Make more QPDFPageObjectHelper methods work with form XObject

13 files changed, 164 insertions, 13 deletions

diff --git a/ChangeLog b/ChangeLog
index d93a4c65..27a457ff 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,17 @@
 2021-01-02  Jay Berkenbilt  <ejb@ql.org>
 
+	* Make QPDFPageObjectHelper methods pipeContents, parseContents,
+	and addContentTokenFilter work with form XObjects.
+
+	* Rename some QPDFPageObjectHelper methods and make them support
+	form XObjects as well as pages. The old names will be preserved
+	from compatibility.
+	- pipePageContents -> pipeContents
+	- parsePageContents -> parseContents
+
+	* Add QPDFObjectHandle::parseAsContents to apply ParserCallbacks
+	to a form XObject.
+
 	* QPDFPageObjectHelper::externalizeInlineImages can be called with
 	form XObjects as well as pages.
 
diff --git a/examples/pdf-parse-content.cc b/examples/pdf-parse-content.cc
index a8cd3290..f1680f72 100644
--- a/examples/pdf-parse-content.cc
+++ b/examples/pdf-parse-content.cc
@@ -89,7 +89,7 @@ int main(int argc, char* argv[])
 
         QPDFPageObjectHelper& page = pages.at(QIntC::to_size(pageno-1));
         ParserCallbacks cb;
-        page.parsePageContents(&cb);
+        page.parseContents(&cb);
     }
     catch (std::exception& e)
     {
diff --git a/fuzz/qpdf_fuzzer.cc b/fuzz/qpdf_fuzzer.cc
index 195cf5bb..070817c6 100644
--- a/fuzz/qpdf_fuzzer.cc
+++ b/fuzz/qpdf_fuzzer.cc
@@ -142,7 +142,7 @@ FuzzHelper::testPages()
         try
         {
             page.coalesceContentStreams();
-            page.parsePageContents(&discard_contents);
+            page.parseContents(&discard_contents);
             page.getImages();
             pldh.getLabelForPage(pageno);
             QPDFObjectHandle page_obj(page.getObjectHandle());
diff --git a/include/qpdf/QPDFObjectHandle.hh b/include/qpdf/QPDFObjectHandle.hh
index 2a5d0392..e310f091 100644
--- a/include/qpdf/QPDFObjectHandle.hh
+++ b/include/qpdf/QPDFObjectHandle.hh
@@ -448,7 +448,7 @@ class QPDFObjectHandle
     void parsePageContents(ParserCallbacks* callbacks);
     QPDF_DLL
     void filterPageContents(TokenFilter* filter, Pipeline* next = 0);
-    // See comments for QPDFPageObjectHelper::pipePageContents.
+    // See comments for QPDFPageObjectHelper::pipeContents.
     QPDF_DLL
     void pipePageContents(Pipeline* p);
     QPDF_DLL
@@ -460,6 +460,10 @@ class QPDFObjectHandle
     // XObject, whose data is in the same format as a content stream.
     QPDF_DLL
     void filterAsContents(TokenFilter* filter, Pipeline* next = 0);
+    // Called on a stream to parse the stream as page contents. This
+    // can be used to parse a form XObject.
+    QPDF_DLL
+    void parseAsContents(ParserCallbacks* callbacks);
 
     // Type-specific factories
     QPDF_DLL
diff --git a/include/qpdf/QPDFPageObjectHelper.hh b/include/qpdf/QPDFPageObjectHelper.hh
index ccf56630..ee813c96 100644
--- a/include/qpdf/QPDFPageObjectHelper.hh
+++ b/include/qpdf/QPDFPageObjectHelper.hh
@@ -182,7 +182,11 @@ class QPDFPageObjectHelper: public QPDFObjectHelper
 
     // Parse a page's contents through ParserCallbacks, described
     // above. This method works whether the contents are a single
-    // stream or an array of streams. Call on a page object.
+    // stream or an array of streams. Call on a page object. Also
+    // works for form XObjects.
+    QPDF_DLL
+    void parseContents(QPDFObjectHandle::ParserCallbacks* callbacks);
+    // Old name
     QPDF_DLL
     void parsePageContents(QPDFObjectHandle::ParserCallbacks* callbacks);
 
@@ -206,14 +210,17 @@ class QPDFPageObjectHelper: public QPDFObjectHelper
 
     // Pipe a page's contents through the given pipeline. This method
     // works whether the contents are a single stream or an array of
-    // streams.
+    // streams. Also works on form XObjects.
+    QPDF_DLL
+    void pipeContents(Pipeline* p);
+    // Old name
     QPDF_DLL
     void pipePageContents(Pipeline* p);
 
     // Attach a token filter to a page's contents. If the page's
     // contents is an array of streams, it is automatically coalesced.
     // The token filter is applied to the page's contents as a single
-    // stream.
+    // stream. Also works on form XObjects.
     QPDF_DLL
     void addContentTokenFilter(
         PointerHolder<QPDFObjectHandle::TokenFilter> token_filter);
diff --git a/libqpdf/QPDFObjectHandle.cc b/libqpdf/QPDFObjectHandle.cc
index 94b81f3d..60419720 100644
--- a/libqpdf/QPDFObjectHandle.cc
+++ b/libqpdf/QPDFObjectHandle.cc
@@ -1669,6 +1669,15 @@ QPDFObjectHandle::parsePageContents(ParserCallbacks* callbacks)
 }
 
 void
+QPDFObjectHandle::parseAsContents(ParserCallbacks* callbacks)
+{
+    std::string description = "object " +
+        QUtil::int_to_string(this->objid) + " " +
+        QUtil::int_to_string(this->generation);
+    this->parseContentStream_internal(description, callbacks);
+}
+
+void
 QPDFObjectHandle::filterPageContents(TokenFilter* filter, Pipeline* next)
 {
     std::string description = "token filter for page object " +
diff --git a/libqpdf/QPDFPageObjectHelper.cc b/libqpdf/QPDFPageObjectHelper.cc
index f7fcd395..3e303f6a 100644
--- a/libqpdf/QPDFPageObjectHelper.cc
+++ b/libqpdf/QPDFPageObjectHelper.cc
@@ -584,7 +584,21 @@ void
 QPDFPageObjectHelper::parsePageContents(
     QPDFObjectHandle::ParserCallbacks* callbacks)
 {
-    this->oh.parsePageContents(callbacks);
+    parseContents(callbacks);
+}
+
+void
+QPDFPageObjectHelper::parseContents(
+    QPDFObjectHandle::ParserCallbacks* callbacks)
+{
+    if (this->oh.isFormXObject())
+    {
+        this->oh.parseAsContents(callbacks);
+    }
+    else
+    {
+        this->oh.parsePageContents(callbacks);
+    }
 }
 
 void
@@ -613,14 +627,34 @@ QPDFPageObjectHelper::filterContents(
 void
 QPDFPageObjectHelper::pipePageContents(Pipeline* p)
 {
-    this->oh.pipePageContents(p);
+    pipeContents(p);
+}
+
+void
+QPDFPageObjectHelper::pipeContents(Pipeline* p)
+{
+    if (this->oh.isFormXObject())
+    {
+        this->oh.pipeStreamData(p, 0, qpdf_dl_specialized);
+    }
+    else
+    {
+        this->oh.pipePageContents(p);
+    }
 }
 
 void
 QPDFPageObjectHelper::addContentTokenFilter(
     PointerHolder<QPDFObjectHandle::TokenFilter> token_filter)
 {
-    this->oh.addContentTokenFilter(token_filter);
+    if (this->oh.isFormXObject())
+    {
+        this->oh.addTokenFilter(token_filter);
+    }
+    else
+    {
+        this->oh.addContentTokenFilter(token_filter);
+    }
 }
 
 class NameWatcher: public QPDFObjectHandle::TokenFilter
diff --git a/manual/qpdf-manual.xml b/manual/qpdf-manual.xml
index 409313ef..c6e00791 100644
--- a/manual/qpdf-manual.xml
+++ b/manual/qpdf-manual.xml
@@ -4893,6 +4893,18 @@ print "\n";
              <function>filterContents</function>
             </para>
            </listitem>
+           <listitem>
+            <para>
+             <function>pipePageContents</function> to
+             <function>pipeContents</function>
+            </para>
+           </listitem>
+           <listitem>
+            <para>
+             <function>parsePageContents</function> to
+             <function>parseContents</function>
+            </para>
+           </listitem>
           </itemizedlist>
          </para>
         </listitem>
diff --git a/qpdf/qpdf.cc b/qpdf/qpdf.cc
index cc02e1fd..c11f2486 100644
--- a/qpdf/qpdf.cc
+++ b/qpdf/qpdf.cc
@@ -3539,7 +3539,7 @@ static void do_check(QPDF& pdf, Options& o, int& exit_code)
             ++pageno;
             try
             {
-                page.parsePageContents(&discard_contents);
+                page.parseContents(&discard_contents);
             }
             catch (QPDFExc& e)
             {
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index 9931e7dc..6919bfcf 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -423,7 +423,7 @@ foreach my $i (@choice_values)
 show_ntests();
 # ----------
 $td->notify("--- Form XObject, underlay, overlay ---");
-$n_tests += 19;
+$n_tests += 20;
 
 $td->runtest("form xobject creation",
              {$td->COMMAND => "test_driver 55 fxo-red.pdf"},
@@ -491,6 +491,11 @@ $td->runtest("foreach",
              {$td->FILE => "nested-form-xobjects.out",
                   $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("page operations on form xobject",
+             {$td->COMMAND => "test_driver 72 nested-form-xobjects.pdf"},
+             {$td->FILE => "page-ops-on-form-xobject.out",
+                  $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/page-ops-on-form-xobject.out b/qpdf/qtest/qpdf/page-ops-on-form-xobject.out
new file mode 100644
index 00000000..f6d05af4
--- /dev/null
+++ b/qpdf/qtest/qpdf/page-ops-on-form-xobject.out
@@ -0,0 +1,47 @@
+--- parseContents ---
+content size: 173
+operator, offset=0, length=2: BT
+name, offset=5, length=3: /F1
+integer, offset=9, length=2: 24
+operator, offset=12, length=2: Tf
+integer, offset=17, length=1: 0
+integer, offset=19, length=3: 320
+operator, offset=23, length=2: Td
+string, offset=28, length=5: (FX1)
+operator, offset=34, length=2: Tj
+operator, offset=37, length=2: ET
+operator, offset=40, length=1: q
+integer, offset=42, length=3: 100
+integer, offset=46, length=1: 0
+integer, offset=48, length=1: 0
+integer, offset=50, length=3: 100
+integer, offset=54, length=3: 0
+integer, offset=58, length=3: 200
+operator, offset=62, length=2: cm
+name, offset=65, length=4: /Im1
+operator, offset=70, length=2: Do
+operator, offset=73, length=1: Q
+operator, offset=75, length=1: q
+integer, offset=77, length=3: 100
+integer, offset=81, length=1: 0
+integer, offset=83, length=1: 0
+integer, offset=85, length=3: 100
+integer, offset=89, length=3: 120
+integer, offset=93, length=3: 200
+operator, offset=97, length=2: cm
+name, offset=100, length=4: /Im2
+operator, offset=105, length=2: Do
+operator, offset=108, length=1: Q
+operator, offset=110, length=1: q
+real, offset=112, length=7: 1.00000
+real, offset=120, length=7: 0.00000
+real, offset=128, length=7: 0.00000
+real, offset=136, length=7: 1.00000
+real, offset=144, length=7: 0.00000
+real, offset=152, length=7: 0.00000
+operator, offset=160, length=2: cm
+name, offset=163, length=4: /Fx1
+operator, offset=168, length=2: Do
+operator, offset=171, length=1: Q
+-EOF-
+test 72 done
diff --git a/qpdf/test_driver.cc b/qpdf/test_driver.cc
index ff55e63f..aef46d87 100644
--- a/qpdf/test_driver.cc
+++ b/qpdf/test_driver.cc
@@ -1463,7 +1463,7 @@ void runtest(int n, char const* filename1, char const* arg2)
         {
             QPDFPageObjectHelper& page(*iter);
             ParserCallbacks cb;
-            page.parsePageContents(&cb);
+            page.parseContents(&cb);
         }
     }
     else if (n == 38)
@@ -2279,6 +2279,27 @@ void runtest(int n, char const* filename1, char const* arg2)
             std::cout << i.first << " -> " << i.second.unparse() << std::endl;
         }
     }
+    else if (n == 72)
+    {
+        // Call some QPDFPageObjectHelper methods on form XObjects.
+        auto page = QPDFPageDocumentHelper(pdf).getAllPages().at(0);
+        auto fx1 = QPDFPageObjectHelper(
+            page.getObjectHandle()
+            .getKey("/Resources")
+            .getKey("/XObject")
+            .getKey("/Fx1"));
+        std::cout << "--- parseContents ---" << std::endl;
+        ParserCallbacks cb;
+        fx1.parseContents(&cb);
+        Pl_Buffer b("buffer");
+        fx1.addContentTokenFilter(new TokenFilter);
+        fx1.pipeContents(&b);
+        std::unique_ptr<Buffer> buf(b.getBuffer());
+        std::string s(
+            reinterpret_cast<char const*>(buf->getBuffer()),
+            buf->getSize());
+        assert(s.find("/bye") != std::string::npos);
+    }
     else
     {
 	throw std::runtime_error(std::string("invalid test ") +
diff --git a/qpdf/test_tokenizer.cc b/qpdf/test_tokenizer.cc
index b8a51b81..8c2b557e 100644
--- a/qpdf/test_tokenizer.cc
+++ b/qpdf/test_tokenizer.cc
@@ -219,7 +219,7 @@ static void process(char const* filename, bool include_ignorable,
     {
         ++pageno;
         Pl_Buffer plb("buffer");
-        (*iter).pipePageContents(&plb);
+        (*iter).pipeContents(&plb);
         PointerHolder<Buffer> content_data = plb.getBuffer();
         BufferInputSource* bis = new BufferInputSource(
             "content data", content_data.getPointer());