time-stamp.el

;;; time-stamp.el --- Maintain last change time stamps in files edited by Emacs  -*- lexical-binding: t -*-

;; Copyright (C) 1989, 1993-1995, 1997, 2000-2024 Free Software
;; Foundation, Inc.

;; This file is part of GNU Emacs.

;; Maintainer: Stephen Gildea <stepheng+emacs@gildea.com>
;; Keywords: tools

;; GNU Emacs is free software: you can redistribute it and/or modify
;; it under the terms of the GNU General Public License as published by
;; the Free Software Foundation, either version 3 of the License, or
;; (at your option) any later version.

;; GNU Emacs is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
;; GNU General Public License for more details.

;; You should have received a copy of the GNU General Public License
;; along with GNU Emacs.  If not, see <https://www.gnu.org/licenses/>.

;;; Commentary:

;; A template in a file can be updated with a new time stamp when
;; you save the file.  For example:
;;     static char *ts = "sdmain.c Time-stamp: <2020-04-18 14:10:21 gildea>";

;; To use time-stamping, add this line to your init file:
;;     (add-hook 'before-save-hook 'time-stamp)
;; Now any time-stamp templates in your files will be updated automatically.

;; See the documentation for the functions `time-stamp'
;; and `time-stamp-toggle-active' for details.

;;; Code:

(defgroup time-stamp nil
  "Maintain last change time stamps in files edited by Emacs."
  :group 'data
  :group 'extensions)


(defcustom time-stamp-format "%Y-%02m-%02d %02H:%02M:%02S %l"
  "Format of the string inserted by \\[time-stamp].
This is a string, used verbatim except for character sequences beginning
with %, as follows.

%:A  weekday name: `Monday'             %#A gives uppercase: `MONDAY'
%3a  abbreviated weekday: `Mon'         %#a gives uppercase: `MON'
%:B  month name: `January'              %#B gives uppercase: `JANUARY'
%3b  abbreviated month: `Jan'           %#b gives uppercase: `JAN'
%02d day of month
%02H 24-hour clock hour
%02I 12-hour clock hour
%02m month number
%02M minute
%#p  `am' or `pm'                       %P  gives uppercase: `AM' or `PM'
%02S seconds
%w   day number of week, Sunday is 0
%02y 2-digit year                       %Y  4-digit year
%Z   time zone name: `EST'              %#Z gives lowercase: `est'
%5z  time zone offset: `-0500' (since Emacs 27; see note below)

Non-date items:
%%   a literal percent character: `%'
%f   file name without directory        %F  absolute file name
%l   login name                         %L  full name of logged-in user
%q   unqualified host name              %Q  fully-qualified host name
%h   mail host name

Decimal digits between the % and the type character specify the
field width.  Strings are truncated on the right.
A leading zero in the field width zero-fills a number.

For example, to get the format used by the `date' command,
use \"%3a %3b %2d %02H:%02M:%02S %Z %Y\".

The values of non-numeric formatted items depend on the locale
setting recorded in `system-time-locale' and `locale-coding-system'.
The examples here are for the default (`C') locale.
`time-stamp-time-zone' controls the time zone used.

The default padding of some formats has changed to be more compatible
with format-time-string.  To be compatible with older versions of Emacs,
specify a padding width (as shown) or use the : modifier to request the
transitional behavior (again, as shown).

The behavior of `%5z' is new in Emacs 27.  If your files might be
edited by older versions of Emacs also, do not use this format yet."
  :type 'string
  :version "27.1")
;;;###autoload(put 'time-stamp-format 'safe-local-variable 'stringp)


(defcustom time-stamp-active t
  "Non-nil to enable time-stamping of buffers by \\[time-stamp].
Can be toggled by \\[time-stamp-toggle-active].

This option does not affect when `time-stamp' is run, only what it
does when it runs.  To activate automatic time-stamping of buffers
when they are saved, either add this line to your init file:
    (add-hook \\='before-save-hook \\='time-stamp)
or customize option `before-save-hook'.

To enable automatic time-stamping for only a specific file, add this
line to a local variables list near the end of the file:
    eval: (add-hook \\='before-save-hook \\='time-stamp nil t)

See also the variable `time-stamp-warn-inactive'."
  :type 'boolean)


(defcustom time-stamp-warn-inactive t
  "Have \\[time-stamp] warn if a buffer did not get time-stamped.
If non-nil, a warning is displayed if `time-stamp-active' has
deactivated time stamping and the buffer contains a template that
otherwise would have been updated."
  :type 'boolean
  :version "19.29")


(defcustom time-stamp-time-zone nil
  "The time zone to be used by \\[time-stamp].
Its format is that of the ZONE argument of the `format-time-string' function."
  :type '(choice (const :tag "Emacs local time" nil)
                 (const :tag "Universal Time" t)
                 (const :tag "system wall clock time" wall)
                 (string :tag "TZ environment variable value")
                 (list :tag "Offset and name"
                       (integer :tag "Offset (seconds east of UTC)")
                       (string :tag "Time zone abbreviation"))
                 (integer :tag "Offset (seconds east of UTC)"))
  :version "20.1")
;;;###autoload(put 'time-stamp-time-zone 'safe-local-variable 'time-stamp-zone-type-p)


;;;###autoload
(defun time-stamp-zone-type-p (zone)
  "Return non-nil if ZONE is of the correct type for a timezone rule.
Valid ZONE values are described in the documentation of `format-time-string'."
  (or (memq zone '(nil t wall))
      (stringp zone)
      (and (consp zone)
           (integerp (car zone))
           (consp (cdr zone))
           (stringp (cadr zone)))
      (integerp zone)))


;;; Do not change time-stamp-line-limit, time-stamp-start,
;;; time-stamp-end, time-stamp-pattern, time-stamp-inserts-lines,
;;; or time-stamp-count in your .emacs or you will be incompatible
;;; with other people's files!  If you must change them, do so only
;;; in the local variables section of the file itself.


(defvar time-stamp-line-limit 8	    ;Do not change!
  "Lines of a file searched; positive counts from start, negative from end.
The patterns `time-stamp-start' and `time-stamp-end' must be found in
the first (last) `time-stamp-line-limit' lines of the file for the
file to be time-stamped by \\[time-stamp].  A value of 0 searches the
entire buffer (use with care).

It may be more convenient to use `time-stamp-pattern' if you set more
than one of `time-stamp-line-limit', `time-stamp-start', `time-stamp-end',
or `time-stamp-format'.

These variables are best changed with file-local variables.
If you were to change `time-stamp-line-limit', `time-stamp-start',
`time-stamp-end', or `time-stamp-pattern' in your init file, you
would be incompatible with other people's files.")
;;;###autoload(put 'time-stamp-line-limit 'safe-local-variable 'integerp)


(defvar time-stamp-start "Time-stamp:[ \t]+\\\\?[\"<]+"    ;Do not change!
  "Regexp after which the time stamp is written by \\[time-stamp].

It may be more convenient to use `time-stamp-pattern' if you set more
than one of `time-stamp-line-limit', `time-stamp-start', `time-stamp-end',
or `time-stamp-format'.

These variables are best changed with file-local variables.
If you were to change `time-stamp-line-limit', `time-stamp-start',
`time-stamp-end', or `time-stamp-pattern' in your init file, you
would be incompatible with other people's files.")
;;;###autoload(put 'time-stamp-start 'safe-local-variable 'stringp)


(defvar time-stamp-end "\\\\?[\">]"    ;Do not change!
  "Regexp marking the text after the time stamp.
\\[time-stamp] deletes the text between the first match of `time-stamp-start'
and the following match of `time-stamp-end', then writes the
time stamp specified by `time-stamp-format' between them.

It may be more convenient to use `time-stamp-pattern' if you set more
than one of `time-stamp-line-limit', `time-stamp-start', `time-stamp-end',
or `time-stamp-format'.

The end text normally starts on the same line as the start text ends,
but if there are any newlines in `time-stamp-format', the same number
of newlines must separate the start and end.  Thus \\[time-stamp] tries
to not change the number of lines in the buffer; `time-stamp-inserts-lines'
controls this behavior.

These variables are best changed with file-local variables.
If you were to change `time-stamp-line-limit', `time-stamp-start',
`time-stamp-end', `time-stamp-pattern', or `time-stamp-inserts-lines' in
your init file, you would be incompatible with other people's files.")
;;;###autoload(put 'time-stamp-end 'safe-local-variable 'stringp)


(defvar time-stamp-inserts-lines nil    ;Do not change!
  "Whether \\[time-stamp] can change the number of lines in a file.
If nil, \\[time-stamp] skips as many lines as there are newlines in
`time-stamp-format' before looking for the `time-stamp-end' pattern,
thus it tries not to change the number of lines in the buffer.
If non-nil, \\[time-stamp] starts looking for the end pattern
immediately after the start pattern.  This behavior can cause
unexpected changes in the buffer if used carelessly, but it is useful
for generating repeated time stamps.

These variables are best changed with file-local variables.
If you were to change `time-stamp-end' or `time-stamp-inserts-lines' in
your init file, you would be incompatible with other people's files.")
;;;###autoload(put 'time-stamp-inserts-lines 'safe-local-variable 'symbolp)


(defvar time-stamp-count 1		;Do not change!
  "How many templates \\[time-stamp] will look for in a buffer.
The same time stamp will be written in each case.

`time-stamp-count' is best changed with a file-local variable.
If you were to change it in your init file, you would be incompatible
with other people's files.")
;;;###autoload(put 'time-stamp-count 'safe-local-variable 'integerp)


(defvar time-stamp-pattern nil		;Do not change!
  "Convenience variable setting all `time-stamp' location and format values.
This string has four parts, each of which is optional.
These four parts set `time-stamp-line-limit', `time-stamp-start',
`time-stamp-format', and `time-stamp-end'.  See the documentation
for each of these variables for details.

The first part is a number followed by a slash; the number sets the number
of lines at the beginning (negative counts from end) of the file searched
for the time stamp.  The number and the slash may be omitted to use the
normal value.

The second part is a regexp identifying the pattern preceding the time stamp.
This part may be omitted to use the normal pattern.

The third part specifies the format of the time stamp inserted.  See
the documentation for `time-stamp-format' for details.  Specify this
part as \"%%\" to use the normal format.

The fourth part is a regexp identifying the pattern following the time stamp.
This part may be omitted to use the normal pattern.

The pattern does not need to match the entire line of the time stamp.

These variables are best changed with file-local variables.
If you were to change `time-stamp-pattern', `time-stamp-line-limit',
`time-stamp-start', or `time-stamp-end' in your init file, you
would be incompatible with other people's files.

See also `time-stamp-count' and `time-stamp-inserts-lines'.

Examples:

\"-10/\" (sets only `time-stamp-line-limit')

\"-9/^Last modified: %%$\" (sets `time-stamp-line-limit',
`time-stamp-start' and `time-stamp-end')

\"@set Time-stamp: %:B %1d, %Y$\" (sets `time-stamp-start',
`time-stamp-format' and `time-stamp-end')

\"newcommand{\\\\\\\\timestamp}{%%}\" (sets `time-stamp-start'
and `time-stamp-end')")
;;;###autoload(put 'time-stamp-pattern 'safe-local-variable 'stringp)



;;;###autoload
(defun time-stamp ()
  "Update any time stamp string(s) in the buffer.
This function looks for a time stamp template and updates it with
the current date, time, and/or other info.

The template, which you manually create on one of the first 8 lines
of the file before running this function, by default can look like
one of the following (your choice):
      Time-stamp: <>
      Time-stamp: \" \"
This function writes the current time between the brackets or quotes,
by default formatted like this:
      Time-stamp: <2020-08-07 17:10:21 gildea>

Although you can run this function manually to update a time stamp
once, usually you want automatic time stamp updating.

A time stamp can be automatically updated with current information
every time you save a file.  To enable time-stamping for all files,
customize option `before-save-hook' or add this line to your init file:
    (add-hook \\='before-save-hook \\='time-stamp)

To enable automatic time-stamping for only a specific file, add
this line to a local variables list near the end of the file:
    eval: (add-hook \\='before-save-hook \\='time-stamp nil t)

If the file has no time-stamp template, this function does nothing.

You can set `time-stamp-pattern' in a file's local variables list
to customize the information in the time stamp and where it is written.

The time stamp is updated only if `time-stamp-active' is non-nil."
  (interactive)
  (let ((line-limit time-stamp-line-limit)
	(ts-start time-stamp-start)
	(ts-format time-stamp-format)
	(ts-end time-stamp-end)
	(ts-count time-stamp-count)
	(format-lines 0)
	(end-lines 1)
	(start nil)
	search-limit)
    (if (stringp time-stamp-pattern)
	(progn
	  (string-match "\\`\\(\\(-?[0-9]+\\)/\\)?\\([^%]+\\)?\\(\\(%[-.,:@+_ #^()0-9]*[A-Za-z%][^%]*\\)*%[-.,:@+_ #^()0-9]*[A-Za-z%]\\)?\\([^%]+\\)?\\'" time-stamp-pattern)
	  (and (match-beginning 2)
	       (setq line-limit
		     (string-to-number (match-string 2 time-stamp-pattern))))
	  (and (match-beginning 3)
	       (setq ts-start (match-string 3 time-stamp-pattern)))
	  (and (match-beginning 4)
	       (not (string-equal (match-string 4 time-stamp-pattern) "%%"))
	       (setq ts-format (match-string 4 time-stamp-pattern)))
	  (and (match-beginning 6)
	       (setq ts-end (match-string 6 time-stamp-pattern)))))
    (cond ((not (integerp line-limit))
	   (setq line-limit 8)
	   (message "time-stamp-line-limit is not an integer")
	   (sit-for 1)))
    (cond ((not (integerp ts-count))
	   (setq ts-count 1)
	   (message "time-stamp-count is not an integer")
	   (sit-for 1))
	  ((< ts-count 1)
	   ;; We need to call time-stamp-once at least once
	   ;; to output any warnings about time-stamp not being active.
	   (setq ts-count 1)))
    ;; Figure out what lines the end should be on.
    (if (stringp ts-format)
	(let ((nl-start 0))
	  (while (string-match "\n" ts-format nl-start)
	    (setq format-lines (1+ format-lines) nl-start (match-end 0)))))
    (let ((nl-start 0))
      (while (string-match "\n" ts-end nl-start)
	(setq end-lines (1+ end-lines) nl-start (match-end 0))))
    ;; Find overall what lines to look at
    (save-excursion
      (save-restriction
	(widen)
	(cond ((> line-limit 0)
	       (goto-char (setq start (point-min)))
	       (forward-line line-limit)
	       (setq search-limit (point)))
	      ((< line-limit 0)
	       (goto-char (setq search-limit (point-max)))
	       (forward-line line-limit)
	       (setq start (point)))
	      (t			;0 => no limit (use with care!)
	       (setq start (point-min))
	       (setq search-limit (point-max))))))
    (while (and start
		(< start search-limit)
		(> ts-count 0))
      (setq start (time-stamp-once start search-limit ts-start ts-end
				   ts-format format-lines end-lines))
      (setq ts-count (1- ts-count))))
  nil)

(defun time-stamp-once (start search-limit ts-start ts-end
			ts-format format-lines end-lines)
  "Update one time stamp.  Internal routine called by \\[time-stamp].
Returns the end point, which is where `time-stamp' begins the next search."
  (let ((case-fold-search nil)
	(end nil)
	end-search-start
	(end-length nil))
    (save-excursion
      (save-restriction
	(widen)
	;; Find the location of the time stamp.
	(while (and (< (goto-char start) search-limit)
		    (not end)
		    (re-search-forward ts-start search-limit 'move))
	  (setq start (point))
	  (if (not time-stamp-inserts-lines)
	      (forward-line format-lines))
	  (setq end-search-start (max start (point)))
	  (if (= (forward-line end-lines) 0)
	      (progn
	       (and (bolp) (backward-char))
	       (let ((line-end (min (point) search-limit)))
		 (if (>= line-end end-search-start)
		     (progn
		      (goto-char end-search-start)
		      (if (re-search-forward ts-end line-end t)
			  (progn
			    (setq end (match-beginning 0))
			    (setq end-length (- (match-end 0) end))))))))))))
    (if end
	(progn
	  ;; do all warnings outside save-excursion
	  (cond
	   ((not time-stamp-active)
	    (if time-stamp-warn-inactive
		;; don't signal an error in a write-file-hook
		(progn
		  (message "Warning: time-stamp-active is off; did not time-stamp buffer.")
		  (sit-for 1))))
	   ((not (and (stringp ts-start)
		      (stringp ts-end)))
	    (message "time-stamp-start or time-stamp-end is not a string")
	    (sit-for 1))
	   (t
	    (let ((new-time-stamp (time-stamp-string ts-format)))
	      (if (and (stringp new-time-stamp)
		       (not (string-equal (buffer-substring start end)
					  new-time-stamp)))
		  (save-excursion
		    (save-restriction
		      (widen)
		      (delete-region start end)
		      (goto-char start)
		      (insert-and-inherit new-time-stamp)
		      (setq end (point))
		      ;; remove any tabs used to format time stamp
		      (if (search-backward "\t" start t)
			  (progn
			    (untabify start end)
			    (setq end (point))))))))))))
    ;; return the location after this time stamp, if there was one
    (and end end-length
	 (+ end end-length))))


;;;###autoload
(defun time-stamp-toggle-active (&optional arg)
  "Toggle `time-stamp-active', setting whether \\[time-stamp] updates a buffer.
With ARG, turn time stamping on if and only if ARG is positive."
  (interactive "P")
  (setq time-stamp-active
	(if (null arg)
	    (not time-stamp-active)
	  (> (prefix-numeric-value arg) 0)))
  (message "time-stamp is now %s." (if time-stamp-active "active" "off")))

(defun time-stamp--format (format time)
  "FORMAT a TIME in zone `time-stamp-time-zone'.
Internal helper used by `time-stamp-string-preprocess'."
  (format-time-string format time time-stamp-time-zone))

(defun time-stamp-string (&optional ts-format time)
  "Generate the new string to be inserted by \\[time-stamp].
Optionally use format TS-FORMAT instead of `time-stamp-format' to
format the string.  Optional second argument TIME is only for testing;
normally the current time is used."
  (if (stringp (or ts-format (setq ts-format time-stamp-format)))
      (time-stamp-string-preprocess ts-format time)))


(defconst time-stamp-no-file "(no file)"
  "String to use when the buffer is not associated with a file.")

;;; time-stamp is transitioning to be more compatible with format-time-string.
;;; This function implements the differences.
;;;      At all times, all the formats recommended in the doc string
;;; of time-stamp-format will work not only in the current version of
;;; Emacs, but in all versions that have been released within the past
;;; two years.
;;;      The : modifier is a temporary conversion feature used to resolve
;;; ambiguous formats--formats that are changing (over time) incompatibly.
(defun time-stamp-string-preprocess (format &optional time)
  "Use a FORMAT to format date, time, file, and user information.
Optional second argument TIME is only for testing.
This is an internal routine implementing extensions to `format-time-string'
and all `time-stamp-format' compatibility."
  (let ((fmt-len (length format))
	(ind 0)
	cur-char
	(result ""))
    (while (< ind fmt-len)
      (setq cur-char (aref format ind))
      (setq
       result
       (concat
        result
        (cond
         ((eq cur-char ?%)
	  (let ((prev-char nil)
		(field-width "")
		field-result
		(alt-form 0)
		(change-case nil)
		(upcase nil)
		(flag-pad-with-spaces nil)
		(flag-pad-with-zeros nil)
		(flag-minimize nil)
		(paren-level 0))
	    ;; eat any additional args to allow for future expansion
	    (while (progn
		     (setq ind (1+ ind))
		     (setq cur-char (if (< ind fmt-len)
				        (aref format ind)
				      ?\0))
		     (or (eq ?. cur-char)
		         (eq ?, cur-char) (eq ?: cur-char) (eq ?@ cur-char)
		         (eq ?- cur-char) (eq ?+ cur-char) (eq ?_ cur-char)
		         (eq ?\s cur-char) (eq ?# cur-char) (eq ?^ cur-char)
		         (and (eq ?\( cur-char)
			      (not (eq prev-char ?\\))
			      (setq paren-level (1+ paren-level)))
		         (if (and (eq ?\) cur-char)
			          (not (eq prev-char ?\\))
			          (> paren-level 0))
			     (setq paren-level (1- paren-level))
		           (and (> paren-level 0)
			        (< ind fmt-len)))
		         (if (and (<= ?0 cur-char) (>= ?9 cur-char))
			     ;; get format width
			     (let ((field-index ind)
			           (first-digit cur-char))
			       (while (progn
				        (setq ind (1+ ind))
				        (setq cur-char (if (< ind fmt-len)
						           (aref format ind)
						         ?\0))
					(and (<= ?0 cur-char)
					     (>= ?9 cur-char))))
			       (setq field-width
				     (substring format field-index ind))
			       (setq ind (1- ind))
			       (setq cur-char first-digit)
			       t))))
	      (setq prev-char cur-char)
	      ;; some characters we actually use
	      (cond ((eq cur-char ?:)
		     (setq alt-form (1+ alt-form)))
		    ((eq cur-char ?#)
		     (setq change-case t))
		    ((eq cur-char ?^)
		     (setq upcase t))
		    ((eq cur-char ?0)
		     (setq flag-pad-with-zeros t))
		    ((eq cur-char ?-)
		     (setq field-width "1" flag-minimize t))
		    ((eq cur-char ?_)
		     (setq field-width "2" flag-pad-with-spaces t))))
	    (setq field-result
	          (cond
	           ((eq cur-char ?%)
	            "%")
	           ((eq cur-char ?a)    ;day of week
                    (if (> alt-form 0)
                        (if (string-equal field-width "")
                            (time-stamp--format "%A" time)
                          "")           ;discourage "%:3a"
                      (if (or change-case upcase)
                          (time-stamp--format "%#a" time)
	                (time-stamp--format "%a" time))))
	           ((eq cur-char ?A)
		    (if (or change-case upcase (not (string-equal field-width
								  "")))
			(time-stamp--format "%#A" time)
	              (time-stamp--format "%A" time)))
	           ((eq cur-char ?b)    ;month name
                    (if (> alt-form 0)
                        (if (string-equal field-width "")
                            (time-stamp--format "%B" time)
                          "")           ;discourage "%:3b"
                      (if (or change-case upcase)
                          (time-stamp--format "%#b" time)
	                (time-stamp--format "%b" time))))
		   ((eq cur-char ?B)
		    (if (or change-case upcase (not (string-equal field-width
								  "")))
			(time-stamp--format "%#B" time)
	              (time-stamp--format "%B" time)))
	           ((eq cur-char ?d)    ;day of month, 1-31
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?H)    ;hour, 0-23
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?I)    ;hour, 1-12
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?m)    ;month number, 1-12
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?M)    ;minute, 0-59
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?p)    ;am or pm
	            (if change-case
                        (time-stamp--format "%#p" time)
                      (time-stamp--format "%p" time)))
	           ((eq cur-char ?P)    ;AM or PM
	            (time-stamp--format "%p" time))
	           ((eq cur-char ?S)    ;seconds, 00-60
	            (time-stamp-do-number cur-char alt-form field-width time))
	           ((eq cur-char ?w)    ;weekday number, Sunday is 0
	            (time-stamp--format "%w" time))
	           ((eq cur-char ?y)    ;year
                    (if (> alt-form 0)
                        (string-to-number (time-stamp--format "%Y" time))
                      (if (or (string-equal field-width "")
                              (<= (string-to-number field-width) 2))
                          (string-to-number (time-stamp--format "%y" time))
                        (time-stamp-conv-warn (format "%%%sy" field-width) "%Y")
                        (string-to-number (time-stamp--format "%Y" time)))))
	           ((eq cur-char ?Y)    ;4-digit year
	            (string-to-number (time-stamp--format "%Y" time)))
	           ((eq cur-char ?z)    ;time zone offset
                    (let ((field-width-num (string-to-number field-width))
                          ;; Handle numeric time zone ourselves, because
                          ;; current-time-zone cannot handle offsets
                          ;; greater than 24 hours.
                          (offset-secs
                           (cond ((numberp time-stamp-time-zone)
                                  time-stamp-time-zone)
                                 ((and (consp time-stamp-time-zone)
                                       (numberp (car time-stamp-time-zone)))
                                  (car time-stamp-time-zone))
                                 ;; interpret text time zone
                                 (t (car (current-time-zone
                                          time time-stamp-time-zone))))))
	              ;; we do our own padding; do not let it be updated further
	              (setq field-width "")
	              (cond (change-case
		             "")        ;discourage %z variations
		            ((and (= alt-form 0)
			          (not flag-minimize)
			          (not flag-pad-with-spaces)
			          (not flag-pad-with-zeros)
			          (= field-width-num 0))
		             (time-stamp-conv-warn "%z" "%#Z")
		             (time-stamp--format "%#Z" time))
			    (t (time-stamp-formatz-from-parsed-options
				flag-minimize
				flag-pad-with-spaces
				flag-pad-with-zeros
				alt-form
				field-width-num
				offset-secs)))))
	           ((eq cur-char ?Z)    ;time zone name
	            (if change-case
	                (time-stamp--format "%#Z" time)
	              (time-stamp--format "%Z" time)))
	           ((eq cur-char ?f)    ;buffer-file-name, base name only
	            (if buffer-file-name
	                (file-name-nondirectory buffer-file-name)
	              time-stamp-no-file))
	           ((eq cur-char ?F)    ;buffer-file-name, absolute name
	            (or buffer-file-name
	                time-stamp-no-file))
	           ((eq cur-char ?s)    ;system name, legacy
	            (system-name))
	           ((eq cur-char ?u)    ;user name, legacy
	            (user-login-name))
	           ((eq cur-char ?U)    ;user full name, legacy
	            (user-full-name))
	           ((eq cur-char ?l)    ;login name
	            (user-login-name))
	           ((eq cur-char ?L)    ;full name of logged-in user
	            (user-full-name))
	           ((eq cur-char ?h)    ;mail host name
	            (or mail-host-address (system-name)))
	           ((eq cur-char ?q)    ;unqualified host name
	            (let ((qualname (system-name)))
	              (if (string-match "\\." qualname)
		          (substring qualname 0 (match-beginning 0))
	                qualname)))
	           ((eq cur-char ?Q)    ;fully-qualified host name
	            (system-name))
	           ))
            (and (numberp field-result)
                 (= alt-form 0)
                 (string-equal field-width "")
                 ;; no width provided; set width for default
                 (setq field-width "02"))
	    (let ((padded-result
	           (format (format "%%%s%c"
			           field-width
			           (if (numberp field-result) ?d ?s))
		           (or field-result ""))))
	      (let* ((initial-length (length padded-result))
		     (desired-length (if (string-equal field-width "")
				         initial-length
				       (string-to-number field-width))))
	        (if (> initial-length desired-length)
		    ;; truncate strings on right
		    (if (and (stringp field-result)
			     (not (eq cur-char ?z))) ;offset does not truncate
		        (substring padded-result 0 desired-length)
                      padded-result)	;numbers don't truncate
	          padded-result)))))
         (t
	  (char-to-string cur-char)))))
      (setq ind (1+ ind)))
    result))

(defun time-stamp-do-number (format-char alt-form field-width time)
  "Handle compatible FORMAT-CHAR where only default width/padding will change.
ALT-FORM is whether `#' was specified.  FIELD-WIDTH is the string
width specification or \"\".  TIME is the time to convert.
This is an internal helper for `time-stamp-string-preprocess'."
  (let ((format-string (concat "%" (char-to-string format-char))))
    (if (and (> alt-form 0) (not (string-equal field-width "")))
	""				;discourage "%:2d" and the like
      (string-to-number (time-stamp--format format-string time)))))


(defvar time-stamp-conversion-warn t
  "Enable warnings about soon-to-be-unsupported forms in `time-stamp-format'.
If nil, these warnings are disabled, which would be a bad idea!
You really need to update your files instead.

The new formats will work with old versions of Emacs.
New formats are being recommended now to allow `time-stamp-format'
to change in the future to be compatible with `format-time-string'.
The new forms being recommended now will continue to work then.")


(defun time-stamp-conv-warn (old-form new-form)
  "Display a warning about a soon-to-be-obsolete format.
Suggests replacing OLD-FORM with NEW-FORM."
  (cond
   (time-stamp-conversion-warn
    (with-current-buffer (get-buffer-create "*Time-stamp-compatibility*")
      (goto-char (point-max))
      (if (bobp)
	  (progn
	    (insert
	     "The formats recognized in time-stamp-format will change in a future release\n"
	     "to be more compatible with the format-time-string function.\n\n"
	     "The following obsolescent time-stamp-format construct(s) were found:\n\n")))
      (insert "\"" old-form "\" -- use " new-form "\n"))
    (display-buffer "*Time-stamp-compatibility*"))))


;;; A principled, expressive implementation of time zone offset
;;; formatting ("%z" and variants).

;;; * Overarching principle for %z

;; The output should be clear and complete.
;;
;; That is,
;; a) it should be unambiguous what offset is represented, and
;; b) it should be possible to exactly recreate the offset.

;;; * Principles for %z

;; - The numeric fields are HHMMSS.
;; - The fixed point is at the left.  The first 2 digits are always
;;   hours, the next 2 (if they exist) minutes, and next 2 (if they
;;   exist) seconds.  "+11" is 11 hours (not 11 minutes, not 11 seconds).
;;   "+1015" is 10 hours 15 minutes (not 10 minutes 15 seconds).
;; - Each of the three numeric fields is two digits.
;;   "+1" and "+100" are illegal.  (Is that 1 hour? 10 hours? 100 hours?)
;; - The MMSS fields may be omitted only if both are 00.  Thus, the width
;;   of the field depends on the data.  (This is similar to how
;;   %B is always long enough to spell the entire month name.)
;; - The SS field may be omitted only if it is 00.
;; - Colons between the numeric fields are an option, unless the hours
;;   field is greater than 99, when colons are needed to prevent ambiguity.
;; - If padding with zeros, we must pad on the right, because the
;;   fixed point is at the left.  (This is similar to how %N,
;;   fractional seconds, must add its zeros on the right.)
;; - After zero-padding has filled out minutes and seconds with zeros,
;;   further padding can be blanks only.
;;   Any additional zeros would be confusing.

;;; * Padding for %z

;; Padding is under-specified, so we had to make choices.
;;
;; Principles guiding our choices:
;;
;; - The syntax should be easy to remember and the effect predictable.
;; - The syntax should enable as many useful effects as possible.
;;
;; Padding choices:
;;
;; - By default, pad with spaces, as other formats with non-digits do.
;;   The "0" flag pads first with zeros, until seconds are filled out.
;; - If padding with spaces, pad on the right.  This is consistent with
;;   how zero-padding works.  Padding on the right also keeps the fixed
;;   point in the same place, as other formats do for any given width.
;; - The %_z format always outputs seconds, allowing all added padding
;;   to be spaces.  Without this rule, there would be no way to
;;   request seconds that worked for both 2- and 3-digit hours.
;; - Conflicting options are rejected, lest users depend
;;   on incidental behavior.
;;
;; Padding combos that make no sense and are thus disallowed:
;;
;; %-:z   - minus minimizes to hours, : expands to minutes
;; %-::z  - minus minimizes to hours, :: expands to seconds
;; %_:z   - underscore requires seconds, : displays minutes
;; %_:::z - underscore requires seconds, ::: minimizes to hours
;;
;; Example padding effects (with offsets of 99 and 100 hours):
;;
;; %-7z   "+99    "   "+100:00"
;;  %7z   "+9900  "   "+100:00"
;; %07z   "+990000"   "+100:00"
;; %_7z   "+990000"   "+100:00:00"
;;
;; %7:::z "+99    "   "+100:00"
;; %7:z   "+99:00 "   "+100:00"
;; %07:z  "+99:00:00" "+100:00"
;; %7::z  "+99:00:00" "+100:00:00"

;;; * ABNF syntax of the offset string produced by %z

;; offset = sign ( hours [minutes [seconds]] /
;;                 hours [":" minutes [":" seconds]] /
;;                 bighours ":" minutes [":" seconds] ) padding
;; sign = "+" / "-"
;; hours = digitpair
;; minutes = digitpair
;; seconds = digitpair
;; digitpair = digit digit
;; digit = "0" / "1" / "2" / "3" / "4" / "5" / "6" / "7" / "8" / "9"
;; bighours = 1*digit digitpair
;; padding = *" "

(defun time-stamp-formatz-from-parsed-options (flag-minimize
                                               flag-pad-spaces-only
                                               flag-pad-zeros-first
                                               colon-count
                                               field-width
                                               offset-secs)
  "Formats a time offset according to a %z variation.

With no flags, the output includes hours and minutes: +-HHMM
unless there is a non-zero seconds part, in which case the seconds
are included: +-HHMMSS

FLAG-MINIMIZE is whether \"-\" was specified.  If non-nil, the
output may be limited to hours if minutes and seconds are zero.

FLAG-PAD-SPACES-ONLY is whether \"_\" was specified.  If non-nil,
seconds must be output, so that any padding can be spaces only.

FLAG-PAD-ZEROS-FIRST is whether \"0\" was specified.  If non-nil,
padding to the requested FIELD-WIDTH (if any) is done by adding
00 seconds before padding with spaces.

COLON-COUNT is the number of colons preceding the \"z\" (0-3).  One or
two colons put that many colons in the output (+-HH:MM or +-HH:MM:SS).
Three colons outputs only hours if minutes and seconds are zero and
includes colon separators if minutes and seconds are output.

FIELD-WIDTH is a whole number giving the minimum number of characters
in the output; 0 specifies no minimum.  Additional characters will be
added on the right if necessary.  The added characters will be spaces
unless FLAG-PAD-ZEROS-FIRST is non-nil.

OFFSET-SECS is the time zone offset (in seconds east of UTC) to be
formatted according to the preceding parameters.

This is an internal function used by `time-stamp'."
  ;; The caller of this function must have already parsed the %z
  ;; format string; this function accepts just the parts of the format.
  ;; `time-stamp-string-preprocess' is the full-fledged parser normally
  ;; used.  The unit test (in time-stamp-tests.el) defines the simpler
  ;; parser `format-time-offset'.
  (let ((hrs (/ (abs offset-secs) 3600))
        (mins (/ (% (abs offset-secs) 3600) 60))
        (secs (% (abs offset-secs) 60))
        (result ""))
    ;; valid option combo?
    (cond
     ((not (or (and flag-minimize (> colon-count 0))
               (and flag-pad-spaces-only (> colon-count 0))
               (and flag-pad-spaces-only flag-minimize)
               (and flag-pad-spaces-only flag-pad-zeros-first)
               (and flag-pad-zeros-first flag-minimize)))
      (setq result (concat result (if (>= offset-secs 0) "+" "-")))
      (setq result (concat result (format "%02d" hrs)))
      ;; Need minutes?
      (cond
       ((or (> hrs 99)
            (> mins 0)
            (> secs 0)
            (not (or flag-minimize (= colon-count 3)))
            (and (> field-width (length result))
                 flag-pad-zeros-first))
        ;; Need colon before minutes?
        (if (or (> colon-count 0)
                (> hrs 99))
            (setq result (concat result ":")))
        (setq result (concat result (format "%02d" mins)))
        ;; Need seconds, too?
        (cond
         ((or (> secs 0)
              (= colon-count 2)
              flag-pad-spaces-only
              (and (> field-width (length result))
                   flag-pad-zeros-first))
          ;; Need colon before seconds?
          (if (or (> colon-count 0)
                  (> hrs 99))
              (setq result (concat result ":")))
          (setq result (concat result (format "%02d" secs)))))))
      ;; Need padding?
      (let ((needed-padding (- field-width (length result))))
        (if (> needed-padding 0)
            (setq result (concat result (make-string needed-padding ?\s)))))))
    result))

(provide 'time-stamp)

;;; time-stamp.el ends here