aboutsummaryrefslogtreecommitdiff
path: root/cmd/links/links.go
diff options
context:
space:
mode:
Diffstat (limited to 'cmd/links/links.go')
-rw-r--r--cmd/links/links.go14
1 files changed, 8 insertions, 6 deletions
diff --git a/cmd/links/links.go b/cmd/links/links.go
index 5f76a6a..bf91f3f 100644
--- a/cmd/links/links.go
+++ b/cmd/links/links.go
@@ -20,11 +20,7 @@ var (
validSchemes = flag.String("schemes", "http,https", "comma-separated list of allowed protocols")
)
-func extractLinks(c *crawl.Crawler, u string, depth int, resp *http.Response, err error) error {
- if err != nil {
- return nil
- }
-
+func extractLinks(c *crawl.Crawler, u string, depth int, resp *http.Response, _ error) error {
links, err := analysis.GetLinks(resp)
if err != nil {
// Not a fatal error, just a bad web page.
@@ -50,7 +46,13 @@ func main() {
crawl.NewSeedScope(seeds),
)
- crawler, err := crawl.NewCrawler("crawldb", seeds, scope, crawl.FetcherFunc(http.Get), crawl.NewRedirectHandler(crawl.HandlerFunc(extractLinks)))
+ crawler, err := crawl.NewCrawler(
+ "crawldb",
+ seeds,
+ scope,
+ crawl.FetcherFunc(http.Get),
+ crawl.HandleRetries(crawl.FollowRedirects(crawl.FilterErrors(crawl.HandlerFunc(extractLinks)))),
+ )
if err != nil {
log.Fatal(err)
}