31 lines
843 B
Python
31 lines
843 B
Python
from imagefind.subtitles import parse_subtitles, timestamp_ms
|
|
from imagefind.text import fts_query, normalize_text, search_tokens
|
|
|
|
|
|
def test_chinese_and_english_tokens_are_stable():
|
|
tokens = search_tokens("红色汽车 Red-Car 订单Ａ12")
|
|
assert "红色" in tokens
|
|
assert "汽车" in tokens
|
|
assert "red-car" in tokens
|
|
assert "a12" in tokens
|
|
assert normalize_text(" ＡＢＣ ") == "abc"
|
|
assert '"红色"' in fts_query("红色")
|
|
|
|
|
|
def test_srt_and_vtt_parser():
|
|
cues = parse_subtitles(
|
|
"""WEBVTT
|
|
|
|
00:00:01.200 --> 00:00:03.400
|
|
<b>你好</b> world
|
|
|
|
00:01:10,000 --> 00:01:11,500
|
|
第二行
|
|
"""
|
|
)
|
|
assert [(cue.start_ms, cue.end_ms, cue.text) for cue in cues] == [
|
|
(1200, 3400, "你好 world"),
|
|
(70000, 71500, "第二行"),
|
|
]
|
|
assert timestamp_ms("1:02:03.045") == 3_723_045
|