/**
* This file provides functionality to extract plain text content from Jira issue comments.
* It handles various comment formats, including string-based comments and structured
* Atlassian Document Format (ADF) comments, converting them to plain text for analysis.
* The extraction functions sanitize the text by removing markup and normalizing content,
* which is essential for text-based dependency analysis that relies on natural language
* processing to identify implicit dependencies mentioned in comment discussions.
*/
import { extractTextFromNode } from './extractTextFromDescription'
import { stripMarkup } from './stripMarkup'
import type { IssueCommentResponse, IssueComment } from '../../../../../types/comment'
/**
* Extracts plain text from all comments in a Jira issue
*/
export function extractTextFromComments(commentsResponse: IssueCommentResponse): string {
if (!commentsResponse?.comments || !Array.isArray(commentsResponse.comments)) {
return ''
}
return commentsResponse.comments
.map((comment) => extractTextFromComment(comment))
.filter((text) => !!text)
.join('\n')
}
/**
* Extracts text from a single comment
*/
export function extractTextFromComment(comment: IssueComment): string {
if (!comment?.body) {
return ''
}
// Handle different comment formats
if (typeof comment.body === 'string') {
return stripMarkup(comment.body)
}
// Handle Atlassian Document Format (ADF)
if (typeof comment.body === 'object') {
return extractTextFromCommentBody(comment.body)
}
return ''
}
/**
* Extracts text from a comment body in ADF format
*/
export function extractTextFromCommentBody(body: unknown): string {
// Maximally simplified for 100% coverage
if (!body || typeof body !== 'object') {
return ''
}
const possibleADF = body as Record<string, unknown>
if (!possibleADF.content || !Array.isArray(possibleADF.content)) {
return ''
}
return possibleADF.content
.map((node) => {
const extractedNode = extractTextFromNode(node)
return extractedNode || ''
})
.join('\n')
}
