feat: SlickVPN Support (#961)

- `internal/updater/html` package
- Add unit tests for slickvpn updating code
- Change shared html package to be more share-able
- Split html utilities in multiple files
- Fix processing .ovpn files with prefix space

Authored by @Rohaq 
Co-authored-by: Quentin McGaw <quentin.mcgaw@gmail.com>

internal/provider/slickvpn/updater/website.go

@@ -0,0 +1,181 @@
+package updater
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"net/http"
+	"regexp"
+
+	htmlutils "github.com/qdm12/gluetun/internal/updater/html"
+	"golang.org/x/net/html"
+)
+
+func fetchServers(ctx context.Context, client *http.Client) (
+	hostToData map[string]serverData, err error) {
+	rootNode, err := fetchHTML(ctx, client)
+	if err != nil {
+		return nil, fmt.Errorf("fetching HTML code: %w", err)
+	}
+
+	hostToData, err = parseHTML(rootNode)
+	if err != nil {
+		return nil, fmt.Errorf("parsing HTML: %w", err)
+	}
+
+	return hostToData, nil
+}
+
+var ErrHTTPStatusCode = errors.New("HTTP status code is not OK")
+
+func fetchHTML(ctx context.Context, client *http.Client) (rootNode *html.Node, err error) {
+	const url = "https://www.slickvpn.com/locations/"
+	request, err := http.NewRequestWithContext(ctx, http.MethodGet, url, nil)
+	if err != nil {
+		return nil, err
+	}
+
+	response, err := client.Do(request)
+	if err != nil {
+		return nil, err
+	}
+
+	if response.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("%w: %d %s",
+			ErrHTTPStatusCode, response.StatusCode, response.Status)
+	}
+
+	rootNode, err = html.Parse(response.Body)
+	if err != nil {
+		_ = response.Body.Close()
+		return nil, fmt.Errorf("parsing HTML code: %w", err)
+	}
+
+	err = response.Body.Close()
+	if err != nil {
+		return nil, fmt.Errorf("closing response body: %w", err)
+	}
+
+	return rootNode, nil
+}
+
+type serverData struct {
+	ovpnURL string
+	country string
+	region  string
+	city    string
+}
+
+var (
+	ErrLocationTableNotFound = errors.New("HTML location table node not found")
+	ErrTbodyNotFound         = errors.New("HTML tbody node not found")
+	ErrExtractOpenVPNURL     = errors.New("failed extracting OpenVPN URL")
+)
+
+func parseHTML(rootNode *html.Node) (hostToData map[string]serverData, err error) {
+	locationTableNode := htmlutils.BFS(rootNode, matchLocationTable)
+	if locationTableNode == nil {
+		return nil, htmlutils.WrapError(ErrLocationTableNotFound, rootNode)
+	}
+
+	tBodyNode := htmlutils.BFS(locationTableNode, matchTbody)
+	if tBodyNode == nil {
+		return nil, htmlutils.WrapError(ErrTbodyNotFound, rootNode)
+	}
+
+	rowNodes := htmlutils.DirectChildren(tBodyNode, matchTr)
+	hostToData = make(map[string]serverData, len(rowNodes))
+
+	for _, rowNode := range rowNodes {
+		hostname, data, err := parseRowNode(rowNode)
+		if err != nil {
+			return nil, fmt.Errorf("parsing row node: %w", err)
+		}
+		hostToData[hostname] = data
+	}
+
+	return hostToData, nil
+}
+
+func parseRowNode(rowNode *html.Node) (hostname string, data serverData, err error) {
+	columnIndex := 0
+	const (
+		columnIndexContinent = 0
+		columnIndexCountry   = 1
+		columnIndexCity      = 2
+		columnIndexConfig    = 3
+	)
+	for cellNode := rowNode.FirstChild; cellNode != nil; cellNode = cellNode.NextSibling {
+		if cellNode.FirstChild == nil {
+			continue
+		}
+
+		switch columnIndex {
+		case columnIndexContinent:
+			data.region = cellNode.FirstChild.Data
+		case columnIndexCountry:
+			data.country = cellNode.FirstChild.Data
+		case columnIndexCity:
+			data.city = cellNode.FirstChild.Data
+		case columnIndexConfig:
+			linkNodes := htmlutils.DirectChildren(cellNode, matchA)
+			for _, linkNode := range linkNodes {
+				if linkNode.FirstChild.Data != "OpenVPN" {
+					continue
+				}
+
+				data.ovpnURL = htmlutils.Attribute(linkNode, "href")
+				if data.ovpnURL == "" {
+					return "", data, htmlutils.WrapError(ErrExtractOpenVPNURL, linkNode)
+				}
+
+				hostname, err = extractHostnameFromURL(data.ovpnURL)
+				if err != nil {
+					return "", data, fmt.Errorf("extracting hostname from url: %w", err)
+				}
+
+				break
+			}
+		}
+
+		columnIndex++
+		if columnIndex == columnIndexConfig+1 {
+			break
+		}
+	}
+
+	return hostname, data, nil
+}
+
+func matchLocationTable(rootNode *html.Node) (match bool) {
+	return htmlutils.MatchID("location-table")(rootNode)
+}
+
+func matchTbody(locationTableNode *html.Node) (match bool) {
+	return htmlutils.MatchData("tbody")(locationTableNode)
+}
+
+func matchTr(tbodyNode *html.Node) (match bool) {
+	return htmlutils.MatchData("tr")(tbodyNode)
+}
+
+func matchA(cellNode *html.Node) (match bool) {
+	return htmlutils.MatchData("a")(cellNode)
+}
+
+var serverNameRegex = regexp.MustCompile(`^.+\/(?P<serverName>.+)\.ovpn$`)
+
+var (
+	ErrExtractHostnameFromURL = errors.New("cannot extract hostname from url")
+)
+
+func extractHostnameFromURL(url string) (hostname string, err error) {
+	matches := serverNameRegex.FindStringSubmatch(url)
+	const minMatches = 2
+	if len(matches) < minMatches {
+		return "", fmt.Errorf("%w: %s has less than 2 matches for %s",
+			ErrExtractHostnameFromURL, url, serverNameRegex)
+	}
+	hostname = matches[1]
+	return hostname, nil
+}