypatch.setattr(task_service.DocumentService, "begin2parse", lambda *_args: None)+ monkeypatch.setattr(task_service, "seed_doc_chunking_counter", lambda *_args: True)+ monkeypatch.setattr(task_service.REDIS_CONN, "queue_product", lambda *_args, **_kwargs: True)++ task_service.queue_tasks(+ {+ "id": "doc-id",+ "name": "resume.pdf",+ "type": FileType.PDF.value,+ "parser_id": parser_id,+ "parser_config": parser_config,+ },+ "bucket",+ "resume.pdf",+ 0,+ )+ return queued_tasks+++def test_queue_tasks_collapses_multiple_page_ranges_for_resume(monkeypatch):+ queued_tasks = _queue_tasks(monkeypatch, "resume", {"pages": [[1, 5], [10, 15]]})++ assert len(queued_tasks) == 1+ assert queued_tasks[0]["from_page"] == 0+ assert queued_tasks[0]["to_page"] == 20+++def test_queue_tasks_keeps_multiple_page_ranges_for_naive(monkeypatch):+ queued_tasks = _queue_tasks(monkeypatch, "naive", {"pages": [[1, 5], [10, 15]], "task_page_size": 12})++ assert len(queued_tasks) > 1