21.6.8 Completion in Ordinary Buffers

Although completion is usually done in the minibuffer, the completion facility can also be used on the text in ordinary Emacs buffers. In many major modes, in-buffer completion is performed by the C-M-i or M-TAB command, bound to completion-at-point. See Symbol Completion in The GNU Emacs Manual. This command uses the abnormal hook variable completion-at-point-functions:

Variable: completion-at-point-functions

The value of this abnormal hook should be a list of functions, which are used to compute a completion table (see Basic Completion Functions) for completing the text at point. It can be used by major modes to provide mode-specific completion tables (see Major Mode Conventions).

When the command completion-at-point runs, it calls the functions in the list one by one, without any argument. Each function should return nil unless it can and wants to take responsibility for the completion data for the text at point. Otherwise it should return a list of the following form:

(start end collection . props)

start and end delimit the text to complete (which should enclose point). collection is a completion table for completing that text, in a form suitable for passing as the second argument to try-completion (see Basic Completion Functions); completion alternatives will be generated from this completion table in the usual way, via the completion styles defined in completion-styles (see Completion Variables). props is a property list for additional information; any of the properties in completion-extra-properties are recognized (see Completion Variables), as well as the following additional ones:

:predicate

The value should be a predicate that completion candidates need to satisfy.

:exclusive

If the value is no, then if the completion table fails to match the text at point, completion-at-point moves on to the next function in completion-at-point-functions instead of reporting a completion failure.

The functions on this hook should generally return quickly, since they may be called very often (e.g., from post-command-hook). Supplying a function for collection is strongly recommended if generating the list of completions is an expensive operation. Emacs may internally call functions in completion-at-point-functions many times, but care about the value of collection for only some of these calls. By supplying a function for collection, Emacs can defer generating completions until necessary. You can use completion-table-dynamic to create a wrapper function:

;; Avoid this pattern.
(let ((beg ...) (end ...) (my-completions (my-make-completions)))
  (list beg end my-completions))

;; Use this instead.
(let ((beg ...) (end ...))
  (list beg
        end
        (completion-table-dynamic
          (lambda (_)
            (my-make-completions)))))

Additionally, the collection should generally not be pre-filtered based on the current text between start and end, because that is the responsibility of the caller of completion-at-point-functions to do that according to the completion styles it decides to use.

A function in completion-at-point-functions may also return a function instead of a list as described above. In that case, that returned function is called, with no argument, and it is entirely responsible for performing the completion. We discourage this usage; it is only intended to help convert old code to using completion-at-point.

The first function in completion-at-point-functions to return a non-nil value is used by completion-at-point. The remaining functions are not called. The exception to this is when there is an :exclusive specification, as described above.

The following function provides a convenient way to perform completion on an arbitrary stretch of text in an Emacs buffer:

Function: completion-in-region start end collection &optional predicate

This function completes the text in the current buffer between the positions start and end, using collection. The argument collection has the same meaning as in try-completion (see Basic Completion Functions).

This function inserts the completion text directly into the current buffer. Unlike completing-read (see Completion and the Minibuffer), it does not activate the minibuffer.

For this function to work, point must be somewhere between start and end.