Add QPDFObjectHandle::parseContentStream method

This method allows parsing of the PDF objects in a content stream or
array of content streams.

2 files changed, 34 insertions, 1 deletions

diff --git a/include/qpdf/QPDFObjectHandle.hh b/include/qpdf/QPDFObjectHandle.hh
index 932a6678..c4a922d1 100644
--- a/include/qpdf/QPDFObjectHandle.hh
+++ b/include/qpdf/QPDFObjectHandle.hh
@@ -71,6 +71,21 @@ class QPDFObjectHandle
         virtual void decryptString(std::string& val) = 0;
     };
 
+    // This class is used by parseContentStream.  Callers must
+    // instantiate a subclass of this with handlers defined to accept
+    // QPDFObjectHandles that are parsed from the stream.
+    class ParserCallbacks
+    {
+      public:
+        QPDF_DLL
+        virtual ~ParserCallbacks()
+        {
+        }
+        virtual void handleObject(QPDFObjectHandle) = 0;
+        virtual void handleEOF() = 0;
+    };
+
+
     QPDF_DLL
     QPDFObjectHandle();
     QPDF_DLL
@@ -138,6 +153,11 @@ class QPDFObjectHandle
                                   StringDecrypter* decrypter,
                                   QPDF* context);
 
+    // Helpers for parsing content streams
+    QPDF_DLL
+    static void parseContentStream(QPDFObjectHandle stream_or_array,
+                                   ParserCallbacks* callbacks);
+
     // Type-specific factories
     QPDF_DLL
     static QPDFObjectHandle newNull();
@@ -571,7 +591,10 @@ class QPDFObjectHandle
         std::string const& object_description,
         QPDFTokenizer& tokenizer, bool& empty,
         StringDecrypter* decrypter, QPDF* context,
-        bool in_array, bool in_dictionary);
+        bool in_array, bool in_dictionary,
+        bool content_stream);
+    static void parseContentStream_internal(
+        QPDFObjectHandle stream, ParserCallbacks* callbacks);
 
     bool initialized;
 
diff --git a/include/qpdf/QPDFTokenizer.hh b/include/qpdf/QPDFTokenizer.hh
index 1835fcb1..081e12d3 100644
--- a/include/qpdf/QPDFTokenizer.hh
+++ b/include/qpdf/QPDFTokenizer.hh
@@ -18,6 +18,8 @@
 class QPDFTokenizer
 {
   public:
+    // Token type tt_eof is only returned of allowEOF() is called on
+    // the tokenizer.  tt_eof was introduced in QPDF version 4.1.
     enum token_type_e
     {
 	tt_bad,
@@ -34,6 +36,7 @@ class QPDFTokenizer
 	tt_null,
 	tt_bool,
 	tt_word,
+        tt_eof,
     };
 
     class Token
@@ -97,6 +100,12 @@ class QPDFTokenizer
     QPDF_DLL
     void allowPoundAnywhereInName();
 
+    // If called, treat EOF as a separate token type instead of an
+    // error.  This was introduced in QPDF 4.1 to facilitate
+    // tokenizing content streams.
+    QPDF_DLL
+    void allowEOF();
+
     // Mode of operation:
 
     // Keep presenting characters and calling getToken() until
@@ -140,6 +149,7 @@ class QPDFTokenizer
 	   st_literal, st_in_hexstring, st_token_ready } state;
 
     bool pound_special_in_name;
+    bool allow_eof;
 
     // Current token accumulation
     token_type_e type;