Add QPDFWriter::setExtraHeaderText

11 files changed, 102 insertions, 2 deletions

diff --git a/ChangeLog b/ChangeLog
index ad86b396..ce6018dc 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,11 @@
 2012-09-06  Jay Berkenbilt  <ejb@ql.org>
 
+	* Add new method QPDFWriter::setExtraHeaderText to add extra text,
+	such as application-specific comments, to near the beginning of a
+	PDF file.  For linearized files, this appears after the
+	linearization parameter dictionary.  For non-linearized files, it
+	appears right after the PDF header and non-ASCII comment.
+
 	* Make it possible to write the same QPDF object with two
 	different QPDFWriter objects that have both called
 	setLinearization(true) by making private method
diff --git a/include/qpdf/QPDFWriter.hh b/include/qpdf/QPDFWriter.hh
index 5ee90262..2c1c32f6 100644
--- a/include/qpdf/QPDFWriter.hh
+++ b/include/qpdf/QPDFWriter.hh
@@ -163,6 +163,18 @@ class QPDFWriter
     QPDF_DLL
     void forcePDFVersion(std::string const&);
 
+    // Provide additional text to insert in the PDF file somewhere
+    // near the beginning of the file.  This can be used to add
+    // comments to the beginning of a PDF file, for example, if those
+    // comments are to be consumed by some other application.  No
+    // checks are performed to ensure that the text inserted here is
+    // valid PDF.  If you want to insert multiline comments, you will
+    // need to include \n in the string yourself and start each line
+    // with %.  An extra newline will be appended if one is not
+    // already present at the end of your text.
+    QPDF_DLL
+    void setExtraHeaderText(std::string const&);
+
     // Cause a static /ID value to be generated.  Use only in test
     // suites.
     QPDF_DLL
@@ -354,6 +366,7 @@ class QPDFWriter
     std::string id2;		// trailer dictionary
     std::string min_pdf_version;
     std::string forced_pdf_version;
+    std::string extra_header_text;
     int encryption_dict_objid;
     std::string cur_data_key;
     std::list<PointerHolder<Pipeline> > to_delete;
diff --git a/libqpdf/QPDFWriter.cc b/libqpdf/QPDFWriter.cc
index 19ab1359..eb08488a 100644
--- a/libqpdf/QPDFWriter.cc
+++ b/libqpdf/QPDFWriter.cc
@@ -196,6 +196,22 @@ QPDFWriter::forcePDFVersion(std::string const& version)
 }
 
 void
+QPDFWriter::setExtraHeaderText(std::string const& text)
+{
+    this->extra_header_text = text;
+    if ((this->extra_header_text.length() > 0) &&
+        (*(this->extra_header_text.rbegin()) != '\n'))
+    {
+        QTC::TC("qpdf", "QPDFWriter extra header text add newline");
+        this->extra_header_text += "\n";
+    }
+    else
+    {
+        QTC::TC("qpdf", "QPDFWriter extra header text no newline");
+    }
+}
+
+void
 QPDFWriter::setStaticID(bool val)
 {
     this->static_id = val;
@@ -1832,6 +1848,12 @@ QPDFWriter::writeHeader()
     // it really should be treated as binary.
     writeString("\n%\xbf\xf7\xa2\xfe\n");
     writeStringQDF("%QDF-1.0\n\n");
+
+    // Note: do not write extra header text here.  Linearized PDFs
+    // must include the entire linearization parameter dictionary
+    // within the first 1024 characters of the PDF file, so for
+    // linearized files, we have to write extra header text after the
+    // linearization parameter dictionary.
 }
 
 void
@@ -2189,7 +2211,9 @@ QPDFWriter::writeLinearized()
 	// space to write real dictionary.  200 characters is enough
 	// space if all numerical values in the parameter dictionary
 	// that contain offsets are 20 digits long plus a few extra
-	// characters for safety.
+	// characters for safety.  The entire linearization parameter
+	// dictionary must appear within the first 1024 characters of
+	// the file.
 
 	qpdf_offset_t pos = this->pipeline->getCount();
 	openObject(lindict_id);
@@ -2225,6 +2249,10 @@ QPDFWriter::writeLinearized()
 	writePad(spaces);
 	writeString("\n");
 
+        // If the user supplied any additional header text, write it
+        // here after the linearization parameter dictionary.
+        writeString(this->extra_header_text);
+
 	// Part 3: first page cross reference table and trailer.
 
 	qpdf_offset_t first_xref_offset = this->pipeline->getCount();
@@ -2396,6 +2424,7 @@ QPDFWriter::writeStandard()
     // Start writing
 
     writeHeader();
+    writeString(this->extra_header_text);
 
     // Put root first on queue.
     QPDFObjectHandle trailer = pdf.getTrailer();
diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index 0e062af4..f9663375 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -240,3 +240,5 @@ QPDFObjectHandle trailing data in parse 0
 qpdf pages encryption password 0
 QPDF_Tokenizer EOF reading token 0
 QPDF_Tokenizer EOF reading appendable token 0
+QPDFWriter extra header text no newline 0
+QPDFWriter extra header text add newline 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index b2ee28eb..d12f0644 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -149,7 +149,7 @@ $td->runtest("remove page we don't have",
 	     $td->NORMALIZE_NEWLINES);
 # ----------
 $td->notify("--- Miscellaneous Tests ---");
-$n_tests += 48;
+$n_tests += 53;
 
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -387,6 +387,22 @@ $td->runtest("EOF reading token",
              {$td->COMMAND => "qpdf --check eof-reading-token.pdf"},
              {$td->FILE => "eof-reading-token.out", $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("extra header text",
+             {$td->COMMAND => "test_driver 32 minimal.pdf"},
+             {$td->FILE => "test-32.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+             {$td->FILE => "a.pdf"},
+             {$td->FILE => "extra-header-no-newline.pdf"});
+$td->runtest("check output",
+             {$td->FILE => "b.pdf"},
+             {$td->FILE => "extra-header-lin-no-newline.pdf"});
+$td->runtest("check output",
+             {$td->FILE => "c.pdf"},
+             {$td->FILE => "extra-header-newline.pdf"});
+$td->runtest("check output",
+             {$td->FILE => "d.pdf"},
+             {$td->FILE => "extra-header-lin-newline.pdf"});
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/extra-header-lin-newline.pdf b/qpdf/qtest/qpdf/extra-header-lin-newline.pdf
new file mode 100644
index 00000000..7a5ff3e6
--- /dev/null
+++ b/qpdf/qtest/qpdf/extra-header-lin-newline.pdf
diff --git a/qpdf/qtest/qpdf/extra-header-lin-no-newline.pdf b/qpdf/qtest/qpdf/extra-header-lin-no-newline.pdf
new file mode 100644
index 00000000..e115c462
--- /dev/null
+++ b/qpdf/qtest/qpdf/extra-header-lin-no-newline.pdf
diff --git a/qpdf/qtest/qpdf/extra-header-newline.pdf b/qpdf/qtest/qpdf/extra-header-newline.pdf
new file mode 100644
index 00000000..427623a7
--- /dev/null
+++ b/qpdf/qtest/qpdf/extra-header-newline.pdf
diff --git a/qpdf/qtest/qpdf/extra-header-no-newline.pdf b/qpdf/qtest/qpdf/extra-header-no-newline.pdf
new file mode 100644
index 00000000..1df11d30
--- /dev/null
+++ b/qpdf/qtest/qpdf/extra-header-no-newline.pdf
diff --git a/qpdf/qtest/qpdf/test-32.out b/qpdf/qtest/qpdf/test-32.out
new file mode 100644
index 00000000..ff8d922b
--- /dev/null
+++ b/qpdf/qtest/qpdf/test-32.out
@@ -0,0 +1,13 @@
+file: a.pdf
+linearized: no
+newline: no
+file: b.pdf
+linearized: yes
+newline: no
+file: c.pdf
+linearized: no
+newline: yes
+file: d.pdf
+linearized: yes
+newline: yes
+test 32 done
diff --git a/qpdf/test_driver.cc b/qpdf/test_driver.cc
index b277cdf1..311097f6 100644
--- a/qpdf/test_driver.cc
+++ b/qpdf/test_driver.cc
@@ -1091,6 +1091,27 @@ void runtest(int n, char const* filename1, char const* filename2)
                       << std::endl;
         }
     }
+    else if (n == 32)
+    {
+        // Extra header text
+        char const* filenames[] = {"a.pdf", "b.pdf", "c.pdf", "d.pdf"};
+        for (int i = 0; i < 4; ++i)
+        {
+            bool linearized = ((i & 1) != 0);
+            bool newline = ((i & 2) != 0);
+            QPDFWriter w(pdf, filenames[i]);
+            w.setStaticID(true);
+            std::cout
+                << "file: " << filenames[i] << std::endl
+                << "linearized: " << (linearized ? "yes" : "no") << std::endl
+                << "newline: " << (newline ? "yes" : "no") << std::endl;
+            w.setLinearization(linearized);
+            w.setExtraHeaderText(newline
+                                 ? "%% Comment with newline\n"
+                                 : "%% Comment\n% No newline");
+            w.write();
+        }
+    }
     else
     {
 	throw std::runtime_error(std::string("invalid test ") +