[
https://issues.apache.org/jira/browse/TIKA-4943?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18125631#comment-18125631
]
ASF GitHub Bot commented on TIKA-4943:
--------------------------------------
henry-lindeman-glean commented on code in PR #3277:
URL: https://github.com/apache/tika/pull/3277#discussion_r4232796206
##########
tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-microsoft-module/src/main/java/org/apache/tika/parser/microsoft/onenote/fsshttpb/MSOneStorePackage.java:
##########
@@ -622,25 +627,471 @@ private void walkObject(RevisionStoreObject object,
}
int objectType = jcidIndex(object);
recordEntityGuid(object, objectType);
- // The title structure of a page (StructureElementChildNodes) appears
above the page
- // body on screen, but is declared after the body child nodes. Emit it
first so the
- // text comes out in visual order.
- for (PropertyAction action : actions) {
- if (action.oneNotePropertyEnum ==
OneNotePropertyEnum.StructureElementChildNodes) {
- processAction(action, objectsById, visited, authorRole,
options, metadata, xhtml,
- depth, resourceInfo);
+ ListStyle listStyle = listStyleForObject(object, actions, objectsById);
+ startObjectStructure(objectType, listStyle, xhtml);
+ try {
+ // Emit the title structure before page body nodes to preserve
visual order.
+ List<PropertyAction> structureActions = new ArrayList<>();
+ for (PropertyAction action : actions) {
+ if (action.oneNotePropertyEnum ==
OneNotePropertyEnum.StructureElementChildNodes) {
+ structureActions.add(action);
+ }
}
+ processStructureChildActions(structureActions, objectsById,
visited, authorRole,
+ options, metadata, xhtml, depth, resourceInfo);
+ List<PropertyAction> remainingActions = new ArrayList<>();
+ for (PropertyAction action : actions) {
+ if (hasPrimaryPicture &&
+ action.oneNotePropertyEnum ==
OneNotePropertyEnum.WebPictureContainer14) {
+ continue;
+ }
+ if (action.oneNotePropertyEnum !=
OneNotePropertyEnum.StructureElementChildNodes) {
+ remainingActions.add(action);
+ }
+ }
+ processStructureChildActions(remainingActions, objectsById,
visited, authorRole,
+ options, metadata, xhtml, depth, resourceInfo);
+ } finally {
+ endObjectStructure(objectType, listStyle, xhtml);
}
- for (PropertyAction action : actions) {
- if (hasPrimaryPicture &&
- action.oneNotePropertyEnum ==
OneNotePropertyEnum.WebPictureContainer14) {
+ }
+
+ private void startObjectStructure(int objectType, ListStyle listStyle,
+ XHTMLContentHandler xhtml) throws
SAXException {
+ switch (objectType) {
+ case OneNoteStructureJcid.TABLE_NODE:
+ xhtml.startElement("table");
+ break;
+ case OneNoteStructureJcid.TABLE_ROW_NODE:
+ xhtml.startElement("tr");
+ break;
+ case OneNoteStructureJcid.TABLE_CELL_NODE:
+ xhtml.startElement("td");
+ break;
+ case OneNoteStructureJcid.OUTLINE_ELEMENT_NODE:
+ if (listStyle != null) {
+ if (listStyle.restartValue == null) {
+ xhtml.startElement("li");
+ } else {
+ xhtml.startElement("li", "value",
listStyle.restartValue.toString());
+ }
+ }
+ break;
+ default:
+ break;
+ }
+ }
+
+ private void endObjectStructure(int objectType, ListStyle listStyle,
+ XHTMLContentHandler xhtml) throws
SAXException {
+ switch (objectType) {
+ case OneNoteStructureJcid.TABLE_NODE:
+ xhtml.endElement("table");
+ break;
+ case OneNoteStructureJcid.TABLE_ROW_NODE:
+ xhtml.endElement("tr");
+ break;
+ case OneNoteStructureJcid.TABLE_CELL_NODE:
+ xhtml.endElement("td");
+ break;
+ case OneNoteStructureJcid.OUTLINE_ELEMENT_NODE:
+ if (listStyle != null) {
+ xhtml.endElement("li");
+ }
+ break;
+ default:
+ break;
+ }
+ }
+
+ private void walkObjectGroupRoots(List<RevisionStoreObjectGroup>
objectGroups,
+ Map<ExGuid, RevisionStoreObject>
objectsById,
+ Set<ExGuid> visited,
OneNoteTreeWalkerOptions options,
+ Metadata metadata, XHTMLContentHandler
xhtml)
+ throws SAXException, TikaException, IOException {
+ Set<ExGuid> structureDescendantIds =
collectStructureDescendantIds(objectsById);
+ List<RevisionStoreObject> deferredObjects = new ArrayList<>();
+ for (RevisionStoreObjectGroup objectGroup : objectGroups) {
+ if (objectGroup == null || objectGroup.objects == null) {
continue;
}
- if (action.oneNotePropertyEnum !=
OneNotePropertyEnum.StructureElementChildNodes) {
+ for (RevisionStoreObject object : objectGroup.objects) {
+ if (object != null && object.objectID != null &&
+ structureDescendantIds.contains(object.objectID)) {
+ deferredObjects.add(object);
+ continue;
+ }
+ walkRootObject(object, objectsById, visited, options,
metadata, xhtml);
+ }
+ }
+ // Cyclic owners may have no independent root; start tables before
list containers.
+ for (RevisionStoreObject object : deferredObjects) {
+ if (!visited.contains(object.objectID) &&
+ jcidIndex(object) == OneNoteStructureJcid.TABLE_NODE) {
+ walkRootObject(object, objectsById, visited, options,
metadata, xhtml);
+ }
+ }
+ for (RevisionStoreObject object : deferredObjects) {
+ if (!visited.contains(object.objectID) && isListContainer(object,
objectsById)) {
+ walkRootObject(object, objectsById, visited, options,
metadata, xhtml);
+ }
+ }
+ // Recover only depth-skipped objects, not intentionally suppressed
renditions.
+ while (!depthSkippedObjects.isEmpty()) {
+ Map.Entry<RevisionStoreObject, Map<AuthorRole,
EmbeddedResourceInfo>> skipped =
+ depthSkippedObjects.entrySet().iterator().next();
+ depthSkippedObjects.remove(skipped.getKey());
+ for (Map.Entry<AuthorRole, EmbeddedResourceInfo> reference :
skipped.getValue().entrySet()) {
+ walkRootObject(skipped.getKey(), objectsById, visited,
reference.getKey(), options,
+ metadata, xhtml, reference.getValue());
+ }
+ }
+ }
+
+ private boolean isListContainer(RevisionStoreObject object,
+ Map<ExGuid, RevisionStoreObject>
objectsById) {
+ int type = jcidIndex(object);
+ if (type == OneNoteStructureJcid.OUTLINE_NODE ||
+ type == OneNoteStructureJcid.OUTLINE_ELEMENT_NODE) {
+ return true;
+ }
+ if (object == null || object.propertySet == null ||
+ object.propertySet.objectSpaceObjectPropSet == null) {
+ return false;
+ }
+ for (PropertyAction action : collectObjectActions(object)) {
+ if (action.isChildReference &&
+ jcidIndex(objectsById.get(action.childReference)) ==
+ OneNoteStructureJcid.OUTLINE_ELEMENT_NODE) {
+ return true;
+ }
+ }
+ return false;
+ }
+
+ private Set<ExGuid> collectStructureDescendantIds(
+ Map<ExGuid, RevisionStoreObject> objectsById) {
+ Set<ExGuid> descendants = new HashSet<>();
+ Set<ExGuid> traversed = new HashSet<>();
+ List<RevisionStoreObject> pending = new ArrayList<>();
+ for (RevisionStoreObject object : objectsById.values()) {
+ if (jcidIndex(object) == OneNoteStructureJcid.TABLE_NODE ||
+ isListContainer(object, objectsById)) {
+ traversed.add(object.objectID);
+ pending.add(object);
+ }
+ }
+ for (int index = 0; index < pending.size(); index++) {
+ RevisionStoreObject object = pending.get(index);
+ if (object.propertySet == null ||
+ object.propertySet.objectSpaceObjectPropSet == null) {
+ continue;
+ }
+ for (PropertyAction action : collectObjectActions(object)) {
+ if (!action.isChildReference || action.childReference == null)
{
+ continue;
+ }
+ ExGuid childId = action.childReference;
+ descendants.add(childId);
+ RevisionStoreObject child = objectsById.get(childId);
+ if (child != null && traversed.add(childId)) {
+ pending.add(child);
+ }
+ }
+ }
+ return descendants;
+ }
+
+ private void walkRootObject(RevisionStoreObject object,
+ Map<ExGuid, RevisionStoreObject> objectsById,
Set<ExGuid> visited,
+ OneNoteTreeWalkerOptions options, Metadata
metadata,
+ XHTMLContentHandler xhtml)
+ throws SAXException, TikaException, IOException {
+ walkRootObject(object, objectsById, visited, AuthorRole.NONE, options,
metadata, xhtml, null);
+ }
+
+ private void walkRootObject(RevisionStoreObject object,
+ Map<ExGuid, RevisionStoreObject> objectsById,
Set<ExGuid> visited,
+ AuthorRole authorRole,
OneNoteTreeWalkerOptions options, Metadata metadata,
+ XHTMLContentHandler xhtml,
EmbeddedResourceInfo resourceInfo)
+ throws SAXException, TikaException, IOException {
+ if (!canWalkObject(object, visited, 0)) {
+ walkObject(object, objectsById, visited, authorRole, options,
metadata, xhtml, 0, resourceInfo);
+ return;
+ }
Review Comment:
here's the first three lines of walkObject:
```java
private void walkObject(RevisionStoreObject object,
Map<ExGuid, RevisionStoreObject> objectsById,
Set<ExGuid> visited,
AuthorRole authorRole, OneNoteTreeWalkerOptions
options,
Metadata metadata, XHTMLContentHandler xhtml,
int depth,
EmbeddedResourceInfo inheritedResourceInfo)
throws SAXException, TikaException, IOException {
if (object == null) {
return;
}
```
> Emit onenote lists and tables as html lists and tables
> ------------------------------------------------------
>
> Key: TIKA-4943
> URL: https://issues.apache.org/jira/browse/TIKA-4943
> Project: Tika
> Issue Type: Improvement
> Components: parser
> Reporter: Henry Lindeman
> Priority: Minor
>
> currently table cells and list items are emitted in p tags
> instead use <tr>/<td> and <ul>/<li>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
