NosillaCast Apple Podcast · Technology with an EVER so slight Apple bias

NC #1077 Eddie Tonkoi Building a Fiction Editing Pipeline with Regex and Python — Part 1, Adam Engst on the Phone App in iOS 26

December 22, 2025·1 hr 11 min·4 clips

Etna keeps a flag-word file so "yoghurt" and "curb" can be marked for review.

1. NosillaCast Apple Podcast #1077 centers on Eddie Tonkoi’s Etna editing pipeline and Adam Engst’s look at the iOS 26 Phone app. 2. Allison Sheridan hosts NosillaCast from podfeet.com, Eddie Tonkoi explains his manuscript workflow, and Adam Engst joins to discuss the Phone app changes. 3. The episode asks what a Mac-based editor and an iOS 26 caller interface are for when the software should support, rather than replace, human judgment. 4. Eddie Tonkoi says Jern Tonkoi writes character-driven queer love stories in Scrivener on an iPad while he handles editing and production on a Mac. 5. He describes a Dropbox-based handoff in which Jern writes, he removes the file from Dropbox for editing, and then the manuscript returns for her review. 6. Eddie says Etna is built because Word, Grammarly, ProWritingAid, LibreOffice, and other tools did not meet his requirements for control, speed, and consistency. 7. His checklist includes suggestions only, whole-manuscript analysis, per-book memory for ignored items, transferability across machines, and a free or non-subscription model. 8. He says the manuscript pipeline starts by exporting a Word .docx file and splitting it into chapters or evenly sized chunks with Python docx. 9. The first script also checks for highlighting, colored text, strikethrough, underlining, square brackets, and straight quotes so editorial comments do not reach print. 10. The second script uses regex to catch structural problems such as double spaces, repeated words, malformed ellipses, and em-dash spacing. 11. Eddie explains that regex is fast and repeatable because it matches patterns, not meaning, and that compiled patterns in Python’s re module make scans effectively instantaneous on a 100,000-word novel. 12. He gives concrete regex examples for multiple spaces and repeated words, including a backreference pattern that catches the same word twice in a row. 13. The spelling stage uses Hunspell, the open-source engine behind LibreOffice, Firefox, Thunderbird, and other tools. 14. Eddie adds a global dictionary, a book-specific dictionary, and flag words so terms like glutes, meself, Bermondsey, Dawlish Warren, yoghurt, and curb can be handled differently. 15. He also runs lighter scripts for l-r swaps, fuzzy matching of proper nouns with RapidFuzz, and compound-word consistency such as timeline versus time line. 16. Eddie emphasizes that Etna never writes to the manuscript automatically, so every correction remains a manual decision by the editor. 17. Adam Engst then turns to iOS 26’s Phone app, where Apple offers Classic view and Unified view instead of forcing one interface on users. 18. He and Allison test how the Unified view merges Calls, Contacts, Keypad, and Search, while hidden filters like missed calls, voicemail, and spam live in the hamburger menu. 19. Listeners who like Python workflows, regex, Scrivener, and iPhone usability details will get the most from this episode. 20. Listeners who want short takes without technical workflow or app-navigation detail may skip it.

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from NosillaCast Apple Podcast