mirror of
https://github.com/webrecorder/browsertrix-crawler.git
synced 2025-10-19 06:23:16 +00:00
don't remove 200 status code
use 'status' instead of 'statusCode'
This commit is contained in:
parent
ebab39cdf5
commit
10a42d658f
2 changed files with 12 additions and 12 deletions
|
@ -91,7 +91,7 @@ type PageEntry = {
|
||||||
text?: string;
|
text?: string;
|
||||||
favIconUrl?: string;
|
favIconUrl?: string;
|
||||||
ts?: string;
|
ts?: string;
|
||||||
statusCode?: number;
|
status?: number;
|
||||||
};
|
};
|
||||||
|
|
||||||
// ============================================================================
|
// ============================================================================
|
||||||
|
@ -720,7 +720,7 @@ self.__bx_behaviors.selectMainBehavior();
|
||||||
if (mime) {
|
if (mime) {
|
||||||
data.mime = mime;
|
data.mime = mime;
|
||||||
}
|
}
|
||||||
data.statusCode = 200;
|
data.status = 200;
|
||||||
logger.info(
|
logger.info(
|
||||||
"Direct fetch successful",
|
"Direct fetch successful",
|
||||||
{ url, ...logDetails },
|
{ url, ...logDetails },
|
||||||
|
@ -789,7 +789,7 @@ self.__bx_behaviors.selectMainBehavior();
|
||||||
|
|
||||||
data.loadState = LoadState.EXTRACTION_DONE;
|
data.loadState = LoadState.EXTRACTION_DONE;
|
||||||
|
|
||||||
if (data.statusCode >= 400) {
|
if (data.status >= 400) {
|
||||||
return;
|
return;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -1587,17 +1587,17 @@ self.__bx_behaviors.selectMainBehavior();
|
||||||
}
|
}
|
||||||
|
|
||||||
// Handle 4xx or 5xx response as a page load error
|
// Handle 4xx or 5xx response as a page load error
|
||||||
const statusCode = resp.status();
|
const status = resp.status();
|
||||||
data.statusCode = statusCode;
|
data.status = status;
|
||||||
if (isChromeError) {
|
if (isChromeError) {
|
||||||
if (failCrawlOnError) {
|
if (failCrawlOnError) {
|
||||||
logger.fatal("Seed Page Load Error, failing crawl", {
|
logger.fatal("Seed Page Load Error, failing crawl", {
|
||||||
statusCode,
|
status,
|
||||||
...logDetails,
|
...logDetails,
|
||||||
});
|
});
|
||||||
} else {
|
} else {
|
||||||
logger.error("Page Crashed on Load", {
|
logger.error("Page Crashed on Load", {
|
||||||
statusCode,
|
status,
|
||||||
...logDetails,
|
...logDetails,
|
||||||
});
|
});
|
||||||
throw new Error("logged");
|
throw new Error("logged");
|
||||||
|
@ -1960,7 +1960,7 @@ self.__bx_behaviors.selectMainBehavior();
|
||||||
mime,
|
mime,
|
||||||
favicon,
|
favicon,
|
||||||
ts,
|
ts,
|
||||||
statusCode,
|
status,
|
||||||
}: PageState) {
|
}: PageState) {
|
||||||
const row: PageEntry = { id: pageid!, url, title, loadState };
|
const row: PageEntry = { id: pageid!, url, title, loadState };
|
||||||
|
|
||||||
|
@ -1972,8 +1972,8 @@ self.__bx_behaviors.selectMainBehavior();
|
||||||
row.mime = mime;
|
row.mime = mime;
|
||||||
}
|
}
|
||||||
|
|
||||||
if (statusCode && statusCode !== 200) {
|
if (status) {
|
||||||
row.statusCode = statusCode;
|
row.status = status;
|
||||||
}
|
}
|
||||||
|
|
||||||
if (this.params.writePagesToRedis) {
|
if (this.params.writePagesToRedis) {
|
||||||
|
|
|
@ -46,7 +46,7 @@ export class PageState {
|
||||||
depth: number;
|
depth: number;
|
||||||
extraHops: number;
|
extraHops: number;
|
||||||
|
|
||||||
statusCode: number;
|
status: number;
|
||||||
|
|
||||||
workerid!: WorkerId;
|
workerid!: WorkerId;
|
||||||
|
|
||||||
|
@ -72,7 +72,7 @@ export class PageState {
|
||||||
this.seedId = redisData.seedId;
|
this.seedId = redisData.seedId;
|
||||||
this.depth = redisData.depth;
|
this.depth = redisData.depth;
|
||||||
this.extraHops = redisData.extraHops || 0;
|
this.extraHops = redisData.extraHops || 0;
|
||||||
this.statusCode = 0;
|
this.status = 0;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue