49 lines
1.7 KiB
Python
49 lines
1.7 KiB
Python
from typing import List, Optional
|
|
|
|
from pydantic import BaseModel, ConfigDict, Field, field_validator, model_validator
|
|
|
|
|
|
class SentenceBoundary(BaseModel):
|
|
model_config = ConfigDict(extra="forbid")
|
|
|
|
index: int = Field(ge=0)
|
|
start_ms: int = Field(ge=0)
|
|
end_ms: int = Field(gt=0)
|
|
text: Optional[str] = None
|
|
|
|
@model_validator(mode="after")
|
|
def validate_range(self):
|
|
if self.end_ms <= self.start_ms:
|
|
raise ValueError("end_ms must be greater than start_ms")
|
|
return self
|
|
|
|
|
|
class SentenceBoundaryDocument(BaseModel):
|
|
model_config = ConfigDict(extra="forbid")
|
|
|
|
video_hash: str
|
|
duration_ms: int = Field(gt=0)
|
|
algorithm_version: str = Field(min_length=1)
|
|
sentences: List[SentenceBoundary]
|
|
|
|
@field_validator("video_hash")
|
|
@classmethod
|
|
def validate_video_hash(cls, value: str) -> str:
|
|
normalized = value.strip().lower()
|
|
if len(normalized) != 64 or any(char not in "0123456789abcdef" for char in normalized):
|
|
raise ValueError("video_hash must be a SHA-256 hex digest")
|
|
return normalized
|
|
|
|
@model_validator(mode="after")
|
|
def validate_sentences(self):
|
|
previous_end = 0
|
|
for expected_index, sentence in enumerate(self.sentences):
|
|
if sentence.index != expected_index:
|
|
raise ValueError("sentence indexes must be contiguous and zero-based")
|
|
if sentence.start_ms < previous_end:
|
|
raise ValueError("sentences must be sorted and non-overlapping")
|
|
if sentence.end_ms > self.duration_ms:
|
|
raise ValueError("sentence end_ms cannot exceed duration_ms")
|
|
previous_end = sentence.end_ms
|
|
return self
|