fix patience iteration error; also makes dramatically faster

debug: patience: mark unique debug

diff_atom_cmp: no special case for ignore whitespace when both atoms empty

debug fix: missing space

cache kd_buf in struct diff_state to avoid repeated allocation + free

fix build with PROFILE=1

patience: make it easy to switch impls for mark_unique

fix myers_divide max-effort cut, wrong x for backwards coordinate

add test019 showing max-effort bug

debug output fixes - fix compilation error with DEBUG 1, after moving patience state out of diff_atom, and after renaming diff_data.d to diff_data.root. - add some missing newlines. - also debug "HIT" on a max-effort cut, like when finding a mid snake, which eases finding all the places where boxes get cut.

fix off-by-one in the off-by-one fix made in bdfcb086

don't require both + and - chunks in diff_chunk_context_load_change()

fix off-by-one access beyond mapped file in diff_data_atomize_text_lines_mmap() Thread 1 received signal SIGSEGV, Segmentation fault. 0x0000013992a89eca in diff_data_atomize_text_lines_mmap (d=0x13b9b455668) \ at /home/stsp/src/got/got/../lib/diff_atomize_text.c:134 134 if (line_end[0] == '\r' (gdb) p pos $1 = (const uint8_t *) 0x13be402006d "" (gdb) p end $2 = (const uint8_t *) 0x13be4023000 <error: Cannot access memory at \ address 0x13be4023000> (gdb) p end-1 $3 = (const uint8_t *) 0x13be4022fff "" (gdb) p line_end $4 = (const uint8_t *) 0x13be4023000 <error: Cannot access memory at \ address 0x13be4023000>

search for prototypes above, not at, the first line touched by the diff chunk

add support for showing function prototypes in unidiff hunk headers

don't hex-escape non-printable ASCII in diff_output_lines(); UTF-8 must pass

fix missing allocaion failure check in diff_algo_patience()

add diff_chunk_context_load_change() for use with diff_output_unidiff_chunk()

revert db941c24; callers would also need private structs for that idea to work

avoid printing the --- +++ diff header in diff_output_unidiff_chunk()

fix another format string issue in results_test.c Cast ARRAYLIST_IDX result to unsigned long and use %lu in the format string. Clang 10.0.1 complained as follows: results_test.c:148:9: warning: format specifies type 'int' but the argument has type 'long' [-Wformat] ARRAYLIST_IDX(c, result->chunks), ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

fix a format string issue in results_test.c Use explicit casts to long long and use %lld in the format string. Clang 10.0.1 complained as follows: results_test.c:78:9: warning: format specifies type 'int' but the argument has type 'long long' [-Wformat] c->left_start ? c->left_start->pos : -1, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

fix a format string issue in array test ARRAYLIST_IDX deals with the head pointer. Explicitly cast to unsigned long and use %lu in the format string. Clang 10.0.1 used to complain as follows: arraylist_test.c:19:2: warning: format specifies type 'int' but the argument has type 'long' [-Wformat] arraylist_test.c:15:24: note: expanded from macro 'dump' printf("[%d] %d\n", ARRAYLIST_IDX(p, list), *p); \ ~~ ^~~~~~~~~~~~~~~~~~~~~~

put functions required for using diff_output_unidiff_chunk() into public API They will be needed to implement 'got stage -p' and 'got revert -p', etc. Chunks stored in diff_result still need post-processing in order to be displayed as expected by diff_output_unidiff_chunk(). Exposing these functions allows API consumers to implement the same chunk merging loop as used internally by output_unidiff() and output_edscript(). (Perhaps, ideally, the public diff API would provide pre-merged chunks at some point in the future? Then this commit could probably be reverted.) Change return values from bool to int to avoid making stdbool.h a requirement for the public diff API.

expose chunk_context_empty() via diff_internal.h and use it in edscript output