(C) Martin Väth martin@mvath.de, mvath.develop@gmail.com
This project is under the GNU Public License 2.0.
A WebExtension which can display/remove duplicate bookmarks, empty folders, or descriptions
Available on AMO here: https://addons.mozilla.org/firefox/addon/bookmark-dupes/
After installing bookmarkdupes, the usage is rather simple:
To open bookmarkdupes, click the extension symbol (duplicate stars) or use the link in the options page of the extension. Do not enable “Expert Mode” (unless you fully understand the consequences, see separate section). Then select what you want to display:
After this, you will be offered the list of bookmarks with checkboxes; in case 1 the numbers indicate the order in which matching bookmarks were added according to the internally stored date. There are also buttons to select/unselect convenient sets of checkboxes.
Finally, there are buttons to remove the selected bookmarks in cases 1-3 or to strip the descriptions of the selected bookmarks in case 4. (The latter has some side effects, see section Known Bugs).
Be aware that removing bookmarks or stripping descriptions is irreversible! It is recommended to make backups of your bookmarks first!
Currently, there is no working version of the extension available for android (see section Known Bugs).
When you reorganized/added/removed bookmarks, make sure to update the displayed list (by pressing the corresponding button) before removing bookmarks or stripping descriptions! In particular, stripping descriptions with an outdated list will move the corresponding bookmarks to their previous location in the bookmark menu!
The extension requires the following permissions for these reasons:
When selecting the checkbox for expert mode, details can be configured to ignore certain bookmarks when calculating the list or under which cases bookmarks are considered to be dupes of each other.
In non-expert mode, two bookmarks are considered to be duplicates of each other if their URL coincides. In expert mode there are custom rules by which the URL which is actually used for comparison is to be modified. The details are as follows.
For every bookmark the rules are applied in the given order. There are 2 types of rules: Filter rules and URL modification rules (there are also disabled rules which are only listed but have no effect).
For both types of rules 4 regular expressions can be specified which are used to determine if the rule applies: If the corresponding regular expression is nonempty, the corresponding condition must be satisfied or the rule will not apply. (In the case of filter rules, at least one of these 4 regular expressions must be nonempty or that filter rules will not apply either.)
The 4 regular expressions refer to the bookmark's name or url, respectively, and the regular expressions must either match or not match, respectively.
Bookmark Menu\0Collection\0Example (where \0 denotes the null character
which can be matched by the regular expression with \0 or \x00).If a filter rule applies, the corresponding bookmark is ignored, i.e., it will not be considered as a duplicate and will neither appear in the list of empty folders nor of all bookmarks.
If a URL modification rule applies, a text replacement will occur: All parts matching a specified regular expression are substituted by a replacement text (which might be empty). The rules for this correspond to the javascript String.prototype.replace function with the global modifier, see e.g. https://wiki.selfhtml.org/wiki/JavaScript/Objekte/String/replace or https://developer.mozilla.org/de/docs/Web/JavaScript/Reference/Global_Objects/String/replace
In particular, the replacement text can contain symbols like
$& or $1 to refer to the whole matched text or to the
match of the first brace in the regular expression, respectively.
The following 5 replacement texts have a special meaning which goes beyond the standard javascript replacement rules.
Note that this special meaning is only active if this is the full replacement text. In other words, to get the special meaning, the replacement text must not contain anything else than these 4-6 characters. (This does not really restrict the functionality, because one can combine several rules to work around that limitation if necessary.)
\L$& is the match in lower case.\U$& is the match in upper case.$URL is the url before any other rules have been applied.$NAME is the bookmark's full name/path with folders separated by the
null character, e.g. “Bookmark Menu\0Collection\0Example”$TITLE is the bookmark's title without the path; for instance, for the
bookmark name “Bookmark Menu | Collection | Example” it is “Example”.Cases 3-5 can be preceeded or followed by $$& to prepend or append the match.
Consider two bookmarks as duplicate if their URL differs only in https: vs. http: at the beginnning.
Consider two bookmarks as duplicate if their URL matches up to the first ? symbol, i.e. only their additional information differs.
Do not consider any bookmark from any folder named Mr. Dupe as a duplicate.
Add all bookmarks to the list of duplicates, whether duplicate or not.
Add all bookmarks from the folder Bookmark Menu | Remove to the list of duplicates (no matter whether they actually are duplicates).
Consider two bookmarks as duplicate if they have the same name (instead of the same URL).
Consider two bookmarks as duplicate if they reside in the same folder.
^http: “by” ^https:Explanation: By replacing everywhere the URL beginnning with http:
by https:, it does not matter whether the actual URL started with http:
or https:. Of course, we could have exchanged the roles of http
and https in our rule.
\?.* “by” (empty string)Explanation: Simply omit the ? symbol and all other symbols following it in all URLs.
\0Mr\. Dupe\0 or \0Mr[.] Dupe\0Explanation: If a bookmark is in a folder named “Mr. Dupe”, its full name (path) will contain the text “\0Mr. Dupe\0”; so match that text. Since the “.” symbol has a special meaning for regular expressions, we have to quote it. This can be done by either \. or by looking for a character class […] which contains only the single symbol ..
.+ “by” http://dummyExplanation: Pretend that every bookmark has the URL http://dummy by replacing all characters (.+) of the original URL by that text.
^Bookmark Menu\0Remove\0 “Replace URL matches” .+ “by” http://dummyAs in 4, but only for bookmarks whose full name starts with the matching path.
This works only if the folder contains at least 2 bookmarks (because otherwise http://dummy is not a duplicate URL). Of course, one might use an actually duplicate URL instead of http://dummy to work around this limitation.
.+ “by” $TITLEExplanation: Replace all bookmark URLs by the corresponding bookmark title when looking for dupes.
.+ “by” $NAME followed by a further replacement rule “Replace URL matches” [^\0]*$ “by” (empty string)Explanation: First replace the URL by its full name path, and then omit the last component of this path by cutting of the longest sequence of non-\0-symbols at the end.
(in alphabetical order)