Support programming excluded patterns in gitrepository spec

-- More coming in this commit message soon

Signed-off-by: Erik Hollensbe <github@hollensbe.org>

api/v1alpha1/gitrepository_types.go

@@ -54,6 +54,13 @@ type GitRepositorySpec struct {
 	// Verify OpenPGP signature for the commit that HEAD points to.
 	// +optional
 	Verification *GitRepositoryVerification `json:"verify,omitempty"`
+
+	// SourceIgnore overrides the set of excluded patterns in the .sourceignore
+	// format (which is the same as .gitignore). If not provided, a default will
+	// be used, consult the documentation for your version to find out what those
+	// are.
+	// +optional
+	SourceIgnore *string `json:"sourceIgnore,omitempty"`
 }
 
 // GitRepositoryRef defines the git ref used for pull and checkout operations.

api/v1alpha1/zz_generated.deepcopy.go

@@ -140,6 +140,11 @@ func (in *GitRepositorySpec) DeepCopyInto(out *GitRepositorySpec) {
 		*out = new(GitRepositoryVerification)
 		**out = **in
 	}
+	if in.SourceIgnore != nil {
+		in, out := &in.SourceIgnore, &out.SourceIgnore
+		*out = new(string)
+		**out = **in
+	}
 }
 
 // DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new GitRepositorySpec.

config/crd/bases/source.fluxcd.io_gitrepositories.yaml

@@ -82,6 +82,12 @@ spec:
                     TODO: Add other useful fields. apiVersion, kind, uid?'
                   type: string
               type: object
+            sourceIgnore:
+              description: SourceIgnore overrides the set of excluded patterns in
+                the .sourceignore format (which is the same as .gitignore). If not
+                provided, a default will be used, consult the documentation for your
+                version to find out what those are.
+              type: string
             timeout:
               description: The timeout for remote git operations like cloning, default
                 to 20s.

controllers/gitrepository_controller.go

@@ -198,8 +198,7 @@ func (r *GitRepositoryReconciler) sync(ctx context.Context, repository sourcev1.
 	defer unlock()
 
 	// archive artifact and check integrity
-	err = r.Storage.Archive(artifact, tmpGit)
+	if err := r.Storage.Archive(artifact, tmpGit, repository.Spec); err != nil {
-	if err != nil {
 		err = fmt.Errorf("storage archive error: %w", err)
 		return sourcev1.GitRepositoryNotReady(repository, sourcev1.StorageOperationFailedReason, err.Error()), err
 	}

controllers/storage.go

@@ -19,6 +19,7 @@ package controllers
 import (
 	"archive/tar"
 	"bufio"
+	"bytes"
 	"compress/gzip"
 	"crypto/sha1"
 	"fmt"
@@ -108,7 +109,7 @@ func (s *Storage) RemoveAllButCurrent(artifact sourcev1.Artifact) error {
 	})
 
 	if len(errors) > 0 {
-		return fmt.Errorf("faild to remove files: %s", strings.Join(errors, " "))
+		return fmt.Errorf("failed to remove files: %s", strings.Join(errors, " "))
 	}
 	return nil
 }
@@ -123,15 +124,17 @@ func (s *Storage) ArtifactExist(artifact sourcev1.Artifact) bool {
 
 // Archive creates a tar.gz to the artifact path from the given dir excluding any VCS specific
 // files and directories, or any of the excludes defined in the excludeFiles.
-func (s *Storage) Archive(artifact sourcev1.Artifact, dir string) error {
+// Returns a modified sourcev1.Artifact and any error.
+func (s *Storage) Archive(artifact sourcev1.Artifact, dir string, spec sourcev1.GitRepositorySpec) error {
 	if _, err := os.Stat(dir); err != nil {
 		return err
 	}
 
-	ps, err := loadExcludePatterns(dir)
+	ps, err := loadExcludePatterns(dir, spec)
 	if err != nil {
 		return err
 	}
+
 	matcher := gitignore.NewMatcher(ps)
 
 	gzFile, err := os.Create(artifact.Path)
@@ -241,27 +244,44 @@ func (s *Storage) Lock(artifact sourcev1.Artifact) (unlock func(), err error) {
 	return mutex.Lock()
 }
 
-func loadExcludePatterns(dir string) ([]gitignore.Pattern, error) {
+func getPatterns(reader io.Reader, path []string) []gitignore.Pattern {
+	ps := []gitignore.Pattern{}
+	scanner := bufio.NewScanner(reader)
+
+	for scanner.Scan() {
+		s := scanner.Text()
+		if !strings.HasPrefix(s, "#") && len(strings.TrimSpace(s)) > 0 {
+			ps = append(ps, gitignore.ParsePattern(s, path))
+		}
+	}
+
+	return ps
+}
+
+// loadExcludePatterns loads the excluded patterns from sourceignore or other
+// sources.
+func loadExcludePatterns(dir string, spec sourcev1.GitRepositorySpec) ([]gitignore.Pattern, error) {
 	path := strings.Split(dir, "/")
+
 	var ps []gitignore.Pattern
 	for _, p := range strings.Split(excludeVCS, ",") {
 		ps = append(ps, gitignore.ParsePattern(p, path))
 	}
-	for _, p := range strings.Split(excludeExt, ",") {
-		ps = append(ps, gitignore.ParsePattern(p, path))
-	}
-	if f, err := os.Open(filepath.Join(dir, excludeFile)); err == nil {
-		defer f.Close()
 
-		scanner := bufio.NewScanner(f)
+	if spec.SourceIgnore == nil {
-		for scanner.Scan() {
+		for _, p := range strings.Split(excludeExt, ",") {
-			s := scanner.Text()
+			ps = append(ps, gitignore.ParsePattern(p, path))
-			if !strings.HasPrefix(s, "#") && len(strings.TrimSpace(s)) > 0 {
-				ps = append(ps, gitignore.ParsePattern(s, path))
-			}
 		}
-	} else if !os.IsNotExist(err) {
+
-		return nil, err
+		if f, err := os.Open(filepath.Join(dir, excludeFile)); err == nil {
+			defer f.Close()
+			ps = append(ps, getPatterns(f, path)...)
+		} else if !os.IsNotExist(err) {
+			return nil, err
+		}
+	} else {
+		ps = append(ps, getPatterns(bytes.NewBufferString(*spec.SourceIgnore), path)...)
 	}
+
 	return ps, nil
 }

docs/api/source.md

@@ -157,6 +157,21 @@ GitRepositoryVerification
 <p>Verify OpenPGP signature for the commit that HEAD points to.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>sourceIgnore</code><br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>SourceIgnore overrides the set of excluded patterns in the .sourceignore
+format (which is the same as .gitignore). If not provided, a default will
+be used, consult the documentation for your version to find out what those
+are.</p>
+</td>
+</tr>
 </table>
 </td>
 </tr>
@@ -666,6 +681,21 @@ GitRepositoryVerification
 <p>Verify OpenPGP signature for the commit that HEAD points to.</p>
 </td>
 </tr>
+<tr>
+<td>
+<code>sourceIgnore</code><br>
+<em>
+string
+</em>
+</td>
+<td>
+<em>(Optional)</em>
+<p>SourceIgnore overrides the set of excluded patterns in the .sourceignore
+format (which is the same as .gitignore). If not provided, a default will
+be used, consult the documentation for your version to find out what those
+are.</p>
+</td>
+</tr>
 </tbody>
 </table>
 </div>