This manual explains installation, usage, and basic customization of TaskPaper mode package. For more advanced customization, hacking and scripting see the Scripting Guide.
You can also use Emacs' help commands to access the usual built-in documentation. Documentation strings to each function are available via C-h f
(describe-function
), individual key bindings can be investigated with C-h k
(describe-key
), and a complete list of key bindings is available using C-h m
(describe-mode
).
The recommended way to install TaskPaper mode is to install the package from MELPA using package.el
. First, configure package.el
and the MELPA repository by adding the following to your init file:
(require 'package)
(add-to-list 'package-archives
'("melpa" . "https://melpa.org/packages/"))
(package-initialize)
Then, after restarting Emacs or evaluating the above statements, issue the following command: M-x package-install RET taskpaper-mode RET
.
Alternatively, if you manage loading packages with use-package.el
then you can automatically install and configure TaskPaper mode by adding a declaration such as this one to your init file:
(use-package taskpaper-mode :ensure t)
Download taskpaper-mode.el
and save the file where Emacs can find it (i.e., a directory in your load-path
). Copy the following into your init file:
(require 'taskpaper-mode)
Files with the .taskpaper
extension use TaskPaper mode by default. If you want other filenames to be associated with TaskPaper mode, add the corresponding entry to the auto-mode-alist
in your init file:
(add-to-list 'auto-mode-alist '("\\.todo\\'" . taskpaper-mode))
In TaskPaper mode each line makes a new item. When you hit RET
TaskPaper mode will auto-indent your items. If you don't need it, simply hit RET
again to remove the indentation. This can greatly speed up your writing.
The following commands auto-format and indent you items as you type.
-
RET
: Create a new item with the same level as the one under cursor (taskpaper-new-item-same-level
). -
M-RET
: Create a new task with the same level as the item under cursor (taskpaper-new-task-same-level
).
The following commands change the type of the current item.
-
C-c C-f p
: Format item under cursor as project (taskpaper-item-format-as-project
). -
C-c C-f t
: Format item under cursor as task (taskpaper-item-format-as-task
). -
C-c C-f n
: Format item under cursor as note (taskpaper-item-format-as-note
).
You can expand and collapse parts of the outline to control the level of detail that you see.
The main folding command C-TAB
(taskpaper-cycle
) changes the visibility of items in the buffer cycling through the most important states. The action of the command depends on the current cursor location.
When point is at the beginning of the buffer, rotate the entire buffer among the two states:
- Overview: Only top-level items are shown.
- Show All: Everything is shown.
When point is in an item, rotate current subtree among the three states:
- Folded: Only the main item is shown.
- Children: The item and its direct children are shown. From this state, you can move to one of the children and zoom in further.
- Subtree: The entire subtree under the item is shown.
When Emacs first visits a TaskPaper file, the global state is set to Show All, i.e., all items are visible. This can be configured through the user option taskpaper-startup-folded
.
The command C-c *
(taskpaper-outline-hide-other
) lets you focus on the current item under cursor. It hides everything except the current item with its ancestors and direct children. Other top-level items are shown to give an overview of the general document structure.
The command C-c C-z
(taskpaper-outline-overview
) switches to Overview state showing only top-level items.
The command C-c C-a
(taskpaper-outline-show-all
) switches to Show All state unfolding all items at all levels (also bound to ESC ESC
).
The following commands jump to other items in the buffer.
-
C-DOWN
: Go to the next item with the same level (taskpaper-outline-forward-same-level
). -
C-UP
: Go to the previous item with the same level (taskpaper-outline-backward-same-level
). -
S-UP
: Go to the parent item (taskpaper-outline-up-level
). -
C-c C-j
: Go to selected item (taskpaper-goto
).
The command C-c C-j
(taskpaper-goto
) prompts the user for an outline path to an item offering standard minibuffer completion for possible target locations. Completion packages like Vertico or Ivy provide faster and more convenient way to select an outline path in minibuffer using regular expressions or fuzzy matching and incremental narrowing of possible selections. Additionally, you can use Imenu to go to a specific project in the buffer.
Commands like C-c C-j
jumping to a different position in the current file automatically save the old position to the mark ring, to be able to return easily. The command C-c [
(taskpaper-mark-ring-goto
) navigates back to a recorded position. Using this command several times in direct succession moves through a ring of previously recorded positions. You can also use the command C-c %
(taskpaper-mark-ring-push
) to manually save the current position to the mark ring.
Outlines allow a document to be organized in a hierarchical structure. The structure editing commands let you easily rearrange the order and hierarchy of items in the outline.
Four main commands are provided for structure editing. The commands work on the current subtree (the current item plus all its children) and are assigned to the four arrow keys pressed with a modifier (META by default) in the following way.
-
M-LEFT
: Promote the subtree under cursor by one level (taskpaper-outline-promote-subtree
). -
M-RIGHT
: Demote the subtree under cursor by one level (taskpaper-outline-demote-subtree
). -
M-UP
: Move the subtree under cursor up past the previous same-level subtree (taskpaper-outline-move-subtree-up
). -
M-DOWN
: Move the subtree under cursor down past the next same-level subtree (taskpaper-outline-move-subtree-down
).
The commands M-LEFT
(taskpaper-outline-promote-subtree
) and M-RIGHT
(taskpaper-outline-demote-subtree
) change the current subtree to a different outline level — i.e., the level of all items in the tree is decreased or increased. Note that the scope of "current subtree" may be changed after a promotion.
The commands M-UP
(taskpaper-outline-move-subtree-up
) and M-DOWN
(taskpaper-outline-move-subtree-down
) move the entire current subtree (folded or not) past the next same-level subtree in the given direction. The cursor moves with the subtree, so these commands can be used repeatedly to "drag" a subtree to the wanted position.
Additionally, the commands TAB
and S-TAB
promote/demote the single item under cursor, or all items in the active region.
TaskPaper mode also provides following additional commands for working with subtrees (folded or not):
-
C-c #
: Narrow buffer to the subtree under cursor (taskpaper-narrow-to-subtree
). PressingC-x n w
(widen
) will remove narrowing from the current buffer. -
C-c C-m
: Mark the subtree under cursor (taskpaper-mark-subtree
). -
C-c C-x v
: Copy all visible items in region to the kill ring and clipboard (taskpaper-outline-copy-visible
). -
C-c C-x C-w
: Cut the subtree under cursor into the kill ring (taskpaper-cut-subtree
). -
C-c C-x M-w
: Copy the subtree under cursor into the kill ring (taskpaper-copy-subtree
). -
C-c C-x C-c
: Duplicate the subtree under cursor (taskpaper-clone-subtree
). -
C-c C-x C-y
: Paste the subtree from the kill ring as child of the current item (taskpaper-paste-subtree
).
In addition to the hierarchical ways of organizing your actions, tags provide another axis to categorize and assess items. You can assign any number of tags to each task, project, or note, adding information about due dates, priorities, people, locations, topics, estimated effort, and more.
To create a tag type the @
symbol preceded by a space and followed by a tag name with no spaces. Tag names may basically contain uppercase and lowercase letters, digits, hyphens, underscores, and dots. Tags can optionally have a value (or list of comma-separated values) in parentheses after the tag name:
@today
@effort(0.5)
@priority(1)
@due(2018-08-19)
@status(in press)
The value text inside can have whitespaces, but no newlines. If you need to include parentheses in the tag value, precede them with a backslash.
If an item has a certain tag, all its subitems will inherit that tag. To limit tag inheritance to specific tags, configure the user option taskpaper-tags-exclude-from-inheritance
.
After @
symbol M-TAB
offers in-buffer completion on tag names. The list of tags is created dynamically from all tags used in the current buffer. If your desktop intercepts the key binding M-TAB
to switch windows, use C-M-i
or ESC TAB
as an alternative or customize your environment.
In addition to the in-buffer completion TaskPaper mode also implements another tag selection method called fast tag selection. This allows you to select your commonly used tags with just a single key press. For this to work you should assign unique, case-sensitive, letters to most of your commonly used tags. You can do this by configuring the user option taskpaper-tag-alist
in your init file:
(setq taskpaper-tag-alist
'(("today" . ?t)
("due(%%)" . ?d)
("due(%%+7d)" . ?s)
("priority(1)" . ?1)
("priority(2)" . ?2)
("priority(3)" . ?3)))
Pressing C-c @
(taskpaper-item-set-tag-fast-select
) will then present you with a special interface, listing all predefined tags with corresponding selection keys.
Tag specifiers in taskpaper-tag-alist
can have value in parentheses. If the tag value begins with %%
the following time string will be expanded into an absolute date in ISO 8601 format. E.g., if today is May 8, 2018, the tag specifier due(%%+2d)
will expand to @due(2018-05-10)
, i.e., in two days from now. See Date and Time Formats section below for date & time formats you can use here. This way you can create handy presets for often used relative dates based on your average routine. If after stripping the leading %%
the tag value is empty, the user will be prompted for time stamp using interactive date & time prompt.
The command C-c C-r
(taskpaper-remove-tag-at-point
) deletes single tag under cursor.
Following tags have special meaning in TaskPaper mode:
@done
: To-do items can be marked as completed by applying the@done
tag to them (see the Completing Actions section).@search
: The@search
tags are used to store search queries on per-document basis (see the Storing Queries section).@project
: The@project
tags are used to store original project hierarchy of archived items (see the Archiving section).
Actionable items can be marked as completed by applying the @done
tag. The command C-c C-d
(taskpaper-item-toggle-done
) toggles done state for item under cursor. Alternatively, you can toggle done state of the task by clicking on the task mark with mouse-1
if the user option taskpaper-pretty-marks
is non-nil. If the user option taskpaper-complete-save-date
is non-nil, current date will be added to the @done
tag. See the documentation string of taskpaper-complete-save-date
for possible options. Additionally, you may specify a list of tags, which will be removed once the item is completed, using the user option taskpaper-tags-to-remove-when-done
. When the item is marked as completed, the hook taskpaper-after-completion-hook
is run.
By default, items tagged with @done
are visually crossed out. You can change the way completed items are displayed by customizing the taskpaper-done-item
face or disable the additional fontification of these items completely by configuring the user option taskpaper-fontify-done-items
.
The Emacs calendar displays a three-month calendar with holidays from different countries and cultures. Following commands provide some integration with the calendar.
-
C-c SPC
: Display current date or date under cursor in calendar (taskpaper-show-in-calendar
). -
C-c >
: Access the calendar at the current date or date under cursor (taskpaper-goto-calendar
). -
C-c <
: Insert a time stamp corresponding to the cursor date in the calendar (taskpaper-date-from-calendar
). -
C-c .
: Prompt for a date and insert a corresponding time stamp (taskpaper-read-date-insert-timestamp
).
The command C-c >
(taskpaper-goto-calendar
) goes to the calendar at the current date. If point is on a tag with value, the command interprets it as date and goes to this date instead. With a C-u
prefix, the command always goes to the current date.
The command C-c .
(taskpaper-read-date-insert-timestamp
) prompts for the date & time to insert at point. You can enter a date using date & time syntax described below. The current interpretation of your input will be displayed live in the minibuffer, right next to your input. If you find this distracting, turn the display off with the user option taskpaper-read-date-display-live
.
Parallel to the minibuffer prompt the current interpretation of your input is tracked in the calendar window (see the user option taskpaper-read-date-popup-calendar
). You can control the calendar from the minibuffer using the following commands:
-
RET
: Choose date at cursor in calendar. -
mouse-1
: Select date by clicking on it. -
>
/<
: Scroll calendar forward/backward by one month. -
C-.
: Go to the current date. -
!
: Mark diary items in calendar.
Dates and times are an essential part of any task management system. You can store dates and times in tag values and perform date search on them (see the Searching section) to remind yourself of e.g., upcoming deadlines or scheduled activities. These examples show the different formats that you can use when entering dates and times in the date & time prompt. The same formats can be used for date & time values in tags.
- Do weekly review @due(Friday 12:30)
- Attend meeting @due(2017-08-11 8am)
If the time string is unparseable, current time is returned.
Note: Using of @due
tag here for storing dates is merely convention. You can use any tag name which makes sense to you, e.g., @deadline
.
Dates resolve to midnight of the given date.
2017
2017-5
2017-05-10
--05-10
2017-W02
2017-W02-5
this week
last month
next quarter
next Friday
last June 5
June
today
tomorrow
yesterday
TaskPaper mode understands English month and weekday abbreviations. If you enter the name of a specific time period, the date will be at its beginning. So June
without date means June first. Week refers to ISO 8601 standard week that starts on Monday, not Sunday.
You can refer to relative dates using common words (today
, tomorrow
, yesterday
, last week
, etc.). Words this
, next
, and last
have specific meanings: this Friday
always means the Friday in this week, next Friday
always means the Friday in the next week, and last Friday
always means the Friday in the last week, regardless of what day today is. Other units work in the same way.
Times are relative to the current date.
6am
6:45pm
12:45
now
Duration offsets are relative to the current time.
+5min
+2h
-1d
-4w
+2m
+2q
+1y
+5 minutes
+2 hours
-1 day
-4 weeks
+2 months
+2 quarters
+1 year
+3 Wed
The shorthands min
, h
, d
, w
, m
, q
, and y
stand for minute, hour, day, week, month, quarter, and year, respectively. Positive numbers stand for the future whereas negative numbers stand for the past. In case of weekdays the date will be the nth such day, e.g., +2 Tue
means the second Tuesday from now.
You can combine dates, times, and duration offsets:
tomorrow at 8am
2017-08-19 21:00
2017 Jan 10 14:25
today +2m +2d 5pm
The time string is evaluated from left to right.
When you need tasks that repeat themselves on a regular basis, you can set a recurring due date. Use relative dates in your tag values, and then do a date-based search (see the Searching section). Here are a few examples of recurring tasks:
- Due everyday @due(today)
- Due on each Tuesday @due(Tue)
- Due on at the Monday start of each week @due(this week)
- Due on the 1st of each month @due(this month)
- Due on the 15th of each month @due(this month +15d)
- Due on the second Friday of each month @due(this month +2 Fri)
- Due on each Monday, Wednesday, and Friday @due(Mon,Wed,Fri)
It doesn't cover all cases, but may work for you. For more flexible recurrence setup see the corresponding scripting example in the Scripting Guide.
Please consider that recurring appointments like any other event that takes place at a specific time would do best in your calendar because you cannot act on them until that moment.
TaskPaper mode auto-creates clickable hyperlinks when it recognizes link text. Below are some examples of the plain links that will be recognized.
~/filename.txt
./filename.txt
../filename.txt
/home/username/filename.txt
/username@host:filename.txt
http://www.example.net
mailto:[email protected]
Absolute file links are starting with /
or ~/
. Relative file links starting with ./
or ../
are relative to the location of your TaskPaper file. Spaces in file links must be protected using backslash, e.g., ./my\ file.txt
. By default, file links to non-existing local files highlight red.
The command C-c C-l
(taskpaper-insert-file-link-at-point
) inserts a plain file link at point offering standard minibuffer completion to select the name of the file. The path to the file is inserted relative to the directory of the current TaskPaper file, if the linked file is in the current directory or in a subdirectory of it, or if the path is written relative to the current directory using ../
. Otherwise an absolute path is used, if possible with ~/
for your home directory. You can force an absolute path with C-u
prefix.
Tip: If you want to sync your TaskPaper files with all linked images and other attachments across different devices via some cloud service like Dropbox, put the TaskPaper files together with the attachments somewhere inside the synced directory and link to them from your TaskPaper documents using the relative link syntax.
TaskPaper mode also has experimental support for inline Markdown links in form [Link description](destination)
:
[URL](http://www.example.net)
[File](/home/username/filename.txt)
[Email](mailto:[email protected])
This descriptive syntax is useful for long links which can interrupt reading. If markup hiding is enabled (see the Customization section), only the description part is displayed as hyperlink making the link more readable. You can hover the mouse pointer over the description text to see the destination part. The destination part can include up to two levels of balanced, nested parentheses. Spaces in file links must be backslash-escaped.
Links inside projects and tasks are supported although not encouraged.
If the point is on a link the command C-c C-o
or mouse-1
(taskpaper-open-link-at-point
) will call the system command to open the URI. Furthermore, it will start composing a message in email links or visit local and remote text files in file links with Emacs and select a suitable application for local non-text files. Classification of files is based on file extension only. For non-specified extensions the system command to open files, like open
on MS Windows and macOS, or the command specified in the mailcaps on GNU/Linux will be used. For more details see the user options taskpaper-file-apps
and taskpaper-open-non-existing-files
.
You can jump to the next/previous link in the buffer using the commands C-c C-x C-n
(taskpaper-next-link
) and C-c C-x C-p
(taskpaper-previous-link
). When reaching the buffer boundaries the search fails once, and then wraps around. Because the key bindings are probably too long you might want to bind this also to M-n
and M-p
.
The command C-c C-x C-v
(taskpaper-toggle-inline-images
) toggles the inline display of linked images within the buffer skipping images larger than specified by max-image-size
. Large images may be scaled down to fit in the buffer by setting the user option taskpaper-max-image-size
. Resizing works in Emacs built with ImageMagick support. You can ask for inline images to be displayed at startup by configuring the user option taskpaper-startup-with-inline-images
.
You can use Markdown syntax for inline emphasis in notes. Text enclosed by single asterisk (*
) or underscore (_
) characters is emphasized. Double *
or _
produces strong emphasis:
Emphasis with *asterisks* or _underscores_.
Strong emphasis with **asterisks** or __underscores__.
You can use whichever style you prefer; the only restriction is that the same character must be used as opening and closing emphasis delimiter. A *
or _
character surrounded by whitespaces, or backslash-escaped, will not trigger emphasis. By default, emphasis will be displayed in italic while strong emphasis will be displayed in bold face.
The span of the emphasis is limited to a single item. Emphasis and strong emphasis can be nested. Inline emphasis inside projects and tasks is supported although not encouraged.
If you often use the asterisk or underscore in a different context, you can disable the interpretation of these characters by configuring the user option taskpaper-use-inline-emphasis
.
The following commands sort same-level items. When point is at the beginning of the buffer, the top-level items are sorted. When point is in an item, the children of the current item are sorted. Sorting is case-insensitive. A C-u
prefix will reverse the sort order.
-
C-c C-s a
: Sort same-level items by item text (taskpaper-sort-by-text
). -
C-c C-s t
: Sort same-level items by item type (taskpaper-sort-by-type
).
When sorting is done, the hook taskpaper-after-sorting-items-hook
is run. When children are sorted, hook functions are called with point on the parent item.
To meet more complex needs you can define your own sorting functions as described in the Scripting Guide.
Filtering hides items that don't match the search creating a sparse tree, so that the entire document is folded as much as possible, but the selected information is made visible along with the outline hierarchy above it to provide minimal context.
The command C-c /
(taskpaper-occur
) prompts for a regular expression and creates a sparse tree with all matches. Each match is also highlighted. You can jump to the next/previous match in the buffer using the commands M-g n
(next-error
) and M-g p
(previous-error
). The highlights disappear when the buffer is changed by an editing command, or after pressing C-c C-c
.
The command C-c C-a
(taskpaper-outline-show-all
) unfold all items at all levels (also bound to ESC ESC
).
You can create a sparse tree based on specific combinations of items' assigned tags and attributes. This is a great tool that lets you focus on the right things at the right time. For example, it allows you to get a filtered view of specific items grouped by same context or see which tasks should be done soon and which are overdue.
TaskPaper mode has a special mode for incremental querying. The I-query mode is entered by pressing C-c C-i
(taskpaper-iquery
). Query results are updated instantly as you type, creating a sparse tree with all matches. The command C-c C-q
(taskpaper-query
) is a non-incremental querying command, which requires you to type the entire query string before searching begins. This form of static, one-time querying (as opposed to incremental, on-the-fly querying) may be preferable in some situations, such as over slow network connections or on excessively large and deeply nested outlines, which may affect search responsiveness. You can limit your searches to a certain subtree or region by narrowing the buffer with C-c #
(taskpaper-narrow-to-subtree
) or C-x n n
(narrow-to-region
) respectively.
If your query has a formatting error the query string highlights red in the minibuffer. In addition to the standard motion and editing commands both static and incremental query modes define some additional key bindings while in minibuffer. Pressing TAB
while editing query string offers completion on attribute names at point (see below). Pressing C-c C-c
clears the query string and displays all items in the outline. Pressing ESC ESC
or C-g
quits the I-query mode while preserving the outline view.
By default, incremental query results are not updated immediately after each character typed. Instead, the query is evaluated after a certain period of idle time. If you would like to change this behavior, customize the user option taskpaper-iquery-delay
. For example, delaying evaluation until the user stops typing is beneficial when working on long documents over a slow network connection. To disable the delay entirely set the taskpaper-iquery-delay
value to zero:
(setq taskpaper-iquery-delay 0)
If the point is on a tag the command C-c C-t
or mouse-1
(taskpaper-query-tag-at-point
) queries the buffer based on that tag. When cursor is on the @search
tag, it evaluates the embedded query stored in the tag value (see the Storing Queries section). For other tags when cursor is on the tag name, it queries for the tag name, otherwise for the name and associated value.
The query language syntax is described below.
Note: Though the query language syntax described here represents a valid subset of search syntax implemented in TaskPaper v3 app (with some minor exceptions), the search behavior is slightly different. TaskPaper mode does not support the item path syntax and set operations in search queries evaluating tag inheritance instead.
Every item in the outline has its own set of attributes. Explicit attributes are associated with assigned tags and have the same names, e.g., the tag @priority(1)
added to an item will be translated to an attribute named "priority" whose value is "1". And vice versa, setting an explicit attribute programmatically will change the corresponding tag. TaskPaper mode also includes some implicit (built-in) read-only attributes, which are not associated with tags and set in other way:
type
: Item's type (project, task, or note)text
: Item's full line of text sans indentation
Search predicates describe what you are looking for. The full predicate pattern looks like this:
@<attribute> <relation> [<modifiers>] <search term>
Predicates start with the attribute, whose value or existence you want to test. Attribute represents tag with the same name or one of the implicit attributes and search term will be compared with the tag value. Pressing TAB
offers completion on attribute names at point in both static and incremental query modes. Attribute names are always treated as case-sensitive.
Relations determine the test that the predicate performs:
=
: True if the attribute and search term are equal<
: True if the attribute is less than the search term>
: True if the attribute is greater than the search term<=
: True if the attribute is less than or equal to the search term>=
: True if the attribute is greater than or equal to the search term!=
: True if the attribute and search term are not equal~=
: True if the attribute matches the search termcontains
: True if the attribute contains the search termbeginswith
: True if the attribute begins with the search termendswith
: True if the attribute ends with the search termmatches
: True if the attribute matches the search term (same as~=
)
The default type of comparison is case-insensitive string comparison. You can change this behavior by providing a modifier after the relation. The available modifiers are:
i
: Case-insensitive string compare (the default)s
: Case-sensitive string comparen
: Numeric compared
: Date comparel
: List compare (can be combined with other modifiers)
Search terms can contain multiple words in sequence. Leading and trailing whitespaces are removed and multiple inter-word whitespaces collapsed into to a single space. If you want to search for an exact word or phrase preserving whitespaces, enclose the search term in double quotes. If some words or symbols are part of the predicate syntax or Boolean operators ("and", "or", "not", "matches", "@", etc.), they must be enclosed in double quotes. To include a double-quote character in a quoted search term, precede it with a backslash. If no search term is provided, attribute's presence will be tested.
In case of numeric compare [n]
both sides of the compare are converted to numbers before comparing. This means "01" will equal to "1", which is not true when doing the default string compare.
In case of date compare [d]
both sides of the compare are converted to time values before comparing. The date & time format is described in the Date and Time Formats reference section.
In case of list compare [l]
both sides are converted to lists (comma-separated) before comparing. Other modifiers can precede the list modifier making the search more specific.
In case of list compare using =
, <
, >
, <=
, >=
, !=
, both sides are compared element by element, from left to right.
In case of string compare contains
will return matches if the right side is a substring of the left side. In case of list compare contains
will return matches if the right side is a subset of the left side. For instance the search predicate @value contains[nl] 10,20
will match an item, which has the tag @value(2,10,18,20,22)
.
In case of string compare matches
will return matches if the left side matches the right side converted to a regular expression. In case of list compare matches
will return matches if the left side is a subset of the right side. For instance the search predicate @value matches[nl] 10,20,22
will match an item, which has the tag @value(10,22)
.
Here are some examples for predicates:
@today
@type = note
@due <=[d] +10d
@due <=[d] next week
@priority >[n] 3
@amount <[n] 1.2e3
@text endswith ?
@text matches " v[0-9]"
@text contains[s] new logo
@value contains[nl] 10,20
@text contains "this is not what I want"
@text contains "\"Winter\" by A. Vivaldi"
You don't need to enter the entire predicate pattern every time you search. Predicates use default values when part of the pattern is missing. Attribute defaults to text
, relation defaults to contains
, and modifier defaults to i
. For example the following predicates are equal:
Inbox
@text Inbox
contains Inbox
@text contains Inbox
@text contains[i] Inbox
You can combine predicates with Boolean and
, or
, and not
operators:
@due <=[d] +14d and not @done or @today
Binary logical and
binds more strongly than or
. Unary logical not
binds more strongly than both and
and or
. You can use parentheses to explicitly denote precedence by grouping parts of your query that should be evaluated first:
@due <=[d] +14d and not (@done or @hold)
It's common to search TaskPaper items based on their type: project, task, or note.
For example you might want to find the "Inbox" project. The default way to do this would be:
@type = project and Inbox
TaskPaper mode adds a shortcut for type based searches. The shortcut version of this search is:
project Inbox
These are the shortcut forms and what they expand to:
project
expands to@type = project and
task
expands to@type = task and
note
expands to@type = note and
You can embed search queries in a TaskPaper document using @search
tags. This is a handy way to manage search queries on per-document basis. Note that the parenthesis escaping rule must be followed:
Searches:
Next action @search(@next and not @done)
Active tasks @search(task and not \(@done or @hold\))
The command C-c ?
(taskpaper-query-read-select
) lets the user select a query offering standard minibuffer completion. Query description will be derived from the item's text. If the user option taskpaper-custom-queries
is set (see below), its entries will be added to the selection candidates. Completion packages like Vertico or Ivy provide faster and more convenient way to select a query in minibuffer using regular expressions or fuzzy matching and incremental narrowing of possible selections.
Fast selection interface allows you to save your commonly used search queries and later select them with just a single key press. For this to work you should assign unique, case-sensitive, letters (or other characters, e.g., numbers) to your saved queries. You can do this by configuring the user option taskpaper-custom-queries
in your init file:
(setq taskpaper-custom-queries
'((?w "Waiting" "@waiting and not @done")
(?d "Due Soon" "@due <=[d] +14d and not @done")
(?o "Overdue" "@due <[d] today and not @done")
(?p "Pending" "@start <=[d] today and @due >=[d] today and not @done")))
The initial value in each item defines the key you have to press. The second parameter is a short description and the last one is the query string to be used for the matching. If the first parameter is a string, it will be used as block title to visually group queries. Pressing C-c !
(taskpaper-query-fast-select
) will then present you with a special interface, listing all predefined queries with corresponding selection keys.
By default, predefined (embedded and custom) queries will be evaluated in non-incremental querying mode. If you want to use incremental mode for predefined queries, configure the user option taskpaper-iquery-default
.
You can configure certain queries to be executed automatically when visiting a TaskPaper file. E.g., you can ask for all leaf notes (notes, which may contain other notes, but no task or project items) to be folded at startup by adding following to your init file:
(add-hook 'taskpaper-mode-hook
(lambda () (taskpaper-query "not @type = note")))
When reorganizing your outline, you may want to refile or to copy some of the items into a different subtree. Cutting, finding the right location, and then pasting the item can be a cumbersome task, especially for large outlines with many sublevels. In addition to other structure editing commands (see the Structure Editing section) the following special commands can be used to simplify this process:
-
C-c C-w
: Move the subtree under cursor to different (possibly invisible) location (taskpaper-refile-subtree
). -
C-c M-w
: Copy the subtree under cursor to different (possibly invisible) location (taskpaper-refile-subtree-copy
).
The commands C-c C-w
and C-c M-w
offer possible target locations via outline path completion. This is the interface also used by the C-c C-j
goto command.
The subtree is filed below the target item as a subitem. Depending on taskpaper-reverse-note-order
setting, it will be either the first or last subitem.
When a project represented by a subtree is finished, you may want to remove the tree from the main outline by moving it to a dedicated archive location. This will keep your working outline clean and give you better visibility of all the remaining uncompleted actions. The archived items are out of the way, but accessible if you ever need to refer back to them.
The command C-c C-x a
(taskpaper-archive-subtree
) archives the subtree starting at the cursor position to the location given by taskpaper-archive-location
. The default archive location is a file in the same directory as the current file, with the name derived by appending _archive.taskpaper
to the current file name without extension. You can also choose what item to file archived items under. For details see the documentation string of the user option taskpaper-archive-location
. The subtree is filed below the target item as a subitem. Depending on taskpaper-reverse-note-order
setting, it will be either the first or last subitem. When the user option taskpaper-archive-save-context
is non-nil, a @project
tag with project hierarchy is added to the archived item.
When archiving the hook taskpaper-archive-hook
runs after successfully archiving a subtree. Hook functions are called with point on the subtree in the original location. At this stage, the subtree has been added to the archive location, but not yet deleted from the original one.
For querying a collection of TaskPaper files, TaskPaper mode includes a powerful Agenda mode. In this mode items from different TaskPaper files can be collected based on search queries and displayed in an organized way in a special Agenda mode buffer. This buffer is read-only, but provides commands to visit the corresponding locations in the original TaskPaper files. In this way, all information is stored only once, removing the risk that your agenda view and agenda files may diverge.
The information to be shown is normally collected from all agenda files, the files listed in the user option taskpaper-agenda-files
:
(setq taskpaper-agenda-files '("~/gtd.taskpaper" "~/projects/"))
If a directory is part of this list, all files with the extension .taskpaper
in this directory will be part of the list. You can customize the user option taskpaper-agenda-file-regexp
to change this behavior. If the user option taskpaper-agenda-skip-unavailable-files
is non-nil, Agenda mode will silently skip unavailable agenda files without issuing an error.
The following commands enter the Agenda mode. The command taskpaper-agenda-search
prompts the user for a search query. The command taskpaper-agenda-select
let the user select a predefined query via the custom query dialog described above. You may consider to assign global key bindings to these commands in your init file:
(global-set-key (kbd "C-c a") #'taskpaper-agenda-search)
(global-set-key (kbd "C-c s") #'taskpaper-agenda-select)
Two user options control how the Agenda mode buffer is displayed and whether the window configuration is restored when the Agenda mode exits: taskpaper-agenda-window-setup
and taskpaper-agenda-restore-windows-after-quit
. For details see the documentation strings of these user options.
Before being inserted into an Agenda mode buffer, the items are sorted. Sorting can be customized using the user option taskpaper-agenda-sorting-predicate
. If the variable is nil
, which is the default setting, agenda items just appear in the sequence in which they are found in the agenda files. The sorting predicate function is called with two string arguments, the items to compare, and should return non-nil if the first item should sort before the second one.
In the example below items will be sorted according to their due dates form earliest to latest. The sorting is done by date & time value (converted to float number of seconds since the 1970 epoch). Items, which have no or empty @due
tag, are assumed to have 2100-12-12 as due date, effectively ending up at the bottom of the sorted list.
(setq taskpaper-agenda-sorting-predicate
(lambda (a b)
(setq a (or (taskpaper-string-get-attribute a "due") "2100-12-12")
b (or (taskpaper-string-get-attribute b "due") "2100-12-12"))
(taskpaper-time< a b)))
Items with equal sort keys maintain their relative order before and after the sort, which means, the taskpaper-agenda-sorting-predicate
option can be accommodated to order items according to multiple criteria. In the following example agenda items will be sorted numerically by priority from highest to lowest (where "1" means the highest) and within the same priority by due date from earliest to latest:
(setq taskpaper-agenda-sorting-predicate
(lambda (a b)
(let ((p1 (or (taskpaper-string-get-attribute a "priority") "99"))
(p2 (or (taskpaper-string-get-attribute b "priority") "99"))
(d1 (or (taskpaper-string-get-attribute a "due") "2100-12-12"))
(d2 (or (taskpaper-string-get-attribute b "due") "2100-12-12")))
(if (taskpaper-num= p1 p2) (taskpaper-time< d1 d2) (taskpaper-num< p1 p2)))))
See the Scripting Guide for the list of API functions, which can be used for attribute evaluation.
Items in an Agenda mode buffer are displayed as a flat list and linked back to the TaskPaper file where they originate. You are not allowed to edit the Agenda mode buffer itself, but commands are provided to show and jump to the original item location.
-
n
orDOWN
: Move to the next line (taskpaper-agenda-next-line
). -
p
orUP
: Move to the previous line (taskpaper-agenda-previous-line
). -
SPC
: Display the original location of the item in another window (taskpaper-agenda-show
). -
TAB
: Go to the original location of the item in another window (taskpaper-agenda-goto
). -
RET
: Go to the original location of the item and delete other windows (taskpaper-agenda-switch-to
). -
F
: Toggle Follow mode (taskpaper-agenda-follow-mode
). In Follow mode, as you move the cursor through the Agenda mode buffer, the other window always shows the corresponding location in the original TaskPaper file. The initial setting for this mode in new Agenda mode buffers can be set with the user optiontaskpaper-agenda-start-with-follow-mode
. -
c
: Display date under cursor in calendar (taskpaper-show-in-calendar
). -
>
: Access calendar for the date under cursor (taskpaper-goto-calendar
).
The command SPC
(taskpaper-agenda-show
) runs the hook taskpaper-agenda-after-show-hook
after an item has been shown from the agenda view. Hook functions are called with point in the buffer where the item originated. Thus, if you want to display only the current item, its ancestors and top-level items, put this in your init file:
(add-hook 'taskpaper-agenda-after-show-hook
#'taskpaper-outline-hide-other)
You can also querying the agenda view to further narrow your search. Following commands and key bindings are defined in an Agenda mode buffer:
-
I
: Query agenda view using I-query mode (taskpaper-iquery-mode
). -
Q
: Query agenda view non-interactively (taskpaper-query
). -
S
: Query agenda view using custom query selection dialog (taskpaper-query-fast-select
). -
t
: Show a filtered view of the items that contain the tag under cursor (taskpaper-query-tag-at-point
). -
/
: Prompt for a regular expression and show a filtered view with all matches highlighted (taskpaper-occur
). -
C-c C-c
: Remove highlights (taskpaper-occur-remove-highlights
). -
a
: Show all items (taskpaper-outline-show-all
).
-
o
: Delete other windows (delete-other-windows
). -
v
: Copy all visible items in region to the kill ring and clipboard (taskpaper-outline-copy-visible
). -
r
: Recreate the agenda view (taskpaper-agenda-redo
). Useful to reflect changes after modification of original TaskPaper files. -
q
: Quit agenda and remove the Agenda mode buffer (taskpaper-agenda-quit
). -
x
: Exit agenda and remove the Agenda mode buffer and all buffers loaded by Emacs for the compilation of the agenda view (taskpaper-agenda-exit
). Buffers created by the user to visit TaskPaper files will not be removed.
All the rest which did not fit elsewhere.
-
M-x taskpaper-save-all-taskpaper-buffers RET
: Save all TaskPaper mode buffers without user confirmation. -
M-x taskpaper-outline-normalize-indentation RET
: Normalize outline indentation. -
M-x taskpaper-mode-version RET
: Show TaskPaper mode version in the echo area. -
M-x taskpaper-mode-browse-manual RET
: Browse TaskPaper mode user's manual.
Although no configuration is necessary there are a few things that can be customized in addition to the options mentioned above. All configurations can be performed either via Emacs' Easy Customization Interface or by modifying Emacs' init files directly.
You may assign special faces to specific tags using the user option taskpaper-tag-faces
. For example:
(setq taskpaper-tag-faces
'(("start" . "green")
("today" . font-lock-warning-face)
("due" . (:foreground "red" :weight bold))))
A string is interpreted as a color. The option taskpaper-faces-easy-properties
determines if that color is interpreted as a foreground or a background color. For more details see the documentation string of the user option taskpaper-tag-faces
.
Note: While using a list with face properties as shown for due
should work, this does not always seem to be the case. If necessary, define a special face and use that.
Additionally, the faces used for syntax highlighting can be modified to your liking by issuing M-x customize-group RET taskpaper-faces RET
.
You can hide inline markup elements like emphasis delimiters by configuring the user option taskpaper-hide-markup
. The underlying buffer content remains unchanged, but the markup elements will be hidden. Markup hiding can be toggled using C-c C-x C-m
(taskpaper-toggle-markup-hiding
).
You can activate the task marks by setting the user option taskpaper-pretty-marks
to non-nil, which makes the task marks appear as UTF-8 characters. This does not change the underlying buffer content, but it overlays the UTF-8 character for display purposes only. The overlay characters for the task marks can be customized using the taskpaper-bullet
and taskpaper-bullet-done
user options. For example, you can set it to an empty ballot box and check mark character respectively:
(setq taskpaper-bullet ?\u25a2)
(setq taskpaper-bullet-done ?\u2713)
Tasks can then be marked as done by clicking on the task mark with mouse-1
.
The adaptive-wrap package is helpful for correct wrapping of indented long-line items. The package provides the adaptive-wrap-prefix-mode
minor mode which sets the wrap-prefix
property on the fly so that single-long-line paragraphs get word-wrapped in a way similar to what you'd get with M-q
using adaptive-fill-mode
, but without actually changing the buffer's text. Since v0.7 it supports single-line paragraphs and can be used in TaskPaper mode as well. The package can be activated globally by putting adaptive-wrap-mode.el
on the load path and adding the following lines to your Emacs init file:
(require 'adaptive-wrap)
(add-hook 'visual-line-mode-hook
(lambda () (adaptive-wrap-prefix-mode 1)))
(global-visual-line-mode 1)
This program 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, or (at your option) any later version.
This program 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 this program. If not, see http://www.gnu.org/licenses/.