all: move checking REUSE annotation after all files listed

Use case: on the first scan, the file result in group unknown.
User then modify the spdxconv.cfg to add or update the match-file-pattern.
The next scan should check again the files in unknown group, in case
its match with updated config.

5 files changed, 30 insertions, 32 deletions

diff --git a/report.go b/report.go
index 64720b3..6a38f8d 100644
--- a/report.go
+++ b/report.go
@@ -109,8 +109,10 @@ func loadReport() (rep *report, err error) {
 			f.group = groupBinary
 			rep.listBinary[f.path] = f
 		case reportGroupUnknown:
-			f.group = groupUnknown
-			rep.listUnknown[f.path] = f
+			// Don't store it, we may need to rescan it later.
+			// Use case: the first scan result in unknown, user
+			// then modify the match-file-pattern and re-scan
+			// again.
 		case reportGroupDone:
 			rep.listDone[f.path] = f
 		}
@@ -124,21 +126,28 @@ func loadReport() (rep *report, err error) {
 
 func (rep *report) scan(conv *SPDXConv, listFile []string) (err error) {
 	var logp = `report.scan`
-	var ok bool
-	for _, file := range listFile {
-		if rep.hasScanned(file) {
+	for _, path := range listFile {
+		if rep.hasScanned(path) {
 			continue
 		}
-		_, ok = rep.listDone[file]
-		if ok {
+		// User may update the REUSE.toml on next scan.
+		if conv.reuse != nil && conv.reuse.isAnnotated(path) {
+			rep.listDone[path] = &file{
+				path: path,
+			}
 			continue
 		}
-
-		f, err := newFile(file, conv.cfg.MaxLineMatch)
+		f, err := newFile(path, conv.cfg.MaxLineMatch)
 		if err != nil {
 			return fmt.Errorf(`%s: %w`, logp, err)
 		}
 		if f.group == groupBinary {
+			f.licenseID = valDefault
+			f.copyrightText = valDefault
+			f.getYearFromSCM(conv)
+			if f.copyrightYear == `` {
+				f.copyrightYear = conv.cfg.CopyrightYear
+			}
 			rep.listBinary[f.path] = f
 			continue
 		}
@@ -166,8 +175,8 @@ func (rep *report) scan(conv *SPDXConv, listFile []string) (err error) {
 	return nil
 }
 
-// hasScanned return true if the file is already reported in regular or
-// binary group before.
+// hasScanned return true if the file is already reported in regular, binary,
+// or done group before.
 func (rep *report) hasScanned(path string) bool {
 	var ok bool
 	_, ok = rep.listBinary[path]
@@ -178,7 +187,10 @@ func (rep *report) hasScanned(path string) bool {
 	if ok {
 		return true
 	}
-	delete(rep.listUnknown, path)
+	_, ok = rep.listDone[path]
+	if ok {
+		return true
+	}
 	return false
 }
 
diff --git a/report_test.go b/report_test.go
index fc8ea65..e77901d 100644
--- a/report_test.go
+++ b/report_test.go
@@ -53,15 +53,8 @@ func TestLoadReport(t *testing.T) {
 				group:         groupBinary,
 			},
 		},
-		listUnknown: map[string]*file{
-			`fileU1`: &file{
-				path:          `fileU1`,
-				licenseID:     valDefault,
-				copyrightText: valDefault,
-				group:         groupUnknown,
-			},
-		},
-		listDone: map[string]*file{},
+		listUnknown: map[string]*file{},
+		listDone:    map[string]*file{},
 	}
 	test.Assert(t, workDir, exp, got)
 }
diff --git a/spdxconv.go b/spdxconv.go
index a5c8545..58f199d 100644
--- a/spdxconv.go
+++ b/spdxconv.go
@@ -331,9 +331,6 @@ func (conv *SPDXConv) scanDir(listDir []string) (listFile []string, err error) {
 			if conv.scm.IsIgnored(relpath) {
 				continue
 			}
-			if conv.reuse != nil && conv.reuse.isAnnotated(relpath) {
-				continue
-			}
 			if de.IsDir() {
 				listDir = append(listDir, relpath)
 				continue
diff --git a/spdxconv_test.go b/spdxconv_test.go
index 6c33645..84b3718 100644
--- a/spdxconv_test.go
+++ b/spdxconv_test.go
@@ -163,6 +163,8 @@ test.json,default,0,2026,default,0,,
 //
 .gitignore,exist,0,,exist,1,#,
 a/b/.gitignore,exist,0,,exist,1,#,
+po/test.po,,0,,,0,,
+po/test.pot,,0,,,0,,
 with_spdx.go,exist,0,,exist,1,//,
 `
 
@@ -360,6 +362,8 @@ func TestSPDXConv_scanDir(t *testing.T) {
 			`go.sum`,
 			`no_copyright_year.md`,
 			`po/included.go`,
+			`po/test.po`,
+			`po/test.pot`,
 			`test.go`,
 			`test.html`,
 			`test.json`,
diff --git a/testdata/Apply_test.txt b/testdata/Apply_test.txt
index b643a3c..a6723e2 100644
--- a/testdata/Apply_test.txt
+++ b/testdata/Apply_test.txt
@@ -76,14 +76,6 @@ delete_line_after = "^(//+|#+|\\*+/|--+>|--+)$"
 //
 //spdxconv:unknown
 //
-with_match_license.txt,,0,,,0,,
-with_match_license_bottom.txt,,0,,,0,,
-with_no_order.txt,,0,,,0,,
-with_spdx.txt,,0,,,0,,
-with_spdx_at_bottom.txt,,0,,,0,,
-with_spdx_license_id_only.txt,,0,,,0,,
-with_spdx_no_order.txt,,0,,,0,,
-without_spdx_license_id.txt,,0,,,0,,
 //
 //spdxconv:done
 //