private static void ExtractLines(PdfReader reader, StringBuilder sb, PdfObject content)
{
var ir = (PRIndirectReference)content;
var value = reader.GetPdfObject(ir.Number);
if (value.IsStream())
{
PRStream stream = (PRStream)value;
var streamBytes = PdfReader.GetStreamBytes(stream);
var tokenizer = new PRTokeniser(new RandomAccessFileOrArray(streamBytes));
try
{
while (tokenizer.NextToken())
{
if (tokenizer.TokenType == PRTokeniser.TK_STRING)
{
string str = tokenizer.StringValue;
sb.Append(str);
}
}
}
finally
{
tokenizer.Close();
}
}
}
/// <summary>
/// Old algorithm designed to work with iTextSharp 4.1.6. Use iTextSharp version >= 5 if possible (license changes were made).
/// </summary>
/// <param name="input"></param>
/// <returns></returns>
internal static string ExtractTextFromPdfBytes(byte[] input)
{
if (input == null || input.Length == 0)
{
return("");
}
var result = new StringBuilder();
var tokeniser = new PRTokeniser(input);
try
{
while (tokeniser.NextToken())
{
var tknType = tokeniser.TokenType;
var tknValue = tokeniser.StringValue.Replace('\0', ' ');
if (tknType == PRTokeniser.TK_STRING)
{
result.Append(tknValue);
}
else
{
switch (tknValue)
{
case "-600":
result.Append(" ");
break;
case "TJ":
result.Append(" ");
break;
}
}
}
}
finally
{
tokeniser.Close();
}
return(result.ToString());
}
public List <DataTable> Load(MemoryStream stream)
{
var tables = new List <DataTable>();
var sb = new StringBuilder();
var reader = new PdfReader(stream);
for (int page = 1; page <= reader.NumberOfPages; page++)
{
var cpage = reader.GetPageN(page);
var content = cpage.Get(PdfName.CONTENTS);
var ir = (PRIndirectReference)content;
var value = reader.GetPdfObject(ir.Number);
if (value.IsStream())
{
PRStream prstream = (PRStream)value;
var streamBytes = PdfReader.GetStreamBytes(prstream);
var tokenizer = new PRTokeniser(new RandomAccessFileOrArray(streamBytes));
try
{
while (tokenizer.NextToken())
{
if (tokenizer.TokenType == PRTokeniser.TK_STRING)
{
string str = tokenizer.StringValue;
sb.AppendLine(str);
}
}
}
finally
{
tokenizer.Close();
}
}
}
Console.WriteLine(sb.ToString());
return(tables);
}
static void Main(string[] args)
{
string pdfPath = "C:\\mypdf.pdf";
PdfReader reader = new PdfReader(pdfPath);
StringBuilder sb = new StringBuilder();
for (int page = 1; page <= reader.NumberOfPages; page++)
{
var cpage = reader.GetPageN(page);
var content = cpage.Get(PdfName.CONTENTS);
var ir = (PRIndirectReference)content;
var value = reader.GetPdfObject(ir.Number);
if (value.IsStream())
{
PRStream stream = (PRStream)value;
var streamBytes = PdfReader.GetStreamBytes(stream);
var tokenizer = new PRTokeniser(new RandomAccessFileOrArray(streamBytes));
try
{
while (tokenizer.NextToken())
{
if (tokenizer.TokenType == PRTokeniser.TK_STRING)
{
string str = tokenizer.StringValue;
sb.Append(str);
}
}
}
finally
{
tokenizer.Close();
}
}
}
Console.Write("PDF Content:" + Environment.NewLine);
Console.Write(sb.ToString());
Console.Write(Environment.NewLine + "--EOF--");
}
private static void ParseCid(String cmapName, AbstractCMap cmap, ICidLocation location, int level)
{
if (level >= MAXLEVEL)
{
return;
}
PRTokeniser inp = location.GetLocation(cmapName);
try {
List <PdfObject> list = new List <PdfObject>();
PdfContentParser cp = new PdfContentParser(inp);
int maxExc = 50;
while (true)
{
try {
cp.Parse(list);
}
catch {
if (--maxExc < 0)
{
break;
}
continue;
}
if (list.Count == 0)
{
break;
}
String last = list[list.Count - 1].ToString();
if (level == 0 && list.Count == 3 && last.Equals(DEF))
{
PdfObject key = list[0];
if (PdfName.REGISTRY.Equals(key))
{
cmap.Registry = list[1].ToString();
}
else if (PdfName.ORDERING.Equals(key))
{
cmap.Ordering = list[1].ToString();
}
else if (CMAPNAME.Equals(key))
{
cmap.Name = list[1].ToString();
}
else if (PdfName.SUPPLEMENT.Equals(key))
{
try {
cmap.Supplement = ((PdfNumber)list[1]).IntValue;
}
catch {}
}
}
else if ((last.Equals(ENDCIDCHAR) || last.Equals(ENDBFCHAR)) && list.Count >= 3)
{
int lmax = list.Count - 2;
for (int k = 0; k < lmax; k += 2)
{
if (list[k] is PdfString)
{
cmap.AddChar((PdfString)list[k], list[k + 1]);
}
}
}
else if ((last.Equals(ENDCIDRANGE) || last.Equals(ENDBFRANGE)) && list.Count >= 4)
{
int lmax = list.Count - 3;
for (int k = 0; k < lmax; k += 3)
{
if (list[k] is PdfString && list[k + 1] is PdfString)
{
cmap.AddRange((PdfString)list[k], (PdfString)list[k + 1], list[k + 2]);
}
}
}
else if (last.Equals(USECMAP) && list.Count == 2 && list[0] is PdfName)
{
ParseCid(PdfName.DecodeName(list[0].ToString()), cmap, location, level + 1);
}
}
}
finally {
inp.Close();
}
}
