Duplicate File Search Syntax Help Needed

[dave9]

I'm hoping for some help with the syntax needed to do the following:

I have a folder, let's call it S:\junk

I want to search for all files (prefer to exclude folder names but it doesn't really matter) in S:\junk that have file name and size duplicates anywhere else that Everywhere is indexing, and show search hits that are only duplicates to those in S:\junk, as well as showing the duplicate file in S:\junk, so I can use the sort by Path column and delete the duplicates in the S:\junk folder, leaving behind anything in S:\junk that is the only copy I have of the file. I hope that made sense.

[void]

You'll need Everything 1.5

Please try the following Everything 1.5 search:

Code: Select all

dupe-from:"s:\junk" dupe:name;size

Finding duplicates in Everything 1.5
dupe-from:
dupe:

[dave9]

Thanks for the quick reply!

I just tried that, running Everything ver. 1.5.0.1371a (x64), and it gave no search hits.

I should mention that I am 100% certain that many files in S:\junk have duplicates in file name and size elsewhere that Everything has indexed, so it's not just that there aren't any duplicates.

[void]

dupe-from: was added in Everything 1.5.0.1384.

[dave9]

It figures! Am I correct in observing that the x64 version of Everything has changed filenames from Everything64.exe to Everything.exe? If so then I just got the newest x64 version.

That code seems to work now!

... at least I think it does, have a lot of files to weed through to be certain. When the "size" syntax is used without a numeric value or measurement unit like KB or MB, does that mean granularity down to the exact byte size or only KB? To ensure the exact same file would or should I add "binary:" to that code?

Thanks!

[void]

Everything64.exe was renamed to Everything.exe

Please delete the old Everything64.exe

When the "size" syntax is used without a numeric value or measurement unit like KB or MB, does that mean granularity down to the exact byte size or only KB? To ensure the exact same file would or should I add "binary:" to that code?

size:123

will match files with a size of exactly 123 bytes.

[dave9]

Yes I was just wondering if using "size" without any numerical value after it, meant an exact byte match or some other rounding off sort of numeric comparison. For example if I had filename.txt that is 1030 bytes and filename.txt that is 1031 bytes, would using just "size" as it was in the code you posted, without a numeric value after "size:", consider them size duplicates? Since you posted only size:123 I must assume it is down to the byte level without having to specify that.

I did fix the everything.exe naming including the shortcut to it and the Tools-Options-General checkmark items that were broken by the file name change, by just un-checkmarking and re-checkmarking them.

[void]

dupe:size

will find files with exactly the same size.

[harryray2]

Is it possible to do this when a file has the same size but a slightly different name?

eg testing one for this.mp4
testing on fo thi.mp4

[NotNull]

No, not possible.

[ChrisGreaves]

Please delete the old Everything64.exe

Thanks Void. I have done that, fixed the shortcut pinned to my task bar, installed .1404a, ...
Cheers, Chris

[dave9]

Is it possible to do this when a file has the same size but a slightly different name?

eg testing one for this.mp4
testing on fo thi.mp4

Clearly based on my creation of this topic, I am not an expert on search syntax, but in the example you provided, you might want (depending on # of similar files you are indexing in the first place) at least one more attribute (is that the right word to use?) of the files, if not more of them to narrow the hits, then use wildcards in the filename, or only attributes and no filename.

For example if you check the size of this.mp4 and find it is 285MB, then could use something like:

Code: Select all

*thi*.mp4 size:285MB

- Or if you check the length of the file (see paragraph below) and it's 42:02 then something like:

Code: Select all

*thi*.mp4 length:42:02

- Or if you only specified size and length for any one of the files, that would also narrow down the search a lot. This is providing you have set the (menu bar) Tools-Options-Indexes-Properties to add Length for the file type extensions you want to index, like add ("Include only files:" pane) *.mp4 or any other video file types you want to do this with. I hope this last sentence is correct, I am only a novice.

Code: Select all

size:285MB length:42:02

If even those two attributes result in too many search hits, pick another attribute of the one example file and add it too, like Bit Rate or whatever other file attributes you had set to add to the indexes as described above through the Tools-Options menu. Once you have these additional attributes being indexed, then you can also add column headers for them to sort by that, by right clicking on any existing column header like name, path, size, etc.

One example how to use bit rate is add the "Total Bit Rate" through the Tools-Options menu as already described for other file attributes to index, then add this to the code line: bitrate:(number) eg bitrate:2000 or whatever bitrate the example file has.

In other words the more attributes you index for the video file types, the more parameters you can use to limit search hits and/or sort hit results to compare files if you know the syntax to do so.

If even that is not enough, you can right click a column header "add columns" and add a new column header for a checksum column to sort, like choose CRC32 to make sure they are bit-exact duplicates, but that is a CPU performance (and network latency if not stored locally on very fast storage) hit to calculate if you don't need it, especially on potentially large files like videos.

There's probably a simpler way to do it, I always choose the hard way.