import gradio as gr
from fastapi import FastAPI, HTTPException
from fastapi.responses import StreamingResponse
from fastapi.staticfiles import StaticFiles
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from PIL import Image
from io import BytesIO
import tempfile
import time
import os
import logging
from huggingface_hub import hf_hub_download
# --- Gemini SDK (v1.x) ---------------------------------
from google import genai
from google.genai import types
# -------------------------------------------------------
# ロギング設定
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
# ---------- データモデル ----------
class GeminiRequest(BaseModel):
text: str
extension_percentage: float = 10.0
temperature: float = 0.5
trim_whitespace: bool = True
style: str = "standard"
class ScreenshotRequest(BaseModel):
html_code: str
extension_percentage: float = 10.0
trim_whitespace: bool = True
style: str = "standard"
# ---------- ユーティリティ ----------
def enhance_font_awesome_layout(html_code):
fa_fix_css = """
"""
if '' in html_code:
return html_code.replace('', f'{fa_fix_css}')
elif '')
if head_end > 0:
return html_code[:head_end] + fa_fix_css + html_code[head_end:]
body_start = html_code.find(' 0:
return html_code[:body_start] + f'{fa_fix_css}' + html_code[body_start:]
return f'{fa_fix_css}' + html_code + ''
def load_system_instruction(style="standard"):
valid_styles = ["standard", "cute", "resort", "cool", "dental"]
if style not in valid_styles:
logger.warning(f"無効なスタイル '{style}'。'standard' を使用")
style = "standard"
local_path = os.path.join(os.path.dirname(__file__), style, "prompt.txt")
if os.path.exists(local_path):
with open(local_path, encoding="utf-8") as f:
return f.read()
try:
file_path = hf_hub_download(
repo_id="tomo2chin2/GURAREKOstlyle",
filename=f"{style}/prompt.txt",
repo_type="dataset"
)
with open(file_path, encoding="utf-8") as f:
return f.read()
except Exception:
file_path = hf_hub_download(
repo_id="tomo2chin2/GURAREKOstlyle",
filename="prompt.txt",
repo_type="dataset"
)
with open(file_path, encoding="utf-8") as f:
return f.read()
# ---------- Gemini HTML 生成 ----------
def generate_html_from_text(text, temperature=0.3, style="standard"):
"""Gemini で HTML を生成。2.5 Flash Preview の場合は thinking_off"""
api_key = os.environ.get("GEMINI_API_KEY")
if not api_key:
raise ValueError("GEMINI_API_KEY が未設定")
model_name = os.environ.get("GEMINI_MODEL", "gemini-1.5-pro")
client = genai.Client(api_key=api_key)
system_instruction = load_system_instruction(style)
prompt = f"{system_instruction}\n\n{text}"
safety_settings = [
{"category": "HARM_CATEGORY_HARASSMENT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
{"category": "HARM_CATEGORY_HATE_SPEECH", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
{"category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
{"category": "HARM_CATEGORY_DANGEROUS_CONTENT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
]
# ---- モデル分岐 ----
if model_name == "gemini-2.5-flash-preview-04-17":
generation_cfg = types.GenerateContentConfig(
temperature=temperature,
top_p=0.7,
top_k=20,
max_output_tokens=8192,
candidate_count=1,
thinking_config=types.ThinkingConfig(thinking_budget=0) # thinking OFF
)
else:
generation_cfg = types.GenerateContentConfig(
temperature=temperature,
top_p=0.7,
top_k=20,
max_output_tokens=8192,
candidate_count=1,
)
response = client.models.generate_content(
model=model_name,
contents=prompt,
config=generation_cfg,
safety_settings=safety_settings
)
raw = response.text
# ```html ... ``` 抽出
start = raw.find("```html")
end = raw.rfind("```")
if start != -1 and end != -1 and start < end:
html_code = raw[start + 7:end].strip()
return enhance_font_awesome_layout(html_code)
logger.warning("```html``` ブロックが見つからず全文返却")
return raw
# ---------- 画像トリミング ----------
def trim_image_whitespace(image, threshold=250, padding=10):
gray = image.convert("L")
data = list(gray.getdata())
w, h = gray.size
pixels = [data[i * w:(i + 1) * w] for i in range(h)]
min_x, min_y, max_x, max_y = w, h, 0, 0
for y in range(h):
for x in range(w):
if pixels[y][x] < threshold:
min_x, min_y = min(min_x, x), min(min_y, y)
max_x, max_y = max(max_x, x), max(max_y, y)
if min_x > max_x or min_y > max_y:
return image
min_x = max(0, min_x - padding)
min_y = max(0, min_y - padding)
max_x = min(w - 1, max_x + padding)
max_y = min(h - 1, max_y + padding)
return image.crop((min_x, min_y, max_x + 1, max_y + 1))
# ---------- HTML → スクリーンショット（Selenium） ----------
def render_fullpage_screenshot(html_code, extension_percentage=6.0, trim_whitespace=True):
tmp_path, driver = None, None
try:
with tempfile.NamedTemporaryFile(suffix=".html", delete=False, mode="w", encoding="utf-8") as f:
tmp_path = f.name
f.write(html_code)
options = Options()
options.add_argument("--headless")
options.add_argument("--no-sandbox")
options.add_argument("--disable-dev-shm-usage")
driver = webdriver.Chrome(options=options)
driver.set_window_size(1200, 1000)
driver.get("file://" + tmp_path)
WebDriverWait(driver, 15).until(EC.presence_of_element_located((By.TAG_NAME, "body")))
time.sleep(3)
total_height = driver.execute_script("return Math.max(document.body.scrollHeight, document.documentElement.scrollHeight)")
viewport = driver.execute_script("return window.innerHeight")
for i in range(max(1, total_height // viewport) + 1):
driver.execute_script(f"window.scrollTo(0, {i * (viewport - 200)})")
time.sleep(0.2)
driver.execute_script("window.scrollTo(0, 0)")
driver.execute_script("document.documentElement.style.overflow='hidden';document.body.style.overflow='hidden'")
dims = driver.execute_script("return {w:document.documentElement.scrollWidth,h:document.documentElement.scrollHeight}")
driver.set_window_size(dims["w"], int(dims["h"] * (1 + extension_percentage / 100)))
time.sleep(1)
png = driver.get_screenshot_as_png()
img = Image.open(BytesIO(png))
return trim_image_whitespace(img, 248, 20) if trim_whitespace else img
except Exception as e:
logger.error(f"Selenium error: {e}", exc_info=True)
return Image.new("RGB", (1, 1))
finally:
if driver:
driver.quit()
if tmp_path and os.path.exists(tmp_path):
os.remove(tmp_path)
# ---------- 統合 ----------
def text_to_screenshot(text, ext, temp=0.3, trim=True, style="standard"):
html = generate_html_from_text(text, temp, style)
return render_fullpage_screenshot(html, ext, trim)
# ---------- FastAPI ----------
app = FastAPI()
app.add_middleware(
CORSMiddleware,
allow_origins=["*"], allow_credentials=True,
allow_methods=["*"], allow_headers=["*"],
)
# Gradio 静的ファイル
gradio_dir = os.path.dirname(gr.__file__)
app.mount("/static", StaticFiles(directory=os.path.join(gradio_dir, "templates/frontend/static")), name="static")
# ---------- API ----------
@app.post("/api/screenshot", response_class=StreamingResponse)
async def api_render(req: ScreenshotRequest):
img = render_fullpage_screenshot(req.html_code, req.extension_percentage, req.trim_whitespace)
buf = BytesIO()
img.save(buf, format="PNG")
buf.seek(0)
return StreamingResponse(buf, media_type="image/png")
@app.post("/api/text-to-screenshot", response_class=StreamingResponse)
async def api_text_screenshot(req: GeminiRequest):
img = text_to_screenshot(req.text, req.extension_percentage, req.temperature, req.trim_whitespace, req.style)
buf = BytesIO()
img.save(buf, format="PNG")
buf.seek(0)
return StreamingResponse(buf, media_type="image/png")
# ---------- Gradio ----------
def process_input(mode, txt, ext, temp, trim, style):
if mode == "HTML入力":
return render_fullpage_screenshot(txt, ext, trim)
return text_to_screenshot(txt, ext, temp, trim, style)
# ---------- Gradio UI ----------
with gr.Blocks(title="Full Page Screenshot (テキスト変換対応)", theme=gr.themes.Base()) as iface:
# 見出し
gr.Markdown(
"HTMLビューア ＆ テキスト→インフォグラフィック変換
",
elem_id="title",
inline=True,
)
gr.Markdown(
"HTML を直接レンダリングするか、テキストを Gemini API でインフォグラフィックに変換して画像取得できます。",
elem_id="subtitle",
)
# === 入力モード選択 ===
with gr.Row():
input_mode = gr.Radio(
["HTML入力", "テキスト入力"],
value="HTML入力",
label="入力モード",
)
# === 共通入力テキスト ===
input_text = gr.Textbox(
lines=15,
label="入力",
placeholder="HTMLコードまたはテキストを入力してください（モードに応じて処理）。",
)
# === スタイル + スライダー類を 2 カラムで ===
with gr.Row():
# 左：スタイル
with gr.Column(scale=1, min_width=180):
style_dropdown = gr.Dropdown(
choices=["standard", "cute", "resort", "cool", "dental"],
value="standard",
label="デザインスタイル",
info="テキスト→HTML 変換時のテーマ",
visible=False, # 初期は非表示
)
# 右：拡張率 & 温度
with gr.Column(scale=3):
extension_percentage = gr.Slider(
0, 30, value=10, step=1,
label="上下高さ拡張率（%）"
)
temperature = gr.Slider(
0.0, 1.0, value=0.5, step=0.1,
label="生成時の温度（低い=一貫性高、高い=創造性高）",
visible=False, # 初期は非表示
)
# === トリミング & ボタン ===
trim_whitespace = gr.Checkbox(
value=True,
label="余白を自動トリミング",
)
submit_btn = gr.Button("生成", variant="primary", size="lg")
# === 出力 ===
output_image = gr.Image(
type="pil",
label="ページ全体のスクリーンショット",
show_label=True,
show_download_button=True,
)
# ---------- 変化イベント ----------
def toggle_controls(mode):
"""テキストモードのときだけ温度とスタイルを表示"""
is_text = mode == "テキスト入力"
return (
gr.update(visible=is_text), # temperature
gr.update(visible=is_text), # style_dropdown
)
input_mode.change(
fn=toggle_controls,
inputs=input_mode,
outputs=[temperature, style_dropdown],
)
submit_btn.click(
fn=process_input,
inputs=[
input_mode,
input_text,
extension_percentage,
temperature,
trim_whitespace,
style_dropdown,
],
outputs=output_image,
)
# 下部にモデル表示
gr.Markdown(
f"""
**使用モデル** : `{os.getenv("GEMINI_MODEL", "gemini-1.5-pro")}`
**API** : `/api/screenshot` / `/api/text-to-screenshot`
""",
elem_id="footnote",
)
# --- Gradio を FastAPI にマウント ---
app = gr.mount_gradio_app(app, iface, path="/")
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=7860)