Files
spyda/internal/crawler.go
2021-02-02 12:07:53 +10:00

82 lines
1.4 KiB
Go

package internal
import (
"time"
log "github.com/sirupsen/logrus"
)
type Crawler interface {
Start()
Crawl(url string) error
}
type crawler struct {
conf *Config
db Store
queue chan string
indexer Indexer
}
func NewCrawler(conf *Config, db Store, indexer Indexer) (Crawler, error) {
return &crawler{
conf: conf,
db: db,
queue: make(chan string),
indexer: indexer,
}, nil
}
func (c *crawler) loop() {
for {
url := <-c.queue
log.Debugf("crawling %s", url)
links, err := GetLinks(url)
if err != nil {
log.WithError(err).Error("error crawling %s", url)
continue
}
for link := range links {
hash := HashURL(link)
if c.db.HasURL(hash) {
log.Debugf("seen %s (skipping)", link)
continue
}
log.Debugf("found %s", link)
metrics.Counter("crawler", "crawled").Inc()
url := NewURL(link)
url.CrawledAt = time.Now()
entry, err := Scrape(c.conf, link)
if err != nil {
log.WithError(err).Error("error scraping %s", link)
} else {
if err := c.indexer.Index(entry); err != nil {
log.WithError(err).Error("error indexing %s", link)
} else {
if err := c.db.SetURL(hash, url); err != nil {
log.WithError(err).Error("error recording url %s", link)
} else {
metrics.Counter("crawler", "scraped").Inc()
}
}
}
}
}
}
func (c *crawler) Crawl(url string) error {
c.queue <- url
return nil
}
func (c *crawler) Start() {
go c.loop()
}