sugar/linux
1
0
Fork 0
Code Activity

vt: update gen_ucs_width_table.py to make tables more space efficient

Browse Source 
Split table ranges into BMP (16-bit) and non-BMP (above 16-bit).
This reduces the corresponding text size by 20-25%.

Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Reviewed-by: Jiri Slaby <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20250417184849.475581-13-nico@fluxnic.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit is contained in:
Nicolas Pitre
2025-04-17 14:45:14 -04:00
committed by Greg Kroah-Hartman
parent ffae2340a6
commit ad934777f0
1 changed files with 49 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

55
drivers/tty/vt/gen_ucs_width_table.py
View File

@@ -194,6 +194,27 @@ def write_tables(zero_width_ranges, double_width_ranges):
double_width_ranges: List of (start, end) ranges for double-width characters
"""
# Function to split ranges into BMP (16-bit) and non-BMP (above 16-bit)
def split_ranges_by_size(ranges):
bmp_ranges = []
non_bmp_ranges = []
for start, end in ranges:
if end <= 0xFFFF:
bmp_ranges.append((start, end))
elif start > 0xFFFF:
non_bmp_ranges.append((start, end))
else:
# Split the range at 0xFFFF
bmp_ranges.append((start, 0xFFFF))
non_bmp_ranges.append((0x10000, end))
return bmp_ranges, non_bmp_ranges
# Split ranges into BMP and non-BMP
zero_width_bmp, zero_width_non_bmp = split_ranges_by_size(zero_width_ranges)
double_width_bmp, double_width_non_bmp = split_ranges_by_size(double_width_ranges)
# Function to generate code point description comments
def get_code_point_comment(start, end):
try:
@@ -221,22 +242,44 @@ def write_tables(zero_width_ranges, double_width_ranges):
* Unicode Version: {unicodedata.unidata_version}
*/
/* Zero-width character ranges */
static const struct ucs_interval ucs_zero_width_ranges[] = {{
/* Zero-width character ranges (BMP - Basic Multilingual Plane, U+0000 to U+FFFF) */
static const struct ucs_interval16 ucs_zero_width_bmp_ranges[] = {{
""")
for start, end in zero_width_ranges:
for start, end in zero_width_bmp:
comment = get_code_point_comment(start, end)
f.write(f"\t{{ 0x{start:04X}, 0x{end:04X} }}, {comment}\n")
f.write("""\
};
/* Zero-width character ranges (non-BMP, U+10000 and above) */
static const struct ucs_interval32 ucs_zero_width_non_bmp_ranges[] = {
""")
for start, end in zero_width_non_bmp:
comment = get_code_point_comment(start, end)
f.write(f"\t{{ 0x{start:05X}, 0x{end:05X} }}, {comment}\n")
f.write("""\
};
/* Double-width character ranges */
static const struct ucs_interval ucs_double_width_ranges[] = {
/* Double-width character ranges (BMP - Basic Multilingual Plane, U+0000 to U+FFFF) */
static const struct ucs_interval16 ucs_double_width_bmp_ranges[] = {
""")
for start, end in double_width_ranges:
for start, end in double_width_bmp:
comment = get_code_point_comment(start, end)
f.write(f"\t{{ 0x{start:04X}, 0x{end:04X} }}, {comment}\n")
f.write("""\
};
/* Double-width character ranges (non-BMP, U+10000 and above) */
static const struct ucs_interval32 ucs_double_width_non_bmp_ranges[] = {
""")
for start, end in double_width_non_bmp:
comment = get_code_point_comment(start, end)
f.write(f"\t{{ 0x{start:05X}, 0x{end:05X} }}, {comment}\n")