Great idea. Only problem is that I can't seem to find a way to output their search results in XML format, which is much easier to do than plucking fields out of the HTML. I sent off an email to the diseasesdatabase.com folks to see if they'd consider implementing such a feature. I'll keep you posted. --David Iberri (talk) 20:11, 1 May 2007 (UTC)[reply]
Got a reply from Malcolm at diseasesdatabase.com. He's very interested in helping us out, but doesn't seem to have the resources to implement exactly what we need at the moment. We might be able to work something out; I'll keep you posted. --David Iberri (talk) 15:03, 6 May 2007 (UTC)[reply]
Malcolm suggests that we send him a list of disease terms that he can batch-run against the diseases database; he'll send the codes back to us in some usable format. I'm not sure this is the best solution because 1) I'm not sure how we'd generate such a list (ideally, we'd have the entire diseases database at our fingertips); 2) I unfortunately can't commit any additional resources to storing the results of the batch run; and 3) screen-scraping isn't pretty but it'd get the job done just fine in this case. Any suggestions? --David Iberri (talk) 16:54, 16 May 2007 (UTC)[reply]
I agree that the batch-run proposal is not a sustainable solution. If the screen-scraping approach isn't too hard, then that's probably the best way forward (I admit I thought you were using screen-scraping for all of them; I didn't realize you were using XML) -- but if it is a hassle to code, don't worry about it; we've survived without an automated infobox generating tool so far, and we could continue to do so. By the way, have you thought about putting your code on sourceforge? --Arcadian18:38, 16 May 2007 (UTC)[reply]
Looks like we might be getting an XML version after all. Malcolm has done some great work and has sent me over some mock-ups that would work just great. The next step is to decide whether he's going to make this available for online queries on-the-fly. Once that happens, the ball will be in my court to implement the template filler, and that shouldn't take more than a day or so. Cheers, David Iberri (talk) 17:48, 18 May 2007 (UTC)[reply]
Would you be able to add another function so that webpages could be cited? Ideally, you'd insert the webaddress and your program would query the page to get whatever information (e.g. the title) that it could, throw in the access date etc and give me something to copy and paste? --SeansPotato Business09:55, 6 May 2007 (UTC)[reply]
(moved from User talk:Diberri/Articles)
Thanks, Dave, for writing a page about PKMζ. I was thinking about doing it, but didn't know how to begin. By the way, since most people would type "PKMzeta," how do you set up a redirect to your page? Thanks again, Todd Sacktor
PubMed citation tool
Resolved
This tool is great. A couple of suggestions:
Could the tool pick up the doi field so users will be able to link to full text?
How about formatting as below:
Currently
{{cite journal |author=Lindström J, Ilanne-Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemiö K, Hämäläinen H, Härkönen P, Keinänen-Kiukaanniemi S, Laakso M, Louheranta A, Mannelin M, Paturi M, Sundvall J, Valle TT, Uusitupa M, Tuomilehto J |title=Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study |journal=Lancet |volume=368 |issue=9548 |pages=1673-9 |year=2006 |pmid=17098085}}
Proposed:
<ref name="pmid17098085">{{cite journal |author=Lindström J, Ilanne-Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemiö K, Hämäläinen H, Härkönen P, Keinänen-Kiukaanniemi S, Laakso M, Louheranta A, Mannelin M, Paturi M, Sundvall J, Valle TT, Uusitupa M, Tuomilehto J |title=Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study |journal=Lancet |volume=368 |issue=9548 |pages=1673-9 |year=2006 |pmid=17098085|doi=10.1016/S0140-6736(06)69701-8}}</ref>
Rationale:
Saves the user time by building the ref tag
Avoids references colliding or being duplicated on an article by proposing a unique name attribute for the ref tag (could substitute isbn, etc for pmid when citing other types of sources)
adds the doi to link to the full text of the article
A suggestion on multiple authors for a specific study.
Resolved
Currently your tool adds all of the authors for any given study. However WP:MEDMOS says "AMA citation guidelines suggest that if there are more than six authors, include only the first three, followed by et al." Meaning that if a study has 7 authors for instance, you should only list the first three and then follow it by "et al". I know that in some cases this might not work however in most medical articles I think it would be best to follow the MOS. Is there any way you can add another option for instance that says "Trim authors" or something that makes the output list only the first three authors and then "et al"?Wikidudeman(talk)03:44, 20 May 2007 (UTC)[reply]
Install the bookmarklets by bookmarking the links below
<a onclick="alert('First, bookmark this link in your web browser to use it.'); return false;" href="javascript:s=;function%20gS(w){for(var%20i=0;i<w.length;i++){gS(w.frames[i])};try{s+=w.document.getSelection()}catch(e){try{s+=w.document.selection.createRange().text}catch(e){s+=}}}gS(top);{void(window.open('http://diberri.dyndns.org/wikipedia/templates/?type=pmid&add_ref_tag=1&id='+s,'diberri',))}">Send PMID to Diberri tool</a>
<a onclick="alert('First, bookmark this link in your web browser to use it.'); return false;" href="javascript:s=;function%20gS(w){for(var%20i=0;i<w.length;i++){gS(w.frames[i])};try{s+=w.document.getSelection()}catch(e){try{s+=w.document.selection.createRange().text}catch(e){s+=}}}gS(top);{void(window.open('http://diberri.dyndns.org/wikipedia/templates/?type=isbn&add_ref_tag=1&id='+s,'diberri',))}">Send ISBN to Diberri tool</a>
After installation,to use the bookmarklet,
highlight the PMID or ISBN-10 of your reference at PubMed, Amazon, etc, or other webpage
click the appropriate link that you saved in your browser's bookmarks
I've added the bookmarklets to the page (see [2]), but haven't made the additions live yet because I can't seem to get them working in Firefox (syntax error involving the "<i>"). I unfortunately don't have any time to do debugging myself. Might you have any fixes for this? Cheers, David Iberri (talk) 21:14, 4 June 2007 (UTC)[reply]
I'm using the bookmarklet (adapted from something at squarefree.com)
I have lots of referencing to do so your tool looks ideal. I am only new to this business of editing, so hope you dont mind some basic guidance/
Whilst doing an edit, how do I access the tool, can I set up a permanent shortcut and how/
I set up a shortcut in Firefox. For example, I have http://diberri.dyndns.org/wikipedia/templates/?type=pmid&id=%s bookmarked and set to use "pmid" as a keyword so that whenever I type "pmid XXXX" in my location bar (where XXXX is a PubMed ID), it gives me a full {{cite journal}} template code to copy and paste into the Wikipedia article I'm editing. There are some folks that use bookmarklets (eg, the one above) to do more along the lines of what you're asking.
Then how do I paste a result into my text edit,ie mark text, right click on text, 'copy' routine, 'back' to edit page,'paste'/or other way?
I keep the Wikipedia article I'm editing open in one tab, then open another tab, type in "pmid XXXX", copy the resulting template code, close the tab, then paste the code into my article.
Thanks for the quick response David, unfortunately some of it goes over my head! I am using XP so some terms such as 'tabs' and 'booklets' are new to me, I gather though the aim is have two sites and swap between?
1 If I am editing I can have the site you detail above in my "favorites" (bookmarks?)sidebar and go to it when referencing, copy and "back' to the Wp page to paste the REF. Next reference I can "back' again to your site, but need to clean the feilds before entering the new ref ID. Is that why you close each time? Perhaps a "Clear' button might assist dumbos like me?
2 I note when I open your page with the string above it has %s in the number bar (which has to be cleared) and says "Invalid-----" I presume that your method somehow replaces the %s with the ID number, so that it opens with the full reference? Can you tell me how to set this up for XP and explorer?
I like the idea of a seperate instruction page, but remember to allow for basic users like me,thanks Jagra06:52, 3 June 2007 (UTC)[reply]
Looking for guidance - using Cite templates on a local installation
Resolved
Hi - I found your terrific tool for generating citations from PubMed, ISBN, etc. First of all, thank you for creating and sharing it.
I was wondering if you could point me in the right direction. I would like to install the "cite_book" capability from Wikipedia on a MediaWiki installation I have. The first question, obviously, is whether it's OK to do that. The harder question is how to do it. Although I've installed and am successfully running Mediawiki on a LINUX server, I'm not deeply enough into the structure and coding to understand what to do. From what I can see, it's a template that refers to a bunch of other templates. If that's the case, then I should be able to get those templates and install them on my system. I haven't been able to find any references or leads to people who might know what has to be done or who might have done it for their own servers. Any ideas?
Your citation tool is the most useful thing I know of in my work on wikipedia - thanks so much for making it! Unfortunately, I've noticed that for me, approximately 25-30% of ISBN's don't work with your citation tool, despite returning the correct book when entered directly into ISBNdb.com. I always try with and without hyphens, and any variant numbers for different editions with no success. My most recent example is 0071391401, a very popular medicine textbook. This is rather frustrating, and I would greatly appreciate it if you could help me out. -RustavoTalk/Contribs04:56, 6 June 2007 (UTC)[reply]
Thanks for bringing this to my attention. I'd wondered why some ISBNs failed but had never had a chance to look into it. Seems I've found the problem and fixed it. Now Harrison's and other books for which ISBNdb lacks author information are now able to be queried via the template filler. See [3] for Harrison's. Cheers, David Iberri (talk) 00:58, 7 June 2007 (UTC)[reply]
Great! I guess the problem must be for medical texts (most of my book sources) which have "editors" rather than "authors" for the book as a whole. I don't suppose there is any way to get the "editors" to show up in the template filler, is there? -RustavoTalk/Contribs06:59, 8 June 2007 (UTC)[reply]
If I recall correctly, when I looked at this problem the other day, the ISBNdb data source I'm using didn't actually provide any editors for the Harrison's text (although if you query the isbndb.com website manually, Braunwald appears as editor under the "Notes" section). I'll take another look at this later today. --David Iberri (talk) 11:14, 8 June 2007 (UTC)[reply]
A search through the ISBNdb developers' docs turned up a separate source for author information. It's much more complete than the other method I was using, so authors should appear more reliably now. For example, here's Harrison's again. Cheers, David Iberri (talk) 17:56, 10 June 2007 (UTC)[reply]
Hi, I had a look at your really helpful reference tool.
I'm interested in developing an extension of your tool. As far as I understand it, you just call a site like ISBNdb.com and parse the result and then reformat it to get the Wikipedia citation templates filled, right?
I'm thinking of some real database, i.e. not just something retrieving the details via the ISBN etc., but having in store all of the information. A definitive advantage should be that people are then able to look for books not just by their ISBN but by title and author. Moreover, helpful additional information like urls etc. could then be added by people.
Would you be interested in participating in developing this database? Right now, I just started. The main question is, of course, how to get all the data in the database. I thought, one might somehow scan Wikipedia pages containing ISBN numbers, then get the data from ISBNdb, as you apparently do. Additionally, people should be able to add or modify existing items. In addition to a database of books, I plan one for authors and publishers (appropriately linked to the books) and possibly also for scientific papers in journals.
Thanks for your comments, Jakob. This is definitely a great idea and a big project to undertake. Rather than solving the problem with a simple wiki (which among other limitations, would be difficult to query programmatically), it may be easier to use something like Wikidata, whenever it's ready for prime-time. Unfortunately my summer is rather booked, so I wouldn't be able to contribute much to the project aside from the occasional email. Best wishes, David Iberri (talk) 16:37, 10 June 2007 (UTC)[reply]
how about a barnstar?
A barnstar for Diberri for his peerless citation tool
I am truly impressed. Also, now that I have buttered you up (but in all sincerity...), I have a request. Can you figure out how to take a reference created by a program like EndNote and reformat it into a wikipedia citation with the vertical dividers, etc? EndNote can export in almost any academic format including all the biomedical, scientific, APA, and Harvard styles, but I can't figure out how to automatically put the vertical dividing lines into it. Any ideas, especially for the citations that don't have a PMID? I am actually planning to go back and reference that hundred articles or so that I wrote off the top of my head-- or at least some of them, but would like to avoid retyping. alteripse01:06, 14 June 2007 (UTC)[reply]
Thanks for the butter. I'm familiar with EndNote but have never used it and don't have a copy handy. Do you have any examples of the citation format the EndNote uses? I'm mostly curious about this vertical divider business, since I can't seem to imagine what that refers to. If you've got something handy, I'd love to see it and figure out whether I can munge it into a WP format. Thanks, David Iberri (talk) 15:49, 18 June 2007 (UTC)[reply]
I'm awarding you this "Da vinci" barnstar for your enhancement of wikipedia through your technical knowledge, especially concerning your amazing citation tool which has helped numerous editors improve medical article citations. Booyah! Wikidudeman(talk)05:21, 19 June 2007 (UTC)[reply]
My references are specific, satisfy multiple fact tags, and pass RS and V. I responded on the article talk page, however I have re-added your primary source, and see no reason that my edits need be reverted en masse. the_undertowtalk19:06, 13 July 2007 (UTC)[reply]
^Ohama K, Nagase H, Ogino K; et al. (1997). "Alpha-fetoprotein (AFP) levels in normal children". European journal of pediatric surgery : official journal of Austrian Association of Pediatric Surgery ... [et al] = Zeitschrift für Kinderchirurgie. 7 (5): 267–9. PMID9402482. {{cite journal}}: Explicit use of et al. in: |author= (help)CS1 maint: multiple names: authors list (link)
The issue was that the journal has an equal sign (=) in it, which breaks the template code. I've replaced it with the equivalent entity (=) and things are working just fine now.[4] Cheers, David Iberri (talk) 15:06, 19 July 2007 (UTC)[reply]
I really like your citation tool. I have one question / comment. I notice that the journal title shows in sentence case (New england journal of medicine), but I think it should be capitalised (New England Journal of Medicine). Is this an easy fix? Cheers, --G71620:52, 30 July 2007 (UTC)[reply]
Hello David,
your hosted HTML2Wiki conversion site is down. I had actually planned an important migration for our corporate Wiki and was gonna use your great tool to help migrate from HTML to Wiki. I see that's it's hosted on dyndns, so probably out of your home. I am currently trying to install the Perl modules, but I am a little bit struggling with that as I am a newbie to Perl. Anyway, if you happen to read this message and you have any chance to get your site online again, I would owe you tons!!
Cheers,
Andreas Rindler11:49, 16 July 2007 (UTC) (Email address removed. -- Dave)[reply]
Hi Andreas, thanks for the message. I also got your email. Things are fixed for the time being. Not sure what's going on with the server, as this is the second time it's bombed in less than 24 hours. Should be safe to use for now though. Cheers, David Iberri (talk) 13:21, 16 July 2007 (UTC)[reply]
Hi Dave—the magic template builder appears to be down again. You must be getting tired of doing the tech support thing, but I assure you that I wouldn't be complaining if your tool wasn't so darn useful :) Fvasconcellos (t·c) 20:59, 22 July 2007 (UTC)[reply]
No worries. It's only aggravating in that I can't get figure out what's going wrong with my server. Hopefully I'll get to the bottom of it soon. Thanks for being patient. :-) Best, David Iberri (talk) 01:57, 23 July 2007 (UTC)[reply]
Hi, just started using the tool for ISBNs as well as PubMeds. Seems some ISBNs it states can't find, eg 3805568274, yet the link to the source used did find this. Regards David RubenTalk21:09, 27 July 2007 (UTC)[reply]
A small issue -- on the template filler, when selecting on an HGNC ID, regardless of the ID chosen, the response is "Could not find a protein matching the given HGNC ID". (The rest of the template filler is working fine.) Thanks again for building and continually improving the tool. --Arcadian02:11, 8 August 2007 (UTC)[reply]
Gotcha. I'll take a look at this in the next week and a half or so. Unfortunately I don't have internet access at my apartment right now. --David Iberri (talk) 15:57, 21 August 2007 (UTC)[reply]
By the way, it appears that they've moved their site from "www.gene.ucl.ac.uk" to "www.genenames.org". I've already updated Template:Gene and Template:Protein, but I wouldn't be surprised if this is the issue that is affecting the template filling tool. --Arcadian19:10, 21 August 2007 (UTC)[reply]
Okay, so I got impatient and checked out the problem using a campus computer. Turns out you were right -- I just needed to update the URL. Things should be working fine now. Thanks for the heads-up. --David Iberri (talk) 19:26, 21 August 2007 (UTC)[reply]
Ah, if only that citation tool was around two years ago. :-) [By the way, you might consider displaying your barnstars on your userpage.] Best wishes. Axl19:24, 14 August 2007 (UTC)[reply]
I noticed that your citation tool was on DynDNS, which makes me think it might be on a personal computer. If you're interested in getting permanent hosting for your excellent tool, let me know. Obviously this would be at no cost. If I'm wrong about the server, apologies - consider this a "thanks" for the tool! Antelantalk15:03, 20 August 2007 (UTC)[reply]
I've actually got it hosted on a Go Daddy server at the moment -- I'd completely forgotten I was using DynDNS as well. :-) Thanks for the offer though! Cheers, David Iberri (talk) 15:56, 21 August 2007 (UTC)[reply]
Hehe. The truth is I have this nervous habit of selecting/deselecting text all over a page. My heart stopped when I clicked that rollback button though. :-) Sorry 'bout that. A totally mindless mistake. --David Iberri (talk) 17:10, 22 August 2007 (UTC)[reply]
Resolved
