Navigate back one page in history.
Fills the element matched by selector
with content and falls back to the last used input if the element can't be found. You probably don't want this; it's used internally for editor.
That said, bind gs fillinput null [Tridactyl](https://addons.mozilla.org/en-US/firefox/addon/tridactyl-vim/) is my favourite add-on
could probably come in handy.
Focus the last used input on the page
focus the nth input on the page, or "special" inputs: "-l": last focussed input "-n": input after last focussed one "-N": input before last focussed one "-p": first password field "-b": biggest input field
Find a likely next/previous link and follow it
If a link or anchor element with rel=rel exists, use that, otherwise fall back to:
1) find the last anchor on the page with innerText matching the appropriate followpagepattern
.
2) call urlincrement with 1 or -1
If you want to support e.g. French:
set followpagepatterns.next ^(next|newer|prochain)\b|»|>>
set followpagepatterns.prev ^(prev(ious)?|older|précédent)\b|»|>>
the relation of the target page to the current page: "next" or "prev"
Navigate forward one page in history.
Hint a page.
bind ;c hint -c [class*="expand"],[class="togg"]
works particularly well on reddit and HN-pipe selector key
e.g, -pipe * href
returns the key. Only makes sense with composite
, e.g, composite hint -pipe * textContent | yank
.-W excmd...
append hint href to excmd and execute, e.g, hint -W exclaim mpv
to open YouTube videos. Use composite hint -pipe | [excmd]
instead.Excepting the custom selector mode and background hint mode, each of these
hint modes is available by default as ;<option character>
, so e.g. ;y
to yank a link's target.
To open a hint in the background, the default bind is F
.
Related settings:
With "short" names, Tridactyl will generate short hints that are never prefixes of each other. With "uniform", Tridactyl will generate hints of uniform length. In either case, the hints are generated from the set in "hintchars".
With "numeric" names, hints are always assigned using sequential integers, and "hintchars" is ignored. This has the disadvantage that some hints are prefixes of others (and you need to hit space or enter to select such a hint). But it has the advantage that the hints tend to be more predictable (e.g., a news site will have the same hints for its boilerplate each time you visit it, even if the number of links in the main body changes).
Lets you execute JavaScript in the page context. If you want to get the result back, use composite js ... | fillcmdline
Some of Tridactyl's functions are accessible here via the tri
object. Just do console.log(tri)
in the web console on the new tab page to see what's available.
Aliased to !js
If you want to pipe an argument to js
, you need to use the "-p" flag and then use the JS_ARG global variable, e.g:
composite get_current_url | js -p alert(JS_ARG)
Calls jumpprev(-n)
Similar to Pentadactyl or vim's jump list.
Open a new page in the current tab.
@param urlarr
- if first word looks like it has a schema, treat as a URI
- else if the first word contains a dot, treat as a domain name
- else if the first word is a key of [[SEARCH_URLS]], treat all following terms as search parameters for that provider
- else treat as search parameters for google
Related settings:
- "searchengine": "google" or any of [[SEARCH_URLS]]
- "historyresults": the n-most-recent results to ask Firefox for before they are sorted by frequency. Reduce this number if you find your results are bad.
Can only open about: or file: URLs if you have the native messenger installed, and on OSX you must set browser
to something that will open Firefox from a terminal pass it commmand line options.
Like open but doesn't make a new entry in history.
Scrolls the document of its first scrollable child element by n lines.
The height of a line is defined by the site's CSS. If Tridactyl can't get it, it'll default to 22 pixels.
Scrolls the document by n pages.
The height of a page is the current height of the window.
Scrolls the window or any scrollable child element by a pixels on the horizontal axis and b pixels on the vertical axis.
If two numbers are given, treat as x and y values to give to window.scrollTo If one number is given, scroll to that percentage along a chosen axis, defaulting to the y-axis
Note that if a
is 0 or 100 and if the document is not scrollable in the given direction, Tridactyl will attempt to scroll the first scrollable element until it reaches the very bottom of that element.
Read text content of elements matching the given selector
the selector to match elements
Cancel current reading and clear pending queue
Arguments:
Read the given text using the browser's text to speech functionality and the settings currently set
the command mode -t read the following args as text -c read the content of elements matching the selector
Blur (unfocus) the active element
If the url of the current document matches one of your search engines, will convert it to a list of arguments that open/tabopen will understand. If the url doesn't match any search engine, returns the url without modifications.
For example, if you have searchurls.gi set to "https://www.google.com/search?q=%s&tbm=isch", using this function on a page you opened using "gi butterflies" will return "gi butterflies".
This is useful when combined with fillcmdline, for example like this: bind O composite url2args | fillcmdline open
.
Note that this might break with search engines that redirect you to other pages/add GET parameters that do not exist in your searchurl.
Increment the current tab URL
the increment step, can be positive or negative
Open a URL made by modifying the current URL
There are several modes:
Text replace mode: urlmodify -t <old> <new>
Replaces the first instance of the text old
with new
.
http://example.com
-> (-t exa peta
) -> http://petample.com
Regex replacment mode: urlmodify -r <regexp> <new> [flags]
Replaces the first match of the regexp
with new
. You can use
flags i
and g
to match case-insensitively and to match
all instances respectively
http://example.com
-> (-r [ea] X g
) -> http://XxXmplX.com
Query replace mode: urlmodify -q <query> <new_val>
Replace the value of a query with a new one:
http://e.com?id=foo
-> (-q id bar
) -> `http://e.com?id=barQuery delete mode: urlmodify -Q <query>
Deletes the given query (and the value if any):
http://e.com?id=foo&page=1
-> (-Q id
) -> http://e.com?page=1
Graft mode: urlmodify -g <graft_point> <new_path_tail>
"Grafts" a new tail on the URL path, possibly removing some of the old tail. Graft point indicates where the old URL is truncated before adding the new path.
graft_point
>= 0 counts path levels, starting from the left
(beginning). 0 will append from the "root", and no existing path will
remain, 1 will keep one path level, and so on.graft_point
< 0 counts from the right (i.e. the end of the current
path). -1 will append to the existing path, -2 will remove the last path
level, and so on.http://website.com/this/is/the/path/component
Graft point: ^ ^ ^ ^ ^ ^
From left: 0 1 2 3 4 5
From right: -6 -5 -4 -3 -2 -1
Examples:
http://e.com/issues/42
-> (-g 0 foo
) -> http://e.com/foo
http://e.com/issues/42
-> (-g 1 foo
) -> http://e.com/issues/foo
http://e.com/issues/42
-> (-g -1 foo
) -> http://e.com/issues/42/foo
http://e.com/issues/42
-> (-g -2 foo
) -> http://e.com/issues/foo
The replace mode:
Go to the parent URL of the current tab's URL
Go to the root domain of the current URL
Opens the current configuration in Firefox's native JSON viewer in the current tab.
NB: Tridactyl cannot run on this page!
The specific key you wish to view (e.g, nmaps).
Shows a list of the current containers in Firefox's native JSON viewer in the current tab.
NB: Tridactyl cannot run on this page!
Display the (HTML) source of the current page.
Behaviour can be changed by the 'viewsource' setting.
If the 'viewsource' setting is set to 'default' rather than 'tridactyl', the url the source of which should be displayed can be given as argument. Otherwise, the source of the current document will be displayed.
Generated from excmds.ts. Don't edit this file!