perseids-project / perseids-client-apps Goto Github PK
View Code? Open in Web Editor NEWA simple flask application to serve input forms.
A simple flask application to serve input forms.
@PonteIneptique am I missing a step in the fabfile to regenerate the messages.mo binary files? I added a new string to the treebank form, and updated the en and fr messages files, but the deployment isn't picking up the new string, just the default from the template, and I'm guessing it's because messages.mo isn't regenerated?
We really need to document the deployment steps for this...
See #12
Currently not done by fab.
the base persids-clients-app repo no longer includes ahab or joth, but the modules.py here still requires them.
not sure what the best way is to manage this -- the deployment scripts look at modules.py to decide whether or not to pull these repos in, but in a manual install if you don't want ahab or joth you have to go in and tweak this file.
Per perseids-project/perseids_docs#259 and discussion with @annakrohn yesterday, for the Digital Millet workflow we want to create a new Perseids input app which allows a user to create a new CTS-driven linked data publication in Perseids. The app should create a JSON-LD data structure using the OA data model which describes the elements of the publication and supplies the initial values. Subsequent editing of the publication will be through the existing application features in Perseids.
The following data types should be able to be created from this from:
We also want to be able to specify a keyword or set of keywords and a bibliography, but need to check with @Marie-ClaireBeaulieu about whether these should target the selected CTS passages or the work as a whole, or the linked data publication itself.
Features which might be nice to have but which might be out of scope and need further discussion wrt to the requirements:
Per alpheios-project/arethusa#759 @gcelano requests that we leave the relation attribute blank rather than setting it to nil when creating a treebank template.
when we process input from a supplied URL we should always trigger the language and type detection to be sure it's sent to the tokenizer with the right params.
we should have the option to normalize text input through the input forms -- most of our services (such as tokenizing splitting) require precombined unicode chars to function properly and we should normalize on input if we can.
found by @vgorman1 -- her μήδε used a combining char (combining \u03b7\u0301) but the code looks for the precombined character (\u03ae).
From alpheios-project/arethusa#748:
When I try to import into Arethusa a treebank file from the Perseus Latin repository (e. g. phi1221.phi007.perseus-lat1.tb.xml from https://github.com/PerseusDL/treebank_data/blob/master/v2.1/Latin/texts/phi1221.phi007.perseus-lat1.tb.xml) using the "Upload Base XML Treebank / from file" button, I get the message: ERROR!! CHANGES NOT SAVED! errorunexpected attribute "oldId". When I change the file, removing the header element (with all its children) and body (I put in the annotator element from one of my exported treebank annotations), the file is read and displayed OK.
The users often want to review or change already annotated trees from the "gold standard". Arethusa is the logical choice of environment to do so. Perhaps this could be achieved with an XSL or XQuery stylesheet layer which would, on import, strip out from the base XML treebank file everything above the sentence element, and add a treebank element that is acceptable to Arethusa.
See
perseids-project/llt-segmenter#1
add support to advanced options for specifying semicolon_delimiter paramter to llt/segtok
not sure how this is working live, but when I run the app locally, it can't find this file.
per @PonteIneptique the solution is to change the following in the importing code (views.py):
from mongo_db_ops import add_to_db
to
from app.mongo_db_ops import add_to_db
the fix in 0b710fd doesn't fully work -- tested with erasing and replacing retrieved CTS text -- resulted in an annotation that was missing a target
I'm not sure if this was taken out intentionally, or something that got lost when the code was merged. The annotation/enter.html looks like it has classes from the markdownify library to enable a widget for entering links into the commentary.
I tried adding it in, but the display gets really wonky then. Needs some effort to fix.
This is currenly only in the digmill branch of this code.
if the tokenizer crashes (such as it does when trying to tokenize 'foveamve' with splitting=true), the user doesn't get any notice of the crash, it just looks like the page hangs. We should catch and display tokenizer crashes.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.