Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Rob Jenson
    @ferthalangur
    Which Ammianus were you vexed by @kylepjohnso?
    Y'all might want to add some of the following to your .gitignore files:
    Luke Hollis
    @lukehollis
    Hey that gitignore looks great!
    Rob Jenson
    @ferthalangur
    It could probably use more Windows and "vim" temp file stuff. :)
    Kyle P. Johnson
    @kylepjohnson
    Ha
    Luke Hollis
    @lukehollis
    To follow up quickly before I'm distracted into other work, here's what I'm seeing of the coptic, accessed via mongo and the query resource:
    Screen Shot 2015-11-23 at 13.15.34.png
    I think we'll also need to set the appropriate header responding from the API
    for the response content encoding
    can research more about how flask does this unless anyone's had to do this before
    But, hey, at least we have the start of a functioning api, queried by our frontend app!
    Kyle P. Johnson
    @kylepjohnson
    Hey guys, put up something to (hopefully) build off of.
    api_json.py. Launched on the server, too.
    Kyle P. Johnson
    @kylepjohnson
    I added the Iliad, plus a 'meta' tag
    Kyle P. Johnson
    @kylepjohnson
    I still don't have an answer to processing these faster, but you can see the file format example here: https://raw.githubusercontent.com/cltk/cltk_api/master/json/greek/perseus/homer__iliad.json
    Kyle P. Johnson
    @kylepjohnson
    Hey, I accidentally killed the process upon exiting the server. Available again
    Luke Hollis
    @lukehollis
    Hey these look great!
    Okay, perfect--I'll start using the frontend to access these
    Would anyone be against merging the ingest branch with master now?
    We have to do JSONP to get the data from the API server
    Luke Hollis
    @lukehollis
    But I can cache the results from these queries with Homer really easily in Mongo and then expose them via our Query resource
    Kyle P. Johnson
    @kylepjohnson
    Luke, you should have permissions to merge. Go for it!
    Luke Hollis
    @lukehollis
    Can also standup our frontend app on the shared server when we're ready
    okay, awesome, will do!
    Kyle P. Johnson
    @kylepjohnson
    Today, as soon as I can, I will re-implement the /authors and /texts routing, which currently isn't working.
    Also, what is JSONP?
    Luke Hollis
    @lukehollis
    Okay, good deal--I'll merge right now then
    Luke Hollis
    @lukehollis
    The only real difference is the the extra jsonp decorator for some of the queries: https://github.com/cltk/cltk_api/blob/master/util/jsonp.py
    Kyle P. Johnson
    @kylepjohnson
    So should I add @jsonp to each flask function?
    I'll let you know how the scraper comes ... my first priority is to get you some good data to build off. Even a successful scrape will require some hand correction
    Kyle P. Johnson
    @kylepjohnson
    Some API improvements:
    @lukehollis please take a look and let me know any changes you want. Also let me know about the @jsonp wrapper :)
    Luke Hollis
    @lukehollis
    That's awesome!
    This all looks great
    I'd been using the @jsonp to get and display the API in the frontend app, so any get methods we want to use on the frontend need @jsonp, I believe...
    kk let me get to work implementing these
    Luke Hollis
    @lukehollis
    Gunicorn looks great!
    Kyle P. Johnson
    @kylepjohnson
    Cool. Not sure if I can get anything done for the rest of the night, but ping me if you need help relaunching the flask app. I will add jsonp to the get methods if you don't first
    Luke Hollis
    @lukehollis
    okay awesome--will start moving to concentrate on the meteor development
    Also, I have been mulling over DB performance. If we want to go the relational route, and enjoy very good speeds, we could write the json data in multiple chunks, into the db. Thus, for Herodotus, we could have tables with it pre-chunked into various parts, eg: herodotus_histories_work and herodotus_histories_book and herodotus_histories_book_chapter and herodotus_histories_book_chapter_section. The API would route the request into the appropriate query, for maximum speed. Not for this first or second or even perhaps third iteration, but could solve speed issues if mongo won't scale
    Luke Hollis
    @lukehollis
    Yeah, that sounds like a really good idea