I'm going to write a database for the mass of data I seem to be accumulating now I've started my masters.
I'm using this thread mainly as a place to put my ideas and to see if they make sense, and of course any suggestions and answers to any questions are greatly appreciated
The data is in the form of spectra which are viewed in a program called Spinworks. Each set of data is in a folder named "yyyy-mm-dd-AB-xx". So far I have close to 50 of these sets of data, most having at least 2 spectrums within them in a subdirectory, some 3, giving over probably over 150 spectrums.
Each set of spectra also has a code that I have given it - ATxxx. The actual print outs (which are only used for basic reference) only have both codes, "yyyy-mm-dd-AB-xx" and "ATxxx" on them. I have a list in the back of my lab book telling me what compound/reaction each ATxxx refers to. With my lab book living in the lab most of the time and the accessing of the spectra being done in the office it makes it a real fucking arse checking which spectra is what. So today I made a spreadsheet containing these fields:
AT Code - ATxxx
Lab book ref - ATrxxx
Details - Compound conc etc
Solvent
Elements analysed - H, F, Fdc, C
Server code - yyyy-mm-dd-AB-xx
Time - xx:xx
Obviously a spreadsheet is only a tool for reference and isn't any use for actually looking at things. So, I thought, why not play about a bit, have all the data in a database with a webinterface where I can easily access my spectra, check out their details, and actually look at them.
As it happens, the NMR machine (this is all NMR data) that prints out the overview spectra saves this as a PDF in the "yyyy-mm-dd-AB-xx" folder, these can obviously be easily converted to JPEGs and resized so I can have a preview on the web interface to make sure it is the right one. The PDF itself can also be linked.
I plan to use PHP and MySQL since I've made a few basic databases before with it (which this clearly is). I'll host it on my server running XP, which means installing the right things there. I've never managed to get Apache and MySQL working on that. Can you use MySQL and PHP with IIS?
One thing that I'm wondering about and if it will work is, from the webinterface I want to be able to open the spectra in this Spinworks software. The "yyyy-mm-dd-AB-xx" directory contains a number of files that are needed so you can actually view and manipulate the spectra. The file you open is the "fid" file, which is just that, a file called "fid", no file extension. Is it possible to open a file from a browser, into an application (I will only use it on a system with Spinworks installed obviously) without saving it, or opening with? As if to open the app pointing it to the fid file, instead of opening the fid file itself.
Magnificent minds of Teh, is this all possible?
Brilliant link, that will definitely be of great use.
Spinworks will indeed open any file extension so a PHP rename would be fine in that sense. Problem being, downloading the fid file probably won't be the answer since it relies on the other files in the folder to work.
At least I assume it does.
Yes, just checked, the fid file on its own does nothing.
I guess if it is a security risk which means it's not possible then my plan is scuppered :(
c:\Program Files\SpinWorks\SpinWorks.exe Z:\Docs\Uni\MPhil\NMR\2008-10-21-AB-26\10\fid
My later posts might explain it better, but I'll explain again.
Clicking a link to a file will give the options to Save As or Open With. Since I don't want to save it, and opening with just sticks the file in a temp directory this isn't what I want. With the success of opening the file being dependent on the other files in the folder it just wouldn't work.
Like I've said. I'd like to click a link that opens a program installed on the computer telling it to open a certain file, that file will be local meaning directory access is fine.
I'll probably make two versions/an option for the webinterface. One for home and one for uni. Where the home version will link to the NMR data saved on my server which is essentially a local link since the drive in question is mapped as Z whereas at uni they're a network address:
\\vol5\vol3\NMR................\2008-10-21-AB-26\10\fid
No, but yes. At home the data is on the server which is on the same network, with the drive mapped to the computers using it.
At uni the computers are on the same network but it is a network address instead of a mapped drive, but a fully accessible directory structure.
Edit : Oh, the data (spectra folders) will be stored both at Uni and Home. Meaning full directory access is possible from both and they will be independant from each other.
Right. I have realised some possible confusion.
Database will be on my XP server in the attic.
The NMR data is on both my personal server at home and the fileserver at uni.
Computers at home have this data accessible via a mapped Z drive.
Computers at uni have this data accessible via \\vol5\vol3\NMR................\2008-10-21-AB-26\10\fid
Hope this clears that up.
<a href="Z:\Docs\Uni\MPhil\NMR\2008-10-21-AB-26\10\fid.my_sw_ext">blah</a>
<a href="file://fdgdfg/dfg/fid.my_sw_ext">home</a> <a href="\\dsmfkdsm\dsfsd\dsf\fid.my_sw_ext">uni</a>for everything, which may or not be desirable?
Right, ok.
I was hoping that giving the files an extension then just associating Spinworks with that extension might help.
Also, having two different links for home and uni is absolutely fine.
I've just opened Z:\Docs\Uni\MPhil\NMR\2008-10-21-AB-26\10\fid.fid (after renaming it) from the Firefox address bar, it gives me the option to open the file with Spinworks and it works as I want it to. So that's good news. As does doing the \\server\d$\Docs\... version.
BUT...Another stumbling block will be that renaming the uni copy and giving it a file extension could be a problem and something I'd like to avoid doing
I'll give it a go without the file extension for the uni link and see if it even offers to open it.
If it works then the extra step isn't really a problem since it'll be the lesser of two evils whilst being part of a huge benefit.
With the database I'll be giving myself a far better way to see what data I have available and it also means not clicking through a directory structure which is about 10 folders deep on a slow network.
Right.
Setup the DB, pushing and pulling data to and from it easy enough.
One problem. I can generate a link which is correct
"file:///Z:/Docs/Uni/MPhil/NMR/2008-10-09-AB-49/10/fid"
But when I click it nothing happens, whereas if I copy and paste it to the address bar and press enter it gives me the open file dialog as I want.
Why does the link refuse to work?
If I give it an extension it makes no difference.
I've just tried it in IE7 and without an extension it opens the dialog box asking to either save or go online to find an app that can do it, but not to pick an app of my own. With an extension you can choose the app, then tell it not to ask again so it just opens any .fid file in Spinworks straight away. For once IE wants to work properly instead of Firefox :(