Make guessFormatFromContents() a bit more private, since it is not supposed

to be used in new code.


git-svn-id: svn://svn.lyx.org/lyx/lyx-devel/trunk@40790 a592a061-630c-0410-9148-cb99ea01b6c8
This commit is contained in:
Georg Baum 2012-02-21 20:46:37 +00:00
parent 26e5f1a8ec
commit de6e4a2937
3 changed files with 188 additions and 186 deletions

View File

@ -182,6 +182,193 @@ Format const * Formats::getFormat(string const & name) const
} }
namespace {
/** Guess the file format name (as in Format::name()) from contents.
* Normally you don't want to use this directly, but rather
* Formats::getFormatFromFile().
*/
string guessFormatFromContents(FileName const & fn)
{
// the different filetypes and what they contain in one of the first lines
// (dots are any characters). (Herbert 20020131)
// AGR Grace...
// BMP BM...
// EPS %!PS-Adobe-3.0 EPSF...
// FIG #FIG...
// FITS ...BITPIX...
// GIF GIF...
// JPG JFIF
// PDF %PDF-...
// PNG .PNG...
// PBM P1... or P4 (B/W)
// PGM P2... or P5 (Grayscale)
// PPM P3... or P6 (color)
// PS %!PS-Adobe-2.0 or 1.0, no "EPSF"!
// SGI \001\332... (decimal 474)
// TGIF %TGIF...
// TIFF II... or MM...
// XBM ..._bits[]...
// XPM /* XPM */ sometimes missing (f.ex. tgif-export)
// ...static char *...
// XWD \000\000\000\151 (0x00006900) decimal 105
//
// GZIP \037\213 http://www.ietf.org/rfc/rfc1952.txt
// ZIP PK... http://www.halyava.ru/document/ind_arch.htm
// Z \037\235 UNIX compress
// paranoia check
if (fn.empty() || !fn.isReadableFile())
return string();
ifstream ifs(fn.toFilesystemEncoding().c_str());
if (!ifs)
// Couldn't open file...
return string();
// gnuzip
static string const gzipStamp = "\037\213";
// PKZIP
static string const zipStamp = "PK";
// ZIP containers (koffice, openoffice.org etc).
static string const nonzipStamp = "\008\0\0\0mimetypeapplication/";
// compress
static string const compressStamp = "\037\235";
// Maximum strings to read
int const max_count = 50;
int count = 0;
string str;
string format;
bool firstLine = true;
while ((count++ < max_count) && format.empty()) {
if (ifs.eof()) {
LYXERR(Debug::GRAPHICS, "filetools(getFormatFromContents)\n"
<< "\tFile type not recognised before EOF!");
break;
}
getline(ifs, str);
string const stamp = str.substr(0, 2);
if (firstLine && str.size() >= 2) {
// at first we check for a zipped file, because this
// information is saved in the first bytes of the file!
// also some graphic formats which save the information
// in the first line, too.
if (prefixIs(str, gzipStamp)) {
format = "gzip";
} else if (stamp == zipStamp &&
!contains(str, nonzipStamp)) {
format = "zip";
} else if (stamp == compressStamp) {
format = "compress";
// the graphics part
} else if (stamp == "BM") {
format = "bmp";
} else if (stamp == "\001\332") {
format = "sgi";
// PBM family
// Don't need to use str.at(0), str.at(1) because
// we already know that str.size() >= 2
} else if (str[0] == 'P') {
switch (str[1]) {
case '1':
case '4':
format = "pbm";
break;
case '2':
case '5':
format = "pgm";
break;
case '3':
case '6':
format = "ppm";
}
break;
} else if ((stamp == "II") || (stamp == "MM")) {
format = "tiff";
} else if (prefixIs(str,"%TGIF")) {
format = "tgif";
} else if (prefixIs(str,"#FIG")) {
format = "fig";
} else if (prefixIs(str,"GIF")) {
format = "gif";
} else if (str.size() > 3) {
int const c = ((str[0] << 24) & (str[1] << 16) &
(str[2] << 8) & str[3]);
if (c == 105) {
format = "xwd";
}
}
firstLine = false;
}
if (!format.empty())
break;
else if (contains(str,"EPSF"))
// dummy, if we have wrong file description like
// %!PS-Adobe-2.0EPSF"
format = "eps";
else if (contains(str, "Grace"))
format = "agr";
else if (contains(str, "JFIF"))
format = "jpg";
else if (contains(str, "%PDF"))
format = "pdf";
else if (contains(str, "PNG"))
format = "png";
else if (contains(str, "%!PS-Adobe")) {
// eps or ps
ifs >> str;
if (contains(str,"EPSF"))
format = "eps";
else
format = "ps";
}
else if (contains(str, "_bits[]"))
format = "xbm";
else if (contains(str, "XPM") || contains(str, "static char *"))
format = "xpm";
else if (contains(str, "BITPIX"))
format = "fits";
}
if (!format.empty()) {
LYXERR(Debug::GRAPHICS, "Recognised Fileformat: " << format);
return format;
}
LYXERR(Debug::GRAPHICS, "filetools(getFormatFromContents)\n"
<< "\tCouldn't find a known format!");
return string();
}
}
string Formats::getFormatFromFile(FileName const & filename) const string Formats::getFormatFromFile(FileName const & filename) const
{ {
if (filename.empty()) if (filename.empty())
@ -217,7 +404,7 @@ string Formats::getFormatFromFile(FileName const & filename) const
} }
#endif #endif
string const format = filename.guessFormatFromContents(); string const format = guessFormatFromContents(filename);
string const ext = getExtension(filename.absFileName()); string const ext = getExtension(filename.absFileName());
if (isZippedFileFormat(format) && !ext.empty()) { if (isZippedFileFormat(format) && !ext.empty()) {
string const & fmt_name = formats.getFormatFromExtension(ext); string const & fmt_name = formats.getFormatFromExtension(ext);

View File

@ -769,185 +769,6 @@ void FileName::changeExtension(string const & extension)
} }
string FileName::guessFormatFromContents() const
{
// the different filetypes and what they contain in one of the first lines
// (dots are any characters). (Herbert 20020131)
// AGR Grace...
// BMP BM...
// EPS %!PS-Adobe-3.0 EPSF...
// FIG #FIG...
// FITS ...BITPIX...
// GIF GIF...
// JPG JFIF
// PDF %PDF-...
// PNG .PNG...
// PBM P1... or P4 (B/W)
// PGM P2... or P5 (Grayscale)
// PPM P3... or P6 (color)
// PS %!PS-Adobe-2.0 or 1.0, no "EPSF"!
// SGI \001\332... (decimal 474)
// TGIF %TGIF...
// TIFF II... or MM...
// XBM ..._bits[]...
// XPM /* XPM */ sometimes missing (f.ex. tgif-export)
// ...static char *...
// XWD \000\000\000\151 (0x00006900) decimal 105
//
// GZIP \037\213 http://www.ietf.org/rfc/rfc1952.txt
// ZIP PK... http://www.halyava.ru/document/ind_arch.htm
// Z \037\235 UNIX compress
// paranoia check
if (empty() || !isReadableFile())
return string();
ifstream ifs(toFilesystemEncoding().c_str());
if (!ifs)
// Couldn't open file...
return string();
// gnuzip
static string const gzipStamp = "\037\213";
// PKZIP
static string const zipStamp = "PK";
// ZIP containers (koffice, openoffice.org etc).
static string const nonzipStamp = "\008\0\0\0mimetypeapplication/";
// compress
static string const compressStamp = "\037\235";
// Maximum strings to read
int const max_count = 50;
int count = 0;
string str;
string format;
bool firstLine = true;
while ((count++ < max_count) && format.empty()) {
if (ifs.eof()) {
LYXERR(Debug::GRAPHICS, "filetools(getFormatFromContents)\n"
<< "\tFile type not recognised before EOF!");
break;
}
getline(ifs, str);
string const stamp = str.substr(0, 2);
if (firstLine && str.size() >= 2) {
// at first we check for a zipped file, because this
// information is saved in the first bytes of the file!
// also some graphic formats which save the information
// in the first line, too.
if (prefixIs(str, gzipStamp)) {
format = "gzip";
} else if (stamp == zipStamp &&
!contains(str, nonzipStamp)) {
format = "zip";
} else if (stamp == compressStamp) {
format = "compress";
// the graphics part
} else if (stamp == "BM") {
format = "bmp";
} else if (stamp == "\001\332") {
format = "sgi";
// PBM family
// Don't need to use str.at(0), str.at(1) because
// we already know that str.size() >= 2
} else if (str[0] == 'P') {
switch (str[1]) {
case '1':
case '4':
format = "pbm";
break;
case '2':
case '5':
format = "pgm";
break;
case '3':
case '6':
format = "ppm";
}
break;
} else if ((stamp == "II") || (stamp == "MM")) {
format = "tiff";
} else if (prefixIs(str,"%TGIF")) {
format = "tgif";
} else if (prefixIs(str,"#FIG")) {
format = "fig";
} else if (prefixIs(str,"GIF")) {
format = "gif";
} else if (str.size() > 3) {
int const c = ((str[0] << 24) & (str[1] << 16) &
(str[2] << 8) & str[3]);
if (c == 105) {
format = "xwd";
}
}
firstLine = false;
}
if (!format.empty())
break;
else if (contains(str,"EPSF"))
// dummy, if we have wrong file description like
// %!PS-Adobe-2.0EPSF"
format = "eps";
else if (contains(str, "Grace"))
format = "agr";
else if (contains(str, "JFIF"))
format = "jpg";
else if (contains(str, "%PDF"))
format = "pdf";
else if (contains(str, "PNG"))
format = "png";
else if (contains(str, "%!PS-Adobe")) {
// eps or ps
ifs >> str;
if (contains(str,"EPSF"))
format = "eps";
else
format = "ps";
}
else if (contains(str, "_bits[]"))
format = "xbm";
else if (contains(str, "XPM") || contains(str, "static char *"))
format = "xpm";
else if (contains(str, "BITPIX"))
format = "fits";
}
if (!format.empty()) {
LYXERR(Debug::GRAPHICS, "Recognised Fileformat: " << format);
return format;
}
LYXERR(Debug::GRAPHICS, "filetools(getFormatFromContents)\n"
<< "\tCouldn't find a known format!");
return string();
}
docstring const FileName::relPath(string const & path) const docstring const FileName::relPath(string const & path) const
{ {
// FIXME UNICODE // FIXME UNICODE

View File

@ -169,12 +169,6 @@ public:
*/ */
void changeExtension(std::string const & extension); void changeExtension(std::string const & extension);
/** Guess the file format name (as in Format::name()) from contents.
Normally you don't want to use this directly, but rather
Formats::getFormatFromFile().
*/
std::string guessFormatFromContents() const;
static FileName fromFilesystemEncoding(std::string const & name); static FileName fromFilesystemEncoding(std::string const & name);
/// (securely) create a temporary file with the given mask. /// (securely) create a temporary file with the given mask.
/// \p mask must be in filesystem encoding, if it contains a /// \p mask must be in filesystem encoding, if it contains a