Preserve form fields when splitting pages (fixes #340)

ChangeLog

@@ -1,5 +1,8 @@
 2021-02-22  Jay Berkenbilt  <ejb@ql.org>
 
+	* From qpdf CLI, --pages and --split-pages will properly preserve
+	interactive form functionality. Fixes #340.
+
 	* Add QPDFAcroFormDocumentHelper::copyFieldsFromForeignPage to
 	copy form fields from a foreign page into the current file.
 

libqpdf/QPDFAcroFormDocumentHelper.cc

@@ -676,10 +676,16 @@ QPDFAcroFormDocumentHelper::copyFieldsFromForeignPage(
     QPDFPageObjectHelper foreign_page,
     QPDFAcroFormDocumentHelper& foreign_afdh)
 {
+    std::set<QPDFObjGen> added;
     for (auto field: foreign_afdh.getFormFieldsForPage(foreign_page))
     {
         auto new_field = this->qpdf.copyForeignObject(
             field.getObjectHandle());
+        auto og = new_field.getObjGen();
+        if (! added.count(og))
+        {
             addFormField(new_field);
+            added.insert(og);
+        }
     }
 }

qpdf/qpdf.cc

@@ -5143,6 +5143,19 @@ static void get_uo_pagenos(UnderOverlay& uo,
     }
 }
 
+static QPDFAcroFormDocumentHelper* get_afdh_for_qpdf(
+    std::map<unsigned long long,
+             PointerHolder<QPDFAcroFormDocumentHelper>>& afdh_map,
+    QPDF* q)
+{
+    auto uid = q->getUniqueId();
+    if (! afdh_map.count(uid))
+    {
+        afdh_map[uid] = new QPDFAcroFormDocumentHelper(*q);
+    }
+    return afdh_map[uid].getPointer();
+}
+
 static void do_under_overlay_for_page(
     QPDF& pdf,
     Options& o,
@@ -5164,12 +5177,7 @@ static void do_under_overlay_for_page(
              PointerHolder<QPDFAcroFormDocumentHelper>> afdh;
     auto make_afdh = [&](QPDFPageObjectHelper& ph) {
         QPDF* q = ph.getObjectHandle().getOwningQPDF();
-        auto uid = q->getUniqueId();
-        if (! afdh.count(uid))
-        {
-            afdh[uid] = new QPDFAcroFormDocumentHelper(*q);
-        }
-        return afdh[uid].getPointer();
+        return get_afdh_for_qpdf(afdh, q);
     };
     auto dest_afdh = make_afdh(dest_page);
 
@@ -5835,6 +5843,9 @@ static void handle_page_specs(QPDF& pdf, Options& o, bool& warnings)
     std::vector<QPDFObjectHandle> new_labels;
     bool any_page_labels = false;
     int out_pageno = 0;
+    std::map<unsigned long long,
+             PointerHolder<QPDFAcroFormDocumentHelper>> afdh_map;
+    auto this_afdh = get_afdh_for_qpdf(afdh_map, &pdf);
     for (std::vector<QPDFPageData>::iterator iter =
              parsed_specs.begin();
          iter != parsed_specs.end(); ++iter)
@@ -5847,6 +5858,7 @@ static void handle_page_specs(QPDF& pdf, Options& o, bool& warnings)
             cis->stayOpen(true);
         }
         QPDFPageLabelDocumentHelper pldh(*page_data.qpdf);
+        auto other_afdh = get_afdh_for_qpdf(afdh_map, page_data.qpdf);
         if (pldh.hasPageLabels())
         {
             any_page_labels = true;
@@ -5891,6 +5903,11 @@ static void handle_page_specs(QPDF& pdf, Options& o, bool& warnings)
                 // of the fact that we are using it.
                 selected_from_orig.insert(pageno);
             }
+            else if (other_afdh->hasAcroForm())
+            {
+                QTC::TC("qpdf", "qpdf copy form fields in pages");
+                this_afdh->copyFieldsFromForeignPage(to_copy, *other_afdh);
+            }
         }
         if (page_data.qpdf->anyWarnings())
         {
@@ -6269,6 +6286,7 @@ static void do_split_pages(QPDF& pdf, Options& o, bool& warnings)
         dh.removeUnreferencedResources();
     }
     QPDFPageLabelDocumentHelper pldh(pdf);
+    QPDFAcroFormDocumentHelper afdh(pdf);
     std::vector<QPDFObjectHandle> const& pages = pdf.getAllPages();
     size_t pageno_len = QUtil::uint_to_string(pages.size()).length();
     size_t num_pages = pages.size();
@@ -6282,6 +6300,11 @@ static void do_split_pages(QPDF& pdf, Options& o, bool& warnings)
         }
         QPDF outpdf;
         outpdf.emptyPDF();
+        PointerHolder<QPDFAcroFormDocumentHelper> out_afdh;
+        if (afdh.hasAcroForm())
+        {
+            out_afdh = new QPDFAcroFormDocumentHelper(outpdf);
+        }
         if (o.suppress_warnings)
         {
             outpdf.setSuppressWarnings(true);
@@ -6290,6 +6313,12 @@ static void do_split_pages(QPDF& pdf, Options& o, bool& warnings)
         {
             QPDFObjectHandle page = pages.at(pageno - 1);
             outpdf.addPage(page, false);
+            if (out_afdh.getPointer())
+            {
+                QTC::TC("qpdf", "qpdf copy form fields in split_pages");
+                out_afdh->copyFieldsFromForeignPage(
+                    QPDFPageObjectHelper(page), afdh);
+            }
         }
         if (pldh.hasPageLabels())
         {

qpdf/qpdf.testcov

@@ -575,3 +575,5 @@ QPDFPageObjectHelper flatten inherit rotate 0
 QPDFAcroFormDocumentHelper copy annotation 3
 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/qtest/qpdf.test

@@ -2414,7 +2414,7 @@ foreach my $f (qw(screen print))
 show_ntests();
 # ----------
 $td->notify("--- Copy Annotations ---");
-$n_tests += 16;
+$n_tests += 21;
 
 $td->runtest("complex copy annotations",
              {$td->COMMAND =>
@@ -2458,6 +2458,28 @@ foreach my $d ([1, "appearances-1.pdf"],
                  {$td->FILE => "test80b$n.pdf"});
 }
 
+$td->runtest("page extraction with fields",
+             {$td->COMMAND =>
+                  "qpdf --static-id --empty" .
+                  " --pages fields-two-pages.pdf -- a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+             {$td->FILE => "a.pdf"},
+             {$td->FILE => "fields-pages-out.pdf"});
+$td->runtest("page splitting with fields",
+             {$td->COMMAND =>
+                  "qpdf --static-id" .
+                  " --split-pages fields-two-pages.pdf a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+for (my $i = 1; $i <= 2; ++$i)
+{
+    $td->runtest("check output",
+                 {$td->FILE => "a-$i.pdf"},
+                 {$td->FILE => "fields-split-$i.pdf"});
+}
+
 show_ntests();
 # ----------
 $td->notify("--- Page Tree Issues ---");