/* Returns 1 on error */
static l_int32
testcomp(const char *filename,
PIX *pix,
l_int32 comptype)
{
l_int32 format, sameformat, sameimage;
FILE *fp;
PIX *pixt;
fp = lept_fopen(filename, "rb");
findFileFormatStream(fp, &format);
sameformat = TRUE;
if (format != comptype) {
fprintf(stderr, "File %s has format %d, not comptype %d\n",
filename, format, comptype);
sameformat = FALSE;
}
lept_fclose(fp);
pixt = pixRead(filename);
pixEqual(pix, pixt, &sameimage);
pixDestroy(&pixt);
if (!sameimage)
fprintf(stderr, "Write/read fail for file %s with format %d\n",
filename, format);
return (!sameformat || !sameimage);
}
/*!
* fileFormatIsTiff()
*
* Input: fp (file stream)
* Return: 1 if file is tiff; 0 otherwise or on error
*/
l_int32
fileFormatIsTiff(FILE *fp)
{
l_int32 format;
PROCNAME("fileFormatIsTiff");
if (!fp)
return ERROR_INT("stream not defined", procName, 0);
findFileFormatStream(fp, &format);
if (format == IFF_TIFF || format == IFF_TIFF_PACKBITS ||
format == IFF_TIFF_RLE || format == IFF_TIFF_G3 ||
format == IFF_TIFF_G4 || format == IFF_TIFF_LZW ||
format == IFF_TIFF_ZIP)
return 1;
else
return 0;
}
/*!
* findFileFormat()
*
* Input: filename
* &format (<return>)
* Return: 0 if OK, 1 on error or if format is not recognized
*/
l_int32
findFileFormat(const char *filename,
l_int32 *pformat)
{
l_int32 ret;
FILE *fp;
PROCNAME("findFileFormat");
if (!pformat)
return ERROR_INT("&format not defined", procName, 1);
*pformat = IFF_UNKNOWN;
if (!filename)
return ERROR_INT("filename not defined", procName, 1);
if ((fp = fopenReadStream(filename)) == NULL)
return ERROR_INT("image file not found", procName, 1);
ret = findFileFormatStream(fp, pformat);
fclose(fp);
return ret;
}
/*!
* pixReadHeader()
*
* Input: filename (with full pathname or in local directory)
* &format (<optional return> file format)
* &w, &h (<optional returns> width and height)
* &bps <optional return> bits/sample
* &spp <optional return> samples/pixel (1, 3 or 4)
* &iscmap (<optional return> 1 if cmap exists; 0 otherwise)
* Return: 0 if OK, 1 on error
*
* Notes:
* (1) This reads the actual headers for jpeg, png, tiff and pnm.
* For bmp and gif, we cheat and read the entire file into a pix,
* from which we extract the "header" information.
*/
l_int32
pixReadHeader(const char *filename,
l_int32 *pformat,
l_int32 *pw,
l_int32 *ph,
l_int32 *pbps,
l_int32 *pspp,
l_int32 *piscmap)
{
l_int32 format, ret, w, h, d, bps, spp, iscmap;
l_int32 type; /* ignored */
FILE *fp;
PIX *pix;
PROCNAME("pixReadHeader");
if (pw) *pw = 0;
if (ph) *ph = 0;
if (pbps) *pbps = 0;
if (pspp) *pspp = 0;
if (piscmap) *piscmap = 0;
if (pformat) *pformat = 0;
iscmap = 0; /* init to false */
if (!filename)
return ERROR_INT("filename not defined", procName, 1);
if ((fp = fopenReadStream(filename)) == NULL)
return ERROR_INT("image file not found", procName, 1);
findFileFormatStream(fp, &format);
fclose(fp);
switch (format)
{
case IFF_BMP: /* cheating: reading the entire file */
if ((pix = pixRead(filename)) == NULL)
return ERROR_INT( "bmp: pix not read", procName, 1);
pixGetDimensions(pix, &w, &h, &d);
if (pixGetColormap(pix))
iscmap = 1;
pixDestroy(&pix);
bps = (d == 32) ? 8 : d;
spp = (d == 32) ? 3 : 1;
break;
case IFF_JFIF_JPEG:
ret = readHeaderJpeg(filename, &w, &h, &spp, NULL, NULL);
bps = 8;
if (ret)
return ERROR_INT( "jpeg: no header info returned", procName, 1);
break;
case IFF_PNG:
ret = readHeaderPng(filename, &w, &h, &bps, &spp, &iscmap);
if (ret)
return ERROR_INT( "png: no header info returned", procName, 1);
break;
case IFF_TIFF:
case IFF_TIFF_PACKBITS:
case IFF_TIFF_RLE:
case IFF_TIFF_G3:
case IFF_TIFF_G4:
case IFF_TIFF_LZW:
case IFF_TIFF_ZIP:
/* Reading page 0 by default; possibly redefine format */
ret = readHeaderTiff(filename, 0, &w, &h, &bps, &spp, NULL, &iscmap,
&format);
if (ret)
return ERROR_INT( "tiff: no header info returned", procName, 1);
break;
case IFF_PNM:
ret = readHeaderPnm(filename, &w, &h, &d, &type, &bps, &spp);
if (ret)
return ERROR_INT( "pnm: no header info returned", procName, 1);
break;
case IFF_GIF: /* cheating: reading the entire file */
if ((pix = pixRead(filename)) == NULL)
return ERROR_INT( "gif: pix not read", procName, 1);
pixGetDimensions(pix, &w, &h, &d);
pixDestroy(&pix);
iscmap = 1; /* always colormapped; max 256 colors */
spp = 1;
bps = d;
break;
case IFF_JP2:
ret = readHeaderJp2k(filename, &w, &h, &bps, &spp);
break;
case IFF_WEBP:
if (readHeaderWebP(filename, &w, &h, &spp))
return ERROR_INT( "webp: no header info returned", procName, 1);
bps = 8;
break;
case IFF_SPIX:
ret = readHeaderSpix(filename, &w, &h, &bps, &spp, &iscmap);
if (ret)
return ERROR_INT( "spix: no header info returned", procName, 1);
break;
case IFF_UNKNOWN:
L_ERROR("unknown format in file %s\n", procName, filename);
return 1;
break;
}
if (pw) *pw = w;
if (ph) *ph = h;
if (pbps) *pbps = bps;
if (pspp) *pspp = spp;
if (piscmap) *piscmap = iscmap;
if (pformat) *pformat = format;
return 0;
}
/*!
* pixReadStream()
*
* Input: fp (file stream)
* hint (bitwise OR of L_HINT_* values for jpeg; use 0 for no hint)
* Return: pix if OK; null on error
*
* Notes:
* (1) The hint only applies to jpeg.
*/
PIX *
pixReadStream(FILE *fp,
l_int32 hint)
{
l_int32 format, ret;
l_uint8 *comment;
PIX *pix;
PROCNAME("pixReadStream");
if (!fp)
return (PIX *)ERROR_PTR("stream not defined", procName, NULL);
pix = NULL;
findFileFormatStream(fp, &format);
switch (format)
{
case IFF_BMP:
if ((pix = pixReadStreamBmp(fp)) == NULL )
return (PIX *)ERROR_PTR( "bmp: no pix returned", procName, NULL);
break;
// case IFF_JFIF_JPEG:
// if ((pix = pixReadStreamJpeg(fp, READ_24_BIT_COLOR, 1, NULL, hint))
// == NULL)
// return (PIX *)ERROR_PTR( "jpeg: no pix returned", procName, NULL);
// ret = fgetJpegComment(fp, &comment);
// if (!ret && comment)
// pixSetText(pix, (char *)comment);
// FREE(comment);
// break;
case IFF_PNG:
if ((pix = pixReadStreamPng(fp)) == NULL)
return (PIX *)ERROR_PTR("png: no pix returned", procName, NULL);
break;
case IFF_TIFF:
case IFF_TIFF_PACKBITS:
case IFF_TIFF_RLE:
case IFF_TIFF_G3:
case IFF_TIFF_G4:
case IFF_TIFF_LZW:
case IFF_TIFF_ZIP:
if ((pix = pixReadStreamTiff(fp, 0)) == NULL) /* page 0 by default */
return (PIX *)ERROR_PTR("tiff: no pix returned", procName, NULL);
break;
case IFF_PNM:
if ((pix = pixReadStreamPnm(fp)) == NULL)
return (PIX *)ERROR_PTR("pnm: no pix returned", procName, NULL);
break;
case IFF_GIF:
if ((pix = pixReadStreamGif(fp)) == NULL)
return (PIX *)ERROR_PTR("gif: no pix returned", procName, NULL);
break;
case IFF_JP2:
if ((pix = pixReadStreamJp2k(fp, 1, NULL, 0)) == NULL)
return (PIX *)ERROR_PTR("jp2: no pix returned", procName, NULL);
break;
case IFF_WEBP:
if ((pix = pixReadStreamWebP(fp)) == NULL)
return (PIX *)ERROR_PTR("webp: no pix returned", procName, NULL);
break;
case IFF_SPIX:
if ((pix = pixReadStreamSpix(fp)) == NULL)
return (PIX *)ERROR_PTR("spix: no pix returned", procName, NULL);
break;
case IFF_UNKNOWN:
return (PIX *)ERROR_PTR( "Unknown format: no pix returned",
procName, NULL);
break;
}
if (pix)
pixSetInputFormat(pix, format);
return pix;
}
// TODO(jbreiden) I hear that you can pull the flate stream out
// of a PNG file and, by mentioning the predictor in the PDF object,
// make most of them work without transcoding. If so that's a big win
// versus what we do now. Try it out.
bool TessPDFRenderer::fileToPDFObj(char *filename, long int objnum,
char **pdf_object,
long int *pdf_object_size) {
char b1[kBasicBufSize];
char b2[kBasicBufSize];
if (!pdf_object_size || !pdf_object)
return false;
*pdf_object = NULL;
*pdf_object_size = 0;
if (!filename)
return false;
FILE *fp = fopen(filename, "rb");
if (!fp)
return false;
int format;
findFileFormatStream(fp, &format);
if (format != IFF_JFIF_JPEG) {
fclose(fp);
return false;
}
fseek(fp, 0, SEEK_END);
long int jpeg_size = ftell(fp);
fseek(fp, 0, SEEK_SET);
int spp, cmyk, w, h;
freadHeaderJpeg(fp, &w, &h, &spp, NULL, &cmyk);
const char *colorspace;
switch (spp) {
case 1:
colorspace = "/DeviceGray";
break;
case 3:
colorspace = "/DeviceRGB";
break;
case 4:
if (cmyk)
colorspace = "/DeviceCMYK";
else
return false;
break;
default:
return false;
}
// IMAGE
snprintf(b1, sizeof(b1),
"%ld 0 obj\n"
"<<\n"
" /Length %ld\n"
" /Subtype /Image\n"
" /ColorSpace %s\n"
" /Width %d\n"
" /Height %d\n"
" /BitsPerComponent 8\n"
" /Filter /DCTDecode\n"
">>\n"
"stream\n", objnum, jpeg_size,
colorspace, w, h);
size_t b1_len = strlen(b1);
snprintf(b2, sizeof(b2),
"\n"
"endstream\n"
"endobj\n");
size_t b2_len = strlen(b2);
*pdf_object_size = b1_len + jpeg_size + b2_len;
*pdf_object = new char[*pdf_object_size];
if (!pdf_object)
return false;
memcpy(*pdf_object, b1, b1_len);
if (static_cast<int>(fread(*pdf_object + b1_len, 1, jpeg_size, fp)) !=
jpeg_size) {
delete[] pdf_object;
return false;
}
memcpy(*pdf_object + b1_len + jpeg_size, b2, b2_len);
fclose(fp);
return true;
}