A duplicate group is a group of files that all match together. Every group has a **reference file** and one or more **duplicate files**. The reference file is the first file of the group. Its mark box is disabled. Below it, and indented, are the duplicate files.
You can mark duplicate files, but you can never mark the reference file of a group. This is a security measure to prevent dupeGuru from deleting not only duplicate files, but their reference. You sure don't want that, do you?
What determines which files are reference and which files are duplicates is first their folder state. A file from a reference folder will always be reference in a duplicate group. If all files are from a normal folder, the size determine which file will be the reference of a duplicate group. dupeGuru assumes that you always want to keep the biggest file, so the biggest files will take the reference position.
Although you can just click on **Edit-->Mark All** and then **Actions-->Send Marked to Recycle bin** to quickly delete all duplicate files in your results, it is always recommended to review all duplicates before deleting them.
To help you reviewing the results, you can bring up the **Details panel**. This panel shows all the details of the currently selected file as well as its reference's details. This is very handy to quickly determine if a duplicate really is a duplicate. You can also double-click on a file to open it with its associated application.
If you have more false duplicates than true duplicates (If your filter hardness is very low), the best way to proceed would be to review duplicates, mark true duplicates and then click on **Actions-->Send Marked to Recycle bin**. If you have more true duplicates than false duplicates, you can instead mark all files that are false duplicates, and use **Actions-->Remove Marked from Results**.
A **marked** duplicate is a duplicate with the little box next to it having a check-mark. A **selected** duplicate is a duplicate being highlighted. The multiple selection actions can be performed in dupeGuru in the standard way (Shift/Command/Control click). You can toggle all selected duplicates' mark state by pressing **space**.
When this mode is enabled, the duplicates are shown without their respective reference file. You can select, mark and sort this list, just like in normal mode.
The dupeGuru results, when in normal mode, are sorted according to duplicate groups' **reference file**. This means that if you want, for example, to mark all duplicates with the "exe" extension, you cannot just sort the results by "Kind" to have all exe duplicates together because a group can be composed of more than one kind of files. That is where Dupes Only mode comes into play. To mark all your "exe" duplicates, you just have to:
* Enable the Dupes Only mode.
* Add the "Kind" column with the "Columns" menu.
* Click on that "Kind" column to sort the list by kind.
* Locate the first duplicate with a "exe" kind.
* Select it.
* Scroll down the list to locate the last duplicate with a "exe" kind.
If you turn this switch on, some columns will display the value relative to the duplicate's reference instead of the absolute values. These delta values will also be displayed in a different color so you can spot them easily. For example, if a duplicate is 1.2 MB and its reference is 1.4 MB, the Size column will display -0.2 MB.
Dupes Only and Delta Values
---------------------------
The Dupes Only mode unveil its true power when you use it with the Delta Values switch turned on. When you turn it on, relative values will be displayed instead of absolute ones. So if, for example, you want to remove from your results all duplicates that are more than 300 KB away from their reference, you could sort the dupes only results by Size, select all duplicates under -300 in the Size column, delete them, and then do the same for duplicates over 300 at the bottom of the list.
dupeGuru supports post-scan filtering. With it, you can narrow down your results so you can perform actions on a subset of it. For example, you could easily mark all duplicates with their filename containing "copy" from your results using the filter.
**Windows:** To use the filtering feature, click on Actions --> Apply Filter, write down the filter you want to apply and click OK. To go back to unfiltered results, click on Actions --> Cancel Filter.
**Mac OS X:** To use the filtering feature, type your filter in the "Filter" search field in the toolbar. To go back to unfiltered result, blank out the field, or click on the "X".
In simple mode (the default mode), whatever you type as the filter is the string used to perform the actual filtering, with the exception of one wildcard: **\***. Thus, if you type "[*]" as your filter, it will match anything with [] brackets in it, whatever is in between those brackets.
For more advanced filtering, you can turn "Use regular expressions when filtering" on. The filtering feature will then use **regular expressions**. A regular expression is a language for matching text. Explaining them is beyond the scope of this document. A good place to start learning it is `regular-expressions.info <http://www.regular-expressions.info>`_.
Matches are case insensitive in both simple and regexp mode.
For the filter to match, your regular expression don't have to match the whole filename, it just have to contain a string matching the expression.
You might notice that not all duplicates in the filtered results will match your filter. That is because as soon as one single duplicate in a group matches the filter, the whole group stays in the results so you can have a better view of the duplicate's context. However, non-matching duplicates are in "reference mode". Therefore, you can perform actions like Mark All and be sure to only mark filtered duplicates.
