feat: Tool mode support for File and Save File#9898
Closed
erichare wants to merge 3 commits into
Closed
Conversation
Contributor
WalkthroughRenames “File” components to “Read File,” renames “Save File” to “Write File,” updates descriptions, and adds tool_mode=True to multiple Output declarations across components and starter project templates. News Aggregator also adds "txt" to file_format and updates code block and code_hash. No core logic or control flow changes. Changes
Sequence Diagram(s)Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested labels
Suggested reviewers
Pre-merge checks and finishing touches✅ Passed checks (3 passed)
✨ Finishing touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
github-actions
Bot
added
enhancement
github-actions
Bot
added
enhancement
|
Collaborator
edwinjosechittilappilly
left a comment
edwinjosechittilappilly
left a comment
There was a problem hiding this comment.
Hi Eric can we catch up on this PR?
In tool mode what would be the input to the file component?
Codecov Report✅ All modified and coverable lines are covered by tests. ❌ Your project status has failed because the head coverage (8.12%) is below the target coverage (10.00%). You can increase the head coverage or adjust the target coverage. Additional details and impacted files@@ Coverage Diff @@
## main #9898 +/- ##
==========================================
- Coverage 22.87% 22.86% -0.01%
==========================================
Files 1086 1086
Lines 39739 39739
Branches 5418 5418
==========================================
- Hits 9089 9088 -1
- Misses 30495 30496 +1
Partials 155 155
Flags with carried forward coverage won't be shown. Click here to find out more. 🚀 New features to boost your workflow:
|
github-actions
Bot
added
enhancement
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 5
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (5)
src/backend/base/langflow/initial_setup/starter_projects/Vector Store RAG.json (4)
2766-3514: Add timeout and better failure surfacing for Docling subprocess.
subprocess.runis invoked without a timeout; a hung Docling conversion will deadlock the component and the UX. Capture and handleTimeoutExpired, and surfacestderrwhen present.Apply this diff inside
_process_docling_in_subprocess:- proc = subprocess.run( # noqa: S603 + proc = subprocess.run( # noqa: S603 [sys.executable, "-u", "-c", child_script], input=json.dumps(args).encode("utf-8"), capture_output=True, check=False, + timeout=300, # fail fast; make configurable if needed ) + # Handle timeout explicitly + # (wrap the run in try/except if you prefer structured handling)And augment error reporting:
- if not proc.stdout: + if not proc.stdout: err_msg = proc.stderr.decode("utf-8", errors="replace") or "no output from child process" return Data(data={"error": f"Docling subprocess error: {err_msg}", "file_path": file_path})
2766-3514: Avoid passing None into rollup_data.If Docling returns
None, the advanced branch appends[advanced_data]which may be[None]. Filter outNonebefore callingrollup_data.- return self.rollup_data(file_list, [advanced_data]) + rows = [advanced_data] if advanced_data is not None else [] + return self.rollup_data(file_list, rows)
3503-3514: Fix brittle markdown return (result.text[0]).
load_files()may return a DataFrame of Data, a single Data, or Message depending on path. Indexingresult.text[0]will break when the shape differs. Build the string robustly.def load_files_markdown(self) -> Message: """Load files using advanced Docling processing and export to Markdown format.""" self.markdown = True - result = self.load_files() - return Message(text=str(result.text[0])) + result = self.load_files() + try: + from lfx.schema.dataframe import DataFrame + from lfx.schema.data import Data + from lfx.schema.message import Message as LFMessage + except Exception: + DataFrame = Data = LFMessage = object # type: ignore + + if isinstance(result, DataFrame): + texts = [getattr(row, "text", "") for row in result if getattr(row, "text", "")] + return Message(text="\n\n".join(texts)) + if isinstance(result, Data): + return Message(text=getattr(result, "text", "")) + if isinstance(result, LFMessage): + return result + return Message(text=str(result))
2683-2686: Starter node metadata is inconsistent (name/description/module).
display_namestill shows “File” and the older description; the code setsdisplay_name = "Read File"and new description.metadata.modulepoints tolangflow.components.datastax.astradb.AstraDBVectorStoreComponentfor the File node, which is incorrect and will break code-loading.- "description": "Loads content from one or more files as a DataFrame.", - "display_name": "File", + "description": "Loads content from files with optional advanced document processing and export using Docling.", + "display_name": "Read File", @@ - "module": "langflow.components.datastax.astradb.AstraDBVectorStoreComponent" + "module": "lfx.components.data.file.FileComponent"Also applies to: 2703-2726
src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json (1)
1250-1275: Add openpyxl to dependencies or keep/guard Excel exportRuntime check shows openpyxl is missing ("No module named 'openpyxl'"); add "openpyxl" to the dependencies block in src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json (around lines 1250–1275) or ensure the Excel-export code path is guarded and documented.
🧹 Nitpick comments (16)
src/lfx/src/lfx/components/processing/save_file.py (1)
53-54: tool_mode on “File Path” — return a bare path for tool consumers.Tooling parses outputs better if the Message text is just the path. Keep human-friendly confirmation in
self.status.Apply:
- # Return the final file path and confirmation message + # Return the final file path only (tool-friendly); keep human-friendly confirmation in status final_path = Path.cwd() / file_path if not file_path.is_absolute() else file_path - return Message(text=f"{confirmation} at {final_path}") + self.status = Message(text=confirmation) + return Message(text=str(final_path))Alternative: add a second non-tool output “Confirmation” for the verbose message.
src/backend/base/langflow/initial_setup/starter_projects/Vector Store RAG.json (3)
2766-3514: Correctexport_formatfor structured branch.You set
EXPORT_FORMAT = "Markdown"and reuse it for structured output, which mislabels the payload.- if result.get("mode") == "markdown": + if result.get("mode") == "markdown": exported_content = str(result.get("text", "")) return Data( text=exported_content, data={"exported_content": exported_content, "export_format": self.EXPORT_FORMAT, **meta}, ) rows = list(result.get("doc", [])) - return Data(data={"doc": rows, "export_format": self.EXPORT_FORMAT, **meta}) + return Data(data={"doc": rows, "export_format": "Structured", **meta})
2766-3514: Relax path “sanitization” (not shelling out).Since you pass args as a list and data via stdin, characters like
$/&/`don’t reach a shell. The current check can block legitimate paths without improving safety.- if not isinstance(args["file_path"], str) or any(c in args["file_path"] for c in [";", "|", "&", "$", "`"]): - return Data(data={"error": "Unsafe file path detected.", "file_path": args["file_path"]}) + if not isinstance(args["file_path"], str) or not args["file_path"].strip(): + return Data(data={"error": "Invalid file path.", "file_path": args.get("file_path")})
2766-3514: Honor the “Processing Concurrency” without the deprecated flag.You still gate concurrency by
use_multithreading. If the knob is deprecated, compute fromconcurrency_multithreadingalone.- concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading) + concurrency = max(1, int(self.concurrency_multithreading or 1))src/backend/base/langflow/initial_setup/starter_projects/Text Sentiment Analysis.json (2)
2474-2474: Make load_files_markdown resilient to return shape.
result.text[0]assumes list-like; will fail iftextis str/None or empty list.Apply this diff:
def load_files_markdown(self) -> Message: """Load files using advanced Docling processing and export to Markdown format.""" self.markdown = True result = self.load_files() - return Message(text=str(result.text[0])) + text = getattr(result, "text", "") + if isinstance(text, (list, tuple)): + text = text[0] if text else "" + return Message(text=str(text))
2416-2417: Align node display name with component rename (“Read File”).Starter node still shows “File”; change to “Read File” for consistency with the embedded class and PR intent.
Apply this diff:
- "display_name": "File", + "display_name": "Read File",src/backend/base/langflow/initial_setup/starter_projects/Document Q&A.json (4)
1333-1333: Make load_files_markdown resilient to return shape.Harden against
textnot being list-like.Apply this diff:
def load_files_markdown(self) -> Message: """Load files using advanced Docling processing and export to Markdown format.""" self.markdown = True result = self.load_files() - return Message(text=str(result.text[0])) + text = getattr(result, "text", "") + if isinstance(text, (list, tuple)): + text = text[0] if text else "" + return Message(text=str(text))
1274-1275: Rename node display name to “Read File”.Keep starter labels consistent with the component rename.
- "display_name": "File", + "display_name": "Read File",
750-753: Update docs note to reflect “Read File” rename.Text still says “In the File component”. Consider “In the Read File component”.
301-302: Fix prompt grammar.“What is this document is about?” → “What is this document about?”
- "value": "What is this document is about?" + "value": "What is this document about?"src/backend/base/langflow/initial_setup/starter_projects/Portfolio Website Code Generator.json (1)
1382-1540: tool_mode on Raw Content added — looks good; align node label with new component nameThe embedded FileComponent code now exposes Raw Content with tool_mode=True. However, the surrounding node still shows display_name "File" (Line 1325) while the component’s display_name is "Read File", which will look inconsistent in the UI. Please sync the node label.
Apply this JSON tweak:
- "display_name": "File", + "display_name": "Read File",src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json (3)
1236-1238: Rename mismatch: “Save File” node label vs code “Write File”The component class is now “Write File” but the node display_name is still “Save File”. Align for consistency.
- "display_name": "Save File", + "display_name": "Write File",
1315-1470: SaveToFile: great tool_mode output; drop duplicate import and harden Excel path
- Duplicate import of get_user_by_id inside _upload_file (imported in the first try and again later). Remove the first occurrence.
- to_excel requires openpyxl; add a clear error if it’s missing.
@@ - try: - from langflow.api.v2.files import upload_user_file - from langflow.services.database.models.user.crud import get_user_by_id + try: + from langflow.api.v2.files import upload_user_file @@ - elif fmt == "excel": - dataframe.to_excel(path, index=False, engine="openpyxl") + elif fmt == "excel": + try: + dataframe.to_excel(path, index=False, engine="openpyxl") + except ModuleNotFoundError as e: + raise ModuleNotFoundError("openpyxl is required for Excel export. Please install it.") from e
1315-1470: Optional: return a structured path instead of a plain messageConsider returning Data(data={"file_path": str(final_path)}) or setting Message.properties with the path so tools can consume it programmatically.
src/lfx/src/lfx/components/data/file.py (2)
215-259: All dynamic outputs now tool_mode=True — good; dedupe tabular extension checksThe logic is correct. Minor DRY: define TABULAR_EXTS once and reuse in update_build_config and update_outputs.
class FileComponent(BaseFileComponent): + TABULAR_EXTS = (".csv", ".xlsx", ".parquet") @@ - allow_advanced = file_count == 1 and not file_path.endswith((".csv", ".xlsx", ".parquet")) + allow_advanced = file_count == 1 and not file_path.endswith(self.TABULAR_EXTS) @@ - if file_path.endswith((".csv", ".xlsx", ".parquet")): + if file_path.endswith(self.TABULAR_EXTS): frontend_node["outputs"].append(
483-497: Path “sanitization” not needed for subprocess argsYou’re not using shell=True; argv is a list, so command injection isn’t a concern. This check can be dropped or replaced with a simple existence/permission check if you want stricter validation.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (8)
src/backend/base/langflow/initial_setup/starter_projects/Document Q&A.json(1 hunks)src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json(3 hunks)src/backend/base/langflow/initial_setup/starter_projects/Portfolio Website Code Generator.json(1 hunks)src/backend/base/langflow/initial_setup/starter_projects/Text Sentiment Analysis.json(1 hunks)src/backend/base/langflow/initial_setup/starter_projects/Vector Store RAG.json(1 hunks)src/lfx/src/lfx/base/data/base_file.py(1 hunks)src/lfx/src/lfx/components/data/file.py(3 hunks)src/lfx/src/lfx/components/processing/save_file.py(2 hunks)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (42)
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 39/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 38/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 27/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 34/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 35/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 40/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 20/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 32/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 21/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 33/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 31/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 16/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 30/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 25/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 28/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 29/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 22/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 24/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 18/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 19/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 17/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 9/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 12/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 1/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 2/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 15/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 5/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 13/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 14/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 8/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 10/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 7/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 11/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 6/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 3/40
- GitHub Check: Run Frontend Tests / Playwright Tests - Shard 4/40
- GitHub Check: Run Backend Tests / Unit Tests - Python 3.10 - Group 3
- GitHub Check: Run Backend Tests / Unit Tests - Python 3.10 - Group 2
- GitHub Check: Run Backend Tests / Unit Tests - Python 3.10 - Group 5
- GitHub Check: Run Backend Tests / Unit Tests - Python 3.10 - Group 1
- GitHub Check: Test Starter Templates
- GitHub Check: Optimize new Python code in this PR
🔇 Additional comments (4)
src/lfx/src/lfx/base/data/base_file.py (1)
183-184: ```shell
#!/bin/bash
set -euo pipefailecho "PWD: $(pwd)"
echo
echo "== Files containing 'class Output' =="
rg -nP --hidden -S 'class\s+Output\b' || trueecho
echo "== Full context (up to 200 lines) around each 'class Output' match =="
files=$(rg -lP --hidden -S 'class\s+Output\b' || true)
if [ -n "$files" ]; then
for f in $files; do
echo "---- $f ----"
rg -nP -U -C200 'class\s+Output\b' "$f" || true
echo
done
else
echo "(no files found)"
fiecho
echo "== All occurrences of 'tool_mode' in repo =="
rg -nP --hidden -S '\btool_mode\b' || trueecho
echo "== All 'Output(' calls under src/ (lines that do NOT contain 'tool_mode=' on the same line) =="
rg -nP --hidden -S 'Output(' src || true
echo
rg -nP --hidden -S 'Output(' src | rg -nP -v 'tool_mode\s*=' || trueecho
echo "Done."</blockquote></details> <details> <summary>src/backend/base/langflow/initial_setup/starter_projects/Vector Store RAG.json (1)</summary><blockquote> `2766-3514`: **Tool mode propagation: looks good.** All dynamically added outputs in `update_outputs` set `tool_mode=True`, matching the PR goal. Please confirm Save/Write File template JSONs also mark the “File Path” output with `tool_mode: true`. </blockquote></details> <details> <summary>src/lfx/src/lfx/components/data/file.py (2)</summary><blockquote> `31-34`: **Rename to “Read File” — LGTM** Public name/docstring updated cleanly. --- `163-164`: **Expose Raw Content in tool mode — LGTM** This makes the default output usable in tool flows. </blockquote></details> </blockquote></details> </details> <!-- This is an auto-generated comment by CodeRabbit for review status -->
| "title_case": false, | ||
| "type": "code", | ||
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\"),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"dataframe\", method=\"load_files_structured\"),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\"),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Output\", name=\"advanced\", method=\"load_files_advanced\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\"),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\"),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(Output(display_name=\"Files\", name=\"dataframe\", method=\"load_files\"))\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | ||
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" |
Contributor
There was a problem hiding this comment.
Add timeout and robust error handling to Docling subprocess call.
Same concern as the other starter: protect against hangs and surface a clear timeout error.
Apply this diff inside the embedded FileComponent code:
- proc = subprocess.run( # noqa: S603
- [sys.executable, "-u", "-c", child_script],
- input=json.dumps(args).encode("utf-8"),
- capture_output=True,
- check=False,
- )
+ try:
+ proc = subprocess.run( # noqa: S603
+ [sys.executable, "-u", "-c", child_script],
+ input=json.dumps(args).encode("utf-8"),
+ capture_output=True,
+ check=False,
+ timeout=getattr(self, "subprocess_timeout", 120),
+ )
+ except subprocess.TimeoutExpired:
+ return Data(data={"error": "Docling subprocess timed out", "file_path": file_path})📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | |
| # Validate file_path to avoid command injection or unsafe input | |
| if not isinstance(args["file_path"], str) or any(c in args["file_path"] for c in [";", "|", "&", "$", "`"]): | |
| return Data(data={"error": "Unsafe file path detected.", "file_path": args["file_path"]}) | |
| try: | |
| proc = subprocess.run( # noqa: S603 | |
| [sys.executable, "-u", "-c", child_script], | |
| input=json.dumps(args).encode("utf-8"), | |
| capture_output=True, | |
| check=False, | |
| timeout=getattr(self, "subprocess_timeout", 120), | |
| ) | |
| except subprocess.TimeoutExpired: | |
| return Data(data={"error": "Docling subprocess timed out", "file_path": file_path}) | |
| if not proc.stdout: | |
| err_msg = proc.stderr.decode("utf-8", errors="replace") or "no output from child process" | |
| return Data(data={"error": f"Docling subprocess error: {err_msg}", "file_path": file_path}) |
🤖 Prompt for AI Agents
In src/backend/base/langflow/initial_setup/starter_projects/Document Q&A.json
around line 1333, the Docling subprocess invocation using subprocess.run has no
timeout or explicit handling for hung/long-running children; add a timeout
argument (reasonable default, e.g., 60s) to subprocess.run and wrap the call in
try/except to catch subprocess.TimeoutExpired and other exceptions; on
TimeoutExpired decode any partial output/stderr and return a Data object with a
clear timeout error and file_path, and on other exceptions return a Data object
with the subprocess error message and stderr included so the caller gets robust,
surfaced errors instead of hanging indefinitely.
Guard None from _process_docling_in_subprocess before rollup.
Avoid passing [None] to rollup_data when the child returns nothing.
Apply this diff:
if self.advanced_mode and self._is_docling_compatible(file_path):
advanced_data: Data | None = self._process_docling_in_subprocess(file_path)
+ if advanced_data is None:
+ return self.rollup_data(
+ file_list,
+ [Data(data={"error": "Docling returned no data", "file_path": file_path})],
+ )
# --- UNNEST: expand each element in `doc` to its own Data row
payload = getattr(advanced_data, "data", {}) or {}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | |
| """Enhanced file component with clearer structure and Docling isolation. | |
| Notes: | |
| ----- | |
| - Functionality is preserved with minimal behavioral changes. | |
| - ALL Docling parsing/export runs in a separate OS process to prevent memory | |
| growth and native library state from impacting the main Langflow process. | |
| - Standard text/structured parsing continues to use existing BaseFileComponent | |
| utilities (and optional threading via `parallel_load_data`). | |
| """ | |
| from __future__ import annotations | |
| import json | |
| import subprocess | |
| import sys | |
| import textwrap | |
| from copy import deepcopy | |
| from typing import Any | |
| from lfx.base.data.base_file import BaseFileComponent | |
| from lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data | |
| from lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput | |
| from lfx.io import BoolInput, FileInput, IntInput, Output | |
| from lfx.schema import DataFrame # noqa: TC001 | |
| from lfx.schema.data import Data | |
| from lfx.schema.message import Message | |
| class FileComponent(BaseFileComponent): | |
| """Read File component with optional Docling processing (isolated in a subprocess).""" | |
| display_name = "Read File" | |
| description = "Loads content from files with optional advanced document processing and export using Docling." | |
| documentation: str = "https://docs.langflow.org/components-data#file" | |
| icon = "file-text" | |
| name = "File" | |
| # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader. | |
| VALID_EXTENSIONS = [ | |
| "adoc", | |
| "asciidoc", | |
| "asc", | |
| "bmp", | |
| "csv", | |
| "dotx", | |
| "dotm", | |
| "docm", | |
| "docx", | |
| "htm", | |
| "html", | |
| "jpeg", | |
| "json", | |
| "md", | |
| "pdf", | |
| "png", | |
| "potx", | |
| "ppsx", | |
| "pptm", | |
| "potm", | |
| "ppsm", | |
| "pptx", | |
| "tiff", | |
| "txt", | |
| "xls", | |
| "xlsx", | |
| "xhtml", | |
| "xml", | |
| "webp", | |
| *TEXT_FILE_TYPES, | |
| ] | |
| # Fixed export settings used when markdown export is requested. | |
| EXPORT_FORMAT = "Markdown" | |
| IMAGE_MODE = "placeholder" | |
| _base_inputs = deepcopy(BaseFileComponent.get_base_inputs()) | |
| for input_item in _base_inputs: | |
| if isinstance(input_item, FileInput) and input_item.name == "path": | |
| input_item.real_time_refresh = True | |
| break | |
| inputs = [ | |
| *_base_inputs, | |
| BoolInput( | |
| name="advanced_mode", | |
| display_name="Advanced Parser", | |
| value=False, | |
| real_time_refresh=True, | |
| info=( | |
| "Enable advanced document processing and export with Docling for PDFs, images, and office documents. " | |
| "Available only for single file processing." | |
| ), | |
| show=False, | |
| ), | |
| DropdownInput( | |
| name="pipeline", | |
| display_name="Pipeline", | |
| info="Docling pipeline to use", | |
| options=["standard", "vlm"], | |
| value="standard", | |
| advanced=True, | |
| ), | |
| DropdownInput( | |
| name="ocr_engine", | |
| display_name="OCR Engine", | |
| info="OCR engine to use. Only available when pipeline is set to 'standard'.", | |
| options=["", "easyocr"], | |
| value="", | |
| show=False, | |
| advanced=True, | |
| ), | |
| StrInput( | |
| name="md_image_placeholder", | |
| display_name="Image placeholder", | |
| info="Specify the image placeholder for markdown exports.", | |
| value="<!-- image -->", | |
| advanced=True, | |
| show=False, | |
| ), | |
| StrInput( | |
| name="md_page_break_placeholder", | |
| display_name="Page break placeholder", | |
| info="Add this placeholder between pages in the markdown output.", | |
| value="", | |
| advanced=True, | |
| show=False, | |
| ), | |
| MessageTextInput( | |
| name="doc_key", | |
| display_name="Doc Key", | |
| info="The key to use for the DoclingDocument column.", | |
| value="doc", | |
| advanced=True, | |
| show=False, | |
| ), | |
| # Deprecated input retained for backward-compatibility. | |
| BoolInput( | |
| name="use_multithreading", | |
| display_name="[Deprecated] Use Multithreading", | |
| advanced=True, | |
| value=True, | |
| info="Set 'Processing Concurrency' greater than 1 to enable multithreading.", | |
| ), | |
| IntInput( | |
| name="concurrency_multithreading", | |
| display_name="Processing Concurrency", | |
| advanced=True, | |
| info="When multiple files are being processed, the number of files to process concurrently.", | |
| value=1, | |
| ), | |
| BoolInput( | |
| name="markdown", | |
| display_name="Markdown Export", | |
| info="Export processed documents to Markdown format. Only available when advanced mode is enabled.", | |
| value=False, | |
| show=False, | |
| ), | |
| ] | |
| outputs = [ | |
| Output(display_name="Raw Content", name="message", method="load_files_message", tool_mode=True), | |
| ] | |
| # ------------------------------ UI helpers -------------------------------------- | |
| def _path_value(self, template: dict) -> list[str]: | |
| """Return the list of currently selected file paths from the template.""" | |
| return template.get("path", {}).get("file_path", []) | |
| def update_build_config( | |
| self, | |
| build_config: dict[str, Any], | |
| field_value: Any, | |
| field_name: str | None = None, | |
| ) -> dict[str, Any]: | |
| """Show/hide Advanced Parser and related fields based on selection context.""" | |
| if field_name == "path": | |
| paths = self._path_value(build_config) | |
| file_path = paths[0] if paths else "" | |
| file_count = len(field_value) if field_value else 0 | |
| # Advanced mode only for single (non-tabular) file | |
| allow_advanced = file_count == 1 and not file_path.endswith((".csv", ".xlsx", ".parquet")) | |
| build_config["advanced_mode"]["show"] = allow_advanced | |
| if not allow_advanced: | |
| build_config["advanced_mode"]["value"] = False | |
| for f in ("pipeline", "ocr_engine", "doc_key", "md_image_placeholder", "md_page_break_placeholder"): | |
| if f in build_config: | |
| build_config[f]["show"] = False | |
| elif field_name == "advanced_mode": | |
| for f in ("pipeline", "ocr_engine", "doc_key", "md_image_placeholder", "md_page_break_placeholder"): | |
| if f in build_config: | |
| build_config[f]["show"] = bool(field_value) | |
| return build_config | |
| def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002 | |
| """Dynamically show outputs based on file count/type and advanced mode.""" | |
| if field_name not in ["path", "advanced_mode"]: | |
| return frontend_node | |
| template = frontend_node.get("template", {}) | |
| paths = self._path_value(template) | |
| if not paths: | |
| return frontend_node | |
| frontend_node["outputs"] = [] | |
| if len(paths) == 1: | |
| file_path = paths[0] if field_name == "path" else frontend_node["template"]["path"]["file_path"][0] | |
| if file_path.endswith((".csv", ".xlsx", ".parquet")): | |
| frontend_node["outputs"].append( | |
| Output( | |
| display_name="Structured Content", | |
| name="dataframe", | |
| method="load_files_structured", | |
| tool_mode=True, | |
| ), | |
| ) | |
| elif file_path.endswith(".json"): | |
| frontend_node["outputs"].append( | |
| Output(display_name="Structured Content", name="json", method="load_files_json", tool_mode=True), | |
| ) | |
| advanced_mode = frontend_node.get("template", {}).get("advanced_mode", {}).get("value", False) | |
| if advanced_mode: | |
| frontend_node["outputs"].append( | |
| Output( | |
| display_name="Structured Output", | |
| name="advanced", | |
| method="load_files_advanced", | |
| tool_mode=True, | |
| ), | |
| ) | |
| frontend_node["outputs"].append( | |
| Output(display_name="Markdown", name="markdown", method="load_files_markdown", tool_mode=True), | |
| ) | |
| frontend_node["outputs"].append( | |
| Output(display_name="File Path", name="path", method="load_files_path", tool_mode=True), | |
| ) | |
| else: | |
| frontend_node["outputs"].append( | |
| Output(display_name="Raw Content", name="message", method="load_files_message", tool_mode=True), | |
| ) | |
| frontend_node["outputs"].append( | |
| Output(display_name="File Path", name="path", method="load_files_path", tool_mode=True), | |
| ) | |
| else: | |
| # Multiple files => DataFrame output; advanced parser disabled | |
| frontend_node["outputs"].append( | |
| Output( | |
| display_name="Files", | |
| name="dataframe", | |
| method="load_files", | |
| tool_mode=True, | |
| ) | |
| ) | |
| return frontend_node | |
| # ------------------------------ Core processing ---------------------------------- | |
| def _is_docling_compatible(self, file_path: str) -> bool: | |
| """Lightweight extension gate for Docling-compatible types.""" | |
| docling_exts = ( | |
| ".adoc", | |
| ".asciidoc", | |
| ".asc", | |
| ".bmp", | |
| ".csv", | |
| ".dotx", | |
| ".dotm", | |
| ".docm", | |
| ".docx", | |
| ".htm", | |
| ".html", | |
| ".jpeg", | |
| ".json", | |
| ".md", | |
| ".pdf", | |
| ".png", | |
| ".potx", | |
| ".ppsx", | |
| ".pptm", | |
| ".potm", | |
| ".ppsm", | |
| ".pptx", | |
| ".tiff", | |
| ".txt", | |
| ".xls", | |
| ".xlsx", | |
| ".xhtml", | |
| ".xml", | |
| ".webp", | |
| ) | |
| return file_path.lower().endswith(docling_exts) | |
| def _process_docling_in_subprocess(self, file_path: str) -> Data | None: | |
| """Run Docling in a separate OS process and map the result to a Data object. | |
| We avoid multiprocessing pickling by launching `python -c \"<script>\"` and | |
| passing JSON config via stdin. The child prints a JSON result to stdout. | |
| """ | |
| if not file_path: | |
| return None | |
| args: dict[str, Any] = { | |
| "file_path": file_path, | |
| "markdown": bool(self.markdown), | |
| "image_mode": str(self.IMAGE_MODE), | |
| "md_image_placeholder": str(self.md_image_placeholder), | |
| "md_page_break_placeholder": str(self.md_page_break_placeholder), | |
| "pipeline": str(self.pipeline), | |
| "ocr_engine": str(self.ocr_engine) if getattr(self, "ocr_engine", "") else None, | |
| } | |
| # The child is a tiny, self-contained script to keep memory/state isolated. | |
| child_script = textwrap.dedent( | |
| r""" | |
| import json, sys | |
| def try_imports(): | |
| # Strategy 1: latest layout | |
| try: | |
| from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| from docling_core.types.doc import ImageRefMode # type: ignore | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "latest" | |
| except Exception: | |
| pass | |
| # Strategy 2: alternative layout | |
| try: | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| try: | |
| from docling_core.types import ConversionStatus, InputFormat # type: ignore | |
| except Exception: | |
| try: | |
| from docling.datamodel import ConversionStatus, InputFormat # type: ignore | |
| except Exception: | |
| class ConversionStatus: SUCCESS = "success" | |
| class InputFormat: | |
| PDF="pdf"; IMAGE="image" | |
| try: | |
| from docling_core.types.doc import ImageRefMode # type: ignore | |
| except Exception: | |
| class ImageRefMode: | |
| PLACEHOLDER="placeholder"; EMBEDDED="embedded" | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "alternative" | |
| except Exception: | |
| pass | |
| # Strategy 3: basic converter only | |
| try: | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| class ConversionStatus: SUCCESS = "success" | |
| class InputFormat: | |
| PDF="pdf"; IMAGE="image" | |
| class ImageRefMode: | |
| PLACEHOLDER="placeholder"; EMBEDDED="embedded" | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "basic" | |
| except Exception as e: | |
| raise ImportError(f"Docling imports failed: {e}") from e | |
| def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine): | |
| if strategy == "latest" and pipeline == "standard": | |
| try: | |
| from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore | |
| from docling.document_converter import PdfFormatOption # type: ignore | |
| pipe = PdfPipelineOptions() | |
| if ocr_engine: | |
| try: | |
| from docling.models.factories import get_ocr_factory # type: ignore | |
| pipe.do_ocr = True | |
| fac = get_ocr_factory(allow_external_plugins=False) | |
| pipe.ocr_options = fac.create_options(kind=ocr_engine) | |
| except Exception: | |
| pipe.do_ocr = False | |
| fmt = {} | |
| if hasattr(input_format, "PDF"): | |
| fmt[getattr(input_format, "PDF")] = PdfFormatOption(pipeline_options=pipe) | |
| if hasattr(input_format, "IMAGE"): | |
| fmt[getattr(input_format, "IMAGE")] = PdfFormatOption(pipeline_options=pipe) | |
| return DocumentConverter(format_options=fmt) | |
| except Exception: | |
| return DocumentConverter() | |
| return DocumentConverter() | |
| def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph): | |
| try: | |
| mode = getattr(ImageRefMode, image_mode.upper(), image_mode) | |
| return document.export_to_markdown( | |
| image_mode=mode, | |
| image_placeholder=img_ph, | |
| page_break_placeholder=pg_ph, | |
| ) | |
| except Exception: | |
| try: | |
| return document.export_to_text() | |
| except Exception: | |
| return str(document) | |
| def to_rows(doc_dict): | |
| rows = [] | |
| for t in doc_dict.get("texts", []): | |
| prov = t.get("prov") or [] | |
| page_no = None | |
| if prov and isinstance(prov, list) and isinstance(prov[0], dict): | |
| page_no = prov[0].get("page_no") | |
| rows.append({ | |
| "page_no": page_no, | |
| "label": t.get("label"), | |
| "text": t.get("text"), | |
| "level": t.get("level"), | |
| }) | |
| return rows | |
| def main(): | |
| cfg = json.loads(sys.stdin.read()) | |
| file_path = cfg["file_path"] | |
| markdown = cfg["markdown"] | |
| image_mode = cfg["image_mode"] | |
| img_ph = cfg["md_image_placeholder"] | |
| pg_ph = cfg["md_page_break_placeholder"] | |
| pipeline = cfg["pipeline"] | |
| ocr_engine = cfg.get("ocr_engine") | |
| meta = {"file_path": file_path} | |
| try: | |
| ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports() | |
| converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine) | |
| try: | |
| res = converter.convert(file_path) | |
| except Exception as e: | |
| print(json.dumps({"ok": False, "error": f"Docling conversion error: {e}", "meta": meta})) | |
| return | |
| ok = False | |
| if hasattr(res, "status"): | |
| try: | |
| ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == "success") | |
| except Exception: | |
| ok = (str(res.status).lower() == "success") | |
| if not ok and hasattr(res, "document"): | |
| ok = getattr(res, "document", None) is not None | |
| if not ok: | |
| print(json.dumps({"ok": False, "error": "Docling conversion failed", "meta": meta})) | |
| return | |
| doc = getattr(res, "document", None) | |
| if doc is None: | |
| print(json.dumps({"ok": False, "error": "Docling produced no document", "meta": meta})) | |
| return | |
| if markdown: | |
| text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph) | |
| print(json.dumps({"ok": True, "mode": "markdown", "text": text, "meta": meta})) | |
| return | |
| # structured | |
| try: | |
| doc_dict = doc.export_to_dict() | |
| except Exception as e: | |
| print(json.dumps({"ok": False, "error": f"Docling export_to_dict failed: {e}", "meta": meta})) | |
| return | |
| rows = to_rows(doc_dict) | |
| print(json.dumps({"ok": True, "mode": "structured", "doc": rows, "meta": meta})) | |
| except Exception as e: | |
| print( | |
| json.dumps({ | |
| "ok": False, | |
| "error": f"Docling processing error: {e}", | |
| "meta": {"file_path": file_path}, | |
| }) | |
| ) | |
| if __name__ == "__main__": | |
| main() | |
| """ | |
| ) | |
| # Validate file_path to avoid command injection or unsafe input | |
| if not isinstance(args["file_path"], str) or any(c in args["file_path"] for c in [";", "|", "&", "$", "`"]): | |
| return Data(data={"error": "Unsafe file path detected.", "file_path": args["file_path"]}) | |
| proc = subprocess.run( # noqa: S603 | |
| [sys.executable, "-u", "-c", child_script], | |
| input=json.dumps(args).encode("utf-8"), | |
| capture_output=True, | |
| check=False, | |
| ) | |
| if not proc.stdout: | |
| err_msg = proc.stderr.decode("utf-8", errors="replace") or "no output from child process" | |
| return Data(data={"error": f"Docling subprocess error: {err_msg}", "file_path": file_path}) | |
| try: | |
| result = json.loads(proc.stdout.decode("utf-8")) | |
| except Exception as e: # noqa: BLE001 | |
| err_msg = proc.stderr.decode("utf-8", errors="replace") | |
| return Data( | |
| data={"error": f"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}", "file_path": file_path}, | |
| ) | |
| if not result.get("ok"): | |
| return Data(data={"error": result.get("error", "Unknown Docling error"), **result.get("meta", {})}) | |
| meta = result.get("meta", {}) | |
| if result.get("mode") == "markdown": | |
| exported_content = str(result.get("text", "")) | |
| return Data( | |
| text=exported_content, | |
| data={"exported_content": exported_content, "export_format": self.EXPORT_FORMAT, **meta}, | |
| ) | |
| rows = list(result.get("doc", [])) | |
| return Data(data={"doc": rows, "export_format": self.EXPORT_FORMAT, **meta}) | |
| def process_files( | |
| self, | |
| file_list: list[BaseFileComponent.BaseFile], | |
| ) -> list[BaseFileComponent.BaseFile]: | |
| """Process input files. | |
| - Single file + advanced_mode => Docling in a separate process. | |
| - Otherwise => standard parsing in current process (optionally threaded). | |
| """ | |
| if not file_list: | |
| msg = "No files to process." | |
| raise ValueError(msg) | |
| def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None: | |
| try: | |
| return parse_text_file_to_data(file_path, silent_errors=silent_errors) | |
| except FileNotFoundError as e: | |
| self.log(f"File not found: {file_path}. Error: {e}") | |
| if not silent_errors: | |
| raise | |
| return None | |
| except Exception as e: | |
| self.log(f"Unexpected error processing {file_path}: {e}") | |
| if not silent_errors: | |
| raise | |
| return None | |
| # Advanced path: only for a single Docling-compatible file | |
| if len(file_list) == 1: | |
| file_path = str(file_list[0].path) | |
| if self.advanced_mode and self._is_docling_compatible(file_path): | |
| advanced_data: Data | None = self._process_docling_in_subprocess(file_path) | |
| if advanced_data is None: | |
| return self.rollup_data( | |
| file_list, | |
| [Data(data={"error": "Docling returned no data", "file_path": file_path})], | |
| ) | |
| # --- UNNEST: expand each element in `doc` to its own Data row | |
| payload = getattr(advanced_data, "data", {}) or {} | |
| doc_rows = payload.get("doc") | |
| if isinstance(doc_rows, list): | |
| rows: list[Data | None] = [ | |
| Data( | |
| data={ | |
| "file_path": file_path, | |
| **(item if isinstance(item, dict) else {"value": item}), | |
| }, | |
| ) | |
| for item in doc_rows | |
| ] | |
| return self.rollup_data(file_list, rows) | |
| # If not structured, keep as-is (e.g., markdown export or error dict) | |
| return self.rollup_data(file_list, [advanced_data]) | |
| # Standard multi-file (or single non-advanced) path | |
| concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading) | |
| file_paths = [str(f.path) for f in file_list] | |
| self.log(f"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.") | |
| my_data = parallel_load_data( | |
| file_paths, | |
| silent_errors=self.silent_errors, | |
| load_function=process_file_standard, | |
| max_concurrency=concurrency, | |
| ) | |
| return self.rollup_data(file_list, my_data) | |
| # ------------------------------ Output helpers ----------------------------------- | |
| def load_files_advanced(self) -> DataFrame: | |
| """Load files using advanced Docling processing and export to an advanced format.""" | |
| self.markdown = False | |
| return self.load_files() | |
| def load_files_markdown(self) -> Message: | |
| """Load files using advanced Docling processing and export to Markdown format.""" | |
| self.markdown = True | |
| result = self.load_files() | |
| return Message(text=str(result.text[0])) |
🤖 Prompt for AI Agents
In src/backend/base/langflow/initial_setup/starter_projects/Document Q&A.json
around line 1333, guard against advanced_data being None before calling
rollup_data: after advanced_data =
self._process_docling_in_subprocess(file_path) check if advanced_data is None
(or getattr(advanced_data, "data", None) is None) and if so create a fallback
Data object (e.g., Data(data={"error": "No data from Docling", "file_path":
file_path})) and pass that fallback into rollup_data instead of None; then
proceed with the existing logic to unnest doc rows or return rollup_data.
| "title_case": false, | ||
| "type": "code", | ||
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\"),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"dataframe\", method=\"load_files_structured\"),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\"),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Output\", name=\"advanced\", method=\"load_files_advanced\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\"),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\"),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\"),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(Output(display_name=\"Files\", name=\"dataframe\", method=\"load_files\"))\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | ||
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" |
Contributor
There was a problem hiding this comment.
Guard None from _process_docling_in_subprocess before rollup.
If advanced_data is None, rollup_data(file_list, [advanced_data]) passes [None] and may break downstream.
Apply this diff:
if self.advanced_mode and self._is_docling_compatible(file_path):
advanced_data: Data | None = self._process_docling_in_subprocess(file_path)
+ if advanced_data is None:
+ return self.rollup_data(
+ file_list,
+ [Data(data={"error": "Docling returned no data", "file_path": file_path})],
+ )
# --- UNNEST: expand each element in `doc` to its own Data row
payload = getattr(advanced_data, "data", {}) or {}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | |
| def process_files( | |
| self, | |
| file_list: list[BaseFileComponent.BaseFile], | |
| ) -> list[BaseFileComponent.BaseFile]: | |
| \"\"\"Process input files. | |
| - Single file + advanced_mode => Docling in a separate process. | |
| - Otherwise => standard parsing in current process (optionally threaded). | |
| \"\"\" | |
| if not file_list: | |
| msg = \"No files to process.\" | |
| raise ValueError(msg) | |
| def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None: | |
| try: | |
| return parse_text_file_to_data(file_path, silent_errors=silent_errors) | |
| except FileNotFoundError as e: | |
| self.log(f\"File not found: {file_path}. Error: {e}\") | |
| if not silent_errors: | |
| raise | |
| return None | |
| except Exception as e: | |
| self.log(f\"Unexpected error processing {file_path}: {e}\") | |
| if not silent_errors: | |
| raise | |
| return None | |
| # Advanced path: only for a single Docling-compatible file | |
| if len(file_list) == 1: | |
| file_path = str(file_list[0].path) | |
| if self.advanced_mode and self._is_docling_compatible(file_path): | |
| advanced_data: Data | None = self._process_docling_in_subprocess(file_path) | |
| if advanced_data is None: | |
| return self.rollup_data( | |
| file_list, | |
| [Data(data={\"error\": \"Docling returned no data\", \"file_path\": file_path})], | |
| ) | |
| # --- UNNEST: expand each element in `doc` to its own Data row | |
| payload = getattr(advanced_data, \"data\", {}) or {} | |
| doc_rows = payload.get(\"doc\") | |
| if isinstance(doc_rows, list): | |
| rows: list[Data | None] = [ | |
| Data( | |
| data={ | |
| \"file_path\": file_path, | |
| **(item if isinstance(item, dict) else {\"value\": item}), | |
| }, | |
| ) | |
| for item in doc_rows | |
| ] | |
| return self.rollup_data(file_list, rows) | |
| # If not structured, keep as-is (e.g., markdown export or error dict) | |
| return self.rollup_data(file_list, [advanced_data]) | |
| # Standard multi-file (or single non-advanced) path | |
| concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading) | |
| file_paths = [str(f.path) for f in file_list] | |
| self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\") | |
| my_data = parallel_load_data( | |
| file_paths, | |
| silent_errors=self.silent_errors, | |
| load_function=process_file_standard, | |
| max_concurrency=concurrency, | |
| ) | |
| return self.rollup_data(file_list, my_data) |
🤖 Prompt for AI Agents
In src/backend/base/langflow/initial_setup/starter_projects/Text Sentiment
Analysis.json around line 2474, guard against advanced_data being None before
calling rollup_data: after advanced_data =
self._process_docling_in_subprocess(file_path) check if advanced_data is None
and handle it (e.g., return self.rollup_data(file_list, []) or return an error
Data row) instead of calling self.rollup_data(file_list, [advanced_data]) to
avoid passing [None] downstream.
Add timeout and robust error handling to Docling subprocess call.
subprocess.run has no timeout; a stuck child will hang the flow. Also surface a clear timeout error.
Apply this diff inside the embedded FileComponent code:
- proc = subprocess.run( # noqa: S603
- [sys.executable, "-u", "-c", child_script],
- input=json.dumps(args).encode("utf-8"),
- capture_output=True,
- check=False,
- )
+ try:
+ proc = subprocess.run( # noqa: S603
+ [sys.executable, "-u", "-c", child_script],
+ input=json.dumps(args).encode("utf-8"),
+ capture_output=True,
+ check=False,
+ timeout=getattr(self, "subprocess_timeout", 120),
+ )
+ except subprocess.TimeoutExpired:
+ return Data(data={"error": "Docling subprocess timed out", "file_path": file_path})📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| "value": "\"\"\"Enhanced file component with clearer structure and Docling isolation.\n\nNotes:\n-----\n- Functionality is preserved with minimal behavioral changes.\n- ALL Docling parsing/export runs in a separate OS process to prevent memory\n growth and native library state from impacting the main Langflow process.\n- Standard text/structured parsing continues to use existing BaseFileComponent\n utilities (and optional threading via `parallel_load_data`).\n\"\"\"\n\nfrom __future__ import annotations\n\nimport json\nimport subprocess\nimport sys\nimport textwrap\nfrom copy import deepcopy\nfrom typing import Any\n\nfrom lfx.base.data.base_file import BaseFileComponent\nfrom lfx.base.data.utils import TEXT_FILE_TYPES, parallel_load_data, parse_text_file_to_data\nfrom lfx.inputs.inputs import DropdownInput, MessageTextInput, StrInput\nfrom lfx.io import BoolInput, FileInput, IntInput, Output\nfrom lfx.schema import DataFrame # noqa: TC001\nfrom lfx.schema.data import Data\nfrom lfx.schema.message import Message\n\n\nclass FileComponent(BaseFileComponent):\n \"\"\"Read File component with optional Docling processing (isolated in a subprocess).\"\"\"\n\n display_name = \"Read File\"\n description = \"Loads content from files with optional advanced document processing and export using Docling.\"\n documentation: str = \"https://docs.langflow.org/components-data#file\"\n icon = \"file-text\"\n name = \"File\"\n\n # Docling-supported/compatible extensions; TEXT_FILE_TYPES are supported by the base loader.\n VALID_EXTENSIONS = [\n \"adoc\",\n \"asciidoc\",\n \"asc\",\n \"bmp\",\n \"csv\",\n \"dotx\",\n \"dotm\",\n \"docm\",\n \"docx\",\n \"htm\",\n \"html\",\n \"jpeg\",\n \"json\",\n \"md\",\n \"pdf\",\n \"png\",\n \"potx\",\n \"ppsx\",\n \"pptm\",\n \"potm\",\n \"ppsm\",\n \"pptx\",\n \"tiff\",\n \"txt\",\n \"xls\",\n \"xlsx\",\n \"xhtml\",\n \"xml\",\n \"webp\",\n *TEXT_FILE_TYPES,\n ]\n\n # Fixed export settings used when markdown export is requested.\n EXPORT_FORMAT = \"Markdown\"\n IMAGE_MODE = \"placeholder\"\n\n _base_inputs = deepcopy(BaseFileComponent.get_base_inputs())\n\n for input_item in _base_inputs:\n if isinstance(input_item, FileInput) and input_item.name == \"path\":\n input_item.real_time_refresh = True\n break\n\n inputs = [\n *_base_inputs,\n BoolInput(\n name=\"advanced_mode\",\n display_name=\"Advanced Parser\",\n value=False,\n real_time_refresh=True,\n info=(\n \"Enable advanced document processing and export with Docling for PDFs, images, and office documents. \"\n \"Available only for single file processing.\"\n ),\n show=False,\n ),\n DropdownInput(\n name=\"pipeline\",\n display_name=\"Pipeline\",\n info=\"Docling pipeline to use\",\n options=[\"standard\", \"vlm\"],\n value=\"standard\",\n advanced=True,\n ),\n DropdownInput(\n name=\"ocr_engine\",\n display_name=\"OCR Engine\",\n info=\"OCR engine to use. Only available when pipeline is set to 'standard'.\",\n options=[\"\", \"easyocr\"],\n value=\"\",\n show=False,\n advanced=True,\n ),\n StrInput(\n name=\"md_image_placeholder\",\n display_name=\"Image placeholder\",\n info=\"Specify the image placeholder for markdown exports.\",\n value=\"<!-- image -->\",\n advanced=True,\n show=False,\n ),\n StrInput(\n name=\"md_page_break_placeholder\",\n display_name=\"Page break placeholder\",\n info=\"Add this placeholder between pages in the markdown output.\",\n value=\"\",\n advanced=True,\n show=False,\n ),\n MessageTextInput(\n name=\"doc_key\",\n display_name=\"Doc Key\",\n info=\"The key to use for the DoclingDocument column.\",\n value=\"doc\",\n advanced=True,\n show=False,\n ),\n # Deprecated input retained for backward-compatibility.\n BoolInput(\n name=\"use_multithreading\",\n display_name=\"[Deprecated] Use Multithreading\",\n advanced=True,\n value=True,\n info=\"Set 'Processing Concurrency' greater than 1 to enable multithreading.\",\n ),\n IntInput(\n name=\"concurrency_multithreading\",\n display_name=\"Processing Concurrency\",\n advanced=True,\n info=\"When multiple files are being processed, the number of files to process concurrently.\",\n value=1,\n ),\n BoolInput(\n name=\"markdown\",\n display_name=\"Markdown Export\",\n info=\"Export processed documents to Markdown format. Only available when advanced mode is enabled.\",\n value=False,\n show=False,\n ),\n ]\n\n outputs = [\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n ]\n\n # ------------------------------ UI helpers --------------------------------------\n\n def _path_value(self, template: dict) -> list[str]:\n \"\"\"Return the list of currently selected file paths from the template.\"\"\"\n return template.get(\"path\", {}).get(\"file_path\", [])\n\n def update_build_config(\n self,\n build_config: dict[str, Any],\n field_value: Any,\n field_name: str | None = None,\n ) -> dict[str, Any]:\n \"\"\"Show/hide Advanced Parser and related fields based on selection context.\"\"\"\n if field_name == \"path\":\n paths = self._path_value(build_config)\n file_path = paths[0] if paths else \"\"\n file_count = len(field_value) if field_value else 0\n\n # Advanced mode only for single (non-tabular) file\n allow_advanced = file_count == 1 and not file_path.endswith((\".csv\", \".xlsx\", \".parquet\"))\n build_config[\"advanced_mode\"][\"show\"] = allow_advanced\n if not allow_advanced:\n build_config[\"advanced_mode\"][\"value\"] = False\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = False\n\n elif field_name == \"advanced_mode\":\n for f in (\"pipeline\", \"ocr_engine\", \"doc_key\", \"md_image_placeholder\", \"md_page_break_placeholder\"):\n if f in build_config:\n build_config[f][\"show\"] = bool(field_value)\n\n return build_config\n\n def update_outputs(self, frontend_node: dict[str, Any], field_name: str, field_value: Any) -> dict[str, Any]: # noqa: ARG002\n \"\"\"Dynamically show outputs based on file count/type and advanced mode.\"\"\"\n if field_name not in [\"path\", \"advanced_mode\"]:\n return frontend_node\n\n template = frontend_node.get(\"template\", {})\n paths = self._path_value(template)\n if not paths:\n return frontend_node\n\n frontend_node[\"outputs\"] = []\n if len(paths) == 1:\n file_path = paths[0] if field_name == \"path\" else frontend_node[\"template\"][\"path\"][\"file_path\"][0]\n if file_path.endswith((\".csv\", \".xlsx\", \".parquet\")):\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Content\",\n name=\"dataframe\",\n method=\"load_files_structured\",\n tool_mode=True,\n ),\n )\n elif file_path.endswith(\".json\"):\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Structured Content\", name=\"json\", method=\"load_files_json\", tool_mode=True),\n )\n\n advanced_mode = frontend_node.get(\"template\", {}).get(\"advanced_mode\", {}).get(\"value\", False)\n if advanced_mode:\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Structured Output\",\n name=\"advanced\",\n method=\"load_files_advanced\",\n tool_mode=True,\n ),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Markdown\", name=\"markdown\", method=\"load_files_markdown\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n frontend_node[\"outputs\"].append(\n Output(display_name=\"Raw Content\", name=\"message\", method=\"load_files_message\", tool_mode=True),\n )\n frontend_node[\"outputs\"].append(\n Output(display_name=\"File Path\", name=\"path\", method=\"load_files_path\", tool_mode=True),\n )\n else:\n # Multiple files => DataFrame output; advanced parser disabled\n frontend_node[\"outputs\"].append(\n Output(\n display_name=\"Files\",\n name=\"dataframe\",\n method=\"load_files\",\n tool_mode=True,\n )\n )\n\n return frontend_node\n\n # ------------------------------ Core processing ----------------------------------\n\n def _is_docling_compatible(self, file_path: str) -> bool:\n \"\"\"Lightweight extension gate for Docling-compatible types.\"\"\"\n docling_exts = (\n \".adoc\",\n \".asciidoc\",\n \".asc\",\n \".bmp\",\n \".csv\",\n \".dotx\",\n \".dotm\",\n \".docm\",\n \".docx\",\n \".htm\",\n \".html\",\n \".jpeg\",\n \".json\",\n \".md\",\n \".pdf\",\n \".png\",\n \".potx\",\n \".ppsx\",\n \".pptm\",\n \".potm\",\n \".ppsm\",\n \".pptx\",\n \".tiff\",\n \".txt\",\n \".xls\",\n \".xlsx\",\n \".xhtml\",\n \".xml\",\n \".webp\",\n )\n return file_path.lower().endswith(docling_exts)\n\n def _process_docling_in_subprocess(self, file_path: str) -> Data | None:\n \"\"\"Run Docling in a separate OS process and map the result to a Data object.\n\n We avoid multiprocessing pickling by launching `python -c \"<script>\"` and\n passing JSON config via stdin. The child prints a JSON result to stdout.\n \"\"\"\n if not file_path:\n return None\n\n args: dict[str, Any] = {\n \"file_path\": file_path,\n \"markdown\": bool(self.markdown),\n \"image_mode\": str(self.IMAGE_MODE),\n \"md_image_placeholder\": str(self.md_image_placeholder),\n \"md_page_break_placeholder\": str(self.md_page_break_placeholder),\n \"pipeline\": str(self.pipeline),\n \"ocr_engine\": str(self.ocr_engine) if getattr(self, \"ocr_engine\", \"\") else None,\n }\n\n # The child is a tiny, self-contained script to keep memory/state isolated.\n child_script = textwrap.dedent(\n r\"\"\"\n import json, sys\n\n def try_imports():\n # Strategy 1: latest layout\n try:\n from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore\n from docling.document_converter import DocumentConverter # type: ignore\n from docling_core.types.doc import ImageRefMode # type: ignore\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"latest\"\n except Exception:\n pass\n # Strategy 2: alternative layout\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n try:\n from docling_core.types import ConversionStatus, InputFormat # type: ignore\n except Exception:\n try:\n from docling.datamodel import ConversionStatus, InputFormat # type: ignore\n except Exception:\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n try:\n from docling_core.types.doc import ImageRefMode # type: ignore\n except Exception:\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"alternative\"\n except Exception:\n pass\n # Strategy 3: basic converter only\n try:\n from docling.document_converter import DocumentConverter # type: ignore\n class ConversionStatus: SUCCESS = \"success\"\n class InputFormat:\n PDF=\"pdf\"; IMAGE=\"image\"\n class ImageRefMode:\n PLACEHOLDER=\"placeholder\"; EMBEDDED=\"embedded\"\n return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, \"basic\"\n except Exception as e:\n raise ImportError(f\"Docling imports failed: {e}\") from e\n\n def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine):\n if strategy == \"latest\" and pipeline == \"standard\":\n try:\n from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore\n from docling.document_converter import PdfFormatOption # type: ignore\n pipe = PdfPipelineOptions()\n if ocr_engine:\n try:\n from docling.models.factories import get_ocr_factory # type: ignore\n pipe.do_ocr = True\n fac = get_ocr_factory(allow_external_plugins=False)\n pipe.ocr_options = fac.create_options(kind=ocr_engine)\n except Exception:\n pipe.do_ocr = False\n fmt = {}\n if hasattr(input_format, \"PDF\"):\n fmt[getattr(input_format, \"PDF\")] = PdfFormatOption(pipeline_options=pipe)\n if hasattr(input_format, \"IMAGE\"):\n fmt[getattr(input_format, \"IMAGE\")] = PdfFormatOption(pipeline_options=pipe)\n return DocumentConverter(format_options=fmt)\n except Exception:\n return DocumentConverter()\n return DocumentConverter()\n\n def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph):\n try:\n mode = getattr(ImageRefMode, image_mode.upper(), image_mode)\n return document.export_to_markdown(\n image_mode=mode,\n image_placeholder=img_ph,\n page_break_placeholder=pg_ph,\n )\n except Exception:\n try:\n return document.export_to_text()\n except Exception:\n return str(document)\n\n def to_rows(doc_dict):\n rows = []\n for t in doc_dict.get(\"texts\", []):\n prov = t.get(\"prov\") or []\n page_no = None\n if prov and isinstance(prov, list) and isinstance(prov[0], dict):\n page_no = prov[0].get(\"page_no\")\n rows.append({\n \"page_no\": page_no,\n \"label\": t.get(\"label\"),\n \"text\": t.get(\"text\"),\n \"level\": t.get(\"level\"),\n })\n return rows\n\n def main():\n cfg = json.loads(sys.stdin.read())\n file_path = cfg[\"file_path\"]\n markdown = cfg[\"markdown\"]\n image_mode = cfg[\"image_mode\"]\n img_ph = cfg[\"md_image_placeholder\"]\n pg_ph = cfg[\"md_page_break_placeholder\"]\n pipeline = cfg[\"pipeline\"]\n ocr_engine = cfg.get(\"ocr_engine\")\n meta = {\"file_path\": file_path}\n\n try:\n ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports()\n converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine)\n try:\n res = converter.convert(file_path)\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling conversion error: {e}\", \"meta\": meta}))\n return\n\n ok = False\n if hasattr(res, \"status\"):\n try:\n ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == \"success\")\n except Exception:\n ok = (str(res.status).lower() == \"success\")\n if not ok and hasattr(res, \"document\"):\n ok = getattr(res, \"document\", None) is not None\n if not ok:\n print(json.dumps({\"ok\": False, \"error\": \"Docling conversion failed\", \"meta\": meta}))\n return\n\n doc = getattr(res, \"document\", None)\n if doc is None:\n print(json.dumps({\"ok\": False, \"error\": \"Docling produced no document\", \"meta\": meta}))\n return\n\n if markdown:\n text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph)\n print(json.dumps({\"ok\": True, \"mode\": \"markdown\", \"text\": text, \"meta\": meta}))\n return\n\n # structured\n try:\n doc_dict = doc.export_to_dict()\n except Exception as e:\n print(json.dumps({\"ok\": False, \"error\": f\"Docling export_to_dict failed: {e}\", \"meta\": meta}))\n return\n\n rows = to_rows(doc_dict)\n print(json.dumps({\"ok\": True, \"mode\": \"structured\", \"doc\": rows, \"meta\": meta}))\n except Exception as e:\n print(\n json.dumps({\n \"ok\": False,\n \"error\": f\"Docling processing error: {e}\",\n \"meta\": {\"file_path\": file_path},\n })\n )\n\n if __name__ == \"__main__\":\n main()\n \"\"\"\n )\n\n # Validate file_path to avoid command injection or unsafe input\n if not isinstance(args[\"file_path\"], str) or any(c in args[\"file_path\"] for c in [\";\", \"|\", \"&\", \"$\", \"`\"]):\n return Data(data={\"error\": \"Unsafe file path detected.\", \"file_path\": args[\"file_path\"]})\n\n proc = subprocess.run( # noqa: S603\n [sys.executable, \"-u\", \"-c\", child_script],\n input=json.dumps(args).encode(\"utf-8\"),\n capture_output=True,\n check=False,\n )\n\n if not proc.stdout:\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\") or \"no output from child process\"\n return Data(data={\"error\": f\"Docling subprocess error: {err_msg}\", \"file_path\": file_path})\n\n try:\n result = json.loads(proc.stdout.decode(\"utf-8\"))\n except Exception as e: # noqa: BLE001\n err_msg = proc.stderr.decode(\"utf-8\", errors=\"replace\")\n return Data(\n data={\"error\": f\"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}\", \"file_path\": file_path},\n )\n\n if not result.get(\"ok\"):\n return Data(data={\"error\": result.get(\"error\", \"Unknown Docling error\"), **result.get(\"meta\", {})})\n\n meta = result.get(\"meta\", {})\n if result.get(\"mode\") == \"markdown\":\n exported_content = str(result.get(\"text\", \"\"))\n return Data(\n text=exported_content,\n data={\"exported_content\": exported_content, \"export_format\": self.EXPORT_FORMAT, **meta},\n )\n\n rows = list(result.get(\"doc\", []))\n return Data(data={\"doc\": rows, \"export_format\": self.EXPORT_FORMAT, **meta})\n\n def process_files(\n self,\n file_list: list[BaseFileComponent.BaseFile],\n ) -> list[BaseFileComponent.BaseFile]:\n \"\"\"Process input files.\n\n - Single file + advanced_mode => Docling in a separate process.\n - Otherwise => standard parsing in current process (optionally threaded).\n \"\"\"\n if not file_list:\n msg = \"No files to process.\"\n raise ValueError(msg)\n\n def process_file_standard(file_path: str, *, silent_errors: bool = False) -> Data | None:\n try:\n return parse_text_file_to_data(file_path, silent_errors=silent_errors)\n except FileNotFoundError as e:\n self.log(f\"File not found: {file_path}. Error: {e}\")\n if not silent_errors:\n raise\n return None\n except Exception as e:\n self.log(f\"Unexpected error processing {file_path}: {e}\")\n if not silent_errors:\n raise\n return None\n\n # Advanced path: only for a single Docling-compatible file\n if len(file_list) == 1:\n file_path = str(file_list[0].path)\n if self.advanced_mode and self._is_docling_compatible(file_path):\n advanced_data: Data | None = self._process_docling_in_subprocess(file_path)\n\n # --- UNNEST: expand each element in `doc` to its own Data row\n payload = getattr(advanced_data, \"data\", {}) or {}\n doc_rows = payload.get(\"doc\")\n if isinstance(doc_rows, list):\n rows: list[Data | None] = [\n Data(\n data={\n \"file_path\": file_path,\n **(item if isinstance(item, dict) else {\"value\": item}),\n },\n )\n for item in doc_rows\n ]\n return self.rollup_data(file_list, rows)\n\n # If not structured, keep as-is (e.g., markdown export or error dict)\n return self.rollup_data(file_list, [advanced_data])\n\n # Standard multi-file (or single non-advanced) path\n concurrency = 1 if not self.use_multithreading else max(1, self.concurrency_multithreading)\n file_paths = [str(f.path) for f in file_list]\n self.log(f\"Starting parallel processing of {len(file_paths)} files with concurrency: {concurrency}.\")\n my_data = parallel_load_data(\n file_paths,\n silent_errors=self.silent_errors,\n load_function=process_file_standard,\n max_concurrency=concurrency,\n )\n return self.rollup_data(file_list, my_data)\n\n # ------------------------------ Output helpers -----------------------------------\n\n def load_files_advanced(self) -> DataFrame:\n \"\"\"Load files using advanced Docling processing and export to an advanced format.\"\"\"\n self.markdown = False\n return self.load_files()\n\n def load_files_markdown(self) -> Message:\n \"\"\"Load files using advanced Docling processing and export to Markdown format.\"\"\"\n self.markdown = True\n result = self.load_files()\n return Message(text=str(result.text[0]))\n" | |
| def _process_docling_in_subprocess(self, file_path: str) -> Data | None: | |
| """Run Docling in a separate OS process and map the result to a Data object. | |
| We avoid multiprocessing pickling by launching `python -c "<script>"` and | |
| passing JSON config via stdin. The child prints a JSON result to stdout. | |
| """ | |
| if not file_path: | |
| return None | |
| args: dict[str, Any] = { | |
| "file_path": file_path, | |
| "markdown": bool(self.markdown), | |
| "image_mode": str(self.IMAGE_MODE), | |
| "md_image_placeholder": str(self.md_image_placeholder), | |
| "md_page_break_placeholder": str(self.md_page_break_placeholder), | |
| "pipeline": str(self.pipeline), | |
| "ocr_engine": str(self.ocr_engine) if getattr(self, "ocr_engine", "") else None, | |
| } | |
| # The child is a tiny, self-contained script to keep memory/state isolated. | |
| child_script = textwrap.dedent( | |
| r""" | |
| import json, sys | |
| def try_imports(): | |
| # Strategy 1: latest layout | |
| try: | |
| from docling.datamodel.base_models import ConversionStatus, InputFormat # type: ignore | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| from docling_core.types.doc import ImageRefMode # type: ignore | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "latest" | |
| except Exception: | |
| pass | |
| # Strategy 2: alternative layout | |
| try: | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| try: | |
| from docling_core.types import ConversionStatus, InputFormat # type: ignore | |
| except Exception: | |
| try: | |
| from docling.datamodel import ConversionStatus, InputFormat # type: ignore | |
| except Exception: | |
| class ConversionStatus: SUCCESS = "success" | |
| class InputFormat: | |
| PDF="pdf"; IMAGE="image" | |
| try: | |
| from docling_core.types.doc import ImageRefMode # type: ignore | |
| except Exception: | |
| class ImageRefMode: | |
| PLACEHOLDER="placeholder"; EMBEDDED="embedded" | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "alternative" | |
| except Exception: | |
| pass | |
| # Strategy 3: basic converter only | |
| try: | |
| from docling.document_converter import DocumentConverter # type: ignore | |
| class ConversionStatus: SUCCESS = "success" | |
| class InputFormat: | |
| PDF="pdf"; IMAGE="image" | |
| class ImageRefMode: | |
| PLACEHOLDER="placeholder"; EMBEDDED="embedded" | |
| return ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, "basic" | |
| except Exception as e: | |
| raise ImportError(f"Docling imports failed: {e}") from e | |
| def create_converter(strategy, input_format, DocumentConverter, pipeline, ocr_engine): | |
| if strategy == "latest" and pipeline == "standard": | |
| try: | |
| from docling.datamodel.pipeline_options import PdfPipelineOptions # type: ignore | |
| from docling.document_converter import PdfFormatOption # type: ignore | |
| pipe = PdfPipelineOptions() | |
| if ocr_engine: | |
| try: | |
| from docling.models.factories import get_ocr_factory # type: ignore | |
| pipe.do_ocr = True | |
| fac = get_ocr_factory(allow_external_plugins=False) | |
| pipe.ocr_options = fac.create_options(kind=ocr_engine) | |
| except Exception: | |
| pipe.do_ocr = False | |
| fmt = {} | |
| if hasattr(input_format, "PDF"): | |
| fmt[getattr(input_format, "PDF")] = PdfFormatOption(pipeline_options=pipe) | |
| if hasattr(input_format, "IMAGE"): | |
| fmt[getattr(input_format, "IMAGE")] = PdfFormatOption(pipeline_options=pipe) | |
| return DocumentConverter(format_options=fmt) | |
| except Exception: | |
| return DocumentConverter() | |
| return DocumentConverter() | |
| def export_markdown(document, ImageRefMode, image_mode, img_ph, pg_ph): | |
| try: | |
| mode = getattr(ImageRefMode, image_mode.upper(), image_mode) | |
| return document.export_to_markdown( | |
| image_mode=mode, | |
| image_placeholder=img_ph, | |
| page_break_placeholder=pg_ph, | |
| ) | |
| except Exception: | |
| try: | |
| return document.export_to_text() | |
| except Exception: | |
| return str(document) | |
| def to_rows(doc_dict): | |
| rows = [] | |
| for t in doc_dict.get("texts", []): | |
| prov = t.get("prov") or [] | |
| page_no = None | |
| if prov and isinstance(prov, list) and isinstance(prov[0], dict): | |
| page_no = prov[0].get("page_no") | |
| rows.append({ | |
| "page_no": page_no, | |
| "label": t.get("label"), | |
| "text": t.get("text"), | |
| "level": t.get("level"), | |
| }) | |
| return rows | |
| def main(): | |
| cfg = json.loads(sys.stdin.read()) | |
| file_path = cfg["file_path"] | |
| markdown = cfg["markdown"] | |
| image_mode = cfg["image_mode"] | |
| img_ph = cfg["md_image_placeholder"] | |
| pg_ph = cfg["md_page_break_placeholder"] | |
| pipeline = cfg["pipeline"] | |
| ocr_engine = cfg.get("ocr_engine") | |
| meta = {"file_path": file_path} | |
| try: | |
| ConversionStatus, InputFormat, DocumentConverter, ImageRefMode, strategy = try_imports() | |
| converter = create_converter(strategy, InputFormat, DocumentConverter, pipeline, ocr_engine) | |
| try: | |
| res = converter.convert(file_path) | |
| except Exception as e: | |
| print(json.dumps({"ok": False, "error": f"Docling conversion error: {e}", "meta": meta})) | |
| return | |
| ok = False | |
| if hasattr(res, "status"): | |
| try: | |
| ok = (res.status == ConversionStatus.SUCCESS) or (str(res.status).lower() == "success") | |
| except Exception: | |
| ok = (str(res.status).lower() == "success") | |
| if not ok and hasattr(res, "document"): | |
| ok = getattr(res, "document", None) is not None | |
| if not ok: | |
| print(json.dumps({"ok": False, "error": "Docling conversion failed", "meta": meta})) | |
| return | |
| doc = getattr(res, "document", None) | |
| if doc is None: | |
| print(json.dumps({"ok": False, "error": "Docling produced no document", "meta": meta})) | |
| return | |
| if markdown: | |
| text = export_markdown(doc, ImageRefMode, image_mode, img_ph, pg_ph) | |
| print(json.dumps({"ok": True, "mode": "markdown", "text": text, "meta": meta})) | |
| return | |
| # structured | |
| try: | |
| doc_dict = doc.export_to_dict() | |
| except Exception as e: | |
| print(json.dumps({"ok": False, "error": f"Docling export_to_dict failed: {e}", "meta": meta})) | |
| return | |
| rows = to_rows(doc_dict) | |
| print(json.dumps({"ok": True, "mode": "structured", "doc": rows, "meta": meta})) | |
| except Exception as e: | |
| print( | |
| json.dumps({ | |
| "ok": False, | |
| "error": f"Docling processing error: {e}", | |
| "meta": {"file_path": file_path}, | |
| }) | |
| ) | |
| if __name__ == "__main__": | |
| main() | |
| """ | |
| ) | |
| # Validate file_path to avoid command injection or unsafe input | |
| if not isinstance(args["file_path"], str) or any(c in args["file_path"] for c in [";", "|", "&", "$", "`"]): | |
| return Data(data={"error": "Unsafe file path detected.", "file_path": args["file_path"]}) | |
| try: | |
| proc = subprocess.run( # noqa: S603 | |
| [sys.executable, "-u", "-c", child_script], | |
| input=json.dumps(args).encode("utf-8"), | |
| capture_output=True, | |
| check=False, | |
| timeout=getattr(self, "subprocess_timeout", 120), | |
| ) | |
| except subprocess.TimeoutExpired: | |
| return Data(data={"error": "Docling subprocess timed out", "file_path": file_path}) | |
| if not proc.stdout: | |
| err_msg = proc.stderr.decode("utf-8", errors="replace") or "no output from child process" | |
| return Data(data={"error": f"Docling subprocess error: {err_msg}", "file_path": file_path}) | |
| try: | |
| result = json.loads(proc.stdout.decode("utf-8")) | |
| except Exception as e: # noqa: BLE001 | |
| err_msg = proc.stderr.decode("utf-8", errors="replace") | |
| return Data( | |
| data={"error": f"Invalid JSON from Docling subprocess: {e}. stderr={err_msg}", "file_path": file_path}, | |
| ) | |
| if not result.get("ok"): | |
| return Data(data={"error": result.get("error", "Unknown Docling error"), **result.get("meta", {})}) | |
| meta = result.get("meta", {}) | |
| if result.get("mode") == "markdown": | |
| exported_content = str(result.get("text", "")) | |
| return Data( | |
| text=exported_content, | |
| data={"exported_content": exported_content, "export_format": self.EXPORT_FORMAT, **meta}, | |
| ) | |
| rows = list(result.get("doc", [])) | |
| return Data(data={"doc": rows, "export_format": self.EXPORT_FORMAT, **meta}) |
Comment on lines
+18
to
+19
| display_name = "Write File" | ||
| description = "Write data to a local file in the selected format." |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
Rename to “Write File” — verify downstream references.
Confirm starter JSONs, docs, and UI strings no longer rely on “Save File,” and that the doc anchor still matches.
Run:
Optionally check docs site:
🏁 Script executed:
#!/bin/bash
# Find lingering references and anchors that may now be stale.
rg -nP -C2 '\bSave File\b|#save-file|/save-file' .Length of output: 15367
🌐 Web query:
Does the Langflow docs still use the anchor "components-processing#save-file" for this component, or has it changed to "write-file"?
💡 Result:
The docs still use the "save-file" anchor (components-processing#save-file). See the Processing components page (Save File section). [1]
Sources
[1] Langflow Documentation — Processing components (Save File).
Rename to "Write File" — update downstream references.
- Docs: rename the "Save File" section/anchor in docs/docs/Components/components-processing.mdx (## Save File at ~line 633) to "Write File" and update the component documentation URL in src/lfx/src/lfx/components/processing/save_file.py (currently points to ...#save-file) to ...#write-file.
- Starter project: change "display_name": "Save File" → "Write File" in src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json (around line 1237).
- Run a global search for "Save File" / "#save-file" / "/save-file" and update any remaining UI strings, links, or anchors to keep code, docs, and starter projects consistent.
🤖 Prompt for AI Agents
In src/lfx/src/lfx/components/processing/save_file.py around lines 18-19, the
component display_name should be changed from "Save File" to "Write File" and
its documentation URL fragment updated from "#save-file" to "#write-file"; also
update the docs file docs/docs/Components/components-processing.mdx (around the
Save File section at ~line 633) by renaming the "Save File" header/anchor to
"Write File" and ensure the anchor id matches, edit
src/backend/base/langflow/initial_setup/starter_projects/News Aggregator.json
(around line 1237) to change the "display_name" value to "Write File", and
finally perform a global repository search for occurrences of "Save File",
"#save-file", and "/save-file" to update UI strings, links, and anchors to the
new "Write File"/"#write-file" variants so code, docs, and starter projects
remain consistent.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request updates the output definitions for several file-related components to consistently include the
tool_mode=Trueparameter. This change ensures that all relevant outputs are flagged for tool mode operation, which likely improves integration or compatibility with downstream tooling.Output configuration updates:
tool_mode=Trueto theOutputfor "Files" in the_base_outputslist ofbase_file.py.tool_mode=Trueto the "Raw Content" output in theFileComponentclass infile.py.update_outputsmethod ofFileComponentto includetool_mode=True, covering structured, raw, markdown, advanced, and path outputs for various file types and modes.tool_mode=Trueto the "File Path" output in theSaveToFileComponentinsave_file.py.Summary by CodeRabbit