public bool ReadStartXref(PdfStream st)
{
if (st.PeekLine(true, true) == "startxref")
{
Utility.TraceLine("Read startxref");
st.ReadLine();
try
{
string sline = st.RegexMatch("^([0-9]+)");
m_xrefLocation = Int32.Parse(sline);
}
catch (Exception except)
{
throw new ParseException("Error reading xref inset: " + except.Message);
}
string line = st.ReadLine(true, true);
while (line != "%%EOF")
{
line = st.ReadLine(true, true);
}
return(true);
}
return(false);
}
private void ReadStream(PdfStream st)
{
if (st.PeekLine() == "stream")
{
Utility.TraceLine("Read stream: ");
st.ReadLine();
while (!st.PeekLine(true, true).EndsWith("endstream") && !st.Eof)
{
Utility.Trace("-");
m_stream.Append(st.ReadLineAsBytes());
}
Utility.TraceLine(" Done");
st.ReadLine();
// Sometimes there is a blank line after the endstream
st.ConsumeWhitespace();
}
}
public bool ReadXref(PdfStream st)
{
if (st.PeekLine(true, true) == "xref")
{
Utility.TraceLine("Reading a XREF");
st.ReadLine();
while (st.RegexMatch("^([0-9]+ [0-9]+)", false) != "")
{
Utility.TraceLine("Found XREF value block");
// Each xref block starts with a line with the starting number of the objects in the
// block, and then a count of the number of objects in the block PDF 1.5 p70
int startat = 0;
int count = 0;
try
{
startat = Int32.Parse(st.RegexMatch("^([0-9]+)", true));
count = Int32.Parse(st.RegexMatch("^[ \t]*([0-9]+)", true));
st.ReadLine();
// The you get one line per object in the block. They contain:
// <byte offset> <generation> <n = inuse, f = free>
Utility.TraceLine("XREF reading " + count + " lines");
for (int i = 0; i < count; i++)
{
string line = st.ReadLine();
long offset = Int64.Parse(line.Substring(0, 10));
int generation = Int32.Parse(line.Substring(11, 5));
string inuse = line.Substring(17, 1);
Utility.TraceLine("XREF line: " + offset + " " + generation + " " + inuse);
if ((inuse == "n") && !m_linflag)
{
// PDF/A requires us to verify these offsets
Object obj = m_objects.Get(startat + i, generation);
if (offset != obj.StartedAt)
{
m_pdfa = false;
Utility.TraceLine("PDF/A: Object offset is incorrect " + offset + " != " +
obj.StartedAt + " (required by section 5.4)");
}
}
}
}
catch (Exception except)
{
throw new ParseException("Error reading xref block starter: " + except.Message);
}
}
}
return(false);
}
public void Parse(PdfStream st)
{
Utility.TraceLine("Read object parse pass for object " + Number + " " + Generation);
try
{
m_dictionary.ReadDictionary(st);
ReadStream(st);
ReadDirectValue(st);
if (st.PeekLine(true, true) != "endobj")
{
throw new ParseException("Object " + Number + " " + Generation + " ended early: \"" +
st.PeekLine(true, true) + "\"");
}
st.ReadLine();
}
catch (ParseException ex)
{
Utility.CrashDump(m_parseStartedAt, st.Position, st);
throw new RuntimeException("Error processing object: " + ex.Message + " for object " + Number + " " + Generation);
}
}
private void ReadDirectValue(PdfStream st)
{
if (m_dictionary.Count == 0)
{
while (st.PeekLine() != "endobj")
{
m_directValue += st.ReadLine();
}
}
else
{
Utility.TraceLine("Not eligible for direct value");
}
}
public bool ReadTrailer(PdfStream st)
{
if (st.PeekLine() == "trailer")
{
st.ReadLine();
m_trailer.ReadDictionary(st);
// Enforce PDF/A requirements for the trailer (section 5.3 of draft standard)
// I deliberately do all of these checks so that we can report _all_ errors
// at once
if (m_trailer.Get("Encrypt") != null)
{
m_pdfa = false;
Utility.TraceLine("PDF/A: Forbidden encrypt dictionary item in section 5.3 document trailer");
}
if (m_trailer.Get("Info") != null)
{
m_pdfa = false;
Utility.TraceLine("PDF/A: Forbidden info dictionary item in section 5.3 document trailer");
}
if (m_trailer.Get("Size") == null)
{
m_pdfa = false;
Utility.TraceLine("PDF/A: Required size dictionary document trailer item is missing");
}
if (m_trailer.Get("ID") == null)
{
m_pdfa = false;
Utility.TraceLine("PDF/A: Required ID dictionary document trailer item is missing");
}
Utility.TraceLine("Read trailer");
return(true);
}
return(false);
}
public void ReadDictionary(PdfStream st)
{
if (st.Expect("<<", false))
{
// Until we get to the end of this dictionary
while (st.PeekBlock(2) != ">>")
{
// Consume any leading whitespace
while (Utility.IsWhite(st.PeekBlock(1)))
{
st.ReadBlock(1);
}
// Name
Utility.TraceLine("Checking for a name");
string name = st.RegexMatch("^(/[^ \t/\\[\\]\\(\\)\\<\\>]+)[ \t]*");
if (name == "")
{
throw new ParseException("Dictionary items must have a name");
}
st.ConsumeWhitespace();
// Value
if (st.PeekBlock(2) == "<<")
{
Utility.TraceLine("Traversing subdictionary");
Dictionary dict = new Dictionary();
dict.ReadDictionary(st);
DictionaryItem di = new DictionaryItem(name, dict);
Add(di);
}
else
{
Utility.TraceLine("Finding the value");
string nameval = st.RegexMatch("^(/[^ \t/\\[\\]\\(\\)\\<\\>]+)", false);
string objrefval = st.RegexMatch("^([0-9]+ [0-9]+ R)", false);
string numval = st.RegexMatch("^(-{0,1}[0-9]+)", false);
string floatval = st.RegexMatch("^(-{0,1}[0-9]+\\.[0-9]+)", false);
// The old version of these:
//string rdbrackets = st.RegexMatch("^(\\([^\\)]*\\)+)[ \t]*", false);
string sqbrackets = st.RegexMatch(@"^(\[.*?[^\\]+?\])", false);
string rdbrackets = st.RegexMatch(@"^(\(.*?[^\\]+?\))", false);
string anbrackets = st.RegexMatch(@"^(\<.*?[^\\]+?\>)", false);
string singleword = st.RegexMatch("^([^ \t]+)", false);
if (nameval != "")
{
DictionaryItem di = new DictionaryItem(name, nameval);
Add(di);
st.ReadBlock(nameval.Length);
}
else if (objrefval != "")
{
DictionaryItem di = new DictionaryItem(name, objrefval);
Add(di);
st.ReadBlock(objrefval.Length);
}
else if (floatval != "")
{
DictionaryItem di = new DictionaryItem(name, floatval);
Add(di);
st.ReadBlock(floatval.Length);
}
else if (numval != "")
{
DictionaryItem di = new DictionaryItem(name, numval);
Add(di);
st.ReadBlock(numval.Length);
}
else if (sqbrackets != "")
{
DictionaryItem di = new DictionaryItem(name, sqbrackets);
Add(di);
st.ReadBlock(sqbrackets.Length);
}
else if (rdbrackets != "")
{
DictionaryItem di = new DictionaryItem(name, rdbrackets);
Add(di);
st.ReadBlock(rdbrackets.Length);
}
else if (anbrackets != "")
{
DictionaryItem di = new DictionaryItem(name, anbrackets);
Add(di);
st.ReadBlock(anbrackets.Length);
}
else if (singleword != "")
{
DictionaryItem di = new DictionaryItem(name, singleword);
Add(di);
st.ReadBlock(singleword.Length);
}
else
{
throw new ParseException("Unknown value format: " + st.PeekLine());
}
}
st.ConsumeWhitespace();
}
// Skip over the >> at the end of the dictionary
st.ReadBlock(2);
st.ConsumeWhitespace();
}
}
