Apertium has moved from SourceForge to GitHub.
If you have any questions, please come and talk to us on #apertium on irc.freenode.net or contact the GitHub migration team.

Task ideas for Google Code-in

From Apertium
(Difference between revisions)
Jump to: navigation, search
(Task ideas)
m (Task ideas)
Line 236: Line 236:
 
|tags=xml, dictionaries, svn
 
|tags=xml, dictionaries, svn
 
|title=Translate 100 sentences of public domain text.
 
|title=Translate 100 sentences of public domain text.
|description=Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data. The languages and text are subject to mentor approval.
+
|description=Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data. This task requires the creation of that data. The languages and text are subject to mentor approval.
 
|multi=yes
 
|multi=yes
 
|beginner=yes
 
|beginner=yes

Revision as of 19:38, 17 September 2018

Contents

This is the task ideas page for Google Code-in, here you can find ideas on interesting tasks that will improve your knowledge of Apertium and help you get into the world of open-source development.

The people column lists people who you should get in contact with to request further information. All tasks are 2 hours maximum estimated amount of time that would be spent on the task by an experienced developer, however:

  1. this is the time expected to take by an experienced developer, you may find that you spend more time on the task because of the learning curve.

Categories:

  • code: Tasks related to writing or refactoring code
  • documentation: Tasks related to creating/editing documents and helping others learn more
  • research: Tasks related to community management, outreach/marketting, or studying problems and recommending solutions
  • quality: Tasks related to testing and ensuring code is of high quality.
  • design: Tasks related to user experience research or user interface design and interaction

Clarification of "multiple task" types

  • multi = number of students who can do a given task
  • dup = number of times a student can do the same task

You can find descriptions of some of the mentors here.

Task ideas

typetitledescriptiontagsmentorsbgnr?multi?duplicates
research Join us on IRC Use an IRC client to log onto our IRC channel and stick around for four hours. irc * yes 150
research, quality Adopt a Wiki page Request for an Apertium Wiki account and adopt a Wiki page by updating and fixing any issues with it for a week. wiki * yes 150
code, design Make source browser headings sticky at bottom of window Make headings that are out of view (either below when at the top, or above when scrolled down) sticky on Apertium source browser, so that it's clear what other headings exist. There is a github issue for this. css, javascript, html, web sushain, JNW no
code, quality Increase test coverage of begiak, our IRC bot, by at least 10% There are many modules without any tests at all, unfortunately. See the associated GitHub issue for more details and discussion. python, bot sushain, JNW, wei2912, Josh no 4 4
code Improve the .logs command of begiak, our IRC bot Currently, the .logs command just links to the root logs. Ideally, it would link to the channel specific logs, support a time being handed to it and have tests. See the associated GitHub issue for more details and discussion. python, bot sushain, JNW, wei2912, Josh no
code Update the awikstats module of begiak, our IRC bot, for GitHub There are a couple steps remaining for this process, mostly small modifications to the existing code which are enumerated in the associated GitHub issue which also contains more context and discussion. python, bot sushain, JNW, wei2912 no
code, research Research and propose a flood control system for begiak, our IRC bot Begiak often floods the channel with notifications from modules such as git. Compile a list of modules which flood Begiak, write a mini-report on the associated GitHub issue and propose changes to be made to the modules. For each module, there should be an issue created with a list of proposed changes, referenced from the main issue. The issue should be added to the associated GitHub project. python, bot sushain, JNW, wei2912 no
code, research, quality Clean up obsolete modules for begiak, our IRC bot Refer to the associated GitHub issue for more details. After this task, the list of modules on Begiak's wiki page should be updated. python, bot, wiki sushain, JNW, wei2912 yes
code Add GitHub issue creation functionality to begiak, our IRC bot Ideally, this would be added to an existing module. If that doesn't make sense, a new module is acceptable as well. The associated GitHub issue includes an example of the command's usage and reply. python, github, bot sushain, JNW, wei2912, Josh no
code Support GitHub modules in apertium-get Unfortunately, the transition to GitHub from SVN made it so this script which is very handy for downloading an Apertium language/pair doesn't fetch the newest packages anymore. This also means that beta.apertium.org is out of date. See the associated GitHub issue issue for more details and discussion. bash, github sushain, Unhammer, wei2912, Josh no
code, quality Add default CI configs to Apertium packages via Apertium-init Currently, some Apertium pairs/language modules use CI but it's very inconsistent and doesn't come by default. Apertium-init is the official way to bootstrap a new Apertium package so if it came with CI support by default, that would be great. See the associated GitHub issue issue for more details and discussion. ci, circleci, yaml sushain, Unhammer, wei2912 no
code Ensure XML produced by Apertium-init has consistent XML declarations Currently, some XML produced by Apertium-init, a script which allows bootstrapping Apertium packages easily, has declarations and some doesn't. Moreover, the declarations are sometimes inconsistent. All XML files should have the same declaration. Note that not all of the XML files in Apertium-init use the .xml file extension. See the associated GitHub issue issue for more details and discussion. xml, python sushain, Unhammer, wei2912 yes
code, quality Make Apertium-init's default Makefiles and config files pass make distcheck Currently, the distcheck target for packages created with Apertium-init, a script which allows bootstrapping Apertium packages easily, does not pass the distcheck target. This task requires fixing this issue. See the associated GitHub issue issue for more details and discussion. python, autotools, make, bash Unhammer, Flammie, Unhammer, Josh no
code, quality Increase Apertium-init test coverage Currently, we have a decent set of tests for the script but there are some more complex behaviors such as GitHub interaction that we don't test. This task requires making substantial improvements to the test coverage numbers. See the associated GitHub issue issue for more details and discussion. python, unittest sushain, Unhammer, wei2912 no
code Ignore .prob files in bilingual modules created by Apertium-init Apertium-init bootstraps Apertium packages and comes with a default gitignore. This gitignore could be improved by making it ignore *.prob files but only for pairs since they are meaningful for language modules. It would be extra cool if we had some tests for this functionality that weren't too contrived. See the associated GitHub issue issue for more details and discussion. python, git sushain, Unhammer, wei2912, Josh no
code Set repository topic on repos created by Apertium-init Apertium-init bootstraps Apertium packages and supports creating an associated GitHub repository. Our source browser and other scripts expect a GitHub repository topic like "apertium-incubator". This task requires creating the incubator topic by default on repo push with an option for custom topics. See the associated GitHub issue issue for more details and discussion. python, github, http sushain, Unhammer, wei2912 no
code Install Apertium and verify that it works See Installation for instructions and if you encounter any issues along the way, document them and/or improve the Wiki! bash ftyers, JNW, Unhammer, anakuz, Josh, fotonzade yes 150
research Write a contrastive grammar Document 6 differences between two (preferably related) languages and where they would need to be addressed in the Apertium pipeline (morph analysis, transfer, etc). Use a grammar book/resource for inspiration. Each difference should have no fewer than 3 examples. Put your work on the Apertium wiki under Language1_and_Language2/Contrastive_grammar. See Farsi_and_English/Pending_tests for an example of a contrastive grammar that a previous GCI student made. wiki, languages Mikel, JNW, Josh, xavivars, fotonzade yes 40
quality Add 200 new entries to a bidix to language pair %AAA%-%BBB% Our translation systems require large lexicons so as to provide production-quality coverage of any input data. This task requires the student to add 200 new words to a bidirectional dictionary. xml, dictionaries, svn Mikel, anakuz, xavivars, fotonzade yes yes
quality Translate 100 sentences of public domain text. Many of our systems benefit from statistical methods used with (ideally public domain) bilingual data. This task requires the creation of that data. The languages and text are subject to mentor approval. xml, dictionaries, svn fotonzade yes yes
quality Disambiguate 500 tokens of text in %AAA% Run some text through a morphological analyser and disambiguate the output. Contact the mentor beforehand to approve the choice of language and text. disambiguation, svn Mikel, anakuz, xavivars, fotonzade yes
code Use SWIG or equivalent to add C++ bindings for text analysis in apertium-python Currently, apertium-python just pipes text through the binaries in each mode file. We would like to directly execute the associated C++ function instead. See the associated GitHub issue for more details and discussion. python, c++, swig sushain no
code Integrate HFST's C++ Python bindings into apertium-python Currently, apertium-python just pipes text through the binaries in each mode file. Where appropriate, i.e. a mode accesses HFST binaries, we would like to directly execute the associated C++ function instead. python, c++, swig sushain no
code Improve the apertium-python Windows installation process Currently, apertium-python requires a complex installation process for Windows (and Linux). The goal is something that works out-of-the-box with pip. See the associated GitHub issue for more details and discussion. python, windows sushain, wei2912 no
documentation, code Setup documentation generation for apertium-python Currently, there are some docstrings attached to functions and constants. This task requires setting up Sphinx/readthedocs for apertium-python so these docs are easily accessible. Types should also be visible and documentation should support being written in Markdown, not RST. See the associated GitHub issue for more details and discussion. python, sphinx sushain, Josh no
code, design Upgrade apertium.org (html-tools) to Bootstrap 4 Currently, we are on a frankensteined version of Bootstrap 3. See the associated GitHub issue for more details and discussion. Note that the frankenstein'd CSS will likely need to be fixed and theme support should be retained (should be simple). javascript, css, web, bootstrap sushain, xavivars no
code, quality Get apertium.org (html-tools) QUnit testing coverage working Currently, we have a QUnit testing framework mostly complete. There are some fixes that need to be made in discussion with the mentor and existing comments and JS coverage checking needs to be added so that we can burn down existing debt. See the associated GitHub PR for more details and discussion. javascript, jquery, web sushain no
code, quality Fix/prevent apertium.org (html-tools)'s recursive website translation Currently, if you try translating Apertium's website with Apertium's website, bad things happen. This 'exploit' is also possible through mutual recursion with another site that offers similar behavior. See the associated GitHub issue for more details and discussion. javascript, jquery, web, bootstrap sushain, Unhammer no
code Convert apertium.org's API (APy)'s language name storage from SQL to TSV Currently, language names that power part of the Apertium HTTP API are stored and updated in SQL. It would be nice if they were stored in a more human readable format like TSV and the SQLite were generated at build time. See the associated GitHub issue for more details and discussion. python, sql, tsv sushain, Unhammer no
code Support unicode without escape sequences in apertium.org's API (APy) Currently, HTTP responses with unicode characters are emitted as \uNNNN by the Apertium API. Ideally, the character could just be decoded. See the associated GitHub issue for more details and discussion. python, api, unicode, json, api, http sushain, Unhammer no
code Make apertium.org (html-tools) fail more gracefully when the API is down Currently, html-tools relies on an API endpoint to translate documents, files, etc. However, when this API is down the interface also breaks! This task requires fixing this breakage. See the associated GitHub issue for more details and discussion. javascript, html, css, web sushain, Unhammer no
code, design Refine the apertium.org (html-tools) dictionary interface Significant progress has been made towards providing a dictionary-style interface within html-tools. This task requires refining the existing PR by de-conflicting it with master and resolving the interface concerns discussed here. See the associated GitHub issue for more details and discussion. javascript, html, css, web sushain no
code, documentation Add a Swagger/OpenAPI specification for apertium.org's API (APy) There's been some work towards this already but it's outdated. This task requires updating it and for bonus points ensuring at build time that all paths are minimally present in the Swagger spec. Furthermore, it would be awesome if a simple HTTP page could be made that loads the spec (e.g. this page for another service). See the associated GitHub issue for more details and discussion. python, api, http, openapi, swagger sushain, xavivars no
code Accept ISO-639-1 codes in apertium-stats-service This task requires making /en-es, /en-spa, etc. work the same as /eng-spa and then adding tests that verify the behavior. See the associated GitHub issue for more details and discussion. rust, api, http sushain no
code Accept ISO-639-1 codes in apertium-stats-service This task requires making /en-es, /en-spa, etc. work the same as /eng-spa. See the associated GitHub issue for more details and discussion. rust, api, http, rest sushain no
code Support listing of packages in apertium-stats-service This information is something that is useful in a lot of different places, for example, our source browser. By having the stats service implement it, everyone doesn't have to write the same code in different languages and the information gets cached. For GCI task credit, the last commit info is not required (another task can be made for that feature). This task requires implementing the initial feature, adding some basic tests and tweaking the swagger spec. One or more of those tasks can be broken into other task(s) if the mentor sees fit and the student requests it. See the associated GitHub issue for more details and discussion. rust, api, http, rest sushain no
code Add configurable timeout support to apertium-stats-service Currently, a stats request has no clear timeout and can take ~forever if the async option is not present. This tasks requires adding a timeout option, adding tests and then tweaking the swagger spec. See the associated GitHub issue for more details and discussion. rust, api, http, rest sushain no
code Surface errors to the client in apertium-stats-service Right now, errors are logged and swallowed. The client never knows what happened. This task requires implementing the feature, adding some basic tests and tweaking the swagger spec. One or more of those tasks can be broken into other task(s) if the mentor sees fit and the student requests it. See the associated GitHub issue for more details and discussion. rust, api, http, rest sushain no
code Include Git SHA in apertium-stats-service's file info Right now, only the SVN revision number is provided but that doesn't help with mapping back on to a SHA in Git/GH for the client. This task requires implementing the feature, adding some basic tests and tweaking the swagger spec. See the associated GitHub issue for more details and discussion. rust, api, http, rest, git, svn sushain no
documentation Create a screencast on how to add new entries to an apertium dictionary Screencasts are popular cool way to create a tutorial. Show a narrated work-flow start-to-end on adding new words to a dictionary, compiling and then using it to translate. This task is probably easiest after completing a "Add 200 new entries to a bidix to language pair" task. xml, dictionaries, screencast Flammie, JNW, anakuz, Josh yes no
documentation Create a screencast on how to disambiguate tokens of text Screencasts are popular cool way to create a tutorial. Show a narrated work-flow start-to-end on disambiguating the words. This task is probably easiest after completing a "Disambiguate 500 tokens" task. disambiguation, screencast Flammie, Josh no
quality Create automated (travis-ci) test to ensure naïve coverage The dictionaries can be tested on Coverage, the idea is to make test that operates on frequency word list to count a coverage of a dictionary and then integrate that to Makefile target check for travis to use. test, python or sh Flammie, wei2912 no no
code Scrape Apertium repo information into json Write a script to scrape information about Apertium's translation pairs as they exist in GitHub repositories into a json file like this one. python, git, json JNW, sushain, wei2912 no no
code, design Integrate globe viewer into language family visualiser interface The family visualiser interface has four info boxes when a language is clicked on, and one of those boxes is empty. The globe viewer provides a globe visualisation of languages that we can translate a given language to and from. This task is to integrate the globe viewer for a specific language into the fourth box in the family visualiser. There is an associated GitHub issue. d3, javascript JNW, sushain no no
documentation Fix (or document the blockers for) five mentions of SVN on the Apertium wiki Apertium recently migrated to GitHub from SVN. There are unfortunately still a lot of pages on the Wiki in need of their references to the Wiki to SVN URLs and SVN in general. This task requires finding five such pages and either outright fixing them or documenting the difficulty involved in fixing the issues. Note Category:GitHub_migration_updates lists articles currently marked for needing migration but is not exhaustive. wiki, github, svn JNW, sushain, wei2912, Josh, xavivars 5
research Document resources for a language Document resources for a language without resources already documented on the Apertium wiki. read more... wiki, languages, grammar JNW,ftyers, Josh, fotonzade yes 10
research tesseract (OCR) interface for apertium languages Find out what it would take to integrate apertium or voikkospell into tesseract OCR (image to text). Document thoroughly available options on the wiki. ocr JNW, Josh no
research Create a UD-Apertium morphology mapping Choose a language that has a Universal Dependencies treebank and tabulate a potential set of Apertium morph labels based on the (universal) UD morph labels. See Apertium's list of symbols and UD's POS and feature tags for the labels. morphology, ud, dependencies JNW,ftyers, fotonzade 5
research Create an Apertium-UD morphology mapping Choose a language that has an Apertium morphological analyser and adapt it to convert the morphology to UD morphology morphology, ud, dependencies JNW, ftyers, fotonzade 5
code,design Paradigm generator browser interface Write a standalone webpage that makes queries (though javascript) to an apertium-apy server to fill in a morphological forms based on morphological tags that are hidden throughout the body of the page. For example, say you have the verb "say", and some tags like inf, past, pres.p3.sg—these forms would get filled in as "say", "said", "says". javascript, html, apy JNW,ftyers
Research Syntactic trees Pick up a text of ~200 words and make it's syntactic annotation as for Universal Dependencies treebank. UD Annotatrix can be used for visualisation. Consult with mentor about the language. UD, trees, markup anakuz, fotonzade

Counts

Last updated by Sushain (talk) 19:22, 15 September 2018 (CEST).

Category Count
code 34
documentation 5
research 2
quality 10
design 4
Total 43
Personal tools