Sample Code
Windows Driver Samples/ GenPrint Print Processor Sample/ C++/ text.cpp/
/*++
Copyright (c) 1990-2003 Microsoft Corporation
All Rights Reserved
Abstract:
Routines to facilitate printing of text jobs.
--*/
#include "local.h"
#define FLAG_CR_STATE 0x1
#define FLAG_TAB_STATE 0x2
#define FLAG_DBCS_SPLIT 0x8
#define FLAG_FF 0x10
#define FLAG_LF 0x20
#define FLAG_CR 0x40
#define FLAG_TRANSLATE_LF 0x80
#define FLAG_TRANSLATE_CR 0x100
const WCHAR gszNoTranslateCRLF[] = L"Winprint_TextNoTranslation";
const WCHAR gszNoTranslateCR[] = L"Winprint_TextNoCRTranslation";
const WCHAR gszTransparency[] = L"Transparency";
/** Prototypes for functions in this file **/
PBYTE
GetTabbedLineFromBuffer(
_In_ PBYTE pSrcBuffer,
_In_ PBYTE pSrcBufferEnd,
_Out_writes_bytes_(CharsPerLine) PBYTE pDestBuffer,
_In_ ULONG CharsPerLine,
_In_ ULONG TabExpansionSize,
_In_ ULONG Encoding,
_Inout_ PULONG pLength,
_Inout_ PULONG pTabBase,
_Inout_ PDWORD pfdwFlags
);
/*++
*******************************************************************
P r i n t T e x t J o b
Routine Description:
Prints a text data job.
Arguments:
pData => Data structure for this job
pDocumentName => Name of this document
Return Value:
TRUE if successful
FALSE if failed - GetLastError() will return reason.
*******************************************************************
--*/
BOOL
PrintTextJob(
IN PPRINTPROCESSORDATA pData,
_In_ IN LPWSTR pDocumentName)
{
DOCINFO DocInfo;
LOGFONT LogFont;
CHARSETINFO CharSetInfo;
HFONT hOldFont, hFont = NULL;
DWORD Copies;
BOOL rc;
DWORD NoRead;
DWORD CurrentLine;
DWORD CurrentCol;
HANDLE hPrinter = NULL;
BYTE *ReadBufferStart = NULL;
PBYTE pLineBuffer = NULL;
PBYTE pReadBuffer = NULL;
PBYTE pReadBufferEnd = NULL;
ULONG CharHeight, CharWidth, CharsPerLine, LinesPerPage;
ULONG PageWidth, PageHeight;
ULONG Length, TabBase;
BOOL ReadAll;
TEXTMETRIC tm;
DWORD fdwFlags;
DWORD Encoding;
DWORD SplitSize;
BOOL ReturnValue = FALSE;
BOOL bAbortDoc = FALSE;
DWORD dwNeeded;
DWORD dwNoTranslate = 0;
DWORD dwNoTranslateCR = 0;
DWORD dwTransparent = 0;
INT iBkMode = OPAQUE;
DocInfo.lpszDocName = pData->pDocument; /* Document name */
DocInfo.lpszOutput = NULL; /* Output file */
DocInfo.lpszDatatype = NULL; /* Datatype */
DocInfo.cbSize = sizeof(DOCINFO); /* Size of the structure */
//
// Go figure out the size of the form on the printer. We do this
// by calling GetTextMetrics, which gives us the font size of the
// printer font, then getting the form size and calculating the
// number of characters that will fit. In other cases we treat it as ANSI text.
// Currently the codepage context is fixed to the system default codepage.
//
Encoding = GetACP();
//
// Create FIXED PITCH font and select
//
hOldFont = 0;
ZeroMemory(&CharSetInfo, sizeof(CHARSETINFO));
if (TranslateCharsetInfo((PDWORD)UIntToPtr(Encoding), &CharSetInfo, TCI_SRCCODEPAGE))
{
ZeroMemory(&LogFont, sizeof(LOGFONT));
LogFont.lfWeight = 400;
LogFont.lfCharSet = (BYTE)CharSetInfo.ciCharset;
LogFont.lfPitchAndFamily = FIXED_PITCH;
hFont = CreateFontIndirect(&LogFont);
hOldFont = (HFONT) SelectObject(pData->hDC, hFont);
}
if (!GetTextMetrics(pData->hDC, &tm)) {
// Essential text processing computation failed
goto Done;
}
CharHeight = tm.tmHeight + tm.tmExternalLeading;
CharWidth = tm.tmAveCharWidth;
if (!CharWidth || !CharHeight) {
// Essential text processing computation failed
goto Done;
}
//
// Calculate most fittable characters' number to one line.
//
PageWidth = GetDeviceCaps(pData->hDC, DESKTOPHORZRES);
PageHeight = GetDeviceCaps(pData->hDC, DESKTOPVERTRES);
CharsPerLine = PageWidth / CharWidth;
LinesPerPage = PageHeight / CharHeight;
if (!CharsPerLine || !LinesPerPage) {
// Essential text processing computation failed
goto Done;
}
/** Allocate a buffer for one line of text **/
pLineBuffer = (PBYTE)AllocSplMem(CharsPerLine + 5);
if (!pLineBuffer) {
goto Done;
}
/** Let the printer know we are starting a new document **/
if (!StartDoc(pData->hDC, (LPDOCINFO)&DocInfo)) {
goto Done;
}
ReadBufferStart = (BYTE *)AllocSplMem(READ_BUFFER_SIZE);
if (!ReadBufferStart) {
goto Done;
}
/** Print the data pData->Copies times **/
Copies = pData->Copies;
while (Copies--) {
/**
Loop, getting data and sending it to the printer. This also
takes care of pausing and cancelling print jobs by checking
the processor's status flags while printing. The way we do
this is to read in some data from the printer. We will then
pull data, one tabbed line at a time from there and print
it. If the last bit of data in the buffer does not make up
a whole line, we call GetTabbedLineFromBuffer() with a non-
zero Length, which indicates that there are chars left
from the previous read.
**/
TabBase = 0;
Length = 0;
fdwFlags = FLAG_TRANSLATE_CR | FLAG_TRANSLATE_LF;
CurrentLine = 0;
CurrentCol = 0;
/**
Open the printer. If it fails, return. This also sets up the
pointer for the ReadPrinter calls.
**/
if (!OpenPrinter(pDocumentName, &hPrinter, NULL)) {
hPrinter = NULL;
bAbortDoc = TRUE;
goto Done;
}
//
// Call GetPrinterData to see if the queue wants no LF/CR processing.
//
if( GetPrinterData( hPrinter,
(LPWSTR)gszNoTranslateCRLF,
NULL,
(PBYTE)&dwNoTranslate,
sizeof( dwNoTranslate ),
&dwNeeded ) == ERROR_SUCCESS ){
if( dwNoTranslate ){
fdwFlags &= ~( FLAG_TRANSLATE_CR | FLAG_TRANSLATE_LF );
}
}
//
// Call GetPrinterData to see if the queue wants no CR processing.
//
if( GetPrinterData( hPrinter,
(LPWSTR)gszNoTranslateCR,
NULL,
(PBYTE)&dwNoTranslateCR,
sizeof( dwNoTranslateCR ),
&dwNeeded ) == ERROR_SUCCESS ){
if( dwNoTranslateCR ){
fdwFlags &= ~FLAG_TRANSLATE_CR;
if( GetPrinterData( hPrinter,
(LPWSTR)gszTransparency,
NULL,
(PBYTE)&dwTransparent,
sizeof( dwTransparent ),
&dwNeeded ) == ERROR_SUCCESS ){
if( dwTransparent ){
iBkMode = SetBkMode( pData->hDC, TRANSPARENT );
}
}
}
}
if (StartPage(pData->hDC) == SP_ERROR) {
bAbortDoc = TRUE;
goto Done;
}
/** ReadAll indicates if we are on the last line of the file **/
ReadAll = FALSE;
/**
This next do loop continues until we have read all of the
data for the print job.
**/
do {
if (fdwFlags & FLAG_DBCS_SPLIT) {
SplitSize = (DWORD)(pReadBufferEnd - pReadBuffer);
memcpy(ReadBufferStart, pReadBuffer, SplitSize);
fdwFlags &= ~FLAG_DBCS_SPLIT;
}
else {
SplitSize = 0;
}
rc = ReadPrinter(hPrinter,
(ReadBufferStart + SplitSize),
(READ_BUFFER_SIZE - SplitSize),
&NoRead);
if (!rc || !NoRead) {
ReadAll = TRUE;
} else {
/** Pick up a pointer to the end of the data **/
pReadBuffer = ReadBufferStart;
pReadBufferEnd = ReadBufferStart + SplitSize + NoRead;
}
/**
This loop will process all the data that we have
just read from the printer.
**/
do {
if (!ReadAll) {
/**
Length on entry holds the length of any
residual chars from the last line that we couldn't
print out because we ran out of characters on
the ReadPrinter buffer.
**/
pReadBuffer = GetTabbedLineFromBuffer(
pReadBuffer,
pReadBufferEnd,
pLineBuffer,
CharsPerLine - CurrentCol,
pData->TabSize,
Encoding,
&Length,
&TabBase,
&fdwFlags );
/**
If pReadBuffer == NULL, then we have
exhausted the read buffer and we need to ReadPrinter
again and save the last line chars. Length holds
the number of characters on this partial line,
so the next time we call ReadPrinter we will
pickup where we left off.
The only time we'll get residual chars is if:
1. The last line ends w/o ff/lf/cr ("Hello\EOF")
In this case we should TextOutA the last line
and then quit.
(In this case, don't break here; go ahead and
print, then we'll break out below in the do..while.)
2. The ReadPrinter last byte is in the middle of a line.
Here we should read the next chunk and add the
new characters at the end of the chars we just read.
(In this case, we should break and leave Length
as it is so we will read again and append to the
buffer, beginning at Length.)
**/
if (!pReadBuffer || (fdwFlags & FLAG_DBCS_SPLIT))
break;
}
/** If the print processor is paused, wait for it to be resumed **/
if (pData->fsStatus & PRINTPROCESSOR_PAUSED) {
WaitForSingleObject(pData->semPaused, INFINITE);
}
/** If the job has been aborted, clean up and leave **/
if (pData->fsStatus & PRINTPROCESSOR_ABORTED) {
ReturnValue = TRUE;
bAbortDoc = TRUE;
goto Done;
}
/** Write the data to the printer **/
/** Make sure Length is not zero **/
/** TextOut will fail if Length == 0 **/
if (Length) {
/**
We may have a number of newlines pending, that
may push us to the next page (or even next-next
page).
**/
while (CurrentLine >= LinesPerPage) {
/**
We need a new page; always defer this to the
last second to prevent extra pages from coming out.
**/
if (EndPage(pData->hDC) == SP_ERROR ||
StartPage(pData->hDC) == SP_ERROR) {
bAbortDoc = TRUE;
goto Done;
}
CurrentLine -= LinesPerPage;
}
#pragma prefast(suppress:__WARNING_ANSI_APICALL, "By design. This function is inherently handling only ANSI text print jobs.")
if (TextOutA(pData->hDC,
CurrentCol * CharWidth,
CurrentLine * CharHeight,
(LPCSTR)pLineBuffer,
Length) == FALSE) {
ODS(("TextOut() failed\n"));
bAbortDoc = TRUE;
goto Done;
}
CurrentCol += Length;
}
/**
Even if the length is zero, increment the line.
Should happen when the character is only 0x0D or 0x0A.
**/
if (fdwFlags & FLAG_CR) {
CurrentCol=0;
fdwFlags &= ~FLAG_CR;
}
if (fdwFlags & FLAG_LF) {
CurrentLine++;
fdwFlags &= ~FLAG_LF;
}
/**
We need a new page. Set the current line to the
end of the page. We could do a End/StartPage
sequence, but this may cause a blank page to get
ejected.
Note: this code will avoid printing out pages that
consist of formfeeds only (if you have a page with a
space in it, that counts as text).
**/
if (fdwFlags & FLAG_FF) {
CurrentLine = LinesPerPage;
CurrentCol = 0;
fdwFlags &= ~FLAG_FF;
}
/**
We have done the text out, so these characters have
been successfully printed. Zero out Length
so these characters won't be printed again
**/
Length = 0;
/**
We only terminate this loop if we run out of chars
or we run out of read buffer.
**/
} while (pReadBuffer && pReadBuffer != pReadBufferEnd);
/** Keep going until we get the last line **/
} while (!ReadAll);
if (EndPage(pData->hDC) == SP_ERROR) {
bAbortDoc = TRUE;
goto Done;
}
/**
Close the printer - we open/close the printer for each
copy so the data pointer will rewind.
**/
ClosePrinter(hPrinter);
hPrinter = NULL;
} /* While copies to print */
/** Let the printer know that we are done printing **/
EndDoc(pData->hDC);
ReturnValue = TRUE;
Done:
if (dwTransparent)
SetBkMode( pData->hDC, iBkMode );
if (hPrinter)
ClosePrinter(hPrinter);
if (bAbortDoc)
AbortDoc(pData->hDC);
if (pLineBuffer)
FreeSplMem(pLineBuffer);
if (hOldFont)
{
SelectObject(pData->hDC, hOldFont);
DeleteObject(hFont);
}
if (ReadBufferStart)
{
FreeSplMem(ReadBufferStart);
}
return ReturnValue;
}
/*++
*******************************************************************
G e t T a b b e d L i n e F r o m B u f f e r
Routine Description:
This routine, given a buffer of text, will pull out a
line of tab-expanded text. This is used for tab
expansion of text data jobs.
Arguments:
pSrcBuffer => Start of source buffer.
pSrcBufferEnd => End of source buffer
pDestBuffer => Start of destination buffer
CharsPerLine => Number of characters on a line
TabExpansionSize=> Number of spaces in a tab
Encoding => Code page
pLength => Length of chars from prev line, rets current
pTabBase => New 0 offset for tabbing
pfdwFlags => State
Return Value:
PBYTE => Place left off in the source buffer. This should
be passed in on the next call. If we ran out of
data in the source, this will be unchanged.
*******************************************************************
--*/
PBYTE
GetTabbedLineFromBuffer(
_In_ PBYTE pSrcBuffer,
_In_ PBYTE pSrcBufferEnd,
_Out_writes_bytes_(CharsPerLine) PBYTE pDestBuffer,
_In_ ULONG CharsPerLine,
_In_ ULONG TabExpansionSize,
_In_ ULONG Encoding,
_Inout_ PULONG pLength,
_Inout_ PULONG pTabBase,
_Inout_ PDWORD pfdwFlags
)
{
ULONG current_pos;
ULONG expand, i;
ULONG TabBase = *pTabBase;
ULONG TabBaseLeft = TabExpansionSize-TabBase;
PBYTE pDestBufferEnd = pDestBuffer + CharsPerLine;
/**
If the tab pushed us past the end of the last line, then we need to
add it back to the next one.
**/
if (TabBase && ( *pfdwFlags & FLAG_TAB_STATE )) {
current_pos = 0;
i=TabBase;
while (i-- && (pDestBuffer < pDestBufferEnd)) {
*pDestBuffer++ = ' ';
current_pos++;
}
/**
If we ran out of room again, return. This means that
the tab expansion size is greater than we can fit on
one line.
**/
if (pDestBuffer >= pDestBufferEnd) {
*pLength = current_pos;
*pTabBase -= CharsPerLine;
//
// We need to move to the next line.
//
*pfdwFlags |= FLAG_LF | FLAG_CR;
return pSrcBuffer;
}
*pfdwFlags &= ~FLAG_TAB_STATE;
} else {
/** We may have some chars from the previous ReadPrinter **/
current_pos = *pLength;
pDestBuffer += current_pos;
}
while (pSrcBuffer < pSrcBufferEnd) {
/** Now process other chars **/
switch (*pSrcBuffer) {
case 0x0C:
/** Found a FF. Quit and indicate we need to start a new page **/
*pTabBase = 0;
*pfdwFlags |= FLAG_FF;
*pfdwFlags &= ~FLAG_CR_STATE;
pSrcBuffer++;
break;
case '\t':
*pfdwFlags &= ~FLAG_CR_STATE;
/**
Handle TAB case. If we are really out of buffer,
then defer now so that the tab will be saved for
the next line.
**/
if (pDestBuffer >= pDestBufferEnd) {
goto ShiftTab;
}
pSrcBuffer++;
/** Figure out how far to expand the tabs **/
expand = TabExpansionSize -
(current_pos + TabBaseLeft) % TabExpansionSize;
/** Expand the tabs **/
for (i = 0; (i < expand) && (pDestBuffer < pDestBufferEnd); i++) {
*pDestBuffer++ = ' ';
}
/**
If we reached the end of our dest buffer,
return and set the number of spaces we have left.
**/
if (pDestBuffer >= pDestBufferEnd) {
*pfdwFlags |= FLAG_TAB_STATE;
goto ShiftTab;
}
/** Update our position counter **/
current_pos += expand;
continue;
case 0x0A:
pSrcBuffer++;
/** If the last char was a CR, ignore this guy **/
if (*pfdwFlags & FLAG_CR_STATE) {
*pfdwFlags &= ~FLAG_CR_STATE;
//
// We are translating CRLF, so if we saw a CR
// immediately before this, then don't do anything.
//
continue;
}
if( *pfdwFlags & FLAG_TRANSLATE_LF ){
//
// If we are translating, then treat a LF as a CRLF pair.
//
*pfdwFlags |= FLAG_LF | FLAG_CR;
/** Found a linefeed. That's it for this line. **/
*pTabBase = 0;
} else {
*pfdwFlags |= FLAG_LF;
}
break;
case 0x0D:
/** Found a carriage return. That's it for this line. **/
*pTabBase = 0;
pSrcBuffer++;
if (*pfdwFlags & FLAG_TRANSLATE_CR) {
//
// If we are translating CRLF, then make the newline
// occur now. This handles the case where we have a
// CR all by itself. Also set the CR flag so if there
// happens to be a LF immediately after this, we don't
// move down another line.
//
*pfdwFlags |= FLAG_CR_STATE | FLAG_LF | FLAG_CR;
} else {
*pfdwFlags |= FLAG_CR;
}
break;
default:
/** Not tab or carriage return, must be simply data **/
*pfdwFlags &= ~FLAG_CR_STATE;
//
// We always check before we are adding a character
// (instead of after) since we may be at the end of a line,
// but we can still process chars like 0x0d 0x0a.
// This happens in MS-DOS printscreen.
//
if (pDestBuffer >= pDestBufferEnd ||
(pDestBuffer + 1 >= pDestBufferEnd) &&
IsDBCSLeadByteEx(Encoding, *pSrcBuffer)) {
ShiftTab:
//
// We must shift the tab over since we are on the
// same line.
//
*pTabBase = (*pTabBase + TabExpansionSize -
(CharsPerLine % TabExpansionSize))
% TabExpansionSize;
*pfdwFlags |= FLAG_LF | FLAG_CR;
break;
}
if (IsDBCSLeadByteEx(Encoding, *pSrcBuffer)) {
// Check if we have trail byte also.
if (pSrcBuffer + 1 >= pSrcBufferEnd) {
*pfdwFlags |= FLAG_DBCS_SPLIT;
break;
}
// Advance source pointer (for lead byte).
*pDestBuffer++ = *pSrcBuffer++;
current_pos++;
}
*pDestBuffer++ = *pSrcBuffer++;
current_pos++;
continue;
}
*pLength = current_pos;
return pSrcBuffer;
}
/** We ran out of source buffer before getting to the EOL **/
*pLength = current_pos;
return NULL;
}
Our Services
-
What our customers say about us?
Read our customer testimonials to find out why our clients keep returning for their projects.
View Testimonials
