QPDFPageObjectHelper: filterPageContents -> filterContents + form XObject

Jay Berkenbilt
1 parent 63ea4619
Showing 5 changed files with 51 additions and 47 deletions
ChangeLog
examples/pdf-count-strings.cc
examples/pdf-filter-tokens.cc
include/qpdf/QPDFPageObjectHelper.hh
libqpdf/QPDFPageObjectHelper.cc
 2020-12-31  Jay Berkenbilt  <ejb@ql.org>
-	* Rename QPDFPageObjectHelper::getPageImages to
-	QPDFPageObjectHelper::getImages and make it support form XObjects
-	as well as pages. The old name will be preserved for
-	compatibility.
+	* Rename some QPDFPageObjectHelper methods and make them support
+	form XObjects as well as pages. The old names will be preserved
+	from compatibility.
+	- getPageImages -> getImages
+	- filterPageContents -> filterContents
 	* Add QPDFObjectHandle::isFormXObject to test whether an object is
 	a form XObject.
 //
 // This example illustrates the use of QPDFObjectHandle::TokenFilter
-// with filterPageContents. See also pdf-filter-tokens.cc for an
-// example that uses QPDFObjectHandle::TokenFilter with
-// addContentTokenFilter.
+// with filterContents. See also pdf-filter-tokens.cc for an example
+// that uses QPDFObjectHandle::TokenFilter with addContentTokenFilter.
 //
 #include <iostream>
@@ -108,14 +107,14 @@ int main(int argc, char* argv[])
             if (pageno % 2)
             {
                 // Ignore output for odd pages.
-                ph.filterPageContents(&counter);
+                ph.filterContents(&counter);
             }
             else
             {
                 // Write output to stdout for even pages.
                 Pl_StdioFile out("stdout", stdout);
                 std::cout << "% Contents of page " << pageno << std::endl;
-                ph.filterPageContents(&counter, &out);
+                ph.filterContents(&counter, &out);
                 std::cout << "\n% end " << pageno << std::endl;
             }
             std::cout << "Page " << pageno
@@ -2,7 +2,7 @@
 // This example illustrates the use of QPDFObjectHandle::TokenFilter
 // with addContentTokenFilter. Please see comments inline for details.
 // See also pdf-count-strings.cc for a use of
-// QPDFObjectHandle::TokenFilter with filterPageContents.
+// QPDFObjectHandle::TokenFilter with filterContents.
 //
 #include <iostream>
@@ -140,15 +140,20 @@ class QPDFPageObjectHelper: public QPDFObjectHelper
     QPDF_DLL
     void parsePageContents(QPDFObjectHandle::ParserCallbacks* callbacks);
-    // Pass a page's contents through the given TokenFilter. If a
-    // pipeline is also provided, it will be the target of the write
-    // methods from the token filter. If a pipeline is not specified,
-    // any output generated by the token filter will be discarded. Use
-    // this interface if you need to pass a page's contents through
-    // filter for work purposes without having that filter
-    // automatically applied to the page's contents, as happens with
-    // addContentTokenFilter. See examples/pdf-count-strings.cc for an
-    // example.
+    // Pass a page's or form XObject's contents through the given
+    // TokenFilter. If a pipeline is also provided, it will be the
+    // target of the write methods from the token filter. If a
+    // pipeline is not specified, any output generated by the token
+    // filter will be discarded. Use this interface if you need to
+    // pass a page's contents through filter for work purposes without
+    // having that filter automatically applied to the page's
+    // contents, as happens with addContentTokenFilter. See
+    // examples/pdf-count-strings.cc for an example.
+    QPDF_DLL
+    void filterContents(QPDFObjectHandle::TokenFilter* filter,
+                        Pipeline* next = 0);
+
+    // Old name -- calls filterContents()
     QPDF_DLL
     void filterPageContents(QPDFObjectHandle::TokenFilter* filter,
                             Pipeline* next = 0);
@@ -265,9 +270,7 @@ class QPDFPageObjectHelper: public QPDFObjectHelper
   private:
     static void
     removeUnreferencedResourcesHelper(
-        QPDFObjectHandle oh, std::set<QPDFObjGen>& seen,
-        std::function<QPDFObjectHandle()> get_resource,
-        std::function<void(QPDFObjectHandle::TokenFilter*)> filter_content);
+        QPDFPageObjectHelper ph, std::set<QPDFObjGen>& seen);
     class Members
     {
@@ -435,7 +435,7 @@ QPDFPageObjectHelper::externalizeInlineImages(size_t min_size)
         QPDFObjectHandle::parse("<< /XObject << >> >>"));
     InlineImageTracker iit(this->oh.getOwningQPDF(), min_size, resources);
     Pl_Buffer b("new page content");
-    filterPageContents(&iit, &b);
+    filterContents(&iit, &b);
     if (iit.any_images)
     {
         getObjectHandle().replaceKey(
@@ -504,7 +504,22 @@ QPDFPageObjectHelper::filterPageContents(
     QPDFObjectHandle::TokenFilter* filter,
     Pipeline* next)
 {
-    this->oh.filterPageContents(filter, next);
+    return filterContents(filter, next);
+}
+
+void
+QPDFPageObjectHelper::filterContents(
+    QPDFObjectHandle::TokenFilter* filter,
+    Pipeline* next)
+{
+    if (this->oh.isFormXObject())
+    {
+        this->oh.filterAsContents(filter, next);
+    }
+    else
+    {
+        this->oh.filterPageContents(filter, next);
+    }
 }
 void
@@ -554,23 +569,21 @@ NameWatcher::handleToken(QPDFTokenizer::Token const&amp; token)
 void
 QPDFPageObjectHelper::removeUnreferencedResourcesHelper(
-    QPDFObjectHandle oh, std::set<QPDFObjGen>& seen,
-    std::function<QPDFObjectHandle()> get_resource,
-    std::function<void(QPDFObjectHandle::TokenFilter*)> filter_content)
+    QPDFPageObjectHelper ph, std::set<QPDFObjGen>& seen)
 {
-    if (seen.count(oh.getObjGen()))
+    if (seen.count(ph.oh.getObjGen()))
     {
         return;
     }
-    seen.insert(oh.getObjGen());
+    seen.insert(ph.oh.getObjGen());
     NameWatcher nw;
     try
     {
-        filter_content(&nw);
+        ph.filterContents(&nw);
     }
     catch (std::exception& e)
     {
-        oh.warnIfPossible(
+        ph.oh.warnIfPossible(
             std::string("Unable to parse content stream: ") + e.what() +
             "; not attempting to remove unreferenced objects from this page");
         return;
@@ -578,7 +591,7 @@ QPDFPageObjectHelper::removeUnreferencedResourcesHelper(
     if (nw.saw_bad)
     {
         QTC::TC("qpdf", "QPDFPageObjectHelper bad token finding names");
-        oh.warnIfPossible(
+        ph.oh.warnIfPossible(
             "Bad token found while scanning content stream; "
             "not attempting to remove unreferenced objects from this page");
         return;
@@ -591,7 +604,7 @@ QPDFPageObjectHelper::removeUnreferencedResourcesHelper(
     std::vector<std::string> to_filter;
     to_filter.push_back("/Font");
     to_filter.push_back("/XObject");
-    QPDFObjectHandle resources = get_resource();
+    QPDFObjectHandle resources = ph.getAttribute("/Resources", true);
     for (std::vector<std::string>::iterator d_iter = to_filter.begin();
          d_iter != to_filter.end(); ++d_iter)
     {
@@ -615,14 +628,7 @@ QPDFPageObjectHelper::removeUnreferencedResourcesHelper(
             {
                 QTC::TC("qpdf", "QPDFPageObjectHelper filter form xobject");
                 removeUnreferencedResourcesHelper(
-                    resource.getDict(), seen,
-                    [&resource]() {
-                        return QPDFPageObjectHelper(resource)
-                            .getAttribute("/Resources", true);
-                    },
-                    [&resource](QPDFObjectHandle::TokenFilter* f) {
-                        resource.filterAsContents(f);
-                    });
+                    QPDFPageObjectHelper(resource), seen);
             }
         }
     }
@@ -632,12 +638,7 @@ void
 QPDFPageObjectHelper::removeUnreferencedResources()
 {
     std::set<QPDFObjGen> seen;
-    removeUnreferencedResourcesHelper(
-        this->oh, seen,
-        [this]() { return this->getAttribute("/Resources", true); },
-        [this](QPDFObjectHandle::TokenFilter* f) {
-            this->filterPageContents(f);
-        });
+    removeUnreferencedResourcesHelper(*this, seen);
 }
 QPDFPageObjectHelper