Files & Archives

What these tools are for

Files & Archives is the housekeeping corner of the site. It covers the jobs that come up when a download will not open or a folder has grown untidy. Zip & Unzip and TAR and GZ Extractor handle the common archive formats. Bulk Renamer tidies a whole batch of names at once, and File Splitter & Joiner cuts a large file into parts and rejoins them later.

The other half of the category is about looking inside things. Word Document Viewer and EPUB Reader open documents without the software that made them, and EPUB Maker goes the other way. PDF Highlights & Comments Extractor pulls annotations out as text. Media Info Viewer, File Type Identifier and Hex Viewer answer the plain question of what a file actually is.

Duplicate File Finder and Folder Tree Generator work on whole folders: one hunts for copies you no longer need, the other writes the structure out as text. Font Converter sits nearby for moving typefaces between formats.

Doing this work in a browser

File work is unusually personal. An archive can hold a contract, a tax folder or an entire project. That is the real argument for doing it in the page: the bytes are read by your own browser and stay there. A deploy check blocks any tool that makes a network call, so there is no upload, no queue and no server-imposed size limit.

The trade-off is worth knowing. A web page cannot go rummaging through your disk, so these tools work only on what you choose to open. Nothing runs in the background and nothing watches a folder. The work also happens in memory, so a very large archive asks a lot of your device; a laptop copes with more than a phone.

Picking the right one

Start from the symptom, not the format. If you do not know what a file is, File Type Identifier looks at the content rather than trusting the extension, which is where most mystery downloads come unstuck. If a file opens but looks wrong, Hex Viewer shows the raw bytes. For video or audio, Media Info Viewer answers what the file name cannot.

Two habits save trouble. When splitting a file, keep every part in one folder and do not rename them; the joiner needs the full set to rebuild the original. Before a bulk rename or a clear-out of duplicates, work on a copy or read the list twice. Those are real changes to your own disk, and the page cannot undo them.

The compression is the browser's, and that draws the edges

There is no compression library on this site. Zip & Unzip, TAR and GZ Extractor and Font Converter all hand that part to the browser, through its CompressionStream and DecompressionStream. ZIP entries go through the raw deflate form in both directions, a .tar.gz through gzip, and a WOFF font through plain deflate, one font table at a time. The tar itself needs no help at all: it is a 512-byte header in front of each file, read straight from the bytes.

Borrowed machinery draws the edges. A ZIP entry that is stored or deflated opens; one packed by some other method is named and refused rather than half-read. A password-protected entry is listed with its size and marked, but it will not come out. Zip64 archives stop with a message saying what they are, that being the layout used once an archive passes 4 GB or holds more than 65,535 files. WOFF2 fails for a related reason. It is Brotli inside, and browsers keep Brotli for their own downloads rather than lending it to a page, so the converter says so and stops.

Font Previewer will still show you a WOFF2, because there the browser's own font engine does the reading and the page only supplies the file. The gaps are on purpose as well. Nothing here opens RAR or 7z, since that would mean shipping a decoder for each format. Both are at least recognised: File Type Identifier and Hex Viewer know their signatures, so you can find out what you are holding before going looking for other software.

Why a stick sold as 16 GB will not take 16 GB

Fit Files onto Discs & Drives is the one tool here that is mostly arithmetic, and the arithmetic is the whole point. A stick sold as 16 GB holds 16,000,000,000 bytes, because the makers count in thousands and the computer counts in 1024s. Your file manager will call that 14.9 GB before anything is even formatted. The disc figures are the real ones too: a 700 MB CD is 737,280,000 bytes, which is 360,000 sectors of 2,048.

Two more things eat into the rest. A file does not take up its own size, it takes up whole blocks. Where those blocks are 128 KB, as exFAT chooses on a volume over 32 GiB, ten thousand small notes cost around 1.2 GB between them however little they say. And FAT32 will not hold a single file of 4 GiB or more. That is a 32-bit field in the format rather than a setting, and it is why a long video copies to a card and stops partway. Files caught by either rule are set aside and listed, not quietly dropped from the plan.

The plan itself is first-fit decreasing: sort largest first, then put each file on the first disc it still fits on. Packing bins has no quick perfect answer, so this is a good answer rather than a proved best one. Best-fit decreasing is offered too, along with keeping whole folders together, and filling in the order given for when the discs have to be written one after another. No file is touched. What you get back is a list of what goes where.

Handing over a folder, and getting a download back

Several tools here take a whole folder rather than one file, through the browser's folder chooser. What arrives is every file with its path below the folder you picked, which is how Folder Compare can match backup-may against backup-june without minding that the top names differ. Folder Tree Generator and Folder Size Analyser stop right there, at names, sizes and dates, and never open a file. That is what makes pointing them at a real project or a full drive reasonable.

None of this writes back into the folder you chose. The browser hands a page copies to read, not permission to change anything, so the results arrive as downloads. Bulk Renamer sets the old and new names side by side and gives you a ZIP with the new ones. Search & Replace Across Files saves changed-files.zip, holding only the files it actually altered. Duplicate File Finder deletes nothing whatever: it saves a ZIP of the keepers and leaves everything where it sits. Putting the results back is your own step.

Those last two look alike and are not. Both finish with SHA-256 from the browser's own crypto, and both dodge it wherever they can, because files of different byte lengths cannot possibly be identical and most pairs differ. The question each answers is different. The finder hunts copies inside one pile and ignores names completely. The compare pairs files by path across two folders, and when it judges by date it allows two seconds of slack, since FAT and exFAT store times in two-second steps and a copied file can honestly come back a moment off.