screencastgen

Constants

System-wide byte limits, default values, and file patterns.

Source: screencastgen/constants.py

---

TTS Limits

Constant	Value	Description
MAX_CHUNK_BYTES	4900	Default max UTF-8 bytes per chunk
MAX_TTS_BYTES	5000	Hard limit for TTS input validation
MAX_SENTENCE_BYTES	850	Max bytes before force-splitting a sentence

---

Defaults

Constant	Value	Description
DEFAULT_LANGUAGE	"en-US"	Default TTS language
DEFAULT_OUTPUT_DIR	"audio"	Default output directory name

---

Video Defaults

Constant	Value	Description
DEFAULT_VIDEO_WIDTH	1280	Default video width (px)
DEFAULT_VIDEO_HEIGHT	720	Default video height (px)
DEFAULT_VIDEO_FPS	24	Default frames per second

---

EPUB Defaults

Constant	Value	Description
EPUB_AUDIO_FORMAT	"mp3"	Audio format inside EPUB
EPUB_CHAPTER_DIR	"chapters"	Chapter directory within EPUB

---

File Patterns

Constant	Pattern	Used By
CHUNK_FILE_PATTERN	audio_chunk_{num:04d}.{ext}	Pipeline Common, Concatenator
VIDEO_CHUNK_FILE_PATTERN	lipsync_chunk_{num:04d}.mp4	Lipsync Pipeline

---

Usage

These constants serve as defaults — each TTSBackend declares its own max_chunk_bytes property that may differ. Text Processing and Pipeline Common reference these for validation.

---

See Also

• Types — TTSBackend protocol with per-backend limits
• Text Processing — Uses byte limits for chunking
• TTS Base — BackendSpec can override limits

This site is open source. Improve this page.