zaubentrucker/muele-marlin
1
0
Fork 0
Code Issues Pull requests Actions Packages Projects Releases Wiki Activity

UTF longname followup (#20115)

Browse source
This commit is contained in:
Jason Smith 2020-11-12 20:26:49 -08:00 committed by Scott Lahteine
parent 26a1cd9a00
commit 3ee61e2084
1 changed files with 31 additions and 32 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

45
Marlin/src/sd/SdBaseFile.cpp
View file

@ -1110,7 +1110,7 @@ int8_t SdBaseFile::readDir(dir_t* dir, char* longFilename) {
// needs static bytes addressing. So here just store full UTF-16LE words to re-convert later. // needs static bytes addressing. So here just store full UTF-16LE words to re-convert later.
uint16_t idx = (n + i) * 2; // This is fixed as FAT LFN always contain UTF-16LE encoding uint16_t idx = (n + i) * 2; // This is fixed as FAT LFN always contain UTF-16LE encoding
longFilename[idx] = utf16_ch & 0xFF; longFilename[idx] = utf16_ch & 0xFF;
longFilename[idx+1] = (utf16_ch >> 8) & 0xFF; longFilename[idx + 1] = (utf16_ch >> 8) & 0xFF;
#else #else
// Replace all multibyte characters to '_' // Replace all multibyte characters to '_'
longFilename[n + i] = (utf16_ch > 0xFF) ? '_' : (utf16_ch & 0xFF); longFilename[n + i] = (utf16_ch > 0xFF) ? '_' : (utf16_ch & 0xFF);
@ -1122,45 +1122,44 @@ int8_t SdBaseFile::readDir(dir_t* dir, char* longFilename) {
} }
} }
// Return if normal file or subdirectory // Post-process normal file or subdirectory longname, if any
if (DIR_IS_FILE_OR_SUBDIR(dir)) { if (DIR_IS_FILE_OR_SUBDIR(dir)) {
#if ENABLED(UTF_FILENAME_SUPPORT) #if ENABLED(UTF_FILENAME_SUPPORT)
// Convert filename from utf-16 to utf-8 as Marlin expects
#if LONG_FILENAME_CHARSIZE > 2 #if LONG_FILENAME_CHARSIZE > 2
// Add warning for developers for currently not supported 3-byte cases (Conversion series of 2-byte // Add warning for developers for currently not supported 3-byte cases (Conversion series of 2-byte
// codepoints to 3-byte in-place will break the rest of filename) // codepoints to 3-byte in-place will break the rest of filename)
#error "Currently filename re-encoding is done in-place. It may break the remaining chars to use 3-byte codepoints." #error "Currently filename re-encoding is done in-place. It may break the remaining chars to use 3-byte codepoints."
#endif #endif
uint16_t currentPos = 0;
LOOP_L_N(i, (LONG_FILENAME_LENGTH / 2)) {
uint16_t idx = i * 2; // This is fixed as FAT LFN always contain UTF-16LE encoding
// Is there a long filename to decode?
if (longFilename) {
// Reset n to the start of the long name
n = 0;
for (uint16_t idx = 0; idx < (LONG_FILENAME_LENGTH) / 2; idx += 2) { // idx is fixed since FAT LFN always contains UTF-16LE encoding
uint16_t utf16_ch = longFilename[idx] | (longFilename[idx + 1] << 8); uint16_t utf16_ch = longFilename[idx] | (longFilename[idx + 1] << 8);
if (0xD800 == (utf16_ch & 0xF800)) // Surrogate pair - encode as '_' if (0xD800 == (utf16_ch & 0xF800)) // Surrogate pair - encode as '_'
longFilename[currentPos++] = '_'; longFilename[n++] = '_';
else if (0 == (utf16_ch & 0xFF80)) // Encode as 1-byte utf-8 char else if (0 == (utf16_ch & 0xFF80)) // Encode as 1-byte UTF-8 char
longFilename[currentPos++] = utf16_ch & 0x007F; longFilename[n++] = utf16_ch & 0x007F;
else if (0 == (utf16_ch & 0xF800)) { // Encode as 2-byte utf-8 char else if (0 == (utf16_ch & 0xF800)) { // Encode as 2-byte UTF-8 char
longFilename[currentPos++] = 0xC0 | ((utf16_ch >> 6) & 0x1F); longFilename[n++] = 0xC0 | ((utf16_ch >> 6) & 0x1F);
longFilename[currentPos++] = 0x80 | (utf16_ch & 0x3F); longFilename[n++] = 0x80 | ( utf16_ch & 0x3F);
} }
else { else {
#if LONG_FILENAME_CHARSIZE > 2 // Encode as 3-byte utf-8 char #if LONG_FILENAME_CHARSIZE > 2 // Encode as 3-byte UTF-8 char
longFilename[currentPos++] = 0xE0 | ((utf16_ch >> 12) & 0x0F); longFilename[n++] = 0xE0 | ((utf16_ch >> 12) & 0x0F);
longFilename[currentPos++] = 0xC0 | ((utf16_ch >> 6) & 0x3F); longFilename[n++] = 0xC0 | ((utf16_ch >> 6) & 0x3F);
longFilename[currentPos++] = 0xC0 | (utf16_ch & 0x3F); longFilename[n++] = 0xC0 | ( utf16_ch & 0x3F);
#else // Encode as '_' #else // Encode as '_'
longFilename[currentPos++] = '_'; longFilename[n++] = '_';
#endif #endif
} }
if (0 == utf16_ch) break; // End of filename if (0 == utf16_ch) break; // End of filename
} } // idx
return currentPos; } // longFilename
#else
return n;
#endif #endif
} return n;
} // DIR_IS_FILE_OR_SUBDIR
} }
} }