Keep only referenced form fields in --pages

Jay Berkenbilt
1 parent 50037fb3
Showing 7 changed files with 99 additions and 11 deletions
include/qpdf/QPDFAcroFormDocumentHelper.hh
libqpdf/QPDFAcroFormDocumentHelper.cc
qpdf/qpdf.cc
qpdf/qpdf.testcov
qpdf/qtest/qpdf.test
qpdf/qtest/qpdf/kept-no-fields.pdf
qpdf/qtest/qpdf/kept-some-fields.pdf
@@ -140,7 +140,7 @@ class QPDFAcroFormDocumentHelper: public QPDFDocumentHelper
     std::vector<QPDFAnnotationObjectHelper>
     getWidgetAnnotationsForPage(QPDFPageObjectHelper);
  
-    // Return form fields for a page.
+    // Return top-level form fields for a page.
     QPDF_DLL
     std::vector<QPDFFormFieldObjectHelper>
     getFormFieldsForPage(QPDFPageObjectHelper);
@@ -210,11 +210,15 @@ class QPDFAcroFormDocumentHelper: public QPDFDocumentHelper
         QPDFAcroFormDocumentHelper* from_afdh = nullptr);
  
     // Copy form fields from a page in a different QPDF object to this
-    // QPDF.
+    // QPDF. If copied_fields is not null, it will be initialized with
+    // the fields that were copied. Items in the vector are objects in
+    // the receiving QPDF (the one associated with this
+    // QPDFAcroFormDocumentHelper).
     QPDF_DLL
     void copyFieldsFromForeignPage(
         QPDFPageObjectHelper foreign_page,
-        QPDFAcroFormDocumentHelper& foreign_afdh);
+        QPDFAcroFormDocumentHelper& foreign_afdh,
+        std::vector<QPDFObjectHandle>* copied_fields = nullptr);
  
   private:
     void analyze();
@@ -135,15 +135,21 @@ QPDFAcroFormDocumentHelper::getWidgetAnnotationsForPage(QPDFPageObjectHelper h)
 std::vector<QPDFFormFieldObjectHelper>
 QPDFAcroFormDocumentHelper::getFormFieldsForPage(QPDFPageObjectHelper ph)
 {
+    std::set<QPDFObjGen> added;
     std::vector<QPDFFormFieldObjectHelper> result;
     auto widget_annotations = getWidgetAnnotationsForPage(ph);
     for (auto annot: widget_annotations)
     {
         auto field = getFieldForAnnotation(annot);
         field = field.getTopLevelField();
-        if (field.getObjectHandle().isDictionary())
+        auto og = field.getObjectHandle().getObjGen();
+        if (! added.count(og))
         {
-            result.push_back(field);
+            added.insert(og);
+            if (field.getObjectHandle().isDictionary())
+            {
+                result.push_back(field);
+            }
         }
     }
     return result;
@@ -674,18 +680,27 @@ QPDFAcroFormDocumentHelper::transformAnnotations(
 void
 QPDFAcroFormDocumentHelper::copyFieldsFromForeignPage(
     QPDFPageObjectHelper foreign_page,
-    QPDFAcroFormDocumentHelper& foreign_afdh)
+    QPDFAcroFormDocumentHelper& foreign_afdh,
+    std::vector<QPDFObjectHandle>* copied_fields)
 {
     std::set<QPDFObjGen> added;
     for (auto field: foreign_afdh.getFormFieldsForPage(foreign_page))
     {
         auto new_field = this->qpdf.copyForeignObject(
             field.getObjectHandle());
+        if (! new_field.isIndirect())
+        {
+            new_field = this->qpdf.makeIndirectObject(new_field);
+        }
         auto og = new_field.getObjGen();
         if (! added.count(og))
         {
             addFormField(new_field);
             added.insert(og);
+            if (copied_fields)
+            {
+                copied_fields->push_back(new_field);
+            }
         }
     }
 }
@@ -5846,6 +5846,7 @@ static void handle_page_specs(QPDF&amp; pdf, Options&amp; o, bool&amp; warnings)
     std::map<unsigned long long,
              PointerHolder<QPDFAcroFormDocumentHelper>> afdh_map;
     auto this_afdh = get_afdh_for_qpdf(afdh_map, &pdf);
+    std::set<QPDFObjGen> referenced_fields;
     for (std::vector<QPDFPageData>::iterator iter =
              parsed_specs.begin();
          iter != parsed_specs.end(); ++iter)
@@ -5906,7 +5907,13 @@ static void handle_page_specs(QPDF&amp; pdf, Options&amp; o, bool&amp; warnings)
             else if (other_afdh->hasAcroForm())
             {
                 QTC::TC("qpdf", "qpdf copy form fields in pages");
-                this_afdh->copyFieldsFromForeignPage(to_copy, *other_afdh);
+                std::vector<QPDFObjectHandle> copied_fields;
+                this_afdh->copyFieldsFromForeignPage(
+                    to_copy, *other_afdh, &copied_fields);
+                for (auto const& cf: copied_fields)
+                {
+                    referenced_fields.insert(cf.getObjGen());
+                }
             }
         }
         if (page_data.qpdf->anyWarnings())
@@ -5929,16 +5936,57 @@ static void handle_page_specs(QPDF&amp; pdf, Options&amp; o, bool&amp; warnings)
  
     // Delete page objects for unused page in primary. This prevents
     // those objects from being preserved by being referred to from
-    // other places, such as the outlines dictionary.
+    // other places, such as the outlines dictionary. Also make sure
+    // we keep form fields from pages we preserved.
     for (size_t pageno = 0; pageno < orig_pages.size(); ++pageno)
     {
-        if (selected_from_orig.count(QIntC::to_int(pageno)) == 0)
+        auto page = orig_pages.at(pageno);
+        if (selected_from_orig.count(QIntC::to_int(pageno)))
+        {
+            for (auto field: this_afdh->getFormFieldsForPage(page))
+            {
+                QTC::TC("qpdf", "qpdf pages keeping field from original");
+                referenced_fields.insert(field.getObjectHandle().getObjGen());
+            }
+        }
+        else
         {
             pdf.replaceObject(
-                orig_pages.at(pageno).getObjectHandle().getObjGen(),
+                page.getObjectHandle().getObjGen(),
                 QPDFObjectHandle::newNull());
         }
     }
+    // Remove unreferenced form fields
+    if (this_afdh->hasAcroForm())
+    {
+        auto acroform = pdf.getRoot().getKey("/AcroForm");
+        auto fields = acroform.getKey("/Fields");
+        if (fields.isArray())
+        {
+            auto new_fields = QPDFObjectHandle::newArray();
+            if (fields.isIndirect())
+            {
+                new_fields = pdf.makeIndirectObject(new_fields);
+            }
+            for (auto const& field: fields.aitems())
+            {
+                if (referenced_fields.count(field.getObjGen()))
+                {
+                    new_fields.appendItem(field);
+                }
+            }
+            if (new_fields.getArrayNItems() > 0)
+            {
+                QTC::TC("qpdf", "qpdf keep some fields in pages");
+                acroform.replaceKey("/Fields", new_fields);
+            }
+            else
+            {
+                QTC::TC("qpdf", "qpdf no more fields in pages");
+                pdf.getRoot().removeKey("/AcroForm");
+            }
+        }
+    }
 }
  
 static void handle_rotations(QPDF& pdf, Options& o)
@@ -577,3 +577,6 @@ QPDFAcroFormDocumentHelper field with parent 3
 QPDFAcroFormDocumentHelper modify ap matrix 0
 qpdf copy form fields in split_pages 0
 qpdf copy form fields in pages 0
+qpdf keep some fields in pages 0
+qpdf pages keeping field from original 0
+qpdf no more fields in pages 0
@@ -2414,7 +2414,7 @@ foreach my $f (qw(screen print))
 show_ntests();
 # ----------
 $td->notify("--- Copy Annotations ---");
-$n_tests += 21;
+$n_tests += 25;
  
 $td->runtest("complex copy annotations",
              {$td->COMMAND =>
@@ -2479,6 +2479,24 @@ for (my $i = 1; $i &lt;= 2; ++$i)
                  {$td->FILE => "split-out-$i.pdf"},
                  {$td->FILE => "fields-split-$i.pdf"});
 }
+$td->runtest("keeping some fields",
+             {$td->COMMAND =>
+                  "qpdf --static-id fields-two-pages.pdf" .
+                  " --pages . 1 minimal.pdf -- a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+             {$td->FILE => "a.pdf"},
+             {$td->FILE => "kept-some-fields.pdf"});
+$td->runtest("not keeping any fields",
+             {$td->COMMAND =>
+                  "qpdf --static-id kept-some-fields.pdf" .
+                  " --pages . 2 -- a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+             {$td->FILE => "a.pdf"},
+             {$td->FILE => "kept-no-fields.pdf"});
  
 show_ntests();
 # ----------