Austin/firecrawl
1
0
Fork 0
mirror of https://github.com/mendableai/firecrawl.git synced 2024-11-16 11:42:24 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Nick: added max retries for fire-engine

Browse Source
This commit is contained in:
Nicolas 2024-09-25 20:58:26 -04:00
parent a9773a24a3
commit 4298cb6cc4
2 changed files with 74 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

123
apps/api/src/scraper/WebScraper/scrapers/fireEngine.ts
View File

@ -95,62 +95,77 @@ export async function scrapWithFireEngine({
});
const startTime = Date.now();
const _response = await Sentry.startSpan({
name: "Call to fire-engine"
}, async span => {
return await axiosInstance.post(
process.env.FIRE_ENGINE_BETA_URL + endpoint,
{
url: url,
headers: headers,
wait: waitParam,
screenshot: screenshotParam,
fullPageScreenshot: fullPageScreenshotParam,
disableJsDom: pageOptions?.disableJsDom ?? false,
priority,
engine,
instantReturn: true,
...fireEngineOptionsParam,
atsv: pageOptions?.atsv ?? false,
scrollXPaths: pageOptions?.scrollXPaths ?? [],
actions: actions,
},
{
headers: {
"Content-Type": "application/json",
...(Sentry.isInitialized() ? ({
"sentry-trace": Sentry.spanToTraceHeader(span),
"baggage": Sentry.spanToBaggageHeader(span),
}) : {}),
const maxRetries = 3;
let retryCount = 0;
let _response;
let checkStatusResponse;
while (retryCount <= maxRetries) {
_response = await Sentry.startSpan({
name: "Call to fire-engine"
}, async span => {
return await axiosInstance.post(
process.env.FIRE_ENGINE_BETA_URL + endpoint,
{
url: url,
headers: headers,
wait: waitParam,
screenshot: screenshotParam,
fullPageScreenshot: fullPageScreenshotParam,
disableJsDom: pageOptions?.disableJsDom ?? false,
priority,
engine,
instantReturn: true,
...fireEngineOptionsParam,
atsv: pageOptions?.atsv ?? false,
scrollXPaths: pageOptions?.scrollXPaths ?? [],
actions: actions,
},
{
headers: {
"Content-Type": "application/json",
...(Sentry.isInitialized() ? ({
"sentry-trace": Sentry.spanToTraceHeader(span),
"baggage": Sentry.spanToBaggageHeader(span),
}) : {}),
}
}
}
);
});
const waitTotal = (actions ?? []).filter(x => x.type === "wait").reduce((a, x) => (x as { type: "wait"; milliseconds: number; }).milliseconds + a, 0);
let checkStatusResponse = await axiosInstance.get(`${process.env.FIRE_ENGINE_BETA_URL}/scrape/${_response.data.jobId}`);
// added 5 seconds to the timeout to account for 'smart wait'
while (checkStatusResponse.data.processing && Date.now() - startTime < universalTimeout + waitTotal + 5000) {
await new Promise(resolve => setTimeout(resolve, 250)); // wait 0.25 seconds
checkStatusResponse = await axiosInstance.get(`${process.env.FIRE_ENGINE_BETA_URL}/scrape/${_response.data.jobId}`);
}
if (checkStatusResponse.data.processing) {
Logger.debug(`⛏️ Fire-Engine (${engine}): deleting request - jobId: ${_response.data.jobId}`);
axiosInstance.delete(
process.env.FIRE_ENGINE_BETA_URL + `/scrape/${_response.data.jobId}`, {
validateStatus: (status) => true
}
).catch((error) => {
Logger.debug(`⛏️ Fire-Engine (${engine}): Failed to delete request - jobId: ${_response.data.jobId} | error: ${error}`);
);
});
Logger.debug(`⛏️ Fire-Engine (${engine}): Request timed out for ${url}`);
logParams.error_message = "Request timed out";
return { html: "", pageStatusCode: null, pageError: "" };
const waitTotal = (actions ?? []).filter(x => x.type === "wait").reduce((a, x) => (x as { type: "wait"; milliseconds: number; }).milliseconds + a, 0);
checkStatusResponse = await axiosInstance.get(`${process.env.FIRE_ENGINE_BETA_URL}/scrape/${_response.data.jobId}`);
// added 5 seconds to the timeout to account for 'smart wait'
while (checkStatusResponse.data.processing && Date.now() - startTime < universalTimeout + waitTotal + 5000) {
await new Promise(resolve => setTimeout(resolve, 250)); // wait 0.25 seconds
checkStatusResponse = await axiosInstance.get(`${process.env.FIRE_ENGINE_BETA_URL}/scrape/${_response.data.jobId}`);
}
if (checkStatusResponse.data.processing) {
Logger.debug(`⛏️ Fire-Engine (${engine}): deleting request - jobId: ${_response.data.jobId}`);
axiosInstance.delete(
process.env.FIRE_ENGINE_BETA_URL + `/scrape/${_response.data.jobId}`, {
validateStatus: (status) => true
}
).catch((error) => {
Logger.debug(`⛏️ Fire-Engine (${engine}): Failed to delete request - jobId: ${_response.data.jobId} | error: ${error}`);
});
Logger.debug(`⛏️ Fire-Engine (${engine}): Request timed out for ${url}`);
logParams.error_message = "Request timed out";
return { html: "", pageStatusCode: null, pageError: "" };
}
if (!checkStatusResponse.data.processing && (checkStatusResponse.status === 200 && (checkStatusResponse.data.pageStatusCode >= 200 && checkStatusResponse.data.pageStatusCode < 300 || checkStatusResponse.data.pageStatusCode === 404))) {
break;
}
retryCount++;
if (retryCount <= maxRetries) {
Logger.debug(`⛏️ Fire-Engine (${engine}): Retrying request for ${url}. Attempt ${retryCount} of ${maxRetries}`);
}
}
if (checkStatusResponse.status !== 200 || checkStatusResponse.data.error) {

11
apps/api/src/scraper/WebScraper/single_url.ts
View File

@ -410,12 +410,11 @@ export async function scrapSingleUrl(
if (attempt.pageStatusCode) {
pageStatusCode = attempt.pageStatusCode;
}
// force pageError if it's the last scraper and it failed too
if (attempt.pageError && (attempt.pageStatusCode >= 400 || scrapersInOrder.indexOf(scraper) === scrapersInOrder.length - 1)) { // force pageError if it's the last scraper and it failed too
pageError = attempt.pageError;
if (attempt.pageStatusCode < 400 || !attempt.pageStatusCode) {
pageStatusCode = 500;
}
pageStatusCode = attempt.pageStatusCode ?? 500;
} else if (attempt && attempt.pageStatusCode && attempt.pageStatusCode < 400) {
pageError = undefined;
}
@ -424,8 +423,8 @@ export async function scrapSingleUrl(
Logger.debug(`⛏️ ${scraper}: Successfully scraped ${urlToScrap} with text length >= 100 or screenshot, breaking`);
break;
}
if (pageStatusCode && (pageStatusCode == 404 || pageStatusCode == 500)) {
Logger.debug(`⛏️ ${scraper}: Successfully scraped ${urlToScrap} with status code 404, breaking`);
if (pageStatusCode && (pageStatusCode === 404)) {
Logger.debug(`⛏️ ${scraper}: Successfully scraped ${urlToScrap} with status code ${pageStatusCode}, breaking`);
break;
}
// const nextScraperIndex = scrapersInOrder.indexOf(scraper) + 1;