| Age | Commit message (Collapse) | Author |
|---|
|
|
|
The pkg-config file put include/term on the host's include path, so
ncurses' term.h could be picked up in place of ours. Hosts now write
<term/term.h> and the package adds only its include directory.
The installed Unicode API had unprefixed names that collide with host
symbols and macros. Rename them: term_ for functions, types, and static
tables, TERM_ for macros and include guards. unicode_data.inc and
tools/generate_unicode.py are renamed to match. The generator's inputs
here are Unicode 15.1 and the checked-in table is 18.0.0, so the table
data was not regenerated. Only the declarations were checked against
the generator.
utf8.h casts the malloc result so it compiles as C++.
header_test.sh now includes <term/term.h> with only include/ on the path.
It also checks that a host which defines unicode_width, UTF8_ACCEPT,
GRAPHEME_MAX_LEN, in_ranges, and similar names compiles with the Unicode
headers as C11 and C++, in both declaration orders.
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
|
|
The snapshot getters crashed on a NULL snapshot, unlike the rest of the
API. They now return NULL, an empty span, 0, or write nothing, as the
terminal functions do. term_snapshot_cluster writes nothing when cps is
NULL or cap is 0.
Document that the view and its palette keep their address for the life
of the snapshot, and that a row pointer lasts only until the next update
or free, since an update can realloc the cell array.
Rename the term_input.h include guard to TERM_INPUT_H to match the file.
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
|
|
Make text queries return allocated strings released with term_text_free, keeping each capture atomic. Align public attributes with the write core, remove snapshot conversion tables, and migrate tests and the Kitty adapter.
|
|
Account for cluster records discarded by snapshot scrolling, propagate snapshot allocation failures with an empty retryable view, and allocate resized screens before mutating visible state. Add bounded-memory and allocation-failure regressions, and update the saved diagnostic and audit report.
|
|
|
|
- Replace sequential colon-counting loop with branchless CTZ bitmask in src/pen.h
- Replace linear strchr with 256-bit ASCII punctuation bitmask in src/selection.c
- Flatten 3D color cube loop and precompute constants in src/palette.c
- Use static 256-byte LUT for hex decoding and bsearch for termcap in src/escape.c
- Replace capacity doubling while loops with branchless CLZ pow2 helpers across core
- Use direct memset/memcpy in src/snapshot.c
- Optimize input digit/byte formatting, alt-prefix path, and make keyboard state optional in src/input.c
|
|
The core now builds, tests, and benchmarks alone, as a Meson subproject
with no dependency on the main tree.
- include/term/ holds term.h and the header-only unicode/. src/ holds
the core and its private headers. The main tree gets only
include/term through term_dep, so it cannot reach a private header.
- The term tests, the term benchmarks, the Kitty parser suite, and
tools/amalgamate.py and tools/generate_unicode.py move with it. The
tests run as term:<name>.
- The main build passes its -march (arch_flag) and b_ndebug to the
subproject. A directory configured before this commit needs
meson configure -Dterm:arch_flag=-march=... once.
- The term tests compile with -UNDEBUG. They do work inside assert.
- The subproject exports term_test_dep with term_api.h, for the main
tree's render, thumbnail, color, and selection tests.
- The Kitty bridge used struct vector2i and has not compiled since the
public API redesign. It builds again. 2 of its 14 tests fail: REP
after a tab, and a cursor report after a resize.
- dirgraph.py drops the term and unicode layers. Its self-test uses pty.
- The symbols test ignores the __odr_asan symbols that ASan adds.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
