significant past updates

(Apr '26)

• [lib v2.1.3] / [PyPI] / [app v1.11.1]
  • fixed VH (Velthuis) transliteration mappings for aspirate consonants and liquid vowels
• [lib v2.1.2] / [PyPI]
  • fixed splitter server URL
• [lib v2.1.1] / [PyPI] / [app v1.11.0]
  • added anunāsika / candrabindu (ँ) support in transliteration, with optional normalization to anusvāra
  • added "preserve anunāsika" setting (off by default) on settings page
  • improved textarea font stack for better Unicode combining-diacritic rendering

(Feb–Mar '26)

• [lib v2.1.0] / [PyPI] / [app v1.10.0]
  • added auto scheme detection ("Auto" option in input-scheme selector detects from input and updates label after processing)
  • added "un-space" setting for Roman output (= equivalent of virāma-avoidance for Indic)
  • fixed ITRANS transliteration bugs in scheme maps; improved IAST normalization
• [app v1.9.1]
  • fixed dark mode meter-table colors; guarded melody player on load
• [app v1.9.0]
  • added clear-text and redo-action buttons
• [lib v2.0.9] / [PyPi]
  • fixed r̥ ordering bug in transliteration scheme maps
• [lib v2.0.8] / [PyPi]
  • fixed Dharmamitra splitter for new upstream API endpoints and formats
• [app v1.8.0]
  • fully redesigned appearance (colors, etc.)
  • added sidebar
  • added dark mode
  • unified text workbench and upload-file modes (same page, sidebar controls)
  • improved management and presentation of settings
• [app v1.7.4]
  • improved API (added JSON mode, Swagger doc)
  • refactored UI with Javascript

(Spring '25)

• [app v1.7.3]
  • fixed two melody player bugs
• [app v1.7.2]
  • added favicon
• [app v1.7.1]
  • scaled up gunicorn workers/threads and nginx/gunicorn timeouts
• [app v1.7.0]
  • added Google Cloud Vision OCR

(Dec '24)

• [lib v2.0.7] / [PyPi] / [app v1.6.11]
  • improved batching for both Dharmamitra and 2018 splitters
• [lib v2.0.6] / [PyPi]
  • fixed Windows encoding bug in config file loading
• [lib v2.0.5] / [PyPi] / [app v1.6.10]
  • updated splitter for new endpoint path and JSON output format
• [lib v2.0.4] / [PyPi] / [app v1.6.9]
  • migrated 2018 splitter server to Digital Ocean
• [lib v2.0.3] / [PyPi] / [app v1.6.8]
  • communicated 2018 splitter server outage to users

(Oct '24)

• [lib v2.0.2] / [PyPi] / [app v1.6.7]
  • added batching to Dharmamitra splitter; extended timeouts for 2018
• [lib v2.0.1] / [PyPi] / [app v1.6.6]
  • cleaned up whole-file splitting (removed buffer file)
• [app v1.6.5]
  • added 502 error handling for oversized Dharmamitra requests
• [lib v2.0.0] / [PyPi] / [app v1.6.4]
  • refactored splitter import interface (breaking change, hence v2)
• [lib v1.5.2] / [PyPi] / [app v1.6.2]
  • fixed punctuation interleaving bug in splitter
• [lib v1.5.1] / [PyPi] / [app v1.6.1]
  • fixed SLP final-character transliteration bug from Indic script

(Sept '24)

• [lib v1.5.0] / [PyPi] / [app v1.6.0]
  • added preserve-compound-hyphen setting for splitting with Dharmamitra model
• [lib v1.4.1] / [PyPi] / [app v1.5.0]
  • added new Dharmamitra splitter-model setting
• [lib v1.3.1] / [PyPi] / [app v1.4.0]
  • added new settings page
  • improved settings reset behavior with separate reset button
  • renamed help page (from "tutorial")
  • reorganized nav bar
  • added avoid-virāma setting for Indic-script transliteration output
  • added preserve-punctuation setting for word splitting

(Aug '24)

• [app PR #10]
  • refactored front end to use Flask templates (factored out shared head/nav)
• [app PR #9]
  • refreshed docs; renamed help page (from "tutorial")

(Jul '24)

• [app PR #8]
  • completed server migration by deactivating on PythonAnywhere (except separate splitter server)
• [app PR #7]
  • massively expanded amount of text UI input box can handle (128mb rather than 4kb)
  • partially fixed upload problem with non-ASCII filenames

(May '24)

• [lib v1.3.0] / [PyPi]
  • added single-pāda samavṛtta identification (API only)
• [lib v1.2.1] / [PyPi]
  • completed PyPi refactor of library
• [app PR #6]
  • completed Dockerization of front end
  • deployed duplicate server on Digital Ocean Droplet (ultimately kept)

(Mar '24)

• [app PR #5]
  • began Dockerization of front end
  • experimented with serving on Heroku (ultimately did not pursue)

(Jan '23)

• [app PR #2]
  • added API endpoints
• [lib v1.2.0] / [PyPi]
  • restructured library for PyPi; also released lib v1.1.0
  • released lib v1.2.0 on Zenodo
• prepared second presentation for 18th WSC (mp4 ↓ of 20-min version of talk)

(Dec '22)

• switched to using new separate splitter_server API
• fixed /ChicagoApteIAST endpoint for https

(May '22)

• added hidden /ChicagoApteIAST endpoint to supplement searching DSAL online Apte (e.g. https://skrutable.info/ChicagoApteIAST?query=deva)

(Jan '22)

• presented project a first time for 18th WSC (pdf ↓ of paper; mp4 ↓ of full 36-min talk)

(Nov '21)

• released lib v1.0.0 on Zenodo / [app v1.0.0]

(Feb–Mar '21)

• implemented basic viṣamavṛtta meter identification
• incorporated recitation melodies with multiple voice options
• wrote up project for 18th WSC (2021/2023)
• implemented basic wrapper for Splitter

(Jan–Feb '21)

• added whole-file analysis to front end (transliteration, meter identification)
• improved pāda re-splitting ("re-split lite", keep midpoint)
• enabled mix-and-match scansion detail
• added many more meters from Apte
• improved various meter identification categories (upājati, ardhasamavṛtta, half-anuṣṭubh, etc.)
• added Prakrit support (ï and ü)
• further expanded transliteration (ITRANS, WX, "IAST reduced")
• created mature front end in Flask as separate but coordinated repo, deployed on PythonAnywhere

(2018–2020)

• created front end prototype with wx Python library
• invented "wiggle iterator" for re-splitting verse lines
• gradually expanded three main functions

(2015–2017)

• started development on Python library with transliteration, scansion, meter identification

ideas for future updates

Transliteration

• add number-exclusion option to transliteration
• add alphabetization function to front-end

Meter

• improve partial meter identification: samavṛtta, ardhasamavṛtta, jāti
• add TSV summary output options for meter identification
• implement metrical material detection (maybe)

Splitting

• fix small bug: Splitter 2018 cuts off buffer-final numbers
• refactor Splitter 2018 server to improve response time (maybe)

General

• add progress bar to estimate processing time for large file inputs
• continue cleaning up code and improving documentation
• implement more comprehensive testing with pytest

And maybe something you have in mind? Gmail me at tyler.g.neill and share your ideas.