Wiki pipeline audit

Changes Made

- page-index.json now includes generated meta/status pages, not just entity pages.
- Wiki page count now comes from the final page index, so wiki pages: matches generated HTML files.
- Page IDs now normalize /, \, and _ to : to prevent duplicate index entries for slash-bearing keys.
- Added page-index validation for duplicate page_id and duplicate output_path.
This commit is contained in:
2026-05-15 09:41:54 +02:00
parent 9b2084344d
commit 505afb5ac5
2 changed files with 128 additions and 9 deletions
+71 -3
View File
@@ -148,7 +148,6 @@ func buildWiki(p *project.Project, nativeResult BuildResult, force bool, progres
pageTitles := map[string]string{}
pageStatuses := map[string]string{}
pageIndex := wikiPageIndex{Version: wikiGeneratorVersion}
pageCount := 0
writePage := func(pageID, content, title, status string) error {
path := filepath.Join(outputDir, wikiPageIDToRelPath(pageID))
@@ -177,7 +176,6 @@ func buildWiki(p *project.Project, nativeResult BuildResult, force bool, progres
EditPolicy: wikiEditPolicy(namespace),
StalePolicy: p.EffectiveConfig().TopData.Wiki.StalePages.Default,
})
pageCount++
return nil
}
@@ -187,6 +185,9 @@ func buildWiki(p *project.Project, nativeResult BuildResult, force bool, progres
if err := generateStatusPages(outputDir, pageStatuses, pageTitles, p.EffectiveConfig().TopData.Wiki.ManagedNamespaces, loadWikiManualSections(p)); err != nil {
return wikiResult{}, err
}
if err := indexUnregisteredWikiPages(rootDir, outputDir, p.EffectiveConfig().TopData.Wiki.StalePages.Default, &pageIndex); err != nil {
return wikiResult{}, err
}
if err := saveWikiPageIndex(filepath.Join(rootDir, "page-index.json"), pageIndex); err != nil {
return wikiResult{}, err
}
@@ -194,6 +195,7 @@ func buildWiki(p *project.Project, nativeResult BuildResult, force bool, progres
if err := os.MkdirAll(rootDir, 0o755); err != nil {
return wikiResult{}, fmt.Errorf("create wiki root dir: %w", err)
}
pageCount := len(pageIndex.Pages)
state = wikiStateDocument{Version: wikiGeneratorVersion, Digest: digest, Pages: pageCount}
if err := saveWikiState(statePath, state); err != nil {
return wikiResult{}, err
@@ -243,6 +245,9 @@ func loadWikiManualSections(p *project.Project) []wikiManualSection {
}
func saveWikiPageIndex(path string, index wikiPageIndex) error {
if err := validateWikiPageIndex(index); err != nil {
return err
}
slices.SortFunc(index.Pages, func(a, b wikiPageIndexEntry) int {
return strings.Compare(a.PageID, b.PageID)
})
@@ -256,6 +261,68 @@ func saveWikiPageIndex(path string, index wikiPageIndex) error {
return os.WriteFile(path, append(raw, '\n'), 0o644)
}
func validateWikiPageIndex(index wikiPageIndex) error {
pageIDs := map[string]string{}
outputPaths := map[string]string{}
for _, entry := range index.Pages {
if previous, ok := pageIDs[entry.PageID]; ok {
return fmt.Errorf("duplicate wiki page-index page_id %q for %s and %s", entry.PageID, previous, entry.OutputPath)
}
if previous, ok := outputPaths[entry.OutputPath]; ok {
return fmt.Errorf("duplicate wiki page-index output_path %q for %s and %s", entry.OutputPath, previous, entry.PageID)
}
pageIDs[entry.PageID] = entry.OutputPath
outputPaths[entry.OutputPath] = entry.PageID
}
return nil
}
func indexUnregisteredWikiPages(rootDir, outputDir, stalePolicy string, index *wikiPageIndex) error {
seen := map[string]struct{}{}
for _, entry := range index.Pages {
seen[entry.PageID] = struct{}{}
}
return filepath.WalkDir(outputDir, func(path string, d fs.DirEntry, err error) error {
if err != nil {
return err
}
if d.IsDir() || !strings.EqualFold(filepath.Ext(path), ".html") {
return nil
}
relToPages, err := filepath.Rel(outputDir, path)
if err != nil {
return err
}
pageID := strings.TrimSuffix(filepath.ToSlash(relToPages), filepath.Ext(relToPages))
pageID = strings.ReplaceAll(pageID, "/", ":")
if _, ok := seen[pageID]; ok {
return nil
}
raw, err := os.ReadFile(path)
if err != nil {
return err
}
namespace := strings.SplitN(pageID, ":", 2)[0]
relToRoot, err := filepath.Rel(rootDir, path)
if err != nil {
relToRoot = filepath.Base(path)
}
index.Pages = append(index.Pages, wikiPageIndexEntry{
PageID: pageID,
Namespace: namespace,
SourceDataset: categoryFromWikiNamespace(namespace),
SourceKey: pageID,
Title: extractHTMLTitle(string(raw), pageID),
Hash: computeManagedHash(string(raw)),
OutputPath: filepath.ToSlash(relToRoot),
EditPolicy: wikiEditPolicy(namespace),
StalePolicy: stalePolicy,
})
seen[pageID] = struct{}{}
return nil
})
}
func loadWikiState(path string) (wikiStateDocument, error) {
raw, err := os.ReadFile(path)
if err != nil {
@@ -1482,7 +1549,8 @@ func wikiPageIDForKey(key string) string {
if namespace == "" {
return ""
}
return namespace + ":" + strings.ReplaceAll(value, "_", ":")
value = strings.NewReplacer("/", ":", "\\", ":", "_", ":").Replace(value)
return namespace + ":" + value
}
func wikiPageIDToRelPath(pageID string) string {