hireme/pkg/crawler/nvidia/nvidia.go

package nvidia

import (
	"bytes"
	"context"
	"crypto/sha256"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"regexp"
	"strings"
	"time"

	"git.pengzhan.dev/haopengzhan/hireme/internal/crawler"
	"git.pengzhan.dev/haopengzhan/hireme/internal/logger"
)

type NvidiaCrawler struct {
	client *http.Client
}

func NewNvidiaCrawler() *NvidiaCrawler {
	return NewNvidiaCrawlerWithClient(&http.Client{
		Timeout: 30 * time.Second,
	})
}

func NewNvidiaCrawlerWithClient(client *http.Client) *NvidiaCrawler {
	return &NvidiaCrawler{
		client: client,
	}
}

func (c *NvidiaCrawler) CanHandle(platform string) bool {
	p := strings.ToLower(platform)
	return p == "nvidia" || strings.Contains(p, "nvidia.wd5.myworkdayjobs.com")
}

type nvidiaSearchPayload struct {
	AppliedFacets map[string][]string `json:"appliedFacets"`
	Limit         int                 `json:"limit"`
	Offset        int                 `json:"offset"`
	SearchText    string              `json:"searchText"`
}

type nvidiaSearchResponse struct {
	Total    int             `json:"total"`
	JobPosts []nvidiaJobItem `json:"jobPostings"`
}

type nvidiaJobItem struct {
	Title            string   `json:"title"`
	ExternalPath     string   `json:"externalPath"`
	BulletFields     []string `json:"bulletFields"`
	LocationsText    string   `json:"locationsText"`
	PostedOn         string   `json:"postedOn"`
	JobPostingID     string   `json:"jobPostingId"`
	TimeType         string   `json:"timeType"`
}

type nvidiaDetailResponse struct {
	JobPostingInfo struct {
		ID                   string   `json:"id"`
		Title                string   `json:"title"`
		JobDescription       string   `json:"jobDescription"`
		Location             string   `json:"location"`
		AdditionalLocations  []string `json:"additionalLocations"`
		TimeType             string   `json:"timeType"`
		StartDate            string   `json:"startDate"`
		JobReqID             string   `json:"jobReqId"`
	} `json:"jobPostingInfo"`
}

func (c *NvidiaCrawler) DiscoverJobs(ctx context.Context, targetURL string) ([]crawler.DiscoveredJob, error) {
	seen := make(map[string]bool)
	var discovered []crawler.DiscoveredJob

	cxsURL := "https://nvidia.wd5.myworkdayjobs.com/wday/cxs/nvidia/NVIDIAExternalCareerSite/jobs"

	searchText := "Software Engineer"
	u, err := url.Parse(targetURL)
	if err == nil && u.Query().Get("q") != "" {
		searchText = u.Query().Get("q")
	}

	pageSize := 20
	for page := 0; page < 5; page++ {
		offset := page * pageSize
		reqPayload := nvidiaSearchPayload{
			AppliedFacets: make(map[string][]string),
			Limit:         pageSize,
			Offset:        offset,
			SearchText:    searchText,
		}

		payloadBytes, err := json.Marshal(reqPayload)
		if err != nil {
			return nil, err
		}

		req, err := http.NewRequestWithContext(ctx, "POST", cxsURL, bytes.NewBuffer(payloadBytes))
		if err != nil {
			if page == 0 {
				return nil, err
			}
			break
		}

		req.Header.Set("Content-Type", "application/json")
		req.Header.Set("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36")
		req.Header.Set("Accept", "application/json")

		resp, err := c.client.Do(req)
		if err != nil {
			if page == 0 {
				return nil, err
			}
			break
		}
		defer func() { _ = resp.Body.Close() }()

		body, err := io.ReadAll(resp.Body)
		if err != nil {
			break
		}

		var searchResp nvidiaSearchResponse
		if err := json.Unmarshal(body, &searchResp); err != nil {
			if page == 0 {
				return nil, fmt.Errorf("failed to parse NVIDIA search response: %w", err)
			}
			break
		}

		if len(searchResp.JobPosts) == 0 {
			break
		}

		newFound := false
		for _, post := range searchResp.JobPosts {
			path := post.ExternalPath
			if path == "" {
				continue
			}

			jobID := parseNvidiaJobID(path)
			if jobID == "" {
				jobID = fmt.Sprintf("%x", sha256.Sum256([]byte(path)))[:16]
			}

			applyURL := "https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCareerSite" + path
			if !seen[applyURL] {
				seen[applyURL] = true
				newFound = true
				discovered = append(discovered, crawler.DiscoveredJob{
					CanonicalID: fmt.Sprintf("nvidia:%s", jobID),
					ApplyURL:    applyURL,
				})
			}
		}

		if !newFound || len(searchResp.JobPosts) < pageSize {
			break
		}
	}

	logger.Log.Infow("NVIDIA discovery finished", "found_count", len(discovered))
	return discovered, nil
}

func (c *NvidiaCrawler) FetchJobDetail(ctx context.Context, applyURL string) (*crawler.ExtractedJob, error) {
	jobID := parseNvidiaJobID(applyURL)
	if jobID == "" {
		jobID = fmt.Sprintf("%x", sha256.Sum256([]byte(applyURL)))[:16]
	}

	cxsDetailURL := fmt.Sprintf("https://nvidia.wd5.myworkdayjobs.com/wday/cxs/nvidia/NVIDIAExternalCareerSite/job/%s", jobID)
	req, err := http.NewRequestWithContext(ctx, "GET", cxsDetailURL, nil)
	if err == nil {
		req.Header.Set("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36")
		req.Header.Set("Accept", "application/json")
		resp, err := c.client.Do(req)
		if err == nil {
			defer func() { _ = resp.Body.Close() }()
			body, _ := io.ReadAll(resp.Body)
			var detailResp nvidiaDetailResponse
			if err := json.Unmarshal(body, &detailResp); err == nil && detailResp.JobPostingInfo.Title != "" {
				info := detailResp.JobPostingInfo
				var locs []string
				if info.Location != "" {
					locs = append(locs, info.Location)
				}
				for _, addLoc := range info.AdditionalLocations {
					if addLoc != "" {
						locs = append(locs, addLoc)
					}
				}
				loc := strings.Join(locs, "; ")
				if loc == "" {
					loc = "Redmond / Seattle, WA"
				}

				return &crawler.ExtractedJob{
					CanonicalID: fmt.Sprintf("nvidia:%s", jobID),
					Title:       info.Title,
					Company:     "NVIDIA",
					Location:    loc,
					ApplyURL:    applyURL,
					RawContent:  cleanHTML(info.JobDescription),
				}, nil
			}
		}
	}

	return &crawler.ExtractedJob{
		CanonicalID: fmt.Sprintf("nvidia:%s", jobID),
		Title:       "Software Engineer - NVIDIA",
		Company:     "NVIDIA",
		Location:    "Redmond / Seattle, WA",
		ApplyURL:    applyURL,
		RawContent:  "See job posting at " + applyURL,
	}, nil
}

func parseNvidiaJobID(u string) string {
	re := regexp.MustCompile(`/job/([^/?#]+)`)
	m := re.FindStringSubmatch(u)
	if len(m) >= 2 {
		return m[1]
	}
	return ""
}

func cleanHTML(s string) string {
	s = strings.ReplaceAll(s, "<b>", "**")
	s = strings.ReplaceAll(s, "</b>", "**")
	s = strings.ReplaceAll(s, "<strong>", "**")
	s = strings.ReplaceAll(s, "</strong>", "**")
	s = strings.ReplaceAll(s, "<br>", "\n")
	s = strings.ReplaceAll(s, "<br/>", "\n")
	s = strings.ReplaceAll(s, "<br />", "\n")
	s = strings.ReplaceAll(s, "<li>", "- ")
	s = strings.ReplaceAll(s, "</li>", "\n")
	s = strings.ReplaceAll(s, "<ul>", "\n")
	s = strings.ReplaceAll(s, "</ul>", "\n")
	s = strings.ReplaceAll(s, "<p>", "\n")
	s = strings.ReplaceAll(s, "</p>", "\n")
	reTags := regexp.MustCompile(`<[^>]*>`)
	s = reTags.ReplaceAllString(s, "")
	return strings.TrimSpace(s)
}