Mercurial > hg > xemacs-beta
annotate nt/paths.h @ 4967:0d4c9d0f6a8d
rewrite dynarr code
-------------------- ChangeLog entries follow: --------------------
src/ChangeLog addition:
2010-02-03 Ben Wing <ben@xemacs.org>
* device-x.c (x_get_resource_prefix):
* device-x.c (Fx_get_resource):
* device-x.c (Fx_get_resource_prefix):
* device-x.c (Fx_put_resource):
* dialog-msw.c:
* dialog-msw.c (handle_question_dialog_box):
* dired-msw.c (mswindows_sort_files):
* dired-msw.c (mswindows_get_files):
* extents.c (extent_fragment_sort_by_priority):
* extents.c (Fset_extent_parent):
* file-coding.c (coding_reader):
* file-coding.c (coding_writer):
* file-coding.c (gzip_convert):
* frame.c (generate_title_string):
* gutter.c (calculate_gutter_size_from_display_lines):
* indent.c (vmotion_1):
* lread.c (read_bit_vector):
* mule-coding.c (iso2022_decode):
* rangetab.c:
* rangetab.c (Fcopy_range_table):
* rangetab.c (Fget_range_table):
* rangetab.c (unified_range_table_copy_data):
* redisplay-msw.c (mswindows_output_string):
* redisplay-output.c (output_display_line):
* redisplay-output.c (redisplay_move_cursor):
* redisplay-output.c (redisplay_clear_bottom_of_window):
* redisplay-tty.c (tty_output_ichar_dynarr):
* redisplay-tty.c (set_foreground_to):
* redisplay-tty.c (set_background_to):
* redisplay-xlike-inc.c (XLIKE_output_string):
* redisplay.c (redisplay_window_text_width_string):
* redisplay.c (redisplay_text_width_string):
* redisplay.c (create_text_block):
* redisplay.c (SET_CURRENT_MODE_CHARS_PIXSIZE):
* redisplay.c (generate_fstring_runes):
* redisplay.c (regenerate_modeline):
* redisplay.c (ensure_modeline_generated):
* redisplay.c (real_current_modeline_height):
* redisplay.c (create_string_text_block):
* redisplay.c (regenerate_window):
* redisplay.c (REGEN_INC_FIND_START_END):
* redisplay.c (point_visible):
* redisplay.c (redisplay_window):
* redisplay.c (mark_glyph_block_dynarr):
* redisplay.c (line_start_cache_start):
* redisplay.c (start_with_line_at_pixpos):
* redisplay.c (update_line_start_cache):
* redisplay.c (glyph_to_pixel_translation):
* redisplay.c (pixel_to_glyph_translation):
* sysdep.c (qxe_readdir):
* text.c (dfc_convert_to_external_format):
* text.c (dfc_convert_to_internal_format):
* toolbar-common.c (common_output_toolbar_button):
* window.c (window_modeline_height):
* window.c (Fwindow_last_line_visible_height):
* window.c (window_displayed_height):
* window.c (window_scroll):
* window.c (get_current_pixel_pos):
Use Dynarr_begin() in place of Dynarr_atp (foo, 0).
* dynarr.c (Dynarr_realloc):
* dynarr.c (Dynarr_lisp_realloc):
* dynarr.c (Dynarr_resize):
* dynarr.c (Dynarr_insert_many):
* dynarr.c (Dynarr_delete_many):
* dynarr.c (Dynarr_memory_usage):
* dynarr.c (stack_like_malloc):
* dynarr.c (stack_like_free):
* lisp.h:
* lisp.h (DECLARE_DYNARR_LISP_IMP):
* lisp.h (XD_DYNARR_DESC):
* lisp.h (Dynarr_pop):
* gutter.c (output_gutter):
* redisplay-output.c (sync_rune_structs):
* redisplay-output.c (redisplay_output_window):
Redo the dynarr code, add greater checks.
Rename the `len', `largest' and `max' members to `len_',
`largest_' and `max_' to try and catch existing places that might
directly modify these values. Make new accessors Dynarr_largest()
and Dynarr_max() and make them and existing Dynarr_length() be
non-lvalues by adding '+ 0' to them; fix a couple of places in the
redisplay code that tried to modify the length directly by setting
Dynarr_length(). Use the accessors whenever possible even in the
dynarr code itself. The accessors also verify that 0 <= len <=
largest <= max. Rename settor function Dynarr_set_size() to
Dynarr_set_length() and use it more consistently; also create
lower-level Dynarr_set_length_1(). This latter function should be
the only function that directly modifies the `len_' member of a
Dynarr, and in the process makes sure that the `largest' value is
kept correct.
Consistently use ERROR_CHECK_STRUCTURES instead of
ERROR_CHECK_TYPES for error-checking code. Reintroduce the
temporarily disabled verification code on the positions of
Dynarr_at(), Dynarr_atp() and Dynarr_atp_past_end().
Also create Dynarr_resize_if() in place of a repeated
code fragment. Clean up all the functions that modify Dynarrs to
use the new macros and functions and verify the correctness of the
Dynarr both before and after the change.
Note that there are two kinds of verification -- one for accessing
and one for modifying. The difference is that the modify
verification additionally checks to make sure that the Dynarr
isn't locked. (This is used in redisplay to check for problems
with reentrancy.)
* lrecord.h: Move XD_DYNARR_DESC to lisp.h, grouping with the dynarr code.
| author | Ben Wing <ben@xemacs.org> |
|---|---|
| date | Wed, 03 Feb 2010 20:51:18 -0600 |
| parents | 4be1180a9e89 |
| children |
| rev | line source |
|---|---|
| 100 | 1 /* Hey Emacs, this is -*- C -*- code! */ |
| 2 | |
| 3 /* Synched up with: Not synched with FSF. */ | |
| 4 | |
| 5 /* Think twice before editing this file. Generated automatically by configure. | |
| 6 | |
| 7 The file startup.el guesses at reasonable values for load-path, exec-path, | |
| 8 and lock-directory. This means that if you move emacs and its associated | |
| 9 sub-tree to a different place in the filesystem, or to a different machine, | |
| 10 you won't have to do anything for it to work. | |
| 11 | |
| 12 If you define the paths in this file then they will take precedence over | |
| 13 any value generated by the heuristic in startup.el. The hardcoded paths | |
| 14 will be checked to see if they are valid, in which case they will be used. | |
| 15 Otherwise the editor will attempt to make its normal guess. | |
| 16 | |
| 17 See the NEWS file for a description of the heuristic used to locate the lisp | |
| 18 and exec directories at startup time. If you are looking at this file | |
| 19 because you are having trouble, then you would be much better off arranging | |
| 20 for those heuristics to succeed than defining the paths in this file. | |
| 21 | |
| 22 ** Let me say that again. If you're editing this file, you're making | |
| 23 ** a mistake. Re-read the section on installation in ../etc/NEWS. | |
| 24 | |
| 25 If it defines anything, this file should define some subset of the following: | |
| 26 | |
| 27 PATH_PREFIX The default value of `prefix-directory'. This is the | |
| 28 default root for everything. | |
| 29 | |
| 30 PATH_LOADSEARCH The default value of `load-path'. | |
| 31 | |
| 32 PATH_EXEC The default value of `exec-directory' and `exec-path'. | |
| 33 (exec-path also contains the value of whatever is in | |
| 34 the PATH environment variable.) | |
| 35 | |
| 36 PATH_DATA The default value of `data-directory'. This | |
| 37 is where architecture-independent files are | |
| 38 searched for. | |
| 39 | |
| 40 PATH_LOCK The name of the directory that contains lock files | |
| 41 with which we record what files are being modified in | |
| 42 Emacs. This directory should be writable by everyone. | |
| 43 If this is specified, the string must end with a slash! | |
| 44 | |
| 45 PATH_SUPERLOCK The name of the file !!!SuperLock!!! in the lock | |
| 46 directory. You probably should let this default... | |
| 47 | |
| 48 PATH_INFO The default value of `Info-directory-list'. | |
| 49 This is where info files are searched for. */ |