While investigating Bug#20759, I discovered other quoting problems:
C-h f mishandled characters like backslash and quote in function names.
This fix changes the behavior so that 'C-h f pcase RET' now
generates "... (\` QPAT) ..." instead of "... (` QPAT) ...",
because '(format "%S" '(` FOO))' returns "(\\` FOO)". A comment
in src/lread.c's read1 function says that the backslash will be
needed starting in Emacs 25, which implies that 'format' is
correct and the old pcase documention was wrong to omit the backslash.
* lisp/emacs-lisp/nadvice.el (advice--make-docstring):
* lisp/help-fns.el (help-fns--signature):
* lisp/help.el (help-add-fundoc-usage):
* lisp/progmodes/elisp-mode.el (elisp-function-argstring):
Use help--make-usage-docstring rather than formatting
help-make-usage.
* lisp/emacs-lisp/pcase.el (pcase--make-docstring):
Return raw docstring.
* lisp/help-fns.el (help-fns--signature): New arg RAW, to return
raw docstring. Take more care to distinguish raw from cooked dstrings.
(describe-function-1): Let help-fns--signature substitute
command keys.
* lisp/help.el (help--docstring-quote): New function.
(help-split-fundoc): Use it, to quote funny characters more
systematically.
(help--make-usage): Rename from help-make-usage, since this
should be private. Leave an obsolete alias for the old name.
(help--make-usage-docstring): New function.
* test/automated/help-fns.el (help-fns-test-funny-names): New test.
* lisp/thingatpt.el (thing-at-point--in-string-p)
(thing-at-point--end-of-sexp, thing-at-point--beginning-of-sexp)
(thing-at-point--read-from-whole-string): Rename from
old versions without "thing-at-point--" prefix.
Keep old versions as obsolete aliases. Update all uses.
(sh-here-doc-open-re): Don't mis-match the <<< operator (bug#20683).
(sh-font-lock-quoted-subshell): Make sure double quotes within single
quotes don't mistakenly end prematurely the surrounding string.
* src/fontset.c (face_for_char): If the character's script is
'symbol', and the font used for ASCII face has a glyph for it, use
the font for the ASCII face instead of searching the fontsets.
This comes instead of NS-specific code that used the current
face's font instead, which is now disabled due to undesirable
consequences. (Bug#20727)
* lisp/progmodes/elisp-mode.el (elisp-completion-at-point): Skip
past `#' to find BEG (bug#20771).
* test/automated/elisp-mode-tests.el
(elisp-completes-functions-after-hash-quote): New test.
* lisp/emacs-lisp/checkdoc.el (checkdoc-package-keywords-flag): New
defcustom.
(checkdoc-list-of-strings-p): Add doc.
(checkdoc-current-buffer): When `checkdoc-package-keywords-flag' is
non-nil, call `checkdoc-package-keywords'.
(checkdoc-get-keywords): New defun.
(checkdoc-package-keywords): New command. Warns if the current file
has package.el-style keywords that aren't in `finder-known-keywords'.
* etc/NEWS: Add entry.
* lisp/emacs-lisp/checkdoc.el (checkdoc-error): When
`checkdoc-diagnostic-buffer' is set to "*warn*", print the warning
to the standard output.
(bug#20754)
That way, 'make change-history' needs to change only ChangeLog.2,
instead of having to change two files.
* ChangeLog.2: Add commit info for range that this file covers.
* Makefile.in (new_commit_regexp): New macro.
(change-history-nocommit): Simplify, by putting what used to be
the gen_origin value into the data (ChangeLog.2) rather than
into the program (gitlog-to-emacslog).
* build-aux/gitlog-to-emacslog (gen_origin): Calculate from
the input file (e.g., ChangeLog.2) rather than by having a
constant in the program. Substitute it into the output.
This incorporates:
2015-06-06 acl-permissions: pacify -Wsuggest-attribute=const
2015-06-05 stdio: Don't redefine gets when using C++
2015-06-05 acl-permissions: port to AIX, C89 HP-UX
2015-06-02 file-has-acl: fix build on Mac OS X 10
2015-06-01 gnulib-tool: concatenate lib_SOURCES to a single line
2015-06-01 pthread_sigmask: discount system version if a simple macro
2015-05-31 readlinkat: avoid OS X 10.10 trailing slash bug
* doc/misc/texinfo.tex, lib/acl-internal.h, lib/get-permissions.c:
* lib/readlinkat.c, lib/set-permissions.c, lib/stdio.in.h:
* m4/acl.m4, m4/pthread_sigmask.m4, m4/readlinkat.m4: Copy from gnulib.
* lib/gnulib.mk: Regenerate.
This fixes bug#20628.
* src/xdisp.c (get_phys_cursor_geometry): Correct the Y
coordinate of a hollow cursor glyph when the original glyph's
ascent is too small.
(get_font_ascent_descent, normal_char_ascent_descent)
(normal_char_height): New functions.
(handle_single_display_spec, append_space_for_newline)
(calc_pixel_width_or_height, produce_stretch_glyph)
(calc_line_height_property): Use normal_char_ascent_descent and
normal_char_height.
(x_produce_glyphs): When font-global values of ascent and descent
are too large, use per-character glyph metrics instead, if
possible. But don't allow the glyph row's ascent and descent
values become smaller than the values from the metrics of the
font's "normal" character.
* src/xftfont.c (xftfont_draw):
* src/w32font.c (w32font_draw): Correct the values of ascent and
descent used to draw glyphless characters' hex code in a box.
* src/xterm.c (x_draw_glyph_string_background):
* src/xdisp.c (x_produce_glyphs):
* src/w32term.c (x_draw_glyph_string_background):
* src/nsterm.m (ns_maybe_dumpglyphs_background): Use FONT_TOO_HIGH
to detect fonts whose global ascent and descent values are too
large to be used in layout decision, and redraw the background
when that happens.
* src/dispextern.h (FONT_TOO_HIGH): New macro.
(get_font_ascent_descent): Add prototype.
* src/xterm.c (x_new_font):
* src/w32term.c (x_new_font):
* src/nsterm.m (x_new_font):
* src/font.c (font_open_entity):
* src/composite.c (composition_gstring_width): Use
get_font_ascent_descent to obtain reasonable values for ascent and
descent of a font.
* lisp/ls-lisp.el (ls-lisp-uid-d-fmt, ls-lisp-uid-s-fmt)
(ls-lisp-gid-d-fmt, ls-lisp-gid-s-fmt): Make the initial values be
correct for when displaying individual files separately, not as
part of listing a directory, in which case these values are not
recomputed by 'ls-lisp-insert-directory', but used verbatim.
