feat: improve gitlab/github readme urlurls

Signed-off-by: Carlos A Becker <caarlos0@users.noreply.github.com>

5 files changed, 134 insertions, 59 deletions

diff --git a/github.go b/github.go
index d3d2084..6b169fa 100644
--- a/github.go
+++ b/github.go
@@ -3,40 +3,29 @@ package main
 import (
 	"encoding/json"
 	"errors"
+	"fmt"
 	"io"
 	"net/http"
 	"net/url"
 	"strings"
 )
 
-// isGitHubURL tests a string to determine if it is a well-structured GitHub URL.
-func isGitHubURL(s string) (string, bool) {
-	if strings.HasPrefix(s, "github.com/") {
-		s = "https://" + s
-	}
-
-	u, err := url.ParseRequestURI(s)
-	if err != nil {
-		return "", false
-	}
-
-	return u.String(), strings.ToLower(u.Host) == "github.com"
-}
-
 // findGitHubREADME tries to find the correct README filename in a repository using GitHub API.
-func findGitHubREADME(s string) (*source, error) {
-	sSplit := strings.Split(s, "/")
-	owner, repo := sSplit[3], sSplit[4]
+func findGitHubREADME(u *url.URL) (*source, error) {
+	owner, repo, ok := strings.Cut(strings.TrimPrefix(u.Path, "/"), "/")
+	if !ok {
+		return nil, fmt.Errorf("invalid url: %s", u.String())
+	}
 
 	type readme struct {
 		DownloadURL string `json:"download_url"`
 	}
 
-	apiURL := "https://api.github.com/repos/" + owner + "/" + repo + "/readme"
+	apiURL := fmt.Sprintf("https://api.%s/repos/%s/%s/readme", u.Hostname(), owner, repo)
 
 	// nolint:bodyclose
 	// it is closed on the caller
-	res, err := http.Get(apiURL)
+	res, err := http.Get(apiURL) // nolint: gosec
 	if err != nil {
 		return nil, err
 	}
@@ -47,8 +36,7 @@ func findGitHubREADME(s string) (*source, error) {
 	}
 
 	var result readme
-	jsonErr := json.Unmarshal(body, &result)
-	if jsonErr != nil {
+	if err := json.Unmarshal(body, &result); err != nil {
 		return nil, err
 	}
 
diff --git a/gitlab.go b/gitlab.go
index 81fcf84..05e1239 100644
--- a/gitlab.go
+++ b/gitlab.go
@@ -3,42 +3,31 @@ package main
 import (
 	"encoding/json"
 	"errors"
+	"fmt"
 	"io"
 	"net/http"
 	"net/url"
 	"strings"
 )
 
-// isGitLabURL tests a string to determine if it is a well-structured GitLab URL.
-func isGitLabURL(s string) (string, bool) {
-	if strings.HasPrefix(s, "gitlab.com/") {
-		s = "https://" + s
-	}
-
-	u, err := url.ParseRequestURI(s)
-	if err != nil {
-		return "", false
-	}
-
-	return u.String(), strings.ToLower(u.Host) == "gitlab.com"
-}
-
 // findGitLabREADME tries to find the correct README filename in a repository using GitLab API.
-func findGitLabREADME(s string) (*source, error) {
-	sSplit := strings.Split(s, "/")
-	owner, repo := sSplit[3], sSplit[4]
+func findGitLabREADME(u *url.URL) (*source, error) {
+	owner, repo, ok := strings.Cut(strings.TrimPrefix(u.Path, "/"), "/")
+	if !ok {
+		return nil, fmt.Errorf("invalid url: %s", u.String())
+	}
 
 	projectPath := url.QueryEscape(owner + "/" + repo)
 
 	type readme struct {
-		ReadmeUrl string `json:"readme_url"`
+		ReadmeURL string `json:"readme_url"`
 	}
 
-	apiURL := "https://gitlab.com/api/v4/projects/" + projectPath
+	apiURL := fmt.Sprintf("https://%s/api/v4/projects/%s", u.Hostname(), projectPath)
 
 	// nolint:bodyclose
 	// it is closed on the caller
-	res, err := http.Get(apiURL)
+	res, err := http.Get(apiURL) // nolint: gosec
 	if err != nil {
 		return nil, err
 	}
@@ -49,23 +38,22 @@ func findGitLabREADME(s string) (*source, error) {
 	}
 
 	var result readme
-	jsonErr := json.Unmarshal(body, &result)
-	if jsonErr != nil {
+	if err := json.Unmarshal(body, &result); err != nil {
 		return nil, err
 	}
 
-	readmeRawUrl := strings.Replace(result.ReadmeUrl, "blob", "raw", -1)
+	readmeRawURL := strings.Replace(result.ReadmeURL, "blob", "raw", -1)
 
 	if res.StatusCode == http.StatusOK {
 		// nolint:bodyclose
 		// it is closed on the caller
-		resp, err := http.Get(readmeRawUrl)
+		resp, err := http.Get(readmeRawURL) // nolint: gosec
 		if err != nil {
 			return nil, err
 		}
 
 		if resp.StatusCode == http.StatusOK {
-			return &source{resp.Body, readmeRawUrl}, nil
+			return &source{resp.Body, readmeRawURL}, nil
 		}
 	}
 
diff --git a/main.go b/main.go
index 17d66e7..5a42f9b 100644
--- a/main.go
+++ b/main.go
@@ -62,19 +62,9 @@ func sourceFromArg(arg string) (*source, error) {
 	}
 
 	// a GitHub or GitLab URL (even without the protocol):
-	if u, ok := isGitHubURL(arg); ok {
-		src, err := findGitHubREADME(u)
-		if err != nil {
-			return nil, err
-		}
-		return src, nil
-	}
-	if u, ok := isGitLabURL(arg); ok {
-		src, err := findGitLabREADME(u)
-		if err != nil {
-			return nil, err
-		}
-		return src, nil
+	src, err := readmeURL(arg)
+	if src != nil || err != nil {
+		return src, err
 	}
 
 	// HTTP(S) URLs:
diff --git a/url.go b/url.go
new file mode 100644
index 0000000..43ca167
--- /dev/null
+++ b/url.go
@@ -0,0 +1,80 @@
+package main
+
+import (
+	"net/url"
+	"strings"
+	"sync"
+)
+
+const (
+	protoGithub = "github://"
+	protoGitlab = "gitlab://"
+	protoHTTPS  = "https://"
+)
+
+var (
+	githubURL *url.URL
+	gitlabURL *url.URL
+	urlsOnce  sync.Once
+)
+
+func init() {
+	urlsOnce.Do(func() {
+		githubURL, _ = url.Parse("https://github.com")
+		gitlabURL, _ = url.Parse("https://gitlab.com")
+	})
+}
+
+func readmeURL(path string) (*source, error) {
+	switch {
+	case strings.HasPrefix(path, protoGithub):
+		if u := githubReadmeURL(path); u != nil {
+			return readmeURL(u.String())
+		}
+		return nil, nil
+	case strings.HasPrefix(path, protoGitlab):
+		if u := gitlabReadmeURL(path); u != nil {
+			return readmeURL(u.String())
+		}
+		return nil, nil
+	}
+
+	if !strings.HasPrefix(path, protoHTTPS) {
+		path = protoHTTPS + path
+	}
+	u, err := url.Parse(path)
+	if err != nil {
+		return nil, err
+	}
+
+	switch {
+	case u.Hostname() == githubURL.Hostname():
+		return findGitHubREADME(u)
+	case u.Hostname() == gitlabURL.Hostname():
+		return findGitLabREADME(u)
+	}
+
+	return nil, nil
+}
+
+func githubReadmeURL(path string) *url.URL {
+	path = strings.TrimPrefix(path, protoGithub)
+	parts := strings.Split(path, "/")
+	if len(parts) != 2 {
+		// custom hostnames are not supported yet
+		return nil
+	}
+	u, _ := url.Parse(githubURL.String())
+	return u.JoinPath(path)
+}
+
+func gitlabReadmeURL(path string) *url.URL {
+	path = strings.TrimPrefix(path, protoGitlab)
+	parts := strings.Split(path, "/")
+	if len(parts) != 2 {
+		// custom hostnames are not supported yet
+		return nil
+	}
+	u, _ := url.Parse(gitlabURL.String())
+	return u.JoinPath(path)
+}
diff --git a/url_test.go b/url_test.go
new file mode 100644
index 0000000..02d1547
--- /dev/null
+++ b/url_test.go
@@ -0,0 +1,29 @@
+package main
+
+import "testing"
+
+func TestURLParser(t *testing.T) {
+	for path, url := range map[string]string{
+		"github.com/charmbracelet/glow":             "https://raw.githubusercontent.com/charmbracelet/glow/master/README.md",
+		"github://charmbracelet/glow":               "https://raw.githubusercontent.com/charmbracelet/glow/master/README.md",
+		"github://caarlos0/dotfiles.fish":           "https://raw.githubusercontent.com/caarlos0/dotfiles.fish/main/README.md",
+		"github://tj/git-extras":                    "https://raw.githubusercontent.com/tj/git-extras/master/Readme.md",
+		"https://github.com/goreleaser/nfpm":        "https://raw.githubusercontent.com/goreleaser/nfpm/main/README.md",
+		"gitlab.com/caarlos0/test":                  "https://gitlab.com/caarlos0/test/-/raw/master/README.md",
+		"gitlab://caarlos0/test":                    "https://gitlab.com/caarlos0/test/-/raw/master/README.md",
+		"https://gitlab.com/terrakok/gitlab-client": "https://gitlab.com/terrakok/gitlab-client/-/raw/develop/Readme.md",
+	} {
+		t.Run(path, func(t *testing.T) {
+			got, err := readmeURL(path)
+			if err != nil {
+				t.Fatalf("expected no error, got %v", err)
+			}
+			if got == nil {
+				t.Fatalf("should not be nil")
+			}
+			if url != got.URL {
+				t.Errorf("expected url for %s to be %s, was %s", path, url, got.URL)
+			}
+		})
+	}
+}