Files
CapCutAPI/pattern/001-words.py
2025-08-12 16:56:15 +08:00

627 lines
20 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import requests
import json
from flask import Flask, request, jsonify, Response
import sys
import time
import json
sys.path.append('/Users/sunguannan/capcutapi')
from example import add_image_impl
PORT=9001 #端口
BASE_URL = f"http://localhost:{PORT}"
draft_folder = "/Users/sunguannan/Movies/JianyingPro/User Data/Projects/com.lveditor.draft"
def make_request(endpoint, data, method='POST'):
"""Send HTTP request to the server and handle the response"""
url = f"{BASE_URL}/{endpoint}"
headers = {'Content-Type': 'application/json'}
try:
if method == 'POST':
response = requests.post(url, data=json.dumps(data), headers=headers)
elif method == 'GET':
response = requests.get(url, params=data, headers=headers)
else:
raise ValueError(f"Unsupported HTTP method: {method}")
response.raise_for_status() # Raise an exception if the request fails
return response.json()
except requests.exceptions.RequestException as e:
print(f"Request error: {e}")
sys.exit(1)
except json.JSONDecodeError:
print("Unable to parse server response")
sys.exit(1)
def save_draft_impl(draft_id, draft_folder):
"""API wrapper for save_draft service"""
data = {
"draft_id": draft_id,
"draft_folder": draft_folder
}
return make_request("save_draft", data)
def query_script_impl(draft_id):
"""API wrapper for query_script service"""
data = {
"draft_id": draft_id
}
return make_request("query_script", data)
def add_text_impl(text, start, end, font, font_color, font_size, track_name, draft_folder="123", draft_id=None,
vertical=False, transform_x=0, transform_y=0, font_alpha=1.0,
border_color=None, border_width=0.0, border_alpha=1.0,
background_color=None, background_alpha=1.0, background_style=None,
background_round_radius=0.0, background_height=0.14, background_width=0.14,
background_horizontal_offset=0.5, background_vertical_offset=0.5,
shadow_enabled=False, shadow_alpha=0.9, shadow_angle=-45.0,
shadow_color="#000000", shadow_distance=5.0, shadow_smoothing=0.15,
bubble_effect_id=None, bubble_resource_id=None,
effect_effect_id=None,
intro_animation=None, intro_duration=0.5,
outro_animation=None, outro_duration=0.5,
width=1080, height=1920,
fixed_width=-1, fixed_height=-1,
text_styles=None):
"""Add text with support for multiple styles, shadows, and backgrounds"""
data = {
"draft_folder": draft_folder,
"text": text,
"start": start,
"end": end,
"font": font,
"font_color": font_color,
"font_size": font_size,
"alpha": font_alpha,
"track_name": track_name,
"vertical": vertical,
"transform_x": transform_x,
"transform_y": transform_y
}
# Add border parameters
if border_color:
data["border_color"] = border_color
data["border_width"] = border_width
data["border_alpha"] = border_alpha
# Add background parameters
if background_color:
data["background_color"] = background_color
data["background_alpha"] = background_alpha
if background_style:
data["background_style"] = background_style
data["background_round_radius"] = background_round_radius
data["background_height"] = background_height
data["background_width"] = background_width
data["background_horizontal_offset"] = background_horizontal_offset
data["background_vertical_offset"] = background_vertical_offset
# Add shadow parameters
if shadow_enabled:
data["shadow_enabled"] = shadow_enabled
data["shadow_alpha"] = shadow_alpha
data["shadow_angle"] = shadow_angle
data["shadow_color"] = shadow_color
data["shadow_distance"] = shadow_distance
data["shadow_smoothing"] = shadow_smoothing
# Add bubble effect parameters
if bubble_effect_id:
data["bubble_effect_id"] = bubble_effect_id
if bubble_resource_id:
data["bubble_resource_id"] = bubble_resource_id
# Add text effect parameters
if effect_effect_id:
data["effect_effect_id"] = effect_effect_id
# Add intro animation parameters
if intro_animation:
data["intro_animation"] = intro_animation
data["intro_duration"] = intro_duration
# Add outro animation parameters
if outro_animation:
data["outro_animation"] = outro_animation
data["outro_duration"] = outro_duration
# Add size parameters
data["width"] = width
data["height"] = height
# Add fixed size parameters
if fixed_width > 0:
data["fixed_width"] = fixed_width
if fixed_height > 0:
data["fixed_height"] = fixed_height
if draft_id:
data["draft_id"] = draft_id
# Add text styles parameters
if text_styles:
data["text_styles"] = text_styles
if draft_id:
data["draft_id"] = draft_id
return make_request("add_text", data)
def group_sentences(corrected_srt, threshold=1.0):
"""按时间间隔分句"""
if not corrected_srt:
return []
sentences = []
current_sentence = [corrected_srt[0]]
for i in range(1, len(corrected_srt)):
prev_end = corrected_srt[i-1]["end"]
curr_start = corrected_srt[i]["start"]
if curr_start - prev_end > threshold:
sentences.append(current_sentence)
current_sentence = [corrected_srt[i]]
else:
current_sentence.append(corrected_srt[i])
sentences.append(current_sentence)
return sentences
def adjust_sentence_timing(sentences, gap_adjust=1, time_precision=3):
"""调整句子间的时间间隔,并保留原始时间"""
def round_time(t):
return round(t, time_precision) if time_precision is not None else t
adjusted_sentences = []
total_offset = 0.0
prev_end = sentences[0][-1]["end"]
# 第一句保持原时间
first_sentence = [
{
"word": w["word"],
"start": w["start"],
"end": w["end"],
"original_start": w["start"],
"original_end": w["end"]
}
for w in sentences[0]
]
adjusted_sentences.append(first_sentence)
for i in range(1, len(sentences)):
sentence = sentences[i]
curr_start = sentence[0]["start"]
natural_gap = curr_start - prev_end
adjusted_gap = natural_gap if gap_adjust == 0 else (1.0 if natural_gap > 1.0 else natural_gap)
move_amount = natural_gap - adjusted_gap
total_offset += move_amount
adjusted_sentence = []
for w in sentence:
adjusted_sentence.append({
"word": w["word"],
"start": round_time(w["start"] - total_offset),
"end": round_time(w["end"] - total_offset),
"original_start": w["start"],
"original_end": w["end"]
})
adjusted_sentences.append(adjusted_sentence)
prev_end = sentence[-1]["end"]
return adjusted_sentences
def split_into_paragraphs(sentence, max_words=5, max_chunk_duration=1.5):
"""把句子按词数和时长分段"""
paragraphs = []
i = 0
n = len(sentence)
while i < n:
paragraph = [sentence[i]]
current_start = sentence[i]["start"]
current_end = sentence[i]["end"]
i += 1
while i < n:
current_word = sentence[i]
is_continuous = abs(current_word["start"] - current_end) < 0.001
if (len(paragraph) >= max_words or
(current_word["end"] - current_start) >= max_chunk_duration or
not is_continuous):
break
paragraph.append(current_word)
current_end = current_word["end"]
i += 1
paragraphs.append(paragraph)
return paragraphs
def build_segments_by_mode(
mode,
paragraph,
track_name,
font,
font_size,
highlight_color,
normal_color,
transform_x,
transform_y,
fixed_width,
shadow_enabled,
shadow_color,
border_color,
border_width,
border_alpha,
background_color,
):
"""根据模式生成字幕片段"""
segments = []
#print("二级代码返回调试fx", fixed_width)
if mode == "word_pop":
# 单词跳出
for w in paragraph:
text_styles = []
word_count = len(w["word"].replace(" ", "")) #统计有多少个字
text_styles.append({
"start": 0,
"end": word_count,
"border": {
"alpha": border_alpha,
"color": border_color,
"width": border_width
}
})
segments.append({
"text": w["word"],
"start": w["start"],
"end": w["end"],
"font": font,
"track_name": track_name,
"font_color": normal_color,
"font_size": font_size,
"transform_x": transform_x,
"transform_y": transform_y,
"shadow_enabled": shadow_enabled,
"fixed_width": fixed_width,
"text_styles": text_styles,
"shadow_color": shadow_color,
"border_color": border_color,
"border_width": border_width,
"border_alpha": border_alpha,
"background_color": background_color,
})
elif mode == "word_highlight":
# 单词高亮:当前词亮,其他灰
paragraph_text = " ".join(w["word"] for w in paragraph)
offsets = []
ci = 0
for w in paragraph:
offsets.append((ci, ci + len(w["word"])))
ci += len(w["word"]) + 1
for idx, w in enumerate(paragraph):
text_styles = []
for k, (s, e) in enumerate(offsets):
color = highlight_color if k == idx else normal_color
text_styles.append({
"start": s,
"end": e,
"style": {
"color": color,
"size": font_size,
},
"border": {
"alpha": border_alpha,
"color": border_color,
"width": border_width
}
})
print("text_styles", text_styles)
segments.append({
"text": paragraph_text,
"start": w["start"],
"end": w["end"],
"font": font,
"track_name": track_name,
"font_color": normal_color,
"font_size": font_size,
"text_styles": text_styles,
"transform_x": transform_x,
"transform_y": transform_y,
"shadow_enabled": shadow_enabled,
"fixed_width": fixed_width,
"shadow_color": shadow_color,
"border_color": border_color,
"border_width": border_width,
"border_alpha": border_alpha,
"background_color": background_color,
})
elif mode == "sentence_fade":
# 句子渐显:已亮过的词继续保持亮
paragraph_text = " ".join(w["word"] for w in paragraph)
offsets = []
ci = 0
for w in paragraph:
offsets.append((ci, ci + len(w["word"])))
ci += len(w["word"]) + 1
for idx, w in enumerate(paragraph):
text_styles = []
for k, (s, e) in enumerate(offsets):
color = highlight_color if k <= idx else normal_color
text_styles.append({
"start": s,
"end": e,
"style": {"color": color, "size": font_size},
"border": {
"alpha": border_alpha,
"color": border_color,
"width": border_width
}
})
segments.append({
"text": paragraph_text,
"start": w["start"],
"end": w["end"],
"font": font,
"track_name": track_name,
"font_color": normal_color,
"font_size": font_size,
"text_styles": text_styles,
"transform_x": transform_x,
"transform_y": transform_y,
"shadow_enabled": shadow_enabled,
"fixed_width": fixed_width,
"shadow_color": shadow_color,
"border_color": border_color,
"border_width": border_width,
"border_alpha": border_alpha,
"background_color": background_color,
})
elif mode == "sentence_pop":
# 句子跳出
text = " ".join(w["word"] for w in paragraph)
start_time = paragraph[0]["start"]
end_time = paragraph[-1]["end"]
text_styles = []
word_count = len(text.replace(" ", "")) #统计有多少个字
text_styles.append({
"start": 0,
"end": word_count,
"border": {
"alpha": border_alpha,
"color": border_color,
"width": border_width
}
})
segments.append({
"text": text,
"start": start_time,
"end": end_time,
"font": font,
"track_name": track_name,
"font_color": normal_color,
"font_size": font_size,
"transform_x": transform_x,
"transform_y": transform_y,
"shadow_enabled": shadow_enabled,
"fixed_width": fixed_width,
"text_styles": text_styles,
"shadow_color": shadow_color,
"border_color": border_color,
"border_width": border_width,
"border_alpha": border_alpha,
"background_color": background_color,
})
else:
raise ValueError(f"未知模式: {mode}")
"""segments.append({
"file_name": file_name,
})"""
return segments
corrected_srt = [{
"word": "Hello",
"start": 0.0,
"end": 0.64,
"confidence": 0.93917525
},
{
"word": "I'm",
"start": 0.64,
"end": 0.79999995,
"confidence": 0.9976464
},
{
"word": "PAWA",
"start": 0.79999995,
"end": 1.36,
"confidence": 0.6848311
},
{
"word": "Nice",
"start": 1.36,
"end": 1.52,
"confidence": 0.9850389
},
{
"word": "To",
"start": 1.52,
"end": 1.68,
"confidence": 0.9926886
},
{
"word": "Meet",
"start": 1.68,
"end": 2.08,
"confidence": 0.9972697
},
{
"word": "You",
"start": 2.08,
"end": 2.72,
"confidence": 0.9845563
},
{
"word": "Enjoy",
"start": 2.72,
"end": 3.04,
"confidence": 0.99794894
},
{
"word": "My",
"start": 3.04,
"end": 3.1999998,
"confidence": 0.9970203
},
{
"word": "Parttern",
"start": 3.1999998,
"end": 3.36,
"confidence": 0.9970235
},
{
"word": "Thank",
"start": 3.36,
"end": 3.6799998,
"confidence": 0.98627764
},
{
"word": "You",
"start": 3.6799998,
"end": 4.0,
"confidence": 0.9939551
},
]
def add_koubo_from_srt(
corrected_srt,
track_name,
mode="word_pop",
font="ZY_Modern",
font_size=32,
highlight_color="#FFD700",
normal_color="#AAAAAA", max_chunk_duration=1.5, max_words=5,
gap_adjust=1,
time_precision=3,
transform_x=0.5,
transform_y=0.3,
fixed_width=-1,
shadow_enabled=True,
shadow_color="#000000",
border_color="#000000",
border_width=0.5,
border_alpha=1.0,
background_color="#000000",
):
"""统一入口:根据 mode 选择字幕效果"""
sentences = group_sentences(corrected_srt)
adjusted_sentences = adjust_sentence_timing(sentences, gap_adjust, time_precision)
all_paragraphs = [split_into_paragraphs(s, max_words, max_chunk_duration) for s in adjusted_sentences]
draft_id_ret = None
for sentence_paragraphs in all_paragraphs:
for paragraph in sentence_paragraphs:
segments = build_segments_by_mode(
mode,
paragraph,
track_name,
font,
font_size,
highlight_color,
normal_color,
transform_x,
transform_y,
fixed_width,
shadow_enabled,
shadow_color,
border_color,
border_width,
border_alpha,
background_color,
)
#print("segments", segments)
for seg in segments:
#print("二级代码返回调试fx", seg)
if draft_id_ret:
seg["draft_id"] = draft_id_ret
print("seg", seg)
res = add_text_impl(**seg)
if draft_id_ret is None and isinstance(res, dict):
try:
draft_id_ret = res["output"]["draft_id"]
except:
pass
return draft_id_ret
colors = {
"shadow_color": "#000000",
"border_color": "#FFD700",
"background_color": "#000000",
"normal_color": "#FFFFFF",
"highlight_color": "#DA70D6" # 紫色
}
draft_id = add_koubo_from_srt(
corrected_srt,
track_name="main_text",
font_size=15,
gap_adjust=0,
transform_x=0,
transform_y=-0.45,# 0=保持原间隔1=调整>1s的间隔
fixed_width = 0.6,
mode="word_highlight",
shadow_enabled=True,
border_width=10,
border_alpha=1.0,
**colors,
font="ZY_Modern", #设置自己的字体,需要在字体库中添加
)
add_image_impl(image_url="https://pic1.imgdb.cn/item/689aff2758cb8da5c81e64a2.png", start = 0, end = 4, draft_id=draft_id)
save_result = save_draft_impl(draft_id, draft_folder)
print(save_result)
"""
# 单词高亮
mode="word_highlight"
# 单词跳出
mode="word_pop"
# 句子渐显
mode="sentence_fade"
# 句子跳出
mode="sentence_pop"
"""