From 79d11ea709786a7b51f12723388d17e77d5f6e0f Mon Sep 17 00:00:00 2001
From: DDDDD12138 <43703884+DDDDD12138@users.noreply.github.com>
Date: Wed, 18 Dec 2024 09:08:06 +0800
Subject: [PATCH] feat: add parameters for JinaReaderTool (#11613)

---
 .../builtin/jina/tools/jina_reader.py         | 13 +++++
 .../builtin/jina/tools/jina_reader.yaml       | 55 +++++++++++++++++++
 2 files changed, 68 insertions(+)

diff --git a/api/core/tools/provider/builtin/jina/tools/jina_reader.py b/api/core/tools/provider/builtin/jina/tools/jina_reader.py
index 0dd55c65291783..756b7272248146 100644
--- a/api/core/tools/provider/builtin/jina/tools/jina_reader.py
+++ b/api/core/tools/provider/builtin/jina/tools/jina_reader.py
@@ -43,6 +43,13 @@ def _invoke(
         if wait_for_selector is not None and wait_for_selector != "":
             headers["X-Wait-For-Selector"] = wait_for_selector
 
+        remove_selector = tool_parameters.get("remove_selector")
+        if remove_selector is not None and remove_selector != "":
+            headers["X-Remove-Selector"] = remove_selector
+
+        if tool_parameters.get("retain_images", False):
+            headers["X-Retain-Images"] = "true"
+
         if tool_parameters.get("image_caption", False):
             headers["X-With-Generated-Alt"] = "true"
 
@@ -59,6 +66,12 @@ def _invoke(
         if tool_parameters.get("no_cache", False):
             headers["X-No-Cache"] = "true"
 
+        if tool_parameters.get("with_iframe", False):
+            headers["X-With-Iframe"] = "true"
+
+        if tool_parameters.get("with_shadow_dom", False):
+            headers["X-With-Shadow-Dom"] = "true"
+
         max_retries = tool_parameters.get("max_retries", 3)
         response = ssrf_proxy.get(
             str(URL(self._jina_reader_endpoint + url)),
diff --git a/api/core/tools/provider/builtin/jina/tools/jina_reader.yaml b/api/core/tools/provider/builtin/jina/tools/jina_reader.yaml
index 589bc3433d9478..012a8c7688cb57 100644
--- a/api/core/tools/provider/builtin/jina/tools/jina_reader.yaml
+++ b/api/core/tools/provider/builtin/jina/tools/jina_reader.yaml
@@ -67,6 +67,33 @@ parameters:
       pt_BR: css selector para aguardar elementos específicos
     llm_description: css selector of the target element to wait for
     form: form
+  - name: remove_selector
+    type: string
+    required: false
+    label:
+      en_US: Excluded Selector
+      zh_Hans: 排除选择器
+      pt_BR: Seletor Excluído
+    human_description:
+      en_US: css selector for remove for specific elements
+      zh_Hans: css 选择器用于排除特定元素
+      pt_BR: seletor CSS para remover elementos específicos
+    llm_description: css selector of the target element to remove for
+    form: form
+  - name: retain_images
+    type: boolean
+    required: false
+    default: false
+    label:
+      en_US: Remove All Images
+      zh_Hans: 删除所有图片
+      pt_BR: Remover todas as imagens
+    human_description:
+      en_US: Removes all images from the response.
+      zh_Hans: 从响应中删除所有图片。
+      pt_BR: Remove todas as imagens da resposta.
+    llm_description: Remove all images
+    form: form
   - name: image_caption
     type: boolean
     required: false
@@ -136,6 +163,34 @@ parameters:
       pt_BR: Ignorar o cache
     llm_description: bypass the cache
     form: form
+  - name: with_iframe
+    type: boolean
+    required: false
+    default: false
+    label:
+      en_US: Enable iframe extraction
+      zh_Hans: 启用 iframe 提取
+      pt_BR: Habilitar extração de iframe
+    human_description:
+      en_US: Extract and process content of all embedded iframes in the DOM tree.
+      zh_Hans: 提取并处理 DOM 树中所有嵌入 iframe 的内容。
+      pt_BR: Extrair e processar o conteúdo de todos os iframes incorporados na árvore DOM.
+    llm_description: Extract content from embedded iframes
+    form: form
+  - name: with_shadow_dom
+    type: boolean
+    required: false
+    default: false
+    label:
+      en_US: Enable Shadow DOM extraction
+      zh_Hans: 启用 Shadow DOM 提取
+      pt_BR: Habilitar extração de Shadow DOM
+    human_description:
+      en_US: Traverse all Shadow DOM roots in the document and extract content.
+      zh_Hans: 遍历文档中所有 Shadow DOM 根并提取内容。
+      pt_BR: Percorra todas as raízes do Shadow DOM no documento e extraia o conteúdo.
+    llm_description: Extract content from Shadow DOM roots
+    form: form
   - name: summary
     type: boolean
     required: false