Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
""" Generic, configuration-driven REST API data source connector.
Connect any REST API as a RAGFlow data source without code changes .
All behaviour — URL , auth , pagination , field mapping — is controlled
via the ` ` RestAPIConnectorConfig ` ` schema exposed by the UI .
"""
from __future__ import annotations
import json
import logging
import re
import time
from datetime import datetime , timezone
from typing import Any , Dict , Generator , Iterable , List , Mapping , Optional
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
from urllib . parse import parse_qs , urljoin , urlparse , urlunparse
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
import ipaddress
import socket
import requests
from pydantic import BaseModel , ConfigDict , Field , HttpUrl , ValidationError , field_validator
logger = logging . getLogger ( __name__ )
from api . utils . common import hash128
from common . data_source . config import INDEX_BATCH_SIZE , DocumentSource
from common . data_source . exceptions import (
ConnectorMissingCredentialError ,
ConnectorValidationError ,
)
from common . data_source . interfaces import (
LoadConnector ,
PollConnector ,
SecondsSinceUnixEpoch ,
)
from common . data_source . models import Document
from common . data_source . utils import rl_requests , retry_builder
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
from common . ssrf_guard import assert_url_is_safe , pin_dns
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
try :
from jsonpath import jsonpath as _jsonpath # type: ignore[import]
except Exception : # pragma: no cover
_jsonpath = None
2026-07-03 12:53:39 +08:00
_FIELD_SEGMENT_RE = re . compile ( r " ^(?P<key>[^ \ [ \ ]]+)( \ [(?P<index> \ d+| \ *) \ ])?$ " )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
_DEFAULT_MAX_PAGES = 1000
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
_REDIRECT_STATUSES = frozenset ( { 301 , 302 , 303 , 307 , 308 } )
_MAX_REDIRECTS = 5
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
class AuthType :
NONE = " none "
API_KEY_HEADER = " api_key_header "
BEARER = " bearer "
BASIC = " basic "
class PaginationType :
NONE = " none "
PAGE = " page "
OFFSET = " offset "
CURSOR = " cursor "
def _text_to_dict ( v : Any ) - > Dict [ str , str ] :
""" Parse a dict, JSON string, or ``key=value`` text (one per line) into a dict.
This is module - level because Pydantic ` ` @field_validator ` ` classmethods
on ` ` RestAPIConnectorConfig ` ` need to call it before any instance exists .
"""
if v is None or v == " " :
return { }
if isinstance ( v , dict ) :
return { str ( k ) : str ( vv ) for k , vv in v . items ( ) }
if isinstance ( v , str ) :
try :
parsed = json . loads ( v )
if isinstance ( parsed , dict ) :
return { str ( k ) : str ( vv ) for k , vv in parsed . items ( ) }
except Exception :
pass
result : Dict [ str , str ] = { }
for line in v . splitlines ( ) :
line = line . strip ( )
if not line or line . startswith ( " # " ) :
continue
if " = " in line :
k , _ , val = line . partition ( " = " )
result [ k . strip ( ) ] = val . strip ( )
return result
return { }
class RestAPIConnectorConfig ( BaseModel ) :
""" Validated schema for the REST API connector configuration. """
model_config = ConfigDict ( extra = " ignore " )
url : HttpUrl
method : str = " GET "
headers : Dict [ str , str ] = Field ( default_factory = dict )
query_params : Dict [ str , str ] = Field ( default_factory = dict )
auth_type : str = AuthType . NONE
auth_config : Dict [ str , Any ] = Field ( default_factory = dict )
items_path : Optional [ str ] = None
id_field : Optional [ str ] = None
content_fields : List [ str ] = Field ( default_factory = list )
metadata_fields : List [ str ] = Field ( default_factory = list )
pagination_type : str = PaginationType . NONE
pagination_config : Dict [ str , Any ] = Field ( default_factory = dict )
poll_timestamp_field : Optional [ str ] = None
request_body : Optional [ Dict [ str , Any ] ] = None
field_type_hints : Dict [ str , str ] = Field ( default_factory = dict )
field_default_values : Dict [ str , Any ] = Field ( default_factory = dict )
content_template : Optional [ str ] = None
batch_size : int = INDEX_BATCH_SIZE
max_pages : int = _DEFAULT_MAX_PAGES
request_delay : float = 0.5
@field_validator ( " headers " , mode = " before " )
@classmethod
def _coerce_headers ( cls , v : Any ) - > Dict [ str , str ] :
return _text_to_dict ( v )
@field_validator ( " query_params " , mode = " before " )
@classmethod
def _coerce_query_params ( cls , v : Any ) - > Dict [ str , str ] :
return _text_to_dict ( v )
@field_validator ( " content_fields " , " metadata_fields " , mode = " before " )
@classmethod
def _coerce_field_list ( cls , v : Any ) - > List [ str ] :
if v is None or v == " " :
return [ ]
if isinstance ( v , str ) :
return [ p . strip ( ) for p in v . split ( " , " ) if p . strip ( ) ]
if isinstance ( v , list ) :
return [ str ( p ) . strip ( ) for p in v if str ( p ) . strip ( ) ]
return [ ]
def normalized_method ( self ) - > str :
m = ( self . method or " GET " ) . upper ( )
if m not in { " GET " , " POST " } :
raise ConnectorValidationError ( f " Unsupported HTTP method ' { m } ' . " )
return m
def normalized_auth_type ( self ) - > str :
if self . auth_type not in { AuthType . NONE , AuthType . API_KEY_HEADER , AuthType . BEARER , AuthType . BASIC } :
raise ConnectorValidationError ( f " Unsupported auth_type ' { self . auth_type } ' . " )
return self . auth_type
def normalized_pagination_type ( self ) - > str :
if self . pagination_type not in { PaginationType . NONE , PaginationType . PAGE , PaginationType . OFFSET , PaginationType . CURSOR } :
raise ConnectorValidationError ( f " Unsupported pagination_type ' { self . pagination_type } ' . " )
return self . pagination_type
def ensure_required_fields ( self ) - > None :
if not self . content_fields :
raise ConnectorValidationError ( " At least one content field must be configured (content_fields). " )
class RestAPIConnector ( LoadConnector , PollConnector ) :
""" Configuration-driven REST API connector.
Implements ` ` LoadConnector ` ` and ` ` PollConnector ` ` to fetch documents
from any REST API using user - provided configuration ( URL , auth ,
pagination , field mapping ) .
"""
@staticmethod
def _validate_url_for_ssrf ( url : str ) - > None :
""" Validate that the URL does not point to localhost or private/internal networks.
Raises :
ConnectorValidationError : If the URL is considered unsafe .
"""
parsed = urlparse ( str ( url ) )
if parsed . scheme not in ( " http " , " https " ) :
msg = f " Unsupported URL scheme for REST API connector: { parsed . scheme !r} . Only http/https are allowed. "
logger . warning ( msg )
raise ConnectorValidationError ( msg )
hostname = parsed . hostname
if not hostname :
msg = " REST API connector URL must include a hostname. "
logger . warning ( msg )
raise ConnectorValidationError ( msg )
# Quick checks for obvious localhost-style hostnames.
lower_host = hostname . lower ( )
if lower_host in ( " localhost " , ) :
msg = f " REST API connector URL hostname { hostname !r} is not allowed (localhost is blocked). "
logger . warning ( msg )
raise ConnectorValidationError ( msg )
try :
addrinfo_list = socket . getaddrinfo ( hostname , None )
except OSError as exc :
# If resolution fails, log and let higher-level validation (if any) decide.
# We do not treat this as an SSRF condition by itself.
logger . info ( " DNS resolution failed for REST API connector URL %r : %s " , url , exc )
return
for family , _ , _ , _ , sockaddr in addrinfo_list :
ip_str = sockaddr [ 0 ]
try :
ip_obj = ipaddress . ip_address ( ip_str )
except ValueError :
# Not an IP address we understand; skip.
logger . debug ( " Skipping non-IP address resolved from %r : %r " , hostname , ip_str )
continue
2026-07-03 12:53:39 +08:00
if ip_obj . is_loopback or ip_obj . is_private or ip_obj . is_link_local or ip_obj . is_reserved or ip_obj . is_multicast :
msg = f " REST API connector URL { url !r} resolves to disallowed address { ip_str } (localhost, private, link-local, reserved, or multicast addresses are blocked). "
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
logger . warning ( msg )
raise ConnectorValidationError ( msg )
logger . debug ( " REST API connector URL %r passed SSRF safety validation. " , url )
def __init__ (
self ,
url : str ,
method : str = " GET " ,
headers : Optional [ Dict [ str , str ] ] = None ,
query_params : Optional [ Dict [ str , str ] ] = None ,
auth_type : str = AuthType . NONE ,
auth_config : Optional [ Dict [ str , Any ] ] = None ,
items_path : Optional [ str ] = None ,
id_field : Optional [ str ] = None ,
content_fields : Optional [ List [ str ] ] = None ,
metadata_fields : Optional [ List [ str ] ] = None ,
pagination_type : str = PaginationType . NONE ,
pagination_config : Optional [ Dict [ str , Any ] ] = None ,
poll_timestamp_field : Optional [ str ] = None ,
batch_size : int = INDEX_BATCH_SIZE ,
max_pages : int = _DEFAULT_MAX_PAGES ,
request_delay : float = 0.5 ,
request_body : Optional [ Dict [ str , Any ] ] = None ,
field_type_hints : Optional [ Dict [ str , str ] ] = None ,
field_default_values : Optional [ Dict [ str , Any ] ] = None ,
content_template : Optional [ str ] = None ,
) - > None :
# Validate URL against SSRF-style targets (localhost, private/internal ranges, etc.)
self . _validate_url_for_ssrf ( url )
parsed = urlparse ( str ( url ) )
self . _base_url = urlunparse ( ( parsed . scheme , parsed . netloc , parsed . path , " " , " " , " " ) )
self . _url_params : Dict [ str , str ] = { }
if parsed . query :
for k , v_list in parse_qs ( parsed . query , keep_blank_values = True ) . items ( ) :
self . _url_params [ k ] = v_list [ - 1 ]
2026-07-03 12:53:39 +08:00
self . _explicit_query_params : Dict [ str , str ] = _text_to_dict ( query_params ) if isinstance ( query_params , str ) else ( query_params or { } )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
self . url = self . _base_url
self . method = ( method or " GET " ) . upper ( )
2026-07-03 12:53:39 +08:00
self . _base_headers : Dict [ str , str ] = _text_to_dict ( headers ) if isinstance ( headers , str ) else ( headers or { } )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
self . auth_type = auth_type or AuthType . NONE
self . auth_config : Dict [ str , Any ] = auth_config or { }
self . items_path = items_path
self . id_field = id_field
self . content_fields : List [ str ] = content_fields or [ ]
self . metadata_fields : List [ str ] = metadata_fields or [ ]
self . pagination_type = pagination_type or PaginationType . NONE
self . pagination_config : Dict [ str , Any ] = pagination_config or { }
2026-07-03 12:53:39 +08:00
self . _static_request_body : Dict [ str , Any ] = request_body if request_body is not None else self . pagination_config . get ( " request_body " ) or { }
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
self . poll_timestamp_field = poll_timestamp_field
self . batch_size = batch_size
self . max_pages = max_pages
self . request_delay = max ( request_delay , 0.0 )
self . field_type_hints : Dict [ str , str ] = field_type_hints or { }
self . field_default_values : Dict [ str , Any ] = field_default_values or { }
self . content_template = content_template
self . _credentials : Dict [ str , Any ] = { }
self . _auth_headers : Dict [ str , str ] = { }
self . _basic_auth : Optional [ requests . auth . HTTPBasicAuth ] = None
# -- Credentials --------------------------------------------------------
def load_credentials ( self , credentials : Dict [ str , Any ] ) - > Dict [ str , Any ] | None :
""" Apply authentication credentials (no network call).
Use ` ` validate_config ( ) ` ` to perform a live connectivity check .
"""
self . _credentials = credentials or { }
self . _build_auth ( )
return None
def _build_auth ( self ) - > None :
""" Derive auth headers / basic-auth object from credentials. """
self . _auth_headers = { }
self . _basic_auth = None
if self . auth_type == AuthType . NONE :
logging . info ( " REST API auth_type=none, no authentication configured. " )
return
if self . auth_type == AuthType . API_KEY_HEADER :
header_name = self . auth_config . get ( " header_name " )
2026-07-03 12:53:39 +08:00
api_key = self . _credentials . get ( " api_key " ) or self . auth_config . get ( " api_key_value " ) or self . auth_config . get ( " api_key " )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
if not header_name or not api_key :
logging . warning (
2026-07-03 12:53:39 +08:00
" REST API auth setup failed: header_name= %s , api_key present= %s , credentials keys= %s , auth_config keys= %s " ,
header_name ,
bool ( api_key ) ,
list ( self . _credentials . keys ( ) ) ,
list ( self . auth_config . keys ( ) ) ,
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
)
2026-07-03 12:53:39 +08:00
raise ConnectorMissingCredentialError ( " REST API (api_key_header) requires ' header_name ' in auth_config and ' api_key ' in credentials " )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
self . _auth_headers [ header_name ] = str ( api_key )
logging . info ( " REST API auth configured: header ' %s ' set. " , header_name )
return
if self . auth_type == AuthType . BEARER :
token = self . _credentials . get ( " token " ) or self . auth_config . get ( " token " )
if not token :
raise ConnectorMissingCredentialError ( " REST API (bearer) requires ' token ' in credentials " )
self . _auth_headers [ " Authorization " ] = f " Bearer { token } "
logging . info ( " REST API auth configured: Bearer token set. " )
return
if self . auth_type == AuthType . BASIC :
username = self . _credentials . get ( " username " ) or self . auth_config . get ( " username " )
password = self . _credentials . get ( " password " ) or self . auth_config . get ( " password " )
if not username or password is None :
raise ConnectorMissingCredentialError ( " REST API (basic) requires ' username ' and ' password ' " )
self . _basic_auth = requests . auth . HTTPBasicAuth ( str ( username ) , str ( password ) )
logging . info ( " REST API auth configured: Basic auth for user ' %s ' . " , username )
return
raise ConnectorValidationError ( f " Unsupported auth_type: { self . auth_type } " )
# -- Config validation (test connection) --------------------------------
@classmethod
def parse_storage_config ( cls , raw : Dict [ str , Any ] ) - > RestAPIConnectorConfig :
""" Parse connector config as stored on the connector row (no network I/O).
` ` credentials ` ` live under ` ` raw ` ` but are excluded from the schema and
must be applied via ` ` load_credentials ` ` separately .
"""
body = { k : v for k , v in raw . items ( ) if k != " credentials " }
try :
cfg = RestAPIConnectorConfig ( * * body )
except ValidationError as exc :
raise ConnectorValidationError ( f " Invalid REST API config: { exc } " ) from exc
cfg . normalized_method ( )
cfg . normalized_auth_type ( )
cfg . normalized_pagination_type ( )
cfg . ensure_required_fields ( )
return cfg
@classmethod
def from_parsed_config (
cls ,
cfg : RestAPIConnectorConfig ,
* ,
max_pages : Optional [ int ] = None ,
) - > RestAPIConnector :
""" Build a connector from validated config (``__init__`` runs SSRF validation). """
return cls (
url = str ( cfg . url ) ,
method = cfg . normalized_method ( ) ,
headers = cfg . headers ,
query_params = cfg . query_params ,
auth_type = cfg . normalized_auth_type ( ) ,
auth_config = cfg . auth_config ,
items_path = cfg . items_path ,
id_field = cfg . id_field ,
content_fields = cfg . content_fields ,
metadata_fields = cfg . metadata_fields ,
pagination_type = cfg . normalized_pagination_type ( ) ,
pagination_config = cfg . pagination_config ,
poll_timestamp_field = cfg . poll_timestamp_field ,
batch_size = cfg . batch_size ,
max_pages = max_pages if max_pages is not None else cfg . max_pages ,
request_delay = cfg . request_delay ,
request_body = cfg . request_body ,
field_type_hints = cfg . field_type_hints ,
field_default_values = cfg . field_default_values ,
content_template = cfg . content_template ,
)
@classmethod
def validate_config (
cls ,
config : Dict [ str , Any ] ,
credentials : Optional [ Dict [ str , Any ] ] = None ,
) - > RestAPIConnectorConfig :
""" Validate config schema and optionally perform a live API call.
Args :
config : Raw config dict from the UI / database .
credentials : Optional credentials dict ; when provided a live
connectivity check is performed .
Returns :
The validated ` ` RestAPIConnectorConfig ` ` instance .
Raises :
ConnectorValidationError : On schema or connectivity failure .
"""
cfg = cls . parse_storage_config ( config )
validation_cap = min ( cfg . max_pages , 10 )
connector = cls . from_parsed_config ( cfg , max_pages = validation_cap )
if credentials is None and cfg . auth_type != AuthType . NONE :
return cfg
if credentials is not None :
connector . load_credentials ( credentials )
else :
connector . _credentials = { }
connector . _build_auth ( )
try :
logging . info ( " Validating REST API connector by fetching first page " )
_ = next ( connector . _page_iter_for_validation ( ) )
except StopIteration :
pass
return cfg
# -- LoadConnector / PollConnector interface -----------------------------
def load_from_state ( self ) - > Generator [ List [ Document ] , None , None ] :
""" Full fetch with pagination. """
return self . _yield_documents ( time_window = None )
2026-07-03 12:53:39 +08:00
def poll_source ( self , start : SecondsSinceUnixEpoch , end : SecondsSinceUnixEpoch ) - > Generator [ List [ Document ] , None , None ] :
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
""" Incremental fetch; filters by ``poll_timestamp_field`` if configured. """
if not self . poll_timestamp_field :
2026-07-03 12:53:39 +08:00
logging . warning ( " poll_source called without poll_timestamp_field; falling back to full fetch with in-memory filtering. " )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
return self . _yield_documents (
time_window = (
datetime . fromtimestamp ( start , tz = timezone . utc ) ,
datetime . fromtimestamp ( end , tz = timezone . utc ) ,
)
)
# -- Document generation ------------------------------------------------
def _yield_documents (
self ,
time_window : tuple [ datetime , datetime ] | None ,
) - > Generator [ List [ Document ] , None , None ] :
batch : List [ Document ] = [ ]
for item in self . _iter_items ( ) :
try :
doc = self . _item_to_document ( item )
except Exception as exc :
logging . warning ( " Failed to convert REST API item to Document: %s " , exc )
continue
if time_window is not None and not self . _doc_in_time_window ( doc , * time_window ) :
continue
batch . append ( doc )
if len ( batch ) > = self . batch_size :
yield batch
batch = [ ]
if batch :
yield batch
# -- Pagination & page fetching -----------------------------------------
def _iter_items ( self ) - > Iterable [ Mapping [ str , Any ] ] :
""" Iterate over raw items across all pages. """
page_count = 0
page = int ( self . pagination_config . get ( " start_page " , 1 ) )
per_page = self . _resolve_page_size ( )
offset = int ( self . pagination_config . get ( " start_offset " , 0 ) )
limit = int ( self . pagination_config . get ( " limit " , per_page ) )
if limit < = 0 :
limit = per_page
cursor : Optional [ str ] = self . pagination_config . get ( " initial_cursor " )
while True :
if page_count > = self . max_pages :
logging . warning ( " REST API connector reached max_pages= %d , stopping. " , self . max_pages )
break
params : Dict [ str , Any ] = { }
if self . pagination_type == PaginationType . PAGE :
self . _apply_page_pagination ( params , page , per_page )
elif self . pagination_type == PaginationType . OFFSET :
self . _apply_offset_pagination ( params , offset , limit )
elif self . pagination_type == PaginationType . CURSOR and cursor is not None :
self . _apply_cursor_pagination ( params , cursor )
if page_count > 0 and self . request_delay > 0 :
time . sleep ( self . request_delay )
try :
response_json = self . _fetch_page ( params )
except ( ConnectorValidationError , ConnectorMissingCredentialError ) :
raise
except Exception as exc :
raise ConnectorValidationError ( f " REST API page fetch failed: { exc } " ) from exc
items = self . _extract_items ( response_json )
if not items :
break
for item in items :
if isinstance ( item , Mapping ) :
yield item
page_count + = 1
if self . pagination_type == PaginationType . NONE :
break
elif self . pagination_type == PaginationType . PAGE :
if len ( items ) < per_page :
break
page + = 1
elif self . pagination_type == PaginationType . OFFSET :
if len ( items ) < limit :
break
offset + = limit
elif self . pagination_type == PaginationType . CURSOR :
next_cursor = self . _extract_next_cursor ( response_json )
if not next_cursor :
break
cursor = next_cursor
def _page_iter_for_validation ( self ) - > Iterable [ Mapping [ str , Any ] ] :
""" Single-page iterator used for connectivity checks. """
params : Dict [ str , Any ] = { }
if self . pagination_type == PaginationType . PAGE :
page = int ( self . pagination_config . get ( " start_page " , 1 ) )
per_page = self . _resolve_page_size ( )
self . _apply_page_pagination ( params , page , per_page )
elif self . pagination_type == PaginationType . OFFSET :
per_page = self . _resolve_page_size ( )
offset = int ( self . pagination_config . get ( " start_offset " , 0 ) )
limit = int ( self . pagination_config . get ( " limit " , per_page ) )
if limit < = 0 :
limit = per_page
self . _apply_offset_pagination ( params , offset , limit )
elif self . pagination_type == PaginationType . CURSOR :
cursor = self . pagination_config . get ( " initial_cursor " )
if cursor is not None :
self . _apply_cursor_pagination ( params , cursor )
response_json = self . _fetch_page ( params = params )
for item in self . _extract_items ( response_json ) :
yield item
@retry_builder (
2026-07-03 12:53:39 +08:00
tries = 5 ,
delay = 1 ,
max_delay = 30 ,
backoff = 2 ,
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
exceptions = ( requests . ConnectionError , requests . Timeout , requests . HTTPError ) ,
)
def _fetch_page ( self , params : Dict [ str , Any ] ) - > Any :
""" Fetch a single page with retry and exponential backoff. """
headers = { * * self . _base_headers , * * self . _auth_headers }
merged : Dict [ str , Any ] = { * * self . _url_params }
merged . update ( self . _explicit_query_params )
merged . update ( params )
url , query_params = self . _build_url_with_templates ( merged )
sensitive = { " authorization " , " apikey " , " api-key " , " x-api-key " }
logging . debug (
" REST API request: %s %s | params= %s | headers= %s " ,
2026-07-03 12:53:39 +08:00
self . method ,
url ,
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
{ k : ( " *** " if k . lower ( ) in sensitive else v ) for k , v in query_params . items ( ) } ,
{ k : ( " *** " if k . lower ( ) in sensitive else v ) for k , v in headers . items ( ) } ,
)
if self . method == " GET " :
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
resp = self . _safe_request (
" GET " ,
url ,
headers = headers ,
params = query_params ,
)
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
elif self . method == " POST " :
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
resp = self . _safe_request (
" POST " ,
url ,
headers = headers ,
params = query_params ,
json_body = self . _static_request_body or { } ,
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
)
else :
raise ConnectorValidationError ( f " Unsupported HTTP method: { self . method } " )
try :
resp . raise_for_status ( )
except requests . HTTPError as exc :
status = exc . response . status_code if exc . response is not None else None
if status in ( 401 , 403 ) :
sensitive = { " authorization " , " apikey " , " api-key " , " x-api-key " }
logging . warning (
2026-07-03 12:53:39 +08:00
" REST API %d for %s %s | auth_type= %s | request header keys= %s | auth_header keys= %s " ,
status ,
self . method ,
resp . url ,
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
self . auth_type ,
[ k for k in headers ] ,
[ k for k in self . _auth_headers ] ,
)
2026-07-03 12:53:39 +08:00
raise ConnectorMissingCredentialError ( f " REST API authentication failed with status { status } " ) from exc
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
if status is not None and 400 < = status < 500 and status != 429 :
logging . warning (
" REST API client error %d for %s %s ; not retrying. " ,
status ,
self . method ,
resp . url ,
)
2026-07-03 12:53:39 +08:00
raise ConnectorValidationError ( f " REST API request failed with non-retriable client error status { status } " ) from exc
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
raise
try :
return resp . json ( )
except ValueError as exc :
raise ConnectorValidationError ( " REST API response is not valid JSON " ) from exc
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
# Headers that carry auth state. Stripped on cross-origin redirects to
# prevent credential exfiltration to a third-party host. (Coderabbit MAJOR #3486038792)
2026-07-03 12:53:39 +08:00
_AUTH_SENSITIVE_HEADER_KEYS = frozenset (
{
" authorization " ,
" proxy-authorization " ,
" apikey " ,
" api-key " ,
" x-api-key " ,
" x-auth-token " ,
}
)
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
def _safe_request (
self ,
method : str ,
url : str ,
* ,
headers : Dict [ str , str ] ,
params : Dict [ str , Any ] ,
body : Any = None ,
json_body : Any = None ,
) - > requests . Response :
""" Issue an HTTP request with per-hop SSRF validation and DNS pinning. """
current_url = url
current_method = method . upper ( )
current_body = body
current_json = json_body
current_params = dict ( params )
# Local auth handle: cleared when crossing origins, even though
# ``self._basic_auth`` may still hold the original credentials.
current_auth = self . _basic_auth
previous_netloc = urlparse ( current_url ) . netloc
for _ in range ( _MAX_REDIRECTS + 1 ) :
# Normalize SSRF validation failures to the connector's documented
# ConnectorValidationError so they don't leak ValueError out of
# _page_iter_for_validation(). (Coderabbit MAJOR #3486038789)
try :
hostname , pin_ip = assert_url_is_safe ( current_url )
except ValueError as exc :
2026-07-03 12:53:39 +08:00
raise ConnectorValidationError ( f " Unsafe REST API URL: { exc } " ) from exc
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
with pin_dns ( hostname , pin_ip ) :
if current_method == " GET " :
resp = rl_requests . get (
current_url ,
headers = headers ,
params = current_params ,
auth = current_auth ,
timeout = 60 ,
allow_redirects = False ,
)
elif current_method == " POST " :
resp = rl_requests . post (
current_url ,
headers = headers ,
params = current_params ,
json = current_json ,
auth = current_auth ,
timeout = 60 ,
allow_redirects = False ,
)
else :
resp = rl_requests . request (
current_method ,
current_url ,
headers = headers ,
params = current_params ,
auth = current_auth ,
timeout = 60 ,
data = current_body ,
json = current_json ,
allow_redirects = False ,
)
if resp . status_code not in _REDIRECT_STATUSES :
return resp
location = resp . headers . get ( " Location " )
if not location :
return resp
current_url = urljoin ( current_url , location )
next_netloc = urlparse ( current_url ) . netloc
# Coderabbit MAJOR #3486038792: strip credentials when the redirect
# crosses to a different origin so a public→private redirect chain
# cannot exfiltrate Bearer/Basic/API-key headers.
if next_netloc and next_netloc != previous_netloc :
2026-07-03 12:53:39 +08:00
headers = { k : v for k , v in headers . items ( ) if k . lower ( ) not in self . _AUTH_SENSITIVE_HEADER_KEYS }
Harden closed-advisory fixes (#16409)
## Summary
- harden reopened advisory fixes across REST connector, invoke, document
downloads, and markdown rendering
- add targeted regression coverage for redirect-safe SSRF handling,
invoke SSRF checks, document access control, and markdown sanitization
- verify each referenced GHSA against the original GitHub advisory text
and align the closed-advisory plan with the implemented remediation
## What changed
- add tenant access checks to document download endpoints to avoid
cross-tenant document disclosure
- add per-hop SSRF validation, DNS pinning, redirect handling, and
redirect limits to the REST API connector
- ensure invoke requests validate and pin the resolved host and never
follow redirects implicitly
- keep the generic rate-limited request path wrapped, not just GET and
POST helpers
- sanitize markdown HTML before rendering in the highlight markdown
component
## Validation
- `cd web && npm test -- --runInBand
src/components/highlight-markdown/__tests__/index.test.tsx`
- `.venv/bin/python -m pytest -q
test/unit_test/data_source/test_rest_api_connector.py`
- targeted `test/testcases/test_web_api/...` unit additions were
reviewed, but the suite cannot be executed end-to-end in this
environment because parent `test/testcases/conftest.py` requires a local
service on `127.0.0.1:9380`
## Notes
- all GHSA entries referenced by the plan were checked against the
original GitHub advisory text, not sampled
- the closed-advisory plan document was updated locally during review,
but is intentionally not included in this PR
2026-06-28 11:17:54 +08:00
current_auth = None
previous_netloc = next_netloc
if resp . status_code in ( 301 , 302 , 303 ) :
current_method = " GET "
current_body = None
current_json = None
# Clear carried params — only the new Location URL's query
# string should apply for the downgraded GET.
current_params = { }
raise ConnectorValidationError ( f " Exceeded { _MAX_REDIRECTS } redirects fetching { url !r} " )
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
def _build_url_with_templates ( self , params : Dict [ str , Any ] ) - > tuple [ str , Dict [ str , Any ] ] :
""" Substitute `` {key} `` placeholders in the URL; return remaining query params. """
url = self . url
query_params = dict ( params )
used_keys : List [ str ] = [ ]
for key , value in list ( query_params . items ( ) ) :
placeholder = " { " + key + " } "
if placeholder in url :
url = url . replace ( placeholder , str ( value ) )
used_keys . append ( key )
for key in used_keys :
query_params . pop ( key , None )
return url , query_params
# -- Pagination helpers -------------------------------------------------
def _resolve_page_size ( self ) - > int :
""" Determine per-page size from config, query params, or batch_size fallback.
Priority : explicit ` ` page_size ` ` in pagination_config > value already
present in user query params for the same param name > batch_size .
"""
explicit = self . pagination_config . get ( " page_size " )
if explicit is not None :
val = int ( explicit )
if val > 0 :
return val
size_param = self . pagination_config . get ( " page_size_param " ) or self . pagination_config . get ( " limit_param " )
if size_param :
for source in ( self . _explicit_query_params , self . _url_params ) :
if size_param in source :
try :
val = int ( source [ size_param ] )
if val > 0 :
return val
except ( ValueError , TypeError ) :
pass
return self . batch_size
def _apply_page_pagination ( self , params : Dict [ str , Any ] , page : int , per_page : int ) - > None :
params [ self . pagination_config . get ( " page_param " , " page " ) ] = page
size_param = self . pagination_config . get ( " page_size_param " )
if size_param :
params [ size_param ] = per_page
def _apply_offset_pagination ( self , params : Dict [ str , Any ] , offset : int , limit : int ) - > None :
params [ self . pagination_config . get ( " offset_param " , " offset " ) ] = offset
limit_param = self . pagination_config . get ( " limit_param " )
if limit_param :
params [ limit_param ] = limit
def _apply_cursor_pagination ( self , params : Dict [ str , Any ] , cursor : str ) - > None :
params [ self . pagination_config . get ( " cursor_param " , " cursor " ) ] = cursor
# -- JSON extraction ----------------------------------------------------
def _extract_items ( self , response_json : Any ) - > List [ Mapping [ str , Any ] ] :
""" Extract the items array from a JSON response. """
if self . items_path and _jsonpath is not None :
try :
matches = _jsonpath ( response_json , self . items_path )
except Exception as exc :
2026-07-03 12:53:39 +08:00
raise ConnectorValidationError ( f " Failed to apply items JSONPath ' { self . items_path } ' : { exc } " ) from exc
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
if not matches :
return [ ]
if len ( matches ) == 1 and isinstance ( matches [ 0 ] , list ) :
items = matches [ 0 ]
else :
items = matches
elif isinstance ( response_json , list ) :
items = response_json
elif isinstance ( response_json , dict ) :
items = [ ]
for key in ( " items " , " results " , " data " , " records " ) :
if key in response_json and isinstance ( response_json [ key ] , list ) :
items = response_json [ key ]
break
else :
for value in response_json . values ( ) :
if isinstance ( value , list ) :
items = value
break
else :
items = [ ]
return [ it for it in items if isinstance ( it , Mapping ) ]
def _extract_next_cursor ( self , response_json : Any ) - > Optional [ str ] :
""" Extract cursor value for cursor-based pagination. """
cursor_path = self . pagination_config . get ( " next_cursor_path " )
if not cursor_path :
field = self . pagination_config . get ( " next_cursor_field " )
if field and isinstance ( response_json , Mapping ) :
value = response_json . get ( field )
return str ( value ) if value is not None else None
return None
if _jsonpath is None :
return None
try :
matches = _jsonpath ( response_json , cursor_path )
except Exception :
return None
if not matches :
return None
return str ( matches [ 0 ] ) if matches [ 0 ] is not None else None
# -- Item → Document mapping --------------------------------------------
def _item_to_document ( self , item : Mapping [ str , Any ] ) - > Document :
""" Map a single API item to a ``Document``. """
raw_id = self . _get_typed_field_value ( self . id_field , item ) if self . id_field else None
if raw_id is None :
raw_id = hash128 ( f " rest_api_item: { repr ( item ) } " )
doc_id = hash128 ( f " rest_api: { raw_id } " )
if self . content_template :
content_text = self . _render_content_template ( item )
else :
parts = [ ]
for field in self . content_fields :
val = self . _get_typed_field_value ( field , item )
if val is not None :
text = self . _strip_html ( self . _coerce_to_text ( val ) )
if text :
parts . append ( text )
content_text = " \n " . join ( parts )
blob = content_text . encode ( " utf-8 " )
metadata : Dict [ str , Any ] = { }
for field in self . metadata_fields :
value = self . _get_typed_field_value ( field , item )
if value is not None :
metadata [ field ] = self . _serialize_metadata_value ( value )
doc_updated_at = self . _extract_timestamp ( item ) or datetime . now ( timezone . utc )
sem = str ( self . _extract_field ( item , self . content_fields [ 0 ] ) if self . content_fields else raw_id )
sem = self . _strip_html ( sem ) . replace ( " \n " , " " ) . replace ( " \r " , " " ) . strip ( ) [ : 100 ] or str ( doc_id )
return Document (
id = doc_id ,
source = DocumentSource . REST_API ,
semantic_identifier = sem ,
extension = " .txt " ,
blob = blob ,
doc_updated_at = doc_updated_at ,
size_bytes = len ( blob ) ,
metadata = metadata or None ,
)
# -- Field extraction ---------------------------------------------------
def _extract_field ( self , item : Mapping [ str , Any ] , path : str ) - > Any :
""" Extract a value using dot-notation with optional array indexing.
Examples : ` ` country . name ` ` , ` ` tags [ 0 ] . label ` ` , ` ` tags [ * ] . label ` `
"""
values = self . _extract_field_values ( item , path )
if not values :
return None
return values [ 0 ] if len ( values ) == 1 else values
def _extract_field_values ( self , item : Mapping [ str , Any ] , path : str ) - > List [ Any ] :
""" Return all raw values for a dot-notation field path with wildcards. """
if not path :
return [ ]
current_values : List [ Any ] = [ item ]
for segment in path . split ( " . " ) :
if not segment :
return [ ]
match = _FIELD_SEGMENT_RE . match ( segment )
key = segment
index : Optional [ str ] = None
if match :
key = match . group ( " key " )
index = match . group ( " index " )
next_values : List [ Any ] = [ ]
for value in current_values :
if not isinstance ( value , Mapping ) :
continue
child = value . get ( key )
if child is None :
continue
if index is None :
next_values . append ( child )
elif not isinstance ( child , list ) :
continue
elif index == " * " :
next_values . extend ( child )
else :
try :
idx = int ( index )
except ValueError :
continue
if 0 < = idx < len ( child ) :
next_values . append ( child [ idx ] )
current_values = next_values
if not current_values :
break
return current_values
def _get_typed_field_value ( self , path : str , item : Mapping [ str , Any ] ) - > Any :
""" Extract a field value, applying type hints, defaults, and array joining. """
values = self . _extract_field_values ( item , path )
if not values :
return self . field_default_values . get ( path )
hint = self . field_type_hints . get ( path )
def _convert ( v : Any ) - > Any :
if hint == " string " :
return " " if v is None else str ( v )
if hint == " number " :
if v is None :
return None
try :
num = float ( v )
return int ( num ) if num . is_integer ( ) else num
except Exception :
return None
if hint == " date " :
if isinstance ( v , datetime ) :
return v . isoformat ( )
dt = self . _parse_datetime ( v )
if dt is not None :
return dt . isoformat ( )
return str ( v ) if v is not None else None
return v
converted = [ _convert ( v ) for v in values ]
non_null = [ v for v in converted if v is not None ]
if not non_null :
return None
if len ( non_null ) == 1 :
return non_null [ 0 ]
return " , " . join ( self . _coerce_to_text ( v ) for v in non_null )
# -- Timestamp parsing --------------------------------------------------
def _extract_timestamp ( self , item : Mapping [ str , Any ] ) - > Optional [ datetime ] :
""" Extract and normalise a timestamp from ``poll_timestamp_field``. """
if not self . poll_timestamp_field :
return None
value = self . _extract_field ( item , self . poll_timestamp_field )
if isinstance ( value , list ) and value :
value = value [ 0 ]
return self . _parse_datetime ( value )
@staticmethod
def _parse_datetime ( value : Any ) - > Optional [ datetime ] :
""" Parse a raw value into a UTC datetime, or return None. """
if value is None :
return None
if isinstance ( value , datetime ) :
return ( value if value . tzinfo else value . replace ( tzinfo = timezone . utc ) ) . astimezone ( timezone . utc )
if isinstance ( value , ( int , float ) ) :
try :
return datetime . fromtimestamp ( float ( value ) , tz = timezone . utc )
except Exception :
return None
if isinstance ( value , str ) :
ts = value . strip ( )
for fmt in ( " % Y- % m- %d T % H: % M: % S. %f Z " , " % Y- % m- %d T % H: % M: % SZ " , " % Y- % m- %d % H: % M: % S " , " % Y- % m- %d " ) :
try :
return datetime . strptime ( ts , fmt ) . replace ( tzinfo = timezone . utc )
except Exception :
continue
try :
dt = datetime . fromisoformat ( ts . replace ( " Z " , " +00:00 " ) . replace ( " " , " T " ) )
return ( dt if dt . tzinfo else dt . replace ( tzinfo = timezone . utc ) ) . astimezone ( timezone . utc )
except Exception :
return None
return None
# -- Content template rendering -----------------------------------------
class _SafeDict ( dict ) :
""" Dict subclass that returns empty string for missing keys in format_map. """
2026-07-03 12:53:39 +08:00
Feature/generic api connector (#13545)
# feat: Add Generic REST API Connector
## What problem does this PR solve?
RAGFlow supports many specific data source connectors (MySQL, Slack,
Google Drive, etc.), but there was no way to connect an arbitrary REST
API as a data source. Users with custom or third-party APIs had to write
a new connector class for each one.
This PR adds a **generic, configuration-driven REST API connector** that
lets users connect any REST API as a data source entirely through the UI
— no code changes needed per API.
---
## Features
### Core Connector (`common/data_source/rest_api_connector.py`)
- Implements `LoadConnector` and `PollConnector` interfaces for full and
incremental sync
- **Configurable authentication:** None, API Key (custom header), Bearer
Token, Basic Auth
- **Pluggable pagination:** Page-based, Offset-based, Cursor-based, or
None
- Smart page-size inference from user's query parameters to avoid
duplicate/conflicting params
- Configurable request delay between pages to prevent API rate limiting
- Auto-detection of the items array in JSON responses (`items`,
`results`, `data`, `records`, or first list found)
- **Advanced field mapping** with dot-notation (`country.name`), array
wildcards (`newsType[*].name`), type hints, and default values
- Optional content template rendering (`"Title: {title}\nBody: {body}"`)
- HTML stripping for content fields
- Stable document IDs via `hash128` from a configurable ID field or
auto-generated from item content
- Pydantic configuration schema with automatic coercion of UI string
inputs to dicts/lists
### Backend Registration (`rag/svr/sync_data_source.py`,
`common/constants.py`, `common/data_source/config.py`)
- `REST_API` sync class wired into RAGFlow's `func_factory`
- Full sync (`load_from_state`) and incremental polling (`poll_source`)
support
- Credentials and config passed from task to connector following
existing patterns (MySQL, SeaFile, etc.)
### Test Connection Endpoint (`api/apps/connector_app.py`)
- `POST /v1/connector/<id>/test` validates config schema,
authentication, and API connectivity without triggering a sync
- Clear error messages for auth failures vs. config issues
### Frontend UI (`web/src/pages/user-setting/data-source/constant/`)
- **Postman-style configuration:** Base URL, Query Parameters (key=value
per line), Auth, Content Fields, Metadata Fields, Pagination Type
- Auth-type-aware form: fields for API key header/value, Bearer token,
or Basic username/password appear only when relevant
- **Advanced Settings** toggle for: Custom Headers, Max Pages, Request
Delay, Poll Timestamp Field, Request Body (POST)
- Connector icon (SVG) and i18n strings (English)
- **"Test Connection"** button to validate before syncing
---
## Controls & Safety
- Configurable max pages safety cap (default: 1000, adjustable in UI)
- Configurable request delay between pages (default: 0.5s, adjustable in
UI)
- Auth errors (401/403) fail immediately without retries; transient
errors retry with exponential backoff
- Diagnostic logging: auth setup confirmation, request details on
failure, content field extraction status
---
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
##Visual Screenshots of Features
<img width="482" height="510" alt="Screenshot 2026-03-11 at 5 19 52 PM"
src="https://github.com/user-attachments/assets/dcb7ab4a-1622-44f3-bb02-d6f0527314c4"
/>
(Connector can be configured within the external data sources tab)
Configuration Parameters:
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 46 PM"
src="https://github.com/user-attachments/assets/5e154e71-4ab5-4872-bfb2-04f02b73c18a"
/>
<img width="661" height="682" alt="Screenshot 2026-03-11 at 5 20 54 PM"
src="https://github.com/user-attachments/assets/00cb14b7-0bcf-4b94-9d71-34e93369ecb2"
/>
Connection can be tested before attaching to dataset:
<img width="981" height="681" alt="Screenshot 2026-03-11 at 5 21 40 PM"
src="https://github.com/user-attachments/assets/aaa6eeeb-89a7-4349-bc34-2423bf8be9ee"
/>
Ingestion tested with API connector (works perfectly fine):
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 22 30 PM"
src="https://github.com/user-attachments/assets/afcd0d58-cadd-4152-badc-d2f14d96fbec"
/>
Search & Retrieval works as well with metadata flow:
<img width="1062" height="705" alt="Screenshot 2026-03-11 at 5 23 05 PM"
src="https://github.com/user-attachments/assets/d41ee935-dcf7-4456-b317-22a76ca032c0"
/>
---------
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-13 17:35:01 +05:00
def __missing__ ( self , key : str ) - > str :
return " "
def _render_content_template ( self , item : Mapping [ str , Any ] ) - > str :
""" Render content using a user-provided template with `` {field} `` placeholders. """
template = self . content_template or " "
values : Dict [ str , str ] = { }
for field_path in set ( self . content_fields + self . metadata_fields ) :
val = self . _get_typed_field_value ( field_path , item )
if val is None :
continue
name = re . sub ( r " \ [ \ d+ \ ]| \ [ \ * \ ] " , " " , field_path ) . replace ( " . " , " _ " )
values [ name ] = self . _coerce_to_text ( val )
try :
rendered = template . format_map ( self . _SafeDict ( values ) )
except Exception as exc :
logging . warning ( " Failed to render content template: %s " , exc )
parts = [ self . _coerce_to_text ( self . _get_typed_field_value ( f , item ) ) for f in self . content_fields ]
rendered = " \n " . join ( p for p in parts if p )
return self . _strip_html ( rendered )
# -- Static helpers -----------------------------------------------------
@staticmethod
def _strip_html ( text : str ) - > str :
""" Remove basic HTML tags and normalise whitespace. """
if " < " not in text or " > " not in text :
return text
cleaned = re . sub ( r " <[^>]+> " , " " , text )
return re . sub ( r " \ s+ " , " " , cleaned ) . strip ( )
@staticmethod
def _coerce_to_text ( value : Any ) - > str :
""" Convert any value to a plain-text string. """
if value is None :
return " "
if isinstance ( value , str ) :
return value
if isinstance ( value , ( int , float , bool ) ) :
return str ( value )
try :
return json . dumps ( value , ensure_ascii = False )
except Exception :
return str ( value )
@staticmethod
def _serialize_metadata_value ( value : Any ) - > Any :
""" Serialise a metadata value for storage. """
if isinstance ( value , datetime ) :
if value . tzinfo is None :
value = value . replace ( tzinfo = timezone . utc )
return value . isoformat ( )
if isinstance ( value , ( int , float , bool , str ) ) :
return value
try :
return json . dumps ( value , ensure_ascii = False )
except Exception :
return str ( value )
@staticmethod
def _doc_in_time_window ( doc : Document , start : datetime , end : datetime ) - > bool :
if not doc . doc_updated_at :
return False
dt = doc . doc_updated_at
dt = ( dt if dt . tzinfo else dt . replace ( tzinfo = timezone . utc ) ) . astimezone ( timezone . utc )
return start < = dt < end