mirror of
https://github.com/farcasclaudiu/Flowise.git
synced 2026-06-29 01:01:11 +03:00
* FIX #2617 Big sites scan error * FIX #2617 Big sites scan error - review fix --------- Co-authored-by: Ahmed Osman <ahmed.osman@evolpe.pl>
This commit is contained in:
@@ -131,7 +131,11 @@ class Cheerio_DocumentLoaders implements INode {
|
|||||||
|
|
||||||
async function cheerioLoader(url: string): Promise<any> {
|
async function cheerioLoader(url: string): Promise<any> {
|
||||||
try {
|
try {
|
||||||
let docs = []
|
let docs: IDocument[] = []
|
||||||
|
if (url.endsWith('.pdf')) {
|
||||||
|
if (process.env.DEBUG === 'true') options.logger.info(`CheerioWebBaseLoader does not support PDF files: ${url}`)
|
||||||
|
return docs
|
||||||
|
}
|
||||||
const loader = new CheerioWebBaseLoader(url, params)
|
const loader = new CheerioWebBaseLoader(url, params)
|
||||||
if (textSplitter) {
|
if (textSplitter) {
|
||||||
docs = await loader.loadAndSplit(textSplitter)
|
docs = await loader.loadAndSplit(textSplitter)
|
||||||
@@ -141,6 +145,7 @@ class Cheerio_DocumentLoaders implements INode {
|
|||||||
return docs
|
return docs
|
||||||
} catch (err) {
|
} catch (err) {
|
||||||
if (process.env.DEBUG === 'true') options.logger.error(`error in CheerioWebBaseLoader: ${err.message}, on page: ${url}`)
|
if (process.env.DEBUG === 'true') options.logger.error(`error in CheerioWebBaseLoader: ${err.message}, on page: ${url}`)
|
||||||
|
return []
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user