github/node-DeepResearch
1
0
Fork 0
mirror of https://github.com/jina-ai/node-DeepResearch.git synced 2025-12-26 06:28:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: md footnote

Browse Source
This commit is contained in:
Han Xiao 2025-03-19 17:43:39 +08:00
parent 835d5b9a0f
commit a3f2120885
1 changed files with 18 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
src/utils/text-tools.ts
View File

@ -15,6 +15,9 @@ export function repairMarkdownFootnotes(
// Regex to catch grouped footnotes like [^1, ^2, ^3] or [^1,^2,^3] // Regex to catch grouped footnotes like [^1, ^2, ^3] or [^1,^2,^3]
const groupedFootnoteRegex = /\[\^(\d+)(?:,\s*\^(\d+))+]/g; const groupedFootnoteRegex = /\[\^(\d+)(?:,\s*\^(\d+))+]/g;
// New regex to catch partially marked footnotes like [^10, 11]
const partialGroupedFootnoteRegex = /\[\^(\d+)(?:,\s*(\d+))+]/g;
// Helper function to format references // Helper function to format references
const formatReferences = (refs: Array<Reference>) => { const formatReferences = (refs: Array<Reference>) => {
return refs.map((ref, i) => { return refs.map((ref, i) => {
@ -34,6 +37,11 @@ export function repairMarkdownFootnotes(
// First case: no references - remove any footnote citations // First case: no references - remove any footnote citations
if (!references?.length) { if (!references?.length) {
return markdownString return markdownString
.replace(partialGroupedFootnoteRegex, (match) => {
// Extract all numbers from the partially marked grouped footnote
const numbers = match.match(/\d+/g) || [];
return numbers.map(num => `[^${num}]`).join(', ');
})
.replace(groupedFootnoteRegex, (match) => { .replace(groupedFootnoteRegex, (match) => {
// Extract all numbers from the grouped footnote // Extract all numbers from the grouped footnote
const numbers = match.match(/\d+/g) || []; const numbers = match.match(/\d+/g) || [];
@ -47,12 +55,16 @@ export function repairMarkdownFootnotes(
.replace(/\[(\d+)\^]/g, (_, num) => `[^${num}]`) .replace(/\[(\d+)\^]/g, (_, num) => `[^${num}]`)
.replace(/\[(\d+)]/g, (_, num) => `[^${num}]`); .replace(/\[(\d+)]/g, (_, num) => `[^${num}]`);
// Fix grouped footnotes // Fix grouped footnotes - both fully marked and partially marked types
processedMarkdown = processedMarkdown.replace(groupedFootnoteRegex, (match) => { processedMarkdown = processedMarkdown
// Extract all numbers from the grouped footnote .replace(groupedFootnoteRegex, (match) => {
const numbers = match.match(/\d+/g) || []; const numbers = match.match(/\d+/g) || [];
return numbers.map(num => `[^${num}]`).join(', '); return numbers.map(num => `[^${num}]`).join(', ');
}); })
.replace(partialGroupedFootnoteRegex, (match) => {
const numbers = match.match(/\d+/g) || [];
return numbers.map(num => `[^${num}]`).join(', ');
});
// Now extract all footnotes from the processed answer // Now extract all footnotes from the processed answer
const footnotes: string[] = []; const footnotes: string[] = [];