Dear en.dogeno.us readers,
I am releasing http://pubmeder.appspot.com/ to public today.
It is built with python and javascript/jQuery, using Google's App Engine platform.
It is my very first web app using python/appengine. Hope there will not be that many bugs to prevent you enjoy pubmeder. I have test it on latest versions of Firefox, Chrome, Safari, Opera and IE7/8. It also works on iPad, Android phones and iPhone. Unlike my previous released PubMed-er for Google Chrome, the goal for http://pubmeder.appspot.com/ is to work on any platform where there is a browser.
However, as a web app initially developed for the Chrome extension I mentioned above, http://pubmeder.appspot.com/ is built from the beginning to work seamlessly with the client browser. With the release of v1.3 extension, I have achieved it!
Please give a try and let me know what do you think?
Sincerely,
Liang
As a beta testing, I will keep introducing new features and improving speed and user experiences. The followings are high on my to-do list:
safari extensionfirefox add-on- high CPU on /list or /api operation (better usage of memcache and __KEY__ query)
tagging for multiple list exportdrag to order the export sequencefull-abstract search(I don't see GAE full text search can be implanted easily, so I use eSearch to achieve some search ability)delete an item by user- better UI
in chatbot, /term gets multiple article titles, /list shows user specific top-5
http://pubmeder.appspot.com/ is pretty easy to use. Let these screen capture to tell you what this web app can do.
As a bonus of using Google App Engine, I have an xmpp robot running as well. Add pubmeder@appspot.com as your contact in gtalk, and send something to it. It will respond and tell you what it is capable of.
Enjoy!
At the end, I post my last 10 days' to-do list here
* loading gif animation
* FAIL - json upload from user client - less data transfer between server/client is better
* user specific interface
* various list browsing, e.g. organized by year, journal
* remove duplicate items or title/authors empty item - cron job
* PMID: 1231234 trigger error in ajax display - escape in django, simplejson dump/load - remove any "
* export formated references list / check and uncheck
* pmid based batch upload
* pass value to web app, merge multiple call with minor change into one - post request inside web app, or apikey based get request
* double click for reverse order sort
* multiple export format - right now cell, nature, science format (the later two need ordering)
* user info the localStorage
* need clean the database when load list - ajax get json output from the web app
* impact factor for sorting - journal
* url based batch enter - need apikey
* max efetch limit, or the ajax limit? or high CPU - efetch default retmax=25, I use 25 to separate new items, execute 25, and send the rest to taskqueue
* taskqueue; split very long pmid list (<2000, 8 per id, 220 items)
* two open windows, open checked, un-checked in another, will not save in the other because the other has it checked - use /api?c=1 to get checked items, no more localStorage
finally, better cron jobs are running, hope no more high-cpu and failure due to overtime
just push a version I made before my half dome trip
it is now having:
drag ordering and better chatbot
add PubMed Instant to the server end, very easy by using eSpell from eUtils