1 files changed, 79 insertions, 0 deletions
diff --git a/parseURL.go b/parseURL.go
new file mode 100644
index 0000000..e8b133b
--- /dev/null
+++ b/parseURL.go
@@ -0,0 +1,79 @@
+package gitpb
+
+import (
+	"fmt"
+	"net/url"
+	"regexp"
+	"strings"
+)
+
+// scpLikeURLRegex is a regex to parse scp-like URLs, e.g., "[email protected]:user/repo.git".
+// It captures the user, host, and path.
+var scpLikeURLRegex = regexp.MustCompile(`^([^@]+)@([^:]+):(.+)$`)
+
+// ParseGitURL parses a Git URL, handling both standard schemes (http, https, ssh)
+// and the scp-like syntax (e.g., [email protected]:user/repo.git).
+func ParseGitURL(rawURL string) (*url.URL, error) {
+	// First, try to parse it as a standard URL.
+	// This will handle http://, https://, ssh://, git://, etc.
+	parsedURL, err := url.Parse(rawURL)
+	if err == nil && parsedURL.Scheme != "" {
+		return parsedURL, nil
+	}
+
+	// If parsing failed or the scheme is missing, it might be an scp-like URL.
+	// Try to match it against our regex.
+	matches := scpLikeURLRegex.FindStringSubmatch(rawURL)
+
+	if len(matches) != 4 {
+		// If it doesn't match, we return the original parsing error or a new one.
+		if err != nil {
+			return nil, err
+		}
+		return nil, fmt.Errorf("invalid or unsupported URL format: %s", rawURL)
+	}
+
+	// If it matches, we manually construct a url.URL struct.
+	// We'll use the "ssh" scheme to represent this connection type.
+	user := matches[1]
+	host := matches[2]
+	path := matches[3]
+
+	// The path might have a ".git" suffix, which we can optionally remove.
+	path = strings.TrimSuffix(path, ".git")
+
+	return &url.URL{
+		Scheme: "ssh",
+		User:   url.User(user),
+		Host:   host,
+		Path:   "/" + path, // Standard URL paths start with a slash.
+	}, nil
+}
+
+/*
+func parsemain() {
+	urlsToTest := []string{
+		// The problematic scp-like syntax
+		"[email protected]:jcarr/jcarr.git",
+		// A standard HTTPS URL
+		"https://gitea.wit.com/wit/bash.git",
+		// A standard SSH URL
+		"ssh://[email protected]/golang/go.git",
+		// An invalid URL
+		"this is not a url",
+	}
+
+	for _, u := range urlsToTest {
+		fmt.Printf("Parsing URL: %s\n", u)
+		parsed, err := ParseGitURL(u)
+		if err != nil {
+			fmt.Printf("  Error: %v\n\n", err)
+			continue
+		}
+		fmt.Printf("  Scheme: %s\n", parsed.Scheme)
+		fmt.Printf("  User:   %s\n", parsed.User.Username())
+		fmt.Printf("  Host:   %s\n", parsed.Host)
+		fmt.Printf("  Path:   %s\n\n", parsed.Path)
+	}
+}
+*/