Anyone using a homelab tool as a "reader" for technical manuals?

[mapleLC]

I have a body of manuals strewn across network folders for computers stuff, EE spec sheets and audio gear.  My wife has cooking magazines. Its all PDF. Its always a hassle to look stuff up.

Is anyone using a tool to reference these files across your network (via browser)?  I see tools specific to comic books like Komga, but nothing a bit more generic.

I think the most important feature is some form of auto indexing.  Its not useful if you have to update it with every document that gets added.  I'm willing to spend some time reformatting my stuff like file or folder organizing, but nothing absurd.

[RoGeorge]

From the past experiences, relying on a tool is the worst idea.  Tools gets discontinued, operating systems change and make old programs useless, and so on.  Best is to rely on the filename and the directories structure, by case.

For example, I keep all the datasheets in a single folder, flat, no structure, no database, with the part number included in the filename, often keep the manufacturer or a very brief 2-3 words function description also in the filename.  Any search is done by whatever tools are in the operating system (I use mostly Linux, with a KDE Plasma desktop).

For the parts in stock, I keep a spreadsheet with links to the local datasheet for each part, one CTRL+click away to open.

- the default file browser (Dolphin) has a filter included, so by typing only a few characters will instantly show only those files containing the typed text
- the file browser has file-search, too, in case of a wildcard search is needed
- in KDE (Linux), there is a default indexing tool called 'baloo', which indexes all files by name and also by content if you want, so by hitting the 'super' key then start typing a few chars will instantly show all files that includes those chars/words, with best matches and most frequently opened at the top, so it's usually hit 'super' continue typing 3-4 letters and press enter to open, or arrow down then open if it's not the first in the search result, this works for programs, too, which is great
- in Linux there is yet another included tool, 'locate' which can be used from the command line for fast searching (has a local database, good for searching in remote and offline storage places, too) though this does not indexes the content like 'baloo', 'locate' only the filenames and paths
- keep shortcuts in file browser to datasheets folder, to books, etc. that makes the most used only one click away

To search, it's usually hit the 'super' key and start typing a name or part of it.  This almost eliminates completly any file browser navigation.  If I have to use the file browser, hit 'super'+'E' to open the file explorer, then single click to the wanted shortcut.

For books, projects, magazines and such, I have folders by categories or topics, it depends.

With all these, often an online search might be faster than searching inside the local desktop.

[kripton2035]

on macos, the content of pdf files is automatically indexed, as are also the network volumes each time they are mounted.
you can also add tags to every document, and categorize your stuff with it.
so when I do a search in the finder operating system tool called "spotlight", every term inside a pdf, or any tag are found this way.
I have gigs of pdf I can easyly search on the title or inside this way.
I don't know how other operating systems handle this situation.
also as everything is indexed, the search is very fast.
this has been for many years and many system versions like that.

[JohanH]

Calibre is good for books and magazines (pdf, epub etc). All other kinds of files I've organized in folders.

[fenugrec]

belt and suspenders : a sane folder hierarchy (I like being able to view related parts sometimes, where a single-folder mess would be completely useless), and https://www.lesbonscomptes.com/recoll/pages/index-recoll.html for indexing.
For the "accross the network" part, I use 2 approaches - a local fileserver , and https://syncthing.net/ when working remotely.

[ferdieCX]

To have a database of books,etc., I use Zotero.
One of the optional fields is for a link pointing to the pdf

[barshatriplee]

You can try docfetcher.

[Shonky]

You could use something like paperless-ngx to index your PDFs. It creates a searchable index for PDFs with the text or good OCR and can do OCR itself, although it's more for personal documents, receipts etc. It also ingests the file and stores it itself so can't be just pointed at an existing store.

Not quite the right tool but I thought worthy of a mention as the search it pretty good.

[kripton2035]

nowadays, you can also install privategpt and give it your local pdf folder to digest
then you could ask your computer for any data inside the pdf folder.

[PlainName]

Calibre is good for books and magazines (pdf, epub etc).

But you can have only one such collection on a machine. That is, you can't have a 'work' books collection and then switch to only 'hobby' collection (or have both in separate instances). Best you can do is have categories or something, but if it's on your PC and in Calibre then it'll be mixed in with every other thing on your PC and in Caliber.

That may not be important to some people but I like to keep some things separate (for backup, security and other purposes). And, actually, why I gave up with the otherwise cool TheBrain - in V9 they switched to a single master database for all projects instead of the previous discrete database per project. Just didn't want Client A's data mixed with Client B's, etc. Calibre suffers the same issue.

[JohnG]

FWIW, I use Zotero on Windows. I used to use Jabref, but when I changed jobs, I was forced into using Word for technical writing, and at the time Jabref's Word support was poor. It may be better now, but I haven't checked.

Both of them can generate bibliographies in multiple formats, and the things you need for a good bibliography turn out to be very useful for general organization as well. In my Zotero database, I can have it automatically name documents when they go into the database by author, title, and year (you can change this how you wish). When they are stored, the originals are untouched. I don't use the cloud features. There's a web clipper that is handy as well.

I store technical papers, books, manuals, datasheets, presentations, etc. I don't think there is a limit on filetype.

Edit: I just looked back at Jabref, and it looks pretty good. I think you can set it up to support arbitrary file types as well.

That being said, I may look at some of the other excellent suggestions in this thread.

John

[Bud]

Nothing can beat well structured folder hierarchy. No tools needed.

[PlainName]

Some things ideally want to be in more than one place, and a folder struct doesn't lend itself to that (well, it does, but you have dupes which may be notionally the same but really slightly different). Maybe links if you use an appropriate filesystem, but I don't think a catalog tool should be OS- or FS-dependent.

OTOH, a folder stuct should be used . I deal with my photos that way - they are pulled of my devices and stored appropriately and then set read-only. Then the catalog thing gets to add them to its database, but it can't move or change them. That way I can use different catalog tools, often at the same time, but the raw data stays the same and is reasonably accessible via ANOther file manager.

[Shock]

Disk folder structure as it allows you to also store art, schematics, release notes, firmware. It's easy to backup and record file integrity, search and add notes on. Auto generate an HTML by script from there when you want a pretty intranet style web view. This system can also be used to create a pseudo inventory system. 100% future proof.

An example of a folder for a piece of TE:

Fluke 87V Manual.pdf
Fluke 87V Manual.txt (dumped table of contents for web view)
Fluke 87V Calibration.pdf
Fluke 87V Service Manual.pdf
Fluke 87V Schematics.pdf
Fluke 87V Brochure.pdf
Fluke 87V Specs.jpg (quick specs and function cheat card)
Fluke 87V Front.jpg
Fluke 87V Back.jpg
Fluke 87V Serial 56120801 Calibration Certificate.jpg
Fluke 87V Serial 56120063 Calibration Certificate.jpg
Fluke Logo.jpg (pretty art for your web view).
Fluke 87V Notes.txt (repair or acquisition notes etc)

That is just an example, if you find that useful please click a thanks.

[rstofer]

I would like keyword indexing with a variable number of keywords per document.  I could find the 741 datasheet with a search on 'op amps', 'analog computing', '741' or some other categories.  The keywords are not predefined and the number of keywords per document can grow without bound (more or less).

I am thinking database...