bissen21/Marlin_Firmware
1
0
Fork 0
Code Issues Pull Requests Projects Releases 4 Wiki Activity

Handle UTF in long filenames (#20087)

Browse Source 
Co-authored-by: Scott Lahteine <thinkyhead@users.noreply.github.com>
This commit is contained in:
LinFor
2020-11-11 23:14:39 +03:00
committed by GitHub
parent 0465e0ae3a
commit 5b9aeb2e5f
5 changed files with 124 additions and 100 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
Marlin/src/sd/SdBaseFile.cpp
View File

@ -1103,19 +1103,67 @@ int8_t SdBaseFile::readDir(dir_t* dir, char* longFilename) {
if (WITHIN(seq, 1, MAX_VFAT_ENTRIES)) {
// TODO: Store the filename checksum to verify if a long-filename-unaware system modified the file table.
n = (seq - 1) * (FILENAME_LENGTH);
LOOP_L_N(i, FILENAME_LENGTH)
longFilename[n + i] = (i < 5) ? VFAT->name1[i] : (i < 11) ? VFAT->name2[i - 5] : VFAT->name3[i - 11];
LOOP_L_N(i, FILENAME_LENGTH) {
uint16_t utf16_ch = (i < 5) ? VFAT->name1[i] : (i < 11) ? VFAT->name2[i - 5] : VFAT->name3[i - 11];
#if ENABLED(UTF_FILENAME_SUPPORT)
// We can't reconvert to UTF-8 here as UTF-8 is variable-size encoding, but joining LFN blocks
// 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
longFilename[idx] = utf16_ch & 0xFF;
longFilename[idx+1] = (utf16_ch >> 8) & 0xFF;
#else
// Replace all multibyte characters to '_'
longFilename[n + i] = (utf16_ch > 0xFF) ? '_' : (utf16_ch & 0xFF);
#endif
}
// If this VFAT entry is the last one, add a NUL terminator at the end of the string
if (VFAT->sequenceNumber & 0x40) longFilename[n + FILENAME_LENGTH] = '\0';
if (VFAT->sequenceNumber & 0x40) longFilename[(n + FILENAME_LENGTH) * LONG_FILENAME_CHARSIZE] = '\0';
}
}
}
// Return if normal file or subdirectory
if (DIR_IS_FILE_OR_SUBDIR(dir)) return n;
if (DIR_IS_FILE_OR_SUBDIR(dir)) {
#if ENABLED(UTF_FILENAME_SUPPORT)
// Convert filename from utf-16 to utf-8 as Marlin expects
#if LONG_FILENAME_CHARSIZE > 2
// 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)
#error "Currently filename re-encoding is done in-place. It may break the remaining chars to use 3-byte codepoints."
#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
uint16_t utf16_ch = longFilename[idx] | (longFilename[idx + 1] << 8);
if (0xD800 == (utf16_ch & 0xF800)) // Surrogate pair - encode as '_'
longFilename[currentPos++] = '_';
else if (0 == (utf16_ch & 0xFF80)) // Encode as 1-byte utf-8 char
longFilename[currentPos++] = utf16_ch & 0x007F;
else if (0 == (utf16_ch & 0xF800)) { // Encode as 2-byte utf-8 char
longFilename[currentPos++] = 0xC0 | ((utf16_ch >> 6) & 0x1F);
longFilename[currentPos++] = 0x80 | (utf16_ch & 0x3F);
}
else {
#if LONG_FILENAME_CHARSIZE > 2 // Encode as 3-byte utf-8 char
longFilename[currentPos++] = 0xE0 | ((utf16_ch >> 12) & 0x0F);
longFilename[currentPos++] = 0xC0 | ((utf16_ch >> 6) & 0x3F);
longFilename[currentPos++] = 0xC0 | (utf16_ch & 0x3F);
#else // Encode as '_'
longFilename[currentPos++] = '_';
#endif
}
if (0 == utf16_ch) break; // End of filename
}
return currentPos;
#else
return n;
#endif
}
}
}
// Read next directory entry into the cache
// Assumes file is correctly positioned
dir_t* SdBaseFile::readDirCache() {

7
Marlin/src/sd/SdFatConfig.h
View File

@ -103,5 +103,10 @@
#define FILENAME_LENGTH 13 // Number of UTF-16 characters per entry
// UTF-8 may use up to 3 bytes to represent single UTF-16 code point.
// We discard 3-byte characters allowing only 2-bytes
// or 1-byte if UTF_FILENAME_SUPPORT disabled.
#define LONG_FILENAME_CHARSIZE TERN(UTF_FILENAME_SUPPORT, 2, 1)
// Total bytes needed to store a single long filename
#define LONG_FILENAME_LENGTH (FILENAME_LENGTH * MAX_VFAT_ENTRIES + 1)
#define LONG_FILENAME_LENGTH (FILENAME_LENGTH * LONG_FILENAME_CHARSIZE * MAX_VFAT_ENTRIES + 1)