From: Nicolas Pitre Date: Thu, 17 Apr 2025 18:45:14 +0000 (-0400) Subject: vt: update gen_ucs_width_table.py to make tables more space efficient X-Git-Url: https://www.infradead.org/git/?a=commitdiff_plain;h=ad934777f0f15c8cea042b6a81deaa7fe53b6dea;p=users%2Fjedix%2Flinux-maple.git vt: update gen_ucs_width_table.py to make tables more space efficient 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 Reviewed-by: Jiri Slaby Link: https://lore.kernel.org/r/20250417184849.475581-13-nico@fluxnic.net Signed-off-by: Greg Kroah-Hartman --- diff --git a/drivers/tty/vt/gen_ucs_width_table.py b/drivers/tty/vt/gen_ucs_width_table.py index 00510444a7272..059ed9a8baa21 100755 --- a/drivers/tty/vt/gen_ucs_width_table.py +++ b/drivers/tty/vt/gen_ucs_width_table.py @@ -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_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_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_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_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")