File content parsing

Function

Split and merge document parsing results.

URI

POST /v1/koosearch/doc-search/parse-result/split

Request Parameters

**Table 1** Request header parameters
Parameter	Mandatory	Type	Description
X-Auth-Token	Yes	String	Parameter description: Token used for API authentication. For how to obtain the token, see section 3.2 "Authentication." Constraints: N/A.

**Table 2** Request body parameters
Parameter	Mandatory	Type	Description
doc	Yes	ParsedDocument object	Document parsing information
mode	No	Integer	0: raw; 1: split by contents; 2: split by section rules; 3: split by length; 4: auto split
rule_regexs	No	Array of strings	Regular expression match
chunk_size	No	Integer	Segment size

**Table 3** ParsedDocument
Parameter	Mandatory	Type	Description
doc_id	Yes	String	Document ID, which is generated based on the UUID
doc_name	Yes	String	Document
doc_type	Yes	String	Document type, for example, PDF or DOCX
preview_file_url	No	String	Preview file address
original_file	No	String	Original document path
file_size	No	Integer	Original document size, in bytes
pages	No	Array of ParsedDocumentPage objects	Document page information
images	No	Array of ParsedDocumentImage objects	Document image information
original_tables	No	Array of OriginalTable objects	Original table information

**Table 4** ParsedDocumentPage
Parameter	Mandatory	Type	Description
page_num	Yes	Integer	Page number, which indicates the sequence number of a page in the document
preview_image_url	No	String	Address of the document page preview image
components	No	Array of ParsedDocumentComponent objects	Paragraph information on the page

**Table 5** ParsedDocumentComponent
Parameter	Mandatory	Type	Description
id	Yes	String	Paragraph ID, which is generated based on the UUID
text	Yes	String	Paragraph Content
component_num	Yes	Integer	Paragraph code, which indicates the sequence number of a paragraph in the document. The value starts from 1.
pdf_coordinate	Yes	Array<Array<Integer>>	Coordinates of a paragraph on the page, corresponding to the upper left, upper right, lower right, and lower left, respectively, for highlighting
original_table_id	No	String	This parameter has a value only when the table is split. It is used to save the original long table to support the small2big feature.

**Table 6** ParsedDocumentImage
Parameter	Mandatory	Type	Description
image_id	Yes	String	Image ID, which is the prefix img- and UUID
url	No	String	Path for uploading the image to OBS
data	No	String	Base64-encoded image data
title	No	String	Image title
desc	No	String	Image description
width	No	Integer	Image width,
height	No	Integer	Image height.

**Table 7** OriginalTable
Parameter	Mandatory	Type	Description
id	Yes	String	Table ID. ParsedDocumentComponent will reference this identifier to avoid storing multiple copies.
content	Yes	String	Form content

Response Parameters

Status code: 200

**Table 8** Response body parameters
Parameter	Type	Description
doc_id	String	Document ID, which is generated based on the UUID
doc_name	String	Document
doc_type	String	Document type, for example, PDF or DOCX
preview_file_url	String	Preview file address
original_file	String	Original document path
file_size	Integer	Original document size, in bytes
pages	Array of ParsedDocumentPage objects	Document page information
images	Array of ParsedDocumentImage objects	Document image information
original_tables	Array of OriginalTable objects	Original table information

**Table 9** ParsedDocumentPage
Parameter	Type	Description
page_num	Integer	Page number, which indicates the sequence number of a page in the document
preview_image_url	String	Address of the document page preview image
components	Array of ParsedDocumentComponent objects	Paragraph information on the page

**Table 10** ParsedDocumentComponent
Parameter	Type	Description
id	String	Paragraph ID, which is generated based on the UUID
text	String	Paragraph Content
component_num	Integer	Paragraph code, which indicates the sequence number of a paragraph in the document. The value starts from 1.
pdf_coordinate	Array<Array<Integer>>	Coordinates of a paragraph on the page, corresponding to the upper left, upper right, lower right, and lower left, respectively, for highlighting
original_table_id	String	This parameter has a value only when the table is split. It is used to save the original long table to support the small2big feature.

**Table 11** ParsedDocumentImage
Parameter	Type	Description
image_id	String	Image ID, which is the prefix img- and UUID
url	String	Path for uploading the image to OBS
data	String	Base64-encoded image data
title	String	Image title
desc	String	Image description
width	Integer	Image width,
height	Integer	Image height.

**Table 12** OriginalTable
Parameter	Type	Description
id	String	Table ID. ParsedDocumentComponent will reference this identifier to avoid storing multiple copies.
content	String	Form content

Status code: 400

**Table 13** Response body parameters
Parameter	Type	Description
error_code	String	Error Code
error_msg	String	Error description

Status code: 401

**Table 14** Response body parameters
Parameter	Type	Description
error_code	String	Error Code
error_msg	String	Error description

Status code: 500

**Table 15** Response body parameters
Parameter	Type	Description
error_code	String	Error Code
error_msg	String	Error description

Example Requests

None

Example Responses

Status code: 200

Document splitting and merging result

{
  "pages" : [ {
    "components" : [ {
      "id" : "393c1d28cd9c40f5ad9f7a2d33dffb80",
      "text" : "1--- Level-1 title 1\n1.1 Level-2 title 1\nContent\n1.2 Level-2 title 2\nContent\n2 --- Level-1 title 2\n2.1 Level-2 title 3\nContent\n2.2 Level-2 title 4\nContent\nlistItemByLevel: stores elements of each level, for example, 0, 1, and 2. 0 indicates a level-1 title. The corresponding value is the level-1 title in the document. New group titles will be cleared.\nlistContextMap: stores the mapping between numId and listContext.\nNew title of itemContext. The member variable number indicates the sequence number of the title. There is a parent-child relationship.\nListItemContext: nb in parent is used to count the number of children.\nThe numId of the level-1 and level-2 titles of this document is the same.\n3 --- Level-1 title 3\n3.1 Level-2 title 1\nContent\n3.2 Level-2 title 2\nContent\n4 --- Level-1 title 4\n4.1 Level-2 title 3\nContent\n4.2 Level-2 title 4\nContent\nlistItemByLevel stores elements of each level, for example, 0, 1, and 2. 0 indicates a level-1 title. The corresponding value is the level-1 title in the document. New group titles will be cleared.\nlistContextMap: stores the mapping between numId and listContext.\nNew title of itemContext. The member variable number indicates the sequence number of the title. There is a parent-child relationship.\nListItemContext: nb in parent is used to count the number of children.\nThe numId of the level-1 and level-2 titles of this document is the same.\n5 --- Level-1 title 5\n5.1 Level-2 title 1\nContent\n5.2 Level-2 title 2\nContent\n6 --- Level-1 title 6\n6.1 Level-2 title 3\nContent\n6.2 Level-2 title 4\nContent\nlistItemByLevel stores elements of each level, for example, 0, 1, and 2. 0 indicates a level-1 title. The corresponding value is the level-1 title in the document. New group titles will be cleared.\nlistContextMap: stores the mapping between numId and listContext.\nNew title of itemContext. The member variable number indicates the sequence number of the title. There is a parent-child relationship.\nListItemContext: nb in parent is used to count the number of children.\nThe numId of the level-1 and level-2 titles of this document is the same.\n7--- Level-1 title 7\n7.1 Level-2 title 1\nContent\n7.2 Level-2 title 2\nContent\n8--- Level-1 title 8\n8.1 Level-2 title 3\nContent\n8.2 Level-2 title 4\nContent\nlistItemByLevel stores elements of each level, for example, 0, 1, and 2. 0 indicates a level-1 title. The corresponding value is the level-1 title in the document. New group titles will be cleared.\nlistContextMap: stores the mapping between numId and listContext.\nNew title of itemContext. The member variable number indicates the sequence number of the title. There is a parent-child relationship.\nListItemContext: nb in parent is used to count the number of children.\nThe numId of the level-1 and level-2 titles of this document is the same.",
      "component_num" : 1
    } ],
    "page_num" : 0
  } ],
  "doc_id" : "844f805a7255437b8c139f4331ec3012",
  "doc_name" : "Test Title No..docx",
  "doc_type" : "DOCX",
  "original_file" : "uni-search/files/729cbd739854470da5426ed26bd900ca/fb9731ab-7085-474f-b6c7-64473586f0f3/c5e7dc40-9d43-49fd-8b5f-12c906ed66d2/d5a4ced94f07050841eb9424f87096af/Test Title No..docx",
  "file_size" : 68621
}