Add QPDF::emptyPDF() and pdf_from_scratch test code

10 files changed, 254 insertions, 5 deletions

diff --git a/ChangeLog b/ChangeLog
index 12609a21..5be2c8ea 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,9 @@
 2012-06-21  Jay Berkenbilt  <ejb@ql.org>
 
+	* Add QPDF::emptyPDF() to create an empty QPDF object suitable for
+	adding pages and other objects to.  pdf_from_scratch.cc is test
+	code that exercises it.
+
 	* make/libtool.mk: Place user-specified CPPFLAGS and LDFLAGS later
 	in the compilation so that if a user installs things in a
 	non-standard place that they have to tell the build about, earlier
diff --git a/TODO b/TODO
index 6a20fa15..7f997b52 100644
--- a/TODO
+++ b/TODO
@@ -13,6 +13,35 @@ Next
     - update README-windows.txt docs to indicate that MSVC 2010 is the
       supported version and to update the information about mingw.
 
+ * Testing for files > 4GB
+
+    - Create a PDF from scratch.  Each page has a page number as text
+      and an image.  The image can be 5000x5000 pixels using 8-bit
+      gray scale.  It will be divided into 10 stripes of 500 pixels
+      each.  The left and right 500 pixels of each stripe will
+      alternate black and white.  The remaining part of the image will
+      have white stripes indicating 1 and black stripes indicating 0
+      with the most-significant bit on top to indicate the page
+      number.  In this way, every page will be unique and will consume
+      approximately 25 megabytes.  Creating 200 pages like this will
+      make a file that is 5 GB.
+
+    - The file will have to have object streams since a regular xref
+      table won't be able to support offsets that large.
+
+    - A separate test program can create this file and do various
+      manipulations on it.  This can be enabled with an environment
+      variable controlled by configure in much the same way image
+      comparison tests are enabled now.  The argument to
+      --enable-large-file-test should be a path that has enough disk
+      space to do the tests, probably enough space for two coipes of
+      the file.  The test program should also have an interactive mode
+      so we can generate the large file and then look at it with a
+      PDF viewer like Adobe Reader.
+
+ * Consider adding an example that uses the page APIs, or update the
+   documentation to refer the user to the test suite.
+
 Soon
 ====
 
@@ -24,8 +53,6 @@ Soon
  * See if I can support the new encryption formats mentioned in the
    open bug on sourceforge.  Check other sourceforge bugs.
 
- * Would be nice to confirm that it's working for > 4GB files.
-
  * Splitting/merging concepts
 
    newPDF() could create a PDF with just a trailer, no pages, and a
diff --git a/include/qpdf/QPDF.hh b/include/qpdf/QPDF.hh
index 6a910e58..ef958f8e 100644
--- a/include/qpdf/QPDF.hh
+++ b/include/qpdf/QPDF.hh
@@ -69,6 +69,16 @@ class QPDF
 			   char const* buf, size_t length,
 			   char const* password = 0);
 
+    // Create a QPDF object for an empty PDF.  This PDF has no pages
+    // or objects other than a minimal trailer, a document catalog,
+    // and a /Pages tree containing zero pages.  Pages and other
+    // objects can be added to the file in the normal way, and the
+    // trailer and document catalog can be mutated.  Calling this
+    // method is equivalent to calling processFile on an equivalent
+    // PDF file.
+    QPDF_DLL
+    void emptyPDF();
+
     // Parameter settings
 
     // By default, warning messages are issued to std::cerr and output
diff --git a/libqpdf/QPDF.cc b/libqpdf/QPDF.cc
index e97da295..b0a8741c 100644
--- a/libqpdf/QPDF.cc
+++ b/libqpdf/QPDF.cc
@@ -17,6 +17,24 @@
 
 std::string QPDF::qpdf_version = "2.3.1";
 
+static char const* EMPTY_PDF =
+    "%PDF-1.3\n"
+    "1 0 obj\n"
+    "<< /Type /Catalog /Pages 2 0 R >>\n"
+    "endobj\n"
+    "2 0 obj\n"
+    "<< /Type /Pages /Kids [] /Count 0 >>\n"
+    "endobj\n"
+    "xref\n"
+    "0 3\n"
+    "0000000000 65535 f \n"
+    "0000000009 00000 n \n"
+    "0000000058 00000 n \n"
+    "trailer << /Size 3 /Root 1 0 R >>\n"
+    "startxref\n"
+    "110\n"
+    "%%EOF\n";
+
 void
 QPDF::InputSource::setLastOffset(qpdf_offset_t offset)
 {
@@ -350,6 +368,12 @@ QPDF::processMemoryFile(char const* description,
 }
 
 void
+QPDF::emptyPDF()
+{
+    processMemoryFile("empty file", EMPTY_PDF, strlen(EMPTY_PDF));
+}
+
+void
 QPDF::setIgnoreXRefStreams(bool val)
 {
     this->ignore_xref_streams = val;
diff --git a/libqpdf/QPDF_optimization.cc b/libqpdf/QPDF_optimization.cc
index 10b56b60..f2d33863 100644
--- a/libqpdf/QPDF_optimization.cc
+++ b/libqpdf/QPDF_optimization.cc
@@ -365,7 +365,7 @@ QPDF::optimizePagesTreeInternal(
 	throw QPDFExc(qpdf_e_damaged_pdf, this->file->getName(),
 		      this->last_object_description,
 		      this->file->getLastOffset(),
-		      "invalid Type in page tree");
+		      "invalid Type " + type + " in page tree");
     }
 }
 
diff --git a/libqpdf/QPDF_pages.cc b/libqpdf/QPDF_pages.cc
index 2b8a831e..89f5f195 100644
--- a/libqpdf/QPDF_pages.cc
+++ b/libqpdf/QPDF_pages.cc
@@ -73,7 +73,7 @@ QPDF::getAllPagesInternal(QPDFObjectHandle cur_pages,
 	throw QPDFExc(qpdf_e_damaged_pdf, this->file->getName(),
 		      this->last_object_description,
 		      this->file->getLastOffset(),
-		      ": invalid Type in page tree");
+		      "invalid Type " + type + " in page tree");
     }
 }
 
diff --git a/qpdf/build.mk b/qpdf/build.mk
index 7bfed548..dfe169ab 100644
--- a/qpdf/build.mk
+++ b/qpdf/build.mk
@@ -1,4 +1,4 @@
-BINS_qpdf = qpdf test_driver
+BINS_qpdf = qpdf test_driver pdf_from_scratch
 CBINS_qpdf = qpdf-ctest
 
 TARGETS_qpdf = $(foreach B,$(BINS_qpdf) $(CBINS_qpdf),qpdf/$(OUTPUT_DIR)/$(call binname,$(B)))
diff --git a/qpdf/pdf_from_scratch.cc b/qpdf/pdf_from_scratch.cc
new file mode 100644
index 00000000..22d1999e
--- /dev/null
+++ b/qpdf/pdf_from_scratch.cc
@@ -0,0 +1,137 @@
+#include <qpdf/QPDF.hh>
+
+#include <qpdf/QUtil.hh>
+#include <qpdf/QTC.hh>
+#include <qpdf/QPDFWriter.hh>
+#include <qpdf/QPDFObjectHandle.hh>
+#include <iostream>
+#include <stdio.h>
+#include <string.h>
+#include <stdlib.h>
+#include <assert.h>
+
+static char const* whoami = 0;
+
+void usage()
+{
+    std::cerr << "Usage: " << whoami << " n" << std::endl;
+    exit(2);
+}
+
+static QPDFObjectHandle createPageContents(QPDF& pdf, std::string const& text)
+{
+    std::string contents = "BT /F1 15 Tf 72 720 Td (" + text + ") Tj ET\n";
+    PointerHolder<Buffer> b = new Buffer(contents.length());
+    unsigned char* bp = b->getBuffer();
+    memcpy(bp, (char*)contents.c_str(), contents.length());
+    return QPDFObjectHandle::newStream(&pdf, b);
+}
+
+QPDFObjectHandle newName(std::string const& name)
+{
+    return QPDFObjectHandle::newName(name);
+}
+
+void runtest(int n)
+{
+    QPDF pdf;
+    pdf.emptyPDF();
+    if (n == 0)
+    {
+        // Create a minimal PDF from scratch.
+
+        std::map<std::string, QPDFObjectHandle> keys;
+        std::vector<QPDFObjectHandle> items;
+
+        keys.clear();
+        keys["/Type"] = newName("/Font");
+        keys["/Subtype"] = newName("/Type1");
+        keys["/Name"] = newName("/F1");
+        keys["/BaseFont"] = newName("/Helvetica");
+        keys["/Encoding"] = newName("/WinAnsiEncoding");
+        QPDFObjectHandle font = pdf.makeIndirectObject(
+            QPDFObjectHandle::newDictionary(keys));
+
+        items.clear();
+        items.push_back(newName("/PDF"));
+        items.push_back(newName("/Text"));
+        QPDFObjectHandle procset = pdf.makeIndirectObject(
+            QPDFObjectHandle::newArray(items));
+
+        QPDFObjectHandle contents = createPageContents(pdf, "First Page");
+
+        items.clear();
+        items.push_back(QPDFObjectHandle::newInteger(0));
+        items.push_back(QPDFObjectHandle::newInteger(0));
+        items.push_back(QPDFObjectHandle::newInteger(612));
+        items.push_back(QPDFObjectHandle::newInteger(792));
+        QPDFObjectHandle mediabox = QPDFObjectHandle::newArray(items);
+
+        keys.clear();
+        keys["/F1"] = font;
+        QPDFObjectHandle rfont = QPDFObjectHandle::newDictionary(keys);
+
+        keys.clear();
+        keys["/ProcSet"] = procset;
+        keys["/Font"] = rfont;
+        QPDFObjectHandle resources = QPDFObjectHandle::newDictionary(keys);
+
+        keys.clear();
+        keys["/Type"] = newName("/Page");
+        keys["/MediaBox"] = mediabox;
+        keys["/Contents"] = contents;
+        keys["/Resources"] = resources;
+        QPDFObjectHandle page = pdf.makeIndirectObject(
+            QPDFObjectHandle::newDictionary(keys));
+
+        pdf.addPage(page, true);
+
+	QPDFWriter w(pdf, "a.pdf");
+	w.setStaticID(true);
+	w.setStreamDataMode(qpdf_s_preserve);
+	w.write();
+    }
+    else
+    {
+	throw std::runtime_error(std::string("invalid test ") +
+				 QUtil::int_to_string(n));
+    }
+
+    std::cout << "test " << n << " done" << std::endl;
+}
+
+int main(int argc, char* argv[])
+{
+    QUtil::setLineBuf(stdout);
+    if ((whoami = strrchr(argv[0], '/')) == NULL)
+    {
+	whoami = argv[0];
+    }
+    else
+    {
+	++whoami;
+    }
+    // For libtool's sake....
+    if (strncmp(whoami, "lt-", 3) == 0)
+    {
+	whoami += 3;
+    }
+
+    if (argc != 2)
+    {
+	usage();
+    }
+
+    try
+    {
+	int n = atoi(argv[1]);
+	runtest(n);
+    }
+    catch (std::exception& e)
+    {
+	std::cerr << e.what() << std::endl;
+	exit(2);
+    }
+
+    return 0;
+}
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index a1a76ce3..4ec68c53 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -352,6 +352,17 @@ $td->runtest("shallow copy a stream",
 
 show_ntests();
 # ----------
+$td->notify("--- PDF From Scratch ---");
+$n_tests += 2;
+
+$td->runtest("basic qpdf from scratch",
+	     {$td->COMMAND => "pdf_from_scratch 0"},
+	     {$td->STRING => "test 0 done\n", $td->EXIT_STATUS => 0},
+	     $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "from-scratch-0.pdf"});
+# ----------
 $td->notify("--- Error Condition Tests ---");
 # $n_tests incremented after initialization of badfiles below.
 
diff --git a/qpdf/qtest/qpdf/from-scratch-0.pdf b/qpdf/qtest/qpdf/from-scratch-0.pdf
new file mode 100644
index 00000000..5750ec0e
--- /dev/null
+++ b/qpdf/qtest/qpdf/from-scratch-0.pdf
@@ -0,0 +1,36 @@
+%PDF-1.3
+%¿÷¢þ
+1 0 obj
+<< /Pages 2 0 R /Type /Catalog >>
+endobj
+2 0 obj
+<< /Count 1 /Kids [ 3 0 R ] /Type /Pages >>
+endobj
+3 0 obj
+<< /Contents 4 0 R /MediaBox [ 0 0 612 792 ] /Parent 2 0 R /Resources << /Font << /F1 5 0 R >> /ProcSet 6 0 R >> /Type /Page >>
+endobj
+4 0 obj
+<< /Length 42 >>
+stream
+BT /F1 15 Tf 72 720 Td (First Page) Tj ET
+endstream
+endobj
+5 0 obj
+<< /BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font >>
+endobj
+6 0 obj
+[ /PDF /Text ]
+endobj
+xref
+0 7
+0000000000 65535 f 
+0000000015 00000 n 
+0000000064 00000 n 
+0000000123 00000 n 
+0000000266 00000 n 
+0000000357 00000 n 
+0000000464 00000 n 
+trailer << /Root 1 0 R /Size 7 /ID [<31415926535897932384626433832795><31415926535897932384626433832795>] >>
+startxref
+494
+%%EOF