paragliding/backend/internal/client/faa.go

package client

import (
	"context"
	"fmt"
	"io"
	"net/http"
	"regexp"
	"strings"
	"time"

	"github.com/scottyah/paragliding/internal/model"
)

const (
	faaTFRListURL = "https://tfr.faa.gov/tfr2/list.jsp"
	faaBaseURL    = "https://tfr.faa.gov"
)

// FAAClient fetches TFR data from the FAA website
type FAAClient struct {
	httpClient *http.Client
}

// NewFAAClient creates a new FAA TFR client
func NewFAAClient() *FAAClient {
	return &FAAClient{
		httpClient: &http.Client{
			Timeout: 15 * time.Second,
		},
	}
}

var (
	anchorRe  = regexp.MustCompile(`(?is)<a\s+([^>]*)>(.*?)</a>`)
	hrefRe    = regexp.MustCompile(`(?i)href="([^"]*)"`)
	dateRe    = regexp.MustCompile(`^\d{2}/\d{2}/\d{4}$`)
	htmlTagRe = regexp.MustCompile(`<[^>]*>`)
)

type anchor struct {
	href string
	text string
}

// FetchTFRs retrieves the list of active TFRs from the FAA website
func (c *FAAClient) FetchTFRs(ctx context.Context) ([]model.TFR, error) {
	req, err := http.NewRequestWithContext(ctx, "GET", faaTFRListURL, nil)
	if err != nil {
		return nil, fmt.Errorf("create request: %w", err)
	}
	req.Header.Set("User-Agent", "ParaglidingWeatherApp/1.0")

	resp, err := c.httpClient.Do(req)
	if err != nil {
		return nil, fmt.Errorf("fetch TFR list: %w", err)
	}
	defer resp.Body.Close()

	if resp.StatusCode != http.StatusOK {
		return nil, fmt.Errorf("TFR list returned status %d", resp.StatusCode)
	}

	body, err := io.ReadAll(resp.Body)
	if err != nil {
		return nil, fmt.Errorf("read response: %w", err)
	}

	return parseTFRList(string(body)), nil
}

// extractAnchors finds all <a> tags in HTML and returns their href and text content
func extractAnchors(html string) []anchor {
	matches := anchorRe.FindAllStringSubmatch(html, -1)
	anchors := make([]anchor, 0, len(matches))
	for _, m := range matches {
		text := htmlTagRe.ReplaceAllString(m[2], "")
		text = strings.TrimSpace(text)
		a := anchor{text: text}
		if hrefMatch := hrefRe.FindStringSubmatch(m[1]); hrefMatch != nil {
			a.href = hrefMatch[1]
		}
		anchors = append(anchors, a)
	}
	return anchors
}

// parseTFRList extracts TFR entries from the FAA list page HTML.
// The page contains a table where each row has anchor tags for:
// date, notam ID (with detail link), facility, state, type, description, zoom link
func parseTFRList(html string) []model.TFR {
	rows := strings.Split(html, "<tr")
	var tfrs []model.TFR

	for _, row := range rows {
		anchors := extractAnchors(row)
		if len(anchors) < 6 {
			continue
		}

		// First anchor should contain a date (mm/dd/yyyy)
		if !dateRe.MatchString(anchors[0].text) {
			continue
		}

		// Build detail URL from the NOTAM anchor's href
		detailURL := anchors[1].href
		if detailURL != "" {
			detailURL = strings.Replace(detailURL, "..", faaBaseURL, 1)
			detailURL = strings.ReplaceAll(detailURL, "\n", "")
			detailURL = strings.ReplaceAll(detailURL, "\r", "")
		}

		desc := anchors[5].text
		desc = strings.ReplaceAll(desc, "\n", "")
		desc = strings.ReplaceAll(desc, "\r", "")
		// Collapse multiple spaces
		for strings.Contains(desc, "  ") {
			desc = strings.ReplaceAll(desc, "  ", " ")
		}

		tfr := model.TFR{
			Date:        anchors[0].text,
			NotamID:     anchors[1].text,
			Facility:    anchors[2].text,
			State:       anchors[3].text,
			Type:        anchors[4].text,
			Description: strings.TrimSpace(desc),
			DetailURL:   detailURL,
		}

		tfrs = append(tfrs, tfr)
	}

	return tfrs
}