more testing, bug fix for linearized aes encrypted files

git-svn-id: svn+q:///qpdf/trunk@824 71b93d88-0707-0410-a8cf-f5a4172ac649

7 files changed, 152 insertions, 50 deletions

diff --git a/ChangeLog b/ChangeLog
index fe79bdfe..13badadb 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,11 @@
 2009-10-18  Jay Berkenbilt  <ejb@ql.org>
 
+	* If forcing version, disable object stream creation and/or
+	encryption if previous specifications are incompatible with new
+	version.  It is still possible that PDF content, compression
+	schemes, etc., may be incompatible with the new version, but at
+	least this way, older viewers will at least have a chance.
+
 	* libqpdf/QPDFWriter.cc (unparseObject): avoid compressing
 	Metadata streams if possible.
 
diff --git a/TODO b/TODO
index 952f5c80..81f05087 100644
--- a/TODO
+++ b/TODO
@@ -1,6 +1,9 @@
 2.1
 ===
 
+ * Really need to handle /Crypt filter for Metadata.  Search for crypt
+   below.
+
  * Update documentation to reflect new command line flags and any
    other relevant changes.  Should read through ChangeLog and the
    manual before releasing 2.1.
@@ -16,9 +19,6 @@
  * Add comments for the security functions that map them back to the
    items in Adobe's products.
 
- * Have force version at least turn off object streams and maybe
-   change security settings?
-
  * Add error codes to QPDFException.  Change the error interface so
    that warnings and errors are pointers that can be queried using
    more C API functions.  We need a way to get a full string as well
@@ -47,49 +47,7 @@
 
     - Update C API for R4 encryption
 
-    - When we write encrypted files, we must remember to omit any
-      encryption filter settings from original streams.
-
-    - test various combinations with and without cleartext-metadata
-      and aes in compression tests
-
-    - figure out a way to test crypt filters defined on a stream
-
-    - test combinations of linearization and v4 encryption
-
-    - would be nice to test strings and streams with different
-      encryption types, but without sample data, we'd have to write
-      them ourselves which is not that useful
-
-    - figure out how to look at the metadata so I can tell whether
-      /EncryptMetadata is working the way it's supposed to
-
-    - Do something with embedded files, but what and how?
 
-    - General notes:
-
-      /CF - keys are crypt filter names, values are are crypt
-      dictionaries
-
-      Individual streams may also have crypt filters.  Filter type
-      /Crypt; /DecodeParms must contain a Crypt filter decode
-      parameters dictionary whose /Name entry specifies the particular
-      filter to be used.  If /Name is missing, use /Identity.
-      /DecodeParms << /Crypt << /Name /XYZ >> >> where /XYZ is
-      /Identity or a key in /CF.
-
-      /Identity means not to encrypt.
-
-      Crypt Dictionaries
-
-	/Type (optional) /CryptFilter
-	/CFM:
-	  /V2 - use rc4
-	  /AESV2 - use aes
-	/Length - supposed to be key length, but the one file I have
-	  has a bogus value for it, so I'm ignoring it.
-
-	We will ignore remaining fields and values.
 
 2.2
 ===
@@ -127,7 +85,13 @@ General
    crypt filters, and there are already special cases in the code to
    handle those.  Most likely, it won't be a problem, but someday
    someone may find a file that qpdf doesn't work on because of crypt
-   filters.
+   filters.  There is an example in the spec of using a crypt filter
+   on a metadata stream.
+
+   When we write encrypted files, we must remember to omit any
+   encryption filter settings from original streams.
+
+   We need a way to test this.
 
  * The second xref stream for linearized files has to be padded only
    because we need file_size as computed in pass 1 to be accurate.  If
diff --git a/include/qpdf/QPDFWriter.hh b/include/qpdf/QPDFWriter.hh
index 8a4e9ebe..e16c5f99 100644
--- a/include/qpdf/QPDFWriter.hh
+++ b/include/qpdf/QPDFWriter.hh
@@ -98,6 +98,12 @@ class DLL_EXPORT QPDFWriter
     // you are sure the PDF file in question has no features of newer
     // versions of PDF or if you are willing to create files that old
     // viewers may try to open but not be able to properly interpret.
+    // If any encryption has been applied to the document either
+    // explicitly or by preserving the encryption of the source
+    // document, forcing the PDF version to a value too low to support
+    // that type of encryption will explicitly disable decryption.
+    // Additionally, forcing to a version below 1.5 will disable
+    // object streams.
     void forcePDFVersion(std::string const&);
 
     // Cause a static /ID value to be generated.  Use only in test
@@ -193,6 +199,7 @@ class DLL_EXPORT QPDFWriter
 	char const* user_password, char const* owner_password,
 	bool allow_accessibility, bool allow_extract,
 	r3_print_e print, r3_modify_e modify);
+    void disableIncompatbleEncryption(float v);
     void setEncryptionParameters(
 	char const* user_password, char const* owner_password,
 	int V, int R, int key_len, std::set<int>& bits_to_clear);
diff --git a/libqpdf/QPDFWriter.cc b/libqpdf/QPDFWriter.cc
index d094aa66..3c1640f7 100644
--- a/libqpdf/QPDFWriter.cc
+++ b/libqpdf/QPDFWriter.cc
@@ -345,6 +345,52 @@ QPDFWriter::copyEncryptionParameters()
 }
 
 void
+QPDFWriter::disableIncompatbleEncryption(float v)
+{
+    if (! this->encrypted)
+    {
+	return;
+    }
+
+    bool disable = false;
+    if (v < 1.3)
+    {
+	disable = true;
+    }
+    else
+    {
+	int V = atoi(encryption_dictionary["/V"].c_str());
+	int R = atoi(encryption_dictionary["/R"].c_str());
+	if (v < 1.4)
+	{
+	    if ((V > 1) || (R > 2))
+	    {
+		disable = true;
+	    }
+	}
+	else if (v < 1.5)
+	{
+	    if ((V > 2) || (R > 3))
+	    {
+		disable = true;
+	    }
+	}
+	else if (v < 1.6)
+	{
+	    if (this->encrypt_use_aes)
+	    {
+		disable = true;
+	    }
+	}
+    }
+    if (disable)
+    {
+	QTC::TC("qpdf", "QPDFWriter forced version disabled encryption");
+	this->encrypted = false;
+    }
+}
+
+void
 QPDFWriter::setEncryptionParametersInternal(
     int V, int R, int key_len, long P,
     std::string const& O, std::string const& U,
@@ -965,7 +1011,7 @@ QPDFWriter::unparseObject(QPDFObjectHandle object, int level,
 		Buffer* buf = bufpl.getBuffer();
 		val = QPDF_String(
 		    std::string((char*)buf->getBuffer(),
-				(size_t)buf->getSize())).unparse();
+				(size_t)buf->getSize())).unparse(true);
 		delete buf;
 	    }
 	    else
@@ -1423,6 +1469,17 @@ QPDFWriter::write()
 	copyEncryptionParameters();
     }
 
+    if (! this->forced_pdf_version.empty())
+    {
+	float v = atof(this->forced_pdf_version.c_str());
+	disableIncompatbleEncryption(v);
+	if (v < 1.5)
+	{
+	    QTC::TC("qpdf", "QPDFWriter forcing object stream disable");
+	    this->object_stream_mode = o_disable;
+	}
+    }
+
     if (this->qdf_mode || this->normalize_content ||
 	(this->stream_data_mode == s_uncompress))
     {
diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index fc4cb383..d1cf3293 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -166,4 +166,5 @@ QPDF_encryption CFM AESV2 0
 QPDF_encryption aes decode string 0
 QPDF_encryption cleartext metadata 0
 QPDF_encryption aes decode stream 0
-QPDF_encryption stream crypt filter 0
+QPDFWriter forcing object stream disable 0
+QPDFWriter forced version disabled encryption 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index b4dc07ad..4ff06406 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -998,14 +998,14 @@ $td->runtest("check linearization",
 	     $td->NORMALIZE_NEWLINES);
 $td->runtest("linearize and encrypt file",
 	     {$td->COMMAND =>
-		  "qpdf --linearize --encrypt user owner 128 --" .
+		  "qpdf --linearize --encrypt user owner 128 --use-aes=y --" .
 		  " lin-special.pdf a.pdf"},
 	     {$td->STRING => "",
 	      $td->EXIT_STATUS => 0});
 $td->runtest("check encryption",
 	     {$td->COMMAND => "qpdf --show-encryption --password=owner a.pdf",
 	      $td->FILTER => "grep -v allowed"},
-	     {$td->STRING => "R = 3\nP = -4\nUser password = user\n",
+	     {$td->STRING => "R = 4\nP = -4\nUser password = user\n",
 	      $td->EXIT_STATUS => 0},
 	     $td->NORMALIZE_NEWLINES);
 $td->runtest("check linearization",
@@ -1015,6 +1015,68 @@ $td->runtest("check linearization",
 	      $td->EXIT_STATUS => 0},
 	     $td->NORMALIZE_NEWLINES);
 
+# Test AES encryption in various ways.
+$n_tests += 14;
+$td->runtest("encrypt with AES",
+	     {$td->COMMAND => "qpdf --encrypt '' '' 128 --use-aes=y --" .
+		  " enc-base.pdf a.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check encryption",
+	     {$td->COMMAND => "qpdf --show-encryption a.pdf",
+	      $td->FILTER => "grep -v allowed"},
+	     {$td->STRING => "R = 4\nP = -4\nUser password = \n",
+	      $td->EXIT_STATUS => 0},
+	     $td->NORMALIZE_NEWLINES);
+$td->runtest("convert original to qdf",
+	     {$td->COMMAND => "qpdf --static-id --no-original-object-ids" .
+		  " --qdf --min-version=1.6 enc-base.pdf a.qdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("convert encrypted to qdf",
+	     {$td->COMMAND => "qpdf --static-id --no-original-object-ids" .
+		  " --qdf a.pdf b.qdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("compare files",
+	     {$td->FILE => 'a.qdf'},
+	     {$td->FILE => 'b.qdf'});
+$td->runtest("linearize with AES and object streams",
+	     {$td->COMMAND => "qpdf --encrypt '' '' 128 --use-aes=y --" .
+		  " --linearize --object-streams=generate enc-base.pdf a.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check encryption",
+	     {$td->COMMAND => "qpdf --show-encryption a.pdf",
+	      $td->FILTER => "grep -v allowed"},
+	     {$td->STRING => "R = 4\nP = -4\nUser password = \n",
+	      $td->EXIT_STATUS => 0},
+	     $td->NORMALIZE_NEWLINES);
+$td->runtest("linearize original",
+	     {$td->COMMAND => "qpdf --linearize --object-streams=generate" .
+		  " enc-base.pdf b.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("convert linearized original to qdf",
+	     {$td->COMMAND => "qpdf --static-id --no-original-object-ids" .
+		  " --qdf --object-streams=generate --min-version=1.6" .
+		  " b.pdf a.qdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("convert encrypted to qdf",
+	     {$td->COMMAND => "qpdf --static-id --no-original-object-ids" .
+		  " --qdf --object-streams=generate a.pdf b.qdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("compare files",
+	     {$td->FILE => 'a.qdf'},
+	     {$td->FILE => 'b.qdf'});
+$td->runtest("force version on aes encrypted",
+	     {$td->COMMAND => "qpdf --force-version=1.4 a.pdf b.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check",
+	     {$td->COMMAND => "qpdf --check b.pdf"},
+	     {$td->FILE => "aes-forced-check.out",
+	      $td->EXIT_STATUS => 0},
+	     $td->NORMALIZE_NEWLINES);
+$td->runtest("make sure there is no xref stream",
+	     {$td->COMMAND => "grep /ObjStm b.pdf | wc -l"},
+	     {$td->REGEXP => "\\s*0\\s*", $td->EXIT_STATUS => 0},
+	     $td->NORMALIZE_NEWLINES);
+
 show_ntests();
 # ----------
 $td->notify("--- Content Preservation Tests ---");
diff --git a/qpdf/qtest/qpdf/aes-forced-check.out b/qpdf/qtest/qpdf/aes-forced-check.out
new file mode 100644
index 00000000..9a71439d
--- /dev/null
+++ b/qpdf/qtest/qpdf/aes-forced-check.out
@@ -0,0 +1,5 @@
+checking b.pdf
+PDF Version: 1.4
+File is not encrypted
+File is not linearized
+No errors found