Find PDF header anywhere in the first 1024 bytes

10 files changed, 137 insertions, 13 deletions

diff --git a/ChangeLog b/ChangeLog
index 88f57f30..a06ffdcf 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2012-12-25  Jay Berkenbilt  <ejb@ql.org>
+
+	* Allow PDF header to appear anywhere in the first 1024 bytes of
+	the file as recommended in the implementation notes of the Adobe
+	version of the PDF spec.
+
 2012-11-20  Jay Berkenbilt  <ejb@ql.org>
 
 	* Add zlib and libpcre to Requires.private in the pkg-config file
diff --git a/TODO b/TODO
index 73ffc087..c8e0ceca 100644
--- a/TODO
+++ b/TODO
@@ -1,12 +1,3 @@
-Next
-====
-
- * Find PDF header in the first 1024 bytes of the file.  Treat the
-   location of the PDF header as offset 0 for purposes of resolving
-   explicit file locations as this is what other implementations
-   appear to do.
-
-
 General
 =======
 
diff --git a/libqpdf/OffsetInputSource.cc b/libqpdf/OffsetInputSource.cc
new file mode 100644
index 00000000..c1ec4102
--- /dev/null
+++ b/libqpdf/OffsetInputSource.cc
@@ -0,0 +1,61 @@
+#include <qpdf/OffsetInputSource.hh>
+
+OffsetInputSource::OffsetInputSource(PointerHolder<InputSource> proxied,
+                                     qpdf_offset_t global_offset) :
+    proxied(proxied),
+    global_offset(global_offset)
+{
+}
+
+OffsetInputSource::~OffsetInputSource()
+{
+}
+
+qpdf_offset_t
+OffsetInputSource::findAndSkipNextEOL()
+{
+    return this->proxied->findAndSkipNextEOL() - this->global_offset;
+}
+
+std::string const&
+OffsetInputSource::getName() const
+{
+    return this->proxied->getName();
+}
+
+qpdf_offset_t
+OffsetInputSource::tell()
+{
+    return this->proxied->tell() - this->global_offset;
+}
+
+void
+OffsetInputSource::seek(qpdf_offset_t offset, int whence)
+{
+    if (whence == SEEK_SET)
+    {
+        this->proxied->seek(offset + global_offset, whence);
+    }
+    else
+    {
+        this->proxied->seek(offset, whence);
+    }
+}
+
+void
+OffsetInputSource::rewind()
+{
+    seek(0, SEEK_SET);
+}
+
+size_t
+OffsetInputSource::read(char* buffer, size_t length)
+{
+    return this->proxied->read(buffer, length);
+}
+
+void
+OffsetInputSource::unreadCh(char ch)
+{
+    this->proxied->unreadCh(ch);
+}
diff --git a/libqpdf/QPDF.cc b/libqpdf/QPDF.cc
index ccbfaf7c..ba96cb64 100644
--- a/libqpdf/QPDF.cc
+++ b/libqpdf/QPDF.cc
@@ -13,6 +13,7 @@
 #include <qpdf/Pl_Discard.hh>
 #include <qpdf/FileInputSource.hh>
 #include <qpdf/BufferInputSource.hh>
+#include <qpdf/OffsetInputSource.hh>
 
 #include <qpdf/QPDFExc.hh>
 #include <qpdf/QPDF_Null.hh>
@@ -213,7 +214,7 @@ QPDF::getWarnings()
 void
 QPDF::parse(char const* password)
 {
-    PCRE header_re("^%PDF-(1.\\d+)\\b");
+    PCRE header_re("\\A((?s).*?)%PDF-(1.\\d+)\\b");
     PCRE eof_re("(?s:startxref\\s+(\\d+)\\s+%%EOF\\b)");
 
     if (password)
@@ -221,11 +222,24 @@ QPDF::parse(char const* password)
 	this->provided_password = password;
     }
 
-    std::string line = this->file->readLine(20);
+    // Find the header anywhere in the first 1024 bytes of the file.
+    char buffer[1044];
+    this->file->read(buffer, sizeof(buffer));
+    std::string line(buffer);
     PCRE::Match m1 = header_re.match(line.c_str());
     if (m1)
     {
-	this->pdf_version = m1.getMatch(1);
+        size_t global_offset = m1.getMatch(1).length();
+        if (global_offset != 0)
+        {
+            // Emperical evidence strongly suggests that when there is
+            // leading material prior to the PDF header, all explicit
+            // offsets in the file are such that 0 points to the
+            // beginning of the header.
+            QTC::TC("qpdf", "QPDF global offset");
+            this->file = new OffsetInputSource(this->file, global_offset);
+        }
+	this->pdf_version = m1.getMatch(2);
 	if (atof(this->pdf_version.c_str()) < 1.2)
 	{
 	    this->tokenizer.allowPoundAnywhereInName();
diff --git a/libqpdf/build.mk b/libqpdf/build.mk
index 6debf107..0ad96a2d 100644
--- a/libqpdf/build.mk
+++ b/libqpdf/build.mk
@@ -12,6 +12,7 @@ SRCS_libqpdf = \
 	libqpdf/FileInputSource.cc \
 	libqpdf/InputSource.cc \
 	libqpdf/MD5.cc \
+	libqpdf/OffsetInputSource.cc \
 	libqpdf/PCRE.cc \
 	libqpdf/Pipeline.cc \
 	libqpdf/Pl_AES_PDF.cc \
diff --git a/libqpdf/qpdf/OffsetInputSource.hh b/libqpdf/qpdf/OffsetInputSource.hh
new file mode 100644
index 00000000..aedc574a
--- /dev/null
+++ b/libqpdf/qpdf/OffsetInputSource.hh
@@ -0,0 +1,29 @@
+#ifndef __QPDF_OFFSETINPUTSOURCE_HH__
+#define __QPDF_OFFSETINPUTSOURCE_HH__
+
+// This class implements an InputSource that proxies for an underlying
+// input source but offset a specific number of bytes.
+
+#include <qpdf/InputSource.hh>
+#include <qpdf/PointerHolder.hh>
+
+class OffsetInputSource: public InputSource
+{
+  public:
+    OffsetInputSource(PointerHolder<InputSource>, qpdf_offset_t global_offset);
+    virtual ~OffsetInputSource();
+
+    virtual qpdf_offset_t findAndSkipNextEOL();
+    virtual std::string const& getName() const;
+    virtual qpdf_offset_t tell();
+    virtual void seek(qpdf_offset_t offset, int whence);
+    virtual void rewind();
+    virtual size_t read(char* buffer, size_t length);
+    virtual void unreadCh(char ch);
+
+  private:
+    PointerHolder<InputSource> proxied;
+    qpdf_offset_t global_offset;
+};
+
+#endif // __QPDF_OFFSETINPUTSOURCE_HH__
diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index 3458297a..937d2b0c 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -243,3 +243,4 @@ QPDF_Tokenizer EOF reading appendable token 0
 QPDFWriter extra header text no newline 0
 QPDFWriter extra header text add newline 0
 QPDF bogus 0 offset 0
+QPDF global offset 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index 16af5832..35645466 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 += 56;
+$n_tests += 57;
 
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -414,6 +414,10 @@ $td->runtest("object with zero offset",
              {$td->COMMAND => "qpdf --check zero-offset.pdf"},
              {$td->FILE => "zero-offset.out", $td->EXIT_STATUS => 3},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("check file with leading junk",
+             {$td->COMMAND => "qpdf --check leading-junk.pdf"},
+             {$td->FILE => "leading-junk.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/leading-junk.out b/qpdf/qtest/qpdf/leading-junk.out
new file mode 100644
index 00000000..58847c9d
--- /dev/null
+++ b/qpdf/qtest/qpdf/leading-junk.out
@@ -0,0 +1,17 @@
+checking leading-junk.pdf
+PDF Version: 1.4
+R = 3
+P = -4
+User password = 
+extract for accessibility: allowed
+extract for any purpose: allowed
+print low resolution: allowed
+print high resolution: allowed
+modify document assembly: allowed
+modify forms: allowed
+modify annotations: allowed
+modify other: allowed
+modify anything: allowed
+File is linearized
+No syntax or stream encoding errors found; the file may still contain
+errors that qpdf cannot detect
diff --git a/qpdf/qtest/qpdf/leading-junk.pdf b/qpdf/qtest/qpdf/leading-junk.pdf
new file mode 100644
index 00000000..2b2a0a2c
--- /dev/null
+++ b/qpdf/qtest/qpdf/leading-junk.pdf