Xhs Writer

# OpenAI API 配置（用于图生图功能）
# 如果不配置，会生成纯文字卡片

OPENAI_BASE_URL=https://api.openai.com
OPENAI_API_KEY=sk-your-key-here
GPT_IMAGE_MODEL_NAME=gpt-image-2
GPT_IMAGE_QUALITY=high

# Byte-compiled / optimized / DLL files
__pycache__/
*.py[codz]
*$py.class

# C extensions
*.so

# Distribution / packaging
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
share/python-wheels/
*.egg-info/
.installed.cfg
*.egg
MANIFEST

# PyInstaller
#  Usually these files are written by a python script from a template
#  before PyInstaller builds the exe, so as to inject date/other infos into it.
*.manifest
*.spec

# Installer logs
pip-log.txt
pip-delete-this-directory.txt

# Unit test / coverage reports
htmlcov/
.tox/
.nox/
.coverage
.coverage.*
.cache
nosetests.xml
coverage.xml
*.cover
*.py.cover
.hypothesis/
.pytest_cache/
cover/

# Translations
*.mo
*.pot

# Django stuff:
*.log
local_settings.py
db.sqlite3
db.sqlite3-journal

# Flask stuff:
instance/
.webassets-cache

# Scrapy stuff:
.scrapy

# Sphinx documentation
docs/_build/

# PyBuilder
.pybuilder/
target/

# Jupyter Notebook
.ipynb_checkpoints

# IPython
profile_default/
ipython_config.py

# pyenv
#   For a library or package, you might want to ignore these files since the code is
#   intended to run in multiple environments; otherwise, check them in:
# .python-version

# pipenv
#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
#   However, in case of collaboration, if having platform-specific dependencies or dependencies
#   having no cross-platform support, pipenv may install dependencies that don't work, or not
#   install all needed dependencies.
#Pipfile.lock

# UV
#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
#   This is especially recommended for binary packages to ensure reproducibility, and is more
#   commonly ignored for libraries.
#uv.lock

# poetry
#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
#   This is especially recommended for binary packages to ensure reproducibility, and is more
#   commonly ignored for libraries.
#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
#poetry.lock
#poetry.toml

# pdm
#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
#pdm.lock
#pdm.toml
.pdm-python
.pdm-build/

# pixi
#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
#pixi.lock
#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
#   in the .venv directory. It is recommended not to include this directory in version control.
.pixi

# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
__pypackages__/

# Celery stuff
celerybeat-schedule
celerybeat.pid

# SageMath parsed files
*.sage.py

# Environments
.env
.envrc
.venv
env/
venv/
ENV/
env.bak/
venv.bak/

# Spyder project settings
.spyderproject
.spyproject

# Rope project settings
.ropeproject

# mkdocs documentation
/site

# mypy
.mypy_cache/
.dmypy.json
dmypy.json

# Pyre type checker
.pyre/

# pytype static type analyzer
.pytype/

# Cython debug symbols
cython_debug/

# PyCharm
#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
#  and can be added to the global gitignore or merged into this file.  For a more nuclear
#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
#.idea/

# Abstra
# Abstra is an AI-powered process automation framework.
# Ignore directories containing user credentials, local state, and settings.
# Learn more at https://abstra.io/docs
.abstra/

# Visual Studio Code
#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
#  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
#  and can be added to the global gitignore or merged into this file. However, if you prefer, 
#  you could uncomment the following to ignore the entire vscode folder
# .vscode/

# Ruff stuff:
.ruff_cache/

# PyPI configuration file
.pypirc

# Cursor
#  Cursor is an AI-powered code editor. `.cursorignore` specifies files/directories to
#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
#  refer to https://docs.cursor.com/context/ignore-files
.cursorignore
.cursorindexingignore

# Marimo
marimo/_static/
marimo/_lsp/
__marimo__/

# Skill-generated content when repo is used in-place
output/
outputs/

interface:
  display_name: "XHS (RedNote) Note Writer"
  short_description: "Generate Xiaohongshu (RedNote) notes -- long-form draft, humanized rewrite, split into 3-9 vertical 3:4 cards (≤80 chars each), caption + 5-8 hashtags. Saves to output/小红书/."
  default_prompt: "帮我用 xhs-writer 写一条关于 [topic] 的小红书笔记，5-7 张卡片。"

# Trust surface (what the user agrees to when installing this skill):
#   - Reads scoped .env only (skill dir ~/.claude/skills/xhs-writer-skill/.env);
#     never walks parent project directories.
#   - Uses Claude built-in tools: WebSearch, WebFetch, Bash, Read, Write, Edit.
#   - Downloads images via curl to output/小红书/.../images/; http/https only.
#   - Writes outputs to <cwd>/output/小红书/<date>/<slug>_<ts>/.
#   - Optional outbound: user-configured text-to-image service (T2I_BASE_URL)
#     only if explicitly set in .env; left blank by default.

#!/bin/bash

##############################################################################
# xhs-writer-skill -- Claude Code Skill 安装脚本
#
# 把当前仓库内容拷贝到 ~/.claude/skills/xhs-writer-skill/
# 并引导配置（可选的）.env。
#
# 用法: bash install_as_skill.sh
##############################################################################

set -e

RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m'

print_info()    { echo -e "${BLUE}(i)  $1${NC}"; }
print_success() { echo -e "${GREEN}[OK] $1${NC}"; }
print_warning() { echo -e "${YELLOW}(!)  $1${NC}"; }
print_error()   { echo -e "${RED}[X] $1${NC}"; }
print_header()  { echo ""; echo "========================================"; echo "$1"; echo "========================================"; echo ""; }

command_exists() { command -v "$1" >/dev/null 2>&1; }

main() {
    print_header "xhs-writer-skill -- 安装"

    SKILL_DIR="$HOME/.claude/skills/xhs-writer-skill"
    print_info "目标目录: $SKILL_DIR"

    if [ -d "$SKILL_DIR" ]; then
        print_warning "Skill 目录已存在: $SKILL_DIR"
        read -p "是否覆盖？(y/N) " -n 1 -r
        echo
        if [[ ! $REPLY =~ ^[Yy]$ ]]; then
            print_info "取消"
            exit 0
        fi
        # 备份用户的 .env
        if [ -f "$SKILL_DIR/.env" ]; then
            cp "$SKILL_DIR/.env" "/tmp/xhs-writer.env.bak"
            print_info "已备份现有 .env 到 /tmp/xhs-writer.env.bak"
        fi
        rm -rf "$SKILL_DIR"
    fi

    print_info "创建 Skill 目录..."
    mkdir -p "$SKILL_DIR"
    print_success "目录已创建"

    print_info "复制项目文件..."
    SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"

    # 拷贝核心文件，排除 .git / outputs / venv / .env / __pycache__
    rsync -a \
        --exclude='.git' \
        --exclude='output' \
        --exclude='outputs' \
        --exclude='venv' \
        --exclude='.venv' \
        --exclude='__pycache__' \
        --exclude='.env' \
        "$SCRIPT_DIR/" "$SKILL_DIR/"

    print_success "文件复制完成"

    # 恢复备份的 .env
    if [ -f "/tmp/xhs-writer.env.bak" ]; then
        mv "/tmp/xhs-writer.env.bak" "$SKILL_DIR/.env"
        print_success "已恢复用户 .env"
    fi

    print_header "配置（可选）"

    if [ -f "$SKILL_DIR/.env" ]; then
        print_info "已存在 .env，跳过"
    else
        cp "$SKILL_DIR/.env.example" "$SKILL_DIR/.env"
        print_success "已生成 $SKILL_DIR/.env"
        print_info "本 skill 开箱即用，无需配置 key。"
        print_info "如需接入 text-to-image 服务出卡片配图，可编辑: $SKILL_DIR/.env"
    fi

    print_header "安装完成"

    print_success "已装到 $SKILL_DIR"
    echo ""
    print_info "下一步："
    print_info "  1. 重启 Claude Code 让 skill 生效"
    print_info "  2. 直接对 Claude 说: '帮我用 xhs-writer 写一篇关于 XXX 的小红书笔记'"
    echo ""
}

trap 'print_error "安装过程出错"; exit 1' ERR

main

                                 Apache License
                           Version 2.0, January 2004
                        http://www.apache.org/licenses/

   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION

   1. Definitions.

      "License" shall mean the terms and conditions for use, reproduction,
      and distribution as defined by Sections 1 through 9 of this document.

      "Licensor" shall mean the copyright owner or entity authorized by
      the copyright owner that is granting the License.

      "Legal Entity" shall mean the union of the acting entity and all
      other entities that control, are controlled by, or are under common
      control with that entity. For the purposes of this definition,
      "control" means (i) the power, direct or indirect, to cause the
      direction or management of such entity, whether by contract or
      otherwise, or (ii) ownership of fifty percent (50%) or more of the
      outstanding shares, or (iii) beneficial ownership of such entity.

      "You" (or "Your") shall mean an individual or Legal Entity
      exercising permissions granted by this License.

      "Source" form shall mean the preferred form for making modifications,
      including but not limited to software source code, documentation
      source, and configuration files.

      "Object" form shall mean any form resulting from mechanical
      transformation or translation of a Source form, including but
      not limited to compiled object code, generated documentation,
      and conversions to other media types.

      "Work" shall mean the work of authorship, whether in Source or
      Object form, made available under the License, as indicated by a
      copyright notice that is included in or attached to the work
      (an example is provided in the Appendix below).

      "Derivative Works" shall mean any work, whether in Source or Object
      form, that is based on (or derived from) the Work and for which the
      editorial revisions, annotations, elaborations, or other modifications
      represent, as a whole, an original work of authorship. For the purposes
      of this License, Derivative Works shall not include works that remain
      separable from, or merely link (or bind by name) to the interfaces of,
      the Work and Derivative Works thereof.

      "Contribution" shall mean any work of authorship, including
      the original version of the Work and any modifications or additions
      to that Work or Derivative Works thereof, that is intentionally
      submitted to Licensor for inclusion in the Work by the copyright owner
      or by an individual or Legal Entity authorized to submit on behalf of
      the copyright owner. For the purposes of this definition, "submitted"
      means any form of electronic, verbal, or written communication sent
      to the Licensor or its representatives, including but not limited to
      communication on electronic mailing lists, source code control systems,
      and issue tracking systems that are managed by, or on behalf of, the
      Licensor for the purpose of discussing and improving the Work, but
      excluding communication that is conspicuously marked or otherwise
      designated in writing by the copyright owner as "Not a Contribution."

      "Contributor" shall mean Licensor and any individual or Legal Entity
      on behalf of whom a Contribution has been received by Licensor and
      subsequently incorporated within the Work.

   2. Grant of Copyright License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      copyright license to reproduce, prepare Derivative Works of,
      publicly display, publicly perform, sublicense, and distribute the
      Work and such Derivative Works in Source or Object form.

   3. Grant of Patent License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      (except as stated in this section) patent license to make, have made,
      use, offer to sell, sell, import, and otherwise transfer the Work,
      where such license applies only to those patent claims licensable
      by such Contributor that are necessarily infringed by their
      Contribution(s) alone or by combination of their Contribution(s)
      with the Work to which such Contribution(s) was submitted. If You
      institute patent litigation against any entity (including a
      cross-claim or counterclaim in a lawsuit) alleging that the Work
      or a Contribution incorporated within the Work constitutes direct
      or contributory patent infringement, then any patent licenses
      granted to You under this License for that Work shall terminate
      as of the date such litigation is filed.

   4. Redistribution. You may reproduce and distribute copies of the
      Work or Derivative Works thereof in any medium, with or without
      modifications, and in Source or Object form, provided that You
      meet the following conditions:

      (a) You must give any other recipients of the Work or
          Derivative Works a copy of this License; and

      (b) You must cause any modified files to carry prominent notices
          stating that You changed the files; and

      (c) You must retain, in the Source form of any Derivative Works
          that You distribute, all copyright, patent, trademark, and
          attribution notices from the Source form of the Work,
          excluding those notices that do not pertain to any part of
          the Derivative Works; and

      (d) If the Work includes a "NOTICE" text file as part of its
          distribution, then any Derivative Works that You distribute must
          include a readable copy of the attribution notices contained
          within such NOTICE file, excluding those notices that do not
          pertain to any part of the Derivative Works, in at least one
          of the following places: within a NOTICE text file distributed
          as part of the Derivative Works; within the Source form or
          documentation, if provided along with the Derivative Works; or,
          within a display generated by the Derivative Works, if and
          wherever such third-party notices normally appear. The contents
          of the NOTICE file are for informational purposes only and
          do not modify the License. You may add Your own attribution
          notices within Derivative Works that You distribute, alongside
          or as an addendum to the NOTICE text from the Work, provided
          that such additional attribution notices cannot be construed
          as modifying the License.

      You may add Your own copyright statement to Your modifications and
      may provide additional or different license terms and conditions
      for use, reproduction, or distribution of Your modifications, or
      for any such Derivative Works as a whole, provided Your use,
      reproduction, and distribution of the Work otherwise complies with
      the conditions stated in this License.

   5. Submission of Contributions. Unless You explicitly state otherwise,
      any Contribution intentionally submitted for inclusion in the Work
      by You to the Licensor shall be under the terms and conditions of
      this License, without any additional terms or conditions.
      Notwithstanding the above, nothing herein shall supersede or modify
      the terms of any separate license agreement you may have executed
      with Licensor regarding such Contributions.

   6. Trademarks. This License does not grant permission to use the trade
      names, trademarks, service marks, or product names of the Licensor,
      except as required for reasonable and customary use in describing the
      origin of the Work and reproducing the content of the NOTICE file.

   7. Disclaimer of Warranty. Unless required by applicable law or
      agreed to in writing, Licensor provides the Work (and each
      Contributor provides its Contributions) on an "AS IS" BASIS,
      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
      implied, including, without limitation, any warranties or conditions
      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
      PARTICULAR PURPOSE. You are solely responsible for determining the
      appropriateness of using or redistributing the Work and assume any
      risks associated with Your exercise of permissions under this License.

   8. Limitation of Liability. In no event and under no legal theory,
      whether in tort (including negligence), contract, or otherwise,
      unless required by applicable law (such as deliberate and grossly
      negligent acts) or agreed to in writing, shall any Contributor be
      liable to You for damages, including any direct, indirect, special,
      incidental, or consequential damages of any character arising as a
      result of this License or out of the use or inability to use the
      Work (including but not limited to damages for loss of goodwill,
      work stoppage, computer failure or malfunction, or any and all
      other commercial damages or losses), even if such Contributor
      has been advised of the possibility of such damages.

   9. Accepting Warranty or Additional Liability. While redistributing
      the Work or Derivative Works thereof, You may choose to offer,
      and charge a fee for, acceptance of support, warranty, indemnity,
      or other liability obligations and/or rights consistent with this
      License. However, in accepting such obligations, You may act only
      on Your own behalf and on Your sole responsibility, not on behalf
      of any other Contributor, and only if You agree to indemnify,
      defend, and hold each Contributor harmless for any liability
      incurred by, or claims asserted against, such Contributor by reason
      of your accepting any such warranty or additional liability.

   END OF TERMS AND CONDITIONS

   APPENDIX: How to apply the Apache License to your work.

      To apply the Apache License to your work, attach the following
      boilerplate notice, with the fields enclosed by brackets "[]"
      replaced with your own identifying information. (Don't include
      the brackets!)  The text should be enclosed in the appropriate
      comment syntax for the file format. We also recommend that a
      file or class name and description of purpose be included on the
      same "printed page" as the copyright notice for easier
      identification within third-party archives.

   Copyright [yyyy] [name of copyright owner]

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.

#!/usr/bin/env python3
"""Scan user-provided materials and emit a materials.json manifest.

Multimodal analysis itself (caption, transcribe) is the AGENT's job using its
own vision/audio tools -- this script only does deterministic pre-work:
  - classify each path by extension (text / image / video / audio / unknown)
  - read basic metadata (size; for images: dimensions; for videos: duration
    + frame extraction via ffmpeg if available)
  - emit skeleton manifest that the agent fills in with captions/usages

Usage:
    python3 analyze_material.py PATH [PATH ...] --out reference/materials.json \
        [--frames-dir reference/frames] [--frames-per-video 6]
"""
from __future__ import annotations
import argparse
import json
import shutil
import subprocess
import sys
from pathlib import Path

TEXT_EXT = {".txt", ".md", ".rtf", ".html", ".htm", ".json", ".csv"}
IMAGE_EXT = {".jpg", ".jpeg", ".png", ".gif", ".webp", ".bmp", ".heic", ".heif"}
VIDEO_EXT = {".mp4", ".mov", ".m4v", ".avi", ".mkv", ".webm"}
AUDIO_EXT = {".mp3", ".wav", ".m4a", ".aac", ".flac", ".ogg"}


def classify(p: Path) -> str:
    ext = p.suffix.lower()
    if ext in TEXT_EXT:
        return "text"
    if ext in IMAGE_EXT:
        return "image"
    if ext in VIDEO_EXT:
        return "video"
    if ext in AUDIO_EXT:
        return "audio"
    return "unknown"


def image_info(p: Path) -> dict:
    try:
        from PIL import Image
    except ImportError:
        return {}
    try:
        with Image.open(p) as im:
            return {"width": im.width, "height": im.height,
                    "aspect": round(im.width / im.height, 3)}
    except Exception as e:
        return {"error": str(e)}


def video_info(p: Path) -> dict:
    if not shutil.which("ffprobe"):
        return {"note": "ffprobe not installed; duration unknown"}
    try:
        out = subprocess.check_output(
            ["ffprobe", "-v", "error", "-show_entries",
             "format=duration:stream=width,height,codec_type",
             "-of", "json", str(p)],
            stderr=subprocess.STDOUT).decode()
        data = json.loads(out)
        dur = float(data.get("format", {}).get("duration", 0))
        v = next((s for s in data.get("streams", [])
                  if s.get("codec_type") == "video"), {})
        return {"duration_sec": round(dur, 2),
                "width": v.get("width"), "height": v.get("height")}
    except Exception as e:
        return {"error": str(e)}


def extract_frames(p: Path, out_dir: Path, n: int) -> list[str]:
    if not shutil.which("ffmpeg"):
        return []
    out_dir.mkdir(parents=True, exist_ok=True)
    info = video_info(p)
    dur = info.get("duration_sec", 0) or 0
    if dur <= 0:
        return []
    frames = []
    stem = p.stem.replace(" ", "_")
    for i in range(n):
        t = dur * (i + 0.5) / n
        out = out_dir / f"{stem}_f{i+1:02d}.jpg"
        try:
            subprocess.run(
                ["ffmpeg", "-y", "-v", "error", "-ss", f"{t:.2f}",
                 "-i", str(p), "-frames:v", "1", "-q:v", "3", str(out)],
                check=True)
            frames.append(str(out))
        except Exception:
            pass
    return frames


def text_preview(p: Path, n: int = 400) -> str:
    try:
        return p.read_text(encoding="utf-8", errors="replace")[:n]
    except Exception as e:
        return f"<read error: {e}>"


def main() -> int:
    ap = argparse.ArgumentParser()
    ap.add_argument("paths", nargs="+")
    ap.add_argument("--out", required=True)
    ap.add_argument("--frames-dir", default=None)
    ap.add_argument("--frames-per-video", type=int, default=6)
    args = ap.parse_args()

    items = []
    for raw in args.paths:
        p = Path(raw).expanduser()
        if not p.exists():
            items.append({"path": raw, "error": "not found"})
            continue
        kind = classify(p)
        entry: dict = {
            "id": f"m{len(items)+1:03d}",
            "path": str(p),
            "kind": kind,
            "size_bytes": p.stat().st_size,
            "caption": "",          # agent fills via vision model
            "usage": "",            # cover / content-N / ending / reference
            "strategy": "",         # use_original | text_on_photo | collage | ai_generated
        }
        if kind == "image":
            entry.update(image_info(p))
        elif kind == "video":
            entry.update(video_info(p))
            if args.frames_dir:
                entry["frames"] = extract_frames(
                    p, Path(args.frames_dir), args.frames_per_video)
        elif kind == "text":
            entry["preview"] = text_preview(p)
        items.append(entry)

    out = Path(args.out)
    out.parent.mkdir(parents=True, exist_ok=True)
    out.write_text(json.dumps({"materials": items}, ensure_ascii=False,
                              indent=2), encoding="utf-8")
    print(f"OK: {out} ({len(items)} item(s))")
    return 0


if __name__ == "__main__":
    sys.exit(main())

#!/usr/bin/env python3
"""Compose 2-4 user-provided images into a single 3:4 card (900x1200 by default).

Layouts:
- 2v : two images stacked vertically (top/bottom)
- 2h : two images side by side (left/right) -- rare for 3:4, use for narrow pairs
- 3  : one big on top + two small on bottom (2:1 + 1:1 + 1:1)
- 4  : 2x2 grid

Usage:
    python3 collage_3x4.py OUTPUT --layout 2v --inputs a.jpg b.jpg [--gap 12] \
        [--bg "#FFFFFF"] [--size 900x1200]
"""
from __future__ import annotations
import argparse
import sys
from pathlib import Path

try:
    from PIL import Image
except ImportError:
    print("ERROR: Pillow required", file=sys.stderr)
    sys.exit(2)


def parse_size(s: str) -> tuple[int, int]:
    w, h = s.lower().split("x")
    return int(w), int(h)


def parse_color(s: str) -> tuple[int, int, int]:
    s = s.lstrip("#")
    return int(s[0:2], 16), int(s[2:4], 16), int(s[4:6], 16)


def fit_cover(img: Image.Image, w: int, h: int) -> Image.Image:
    src_w, src_h = img.size
    tgt = w / h
    cur = src_w / src_h
    if cur > tgt:
        new_w = int(src_h * tgt)
        off = (src_w - new_w) // 2
        img = img.crop((off, 0, off + new_w, src_h))
    else:
        new_h = int(src_w / tgt)
        off = (src_h - new_h) // 2
        img = img.crop((0, off, src_w, off + new_h))
    return img.resize((w, h), Image.LANCZOS)


def main() -> int:
    ap = argparse.ArgumentParser()
    ap.add_argument("output")
    ap.add_argument("--layout", required=True,
                    choices=("2v", "2h", "3", "4"))
    ap.add_argument("--inputs", nargs="+", required=True)
    ap.add_argument("--gap", type=int, default=12)
    ap.add_argument("--bg", default="#FFFFFF")
    ap.add_argument("--size", default="900x1200")
    args = ap.parse_args()

    W, H = parse_size(args.size)
    canvas = Image.new("RGB", (W, H), parse_color(args.bg))
    gap = args.gap
    imgs = [Image.open(p).convert("RGB") for p in args.inputs]

    need = {"2v": 2, "2h": 2, "3": 3, "4": 4}[args.layout]
    if len(imgs) != need:
        print(f"ERROR: layout {args.layout} needs {need} inputs, got {len(imgs)}",
              file=sys.stderr)
        return 1

    if args.layout == "2v":
        h_each = (H - gap) // 2
        canvas.paste(fit_cover(imgs[0], W, h_each), (0, 0))
        canvas.paste(fit_cover(imgs[1], W, h_each), (0, h_each + gap))
    elif args.layout == "2h":
        w_each = (W - gap) // 2
        canvas.paste(fit_cover(imgs[0], w_each, H), (0, 0))
        canvas.paste(fit_cover(imgs[1], w_each, H), (w_each + gap, 0))
    elif args.layout == "3":
        top_h = int(H * 2 / 3) - gap // 2
        bot_h = H - top_h - gap
        w_each = (W - gap) // 2
        canvas.paste(fit_cover(imgs[0], W, top_h), (0, 0))
        canvas.paste(fit_cover(imgs[1], w_each, bot_h), (0, top_h + gap))
        canvas.paste(fit_cover(imgs[2], w_each, bot_h),
                     (w_each + gap, top_h + gap))
    else:  # 4
        w_each = (W - gap) // 2
        h_each = (H - gap) // 2
        positions = [(0, 0), (w_each + gap, 0),
                     (0, h_each + gap), (w_each + gap, h_each + gap)]
        for im, pos in zip(imgs, positions):
            canvas.paste(fit_cover(im, w_each, h_each), pos)

    Path(args.output).parent.mkdir(parents=True, exist_ok=True)
    canvas.save(args.output, quality=92)
    print(f"OK: {args.output} ({W}x{H}, layout={args.layout})")
    return 0


if __name__ == "__main__":
    sys.exit(main())

#!/usr/bin/env python3
"""Crop edge watermark off an image (deterministic wrapper over PIL).

Usage:
    python3 crop_watermark.py INPUT OUTPUT --edge bottom --pixels 40
    python3 crop_watermark.py INPUT OUTPUT --box LEFT TOP RIGHT BOTTOM
"""
from __future__ import annotations
import argparse
import sys
from pathlib import Path

try:
    from PIL import Image
except ImportError:
    print("ERROR: Pillow required", file=sys.stderr)
    sys.exit(2)


def main() -> int:
    ap = argparse.ArgumentParser()
    ap.add_argument("input")
    ap.add_argument("output")
    ap.add_argument("--edge", choices=("top", "bottom", "left", "right"))
    ap.add_argument("--pixels", type=int, default=40)
    ap.add_argument("--box", nargs=4, type=int,
                    metavar=("L", "T", "R", "B"),
                    help="absolute crop box (overrides --edge)")
    args = ap.parse_args()

    img = Image.open(args.input)
    w, h = img.size

    if args.box:
        box = tuple(args.box)
    elif args.edge == "bottom":
        box = (0, 0, w, max(0, h - args.pixels))
    elif args.edge == "top":
        box = (0, min(h, args.pixels), w, h)
    elif args.edge == "left":
        box = (min(w, args.pixels), 0, w, h)
    elif args.edge == "right":
        box = (0, 0, max(0, w - args.pixels), h)
    else:
        print("ERROR: provide --edge or --box", file=sys.stderr)
        return 2

    out = img.crop(box)
    Path(args.output).parent.mkdir(parents=True, exist_ok=True)
    out.save(args.output, quality=92)
    print(f"OK: {args.output} ({w}x{h} -> {out.size[0]}x{out.size[1]})")
    return 0


if __name__ == "__main__":
    sys.exit(main())

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
gpt-image-2 图片生成器

调用 OpenAI 兼容的 /v1/chat/completions 端点，model 设为 gpt-image-2
（聚灵等中转站把图片模型挂在 chat completions 上，多模态返回）。

如果 base_url 是 OpenAI 官方或支持 Images API 的中转站，
也可以走 /v1/images/generations ---- 由 GPT_IMAGE_ENDPOINT 切换。
"""

from __future__ import annotations

import base64
import os
import re
from pathlib import Path
from typing import Any, Dict, Optional

import requests

# .env 加载由顶层入口（generate_ppt.py 的 find_and_load_env）统一负责。
# 本模块不调 load_dotenv()，避免 import 时无意识加载父目录 / cwd 的 .env，
# 与 SKILL.md 声明的"scoped .env loading"保持一致。


# 16:9 横版用 1536x1024，竖版 9:16 用 1024x1536，方图用 1024x1024
ASPECT_TO_SIZE = {
    "16:9": "1536x1024",
    "9:16": "1024x1536",
    "1:1": "1024x1024",
}

REQUEST_TIMEOUT_SECS = 600  # 图片生成可能需要 1-3 分钟
MAX_RETRIES = 3  # 524/超时/连接断开等瞬态错误的重试次数
RETRY_DELAY_SECS = 5
MAX_ASPECT_RETRIES = 2  # 比例不合格自动重生次数
ASPECT_TOLERANCE = 0.15  # 比例偏差容忍度（±15%）

# 期望的宽高比（width / height）
ASPECT_RATIO_VALUES = {
    "16:9": 16 / 9,
    "9:16": 9 / 16,
    "1:1": 1.0,
}


def read_png_dimensions(path: str) -> tuple:
    """从 PNG header 读宽高，不依赖 PIL。失败返回 (0, 0)。"""
    import struct
    try:
        with open(path, "rb") as f:
            head = f.read(24)
        if head[:8] != b"\x89PNG\r\n\x1a\n":
            return 0, 0
        # IHDR chunk: 16 字节偏移开始 width(4) + height(4) big-endian
        width, height = struct.unpack(">II", head[16:24])
        return width, height
    except Exception:
        return 0, 0


def aspect_acceptable(width: int, height: int, target: str, tolerance: float = ASPECT_TOLERANCE) -> bool:
    """检查实际宽高比是否在目标比例的容差范围内。"""
    if not (width and height):
        return True  # 读不到就放过，不阻塞流程
    expected = ASPECT_RATIO_VALUES.get(target)
    if expected is None:
        return True
    actual = width / height
    deviation = abs(actual - expected) / expected
    return deviation <= tolerance


class GptImage2Generator:
    """gpt-image-2 图片生成器"""

    def __init__(self, aspect_ratio: str = "16:9") -> None:
        self.base_url = os.getenv("OPENAI_BASE_URL", "https://api.openai.com").rstrip("/")
        self.api_key = os.getenv("OPENAI_API_KEY", "")
        self.model_name = os.getenv("GPT_IMAGE_MODEL_NAME", "gpt-image-2")
        self.quality = os.getenv("GPT_IMAGE_QUALITY", "high")
        # endpoint: chat | images | auto（auto 先 images 不行再 chat）
        self.endpoint = os.getenv("GPT_IMAGE_ENDPOINT", "chat").lower()

        if not self.api_key:
            raise ValueError("缺少 OPENAI_API_KEY，请在 .env 中配置")

        self.aspect_ratio = aspect_ratio
        self.default_size = ASPECT_TO_SIZE.get(aspect_ratio, "1536x1024")

        print(
            f"🎨 初始化 gpt-image-2 生成器 "
            f"(model={self.model_name}, size={self.default_size}, "
            f"quality={self.quality}, endpoint={self.endpoint})"
        )

    # ---------- 通用工具 ----------

    def _save_b64(self, b64: str, output_path: str) -> None:
        if "," in b64 and b64.startswith("data:"):
            b64 = b64.split(",", 1)[1]
        with open(output_path, "wb") as f:
            f.write(base64.b64decode(b64))

    def _download_url(self, url: str, output_path: str) -> None:
        # 安全提示：只接受 http/https；下载前打印 host，便于用户识别异常域；
        # 设最大 50MB 上限避免恶意大文件；非 image/* Content-Type 警告但仍写盘。
        from urllib.parse import urlparse
        parsed = urlparse(url)
        if parsed.scheme not in ("http", "https"):
            raise ValueError(f"拒绝下载非 http(s) 协议的 URL: {parsed.scheme}")
        print(f"📥 下载图片 host={parsed.netloc} path={parsed.path[:80]}")
        resp = requests.get(url, stream=True, timeout=REQUEST_TIMEOUT_SECS)
        resp.raise_for_status()
        ctype = resp.headers.get("content-type", "")
        if ctype and not ctype.startswith("image/"):
            print(f"(!)  非 image Content-Type: {ctype}（仍尝试写盘，请人工核对）")
        MAX_BYTES = 50 * 1024 * 1024
        written = 0
        with open(output_path, "wb") as f:
            for chunk in resp.iter_content(chunk_size=8192):
                written += len(chunk)
                if written > MAX_BYTES:
                    f.close()
                    os.remove(output_path)
                    raise ValueError(f"下载超过 {MAX_BYTES} 字节上限，已丢弃 {url[:80]}")
                f.write(chunk)

    def _save_payload(self, payload: str, output_path: str) -> None:
        """根据 payload 形态（b64 / data url / 普通 url）落盘。"""
        if payload.startswith("data:image/") or (len(payload) > 200 and "/" in payload[:200] is False):
            # data:image/...;base64,xxxxx 或裸 base64
            self._save_b64(payload, output_path)
        elif payload.startswith("http"):
            self._download_url(payload, output_path)
        else:
            # 兜底当 base64
            self._save_b64(payload, output_path)

    # ---------- 多模态响应解析（chat completions 走这里）----------

    def _extract_image(self, content: Any) -> str:
        """从 chat completions 的 content 中提取图片 payload（b64 或 URL）。"""
        if isinstance(content, list):
            for part in content:
                if not isinstance(part, dict):
                    continue
                if part.get("type") == "image_url":
                    img_url = part.get("image_url", {}).get("url", "")
                    if img_url:
                        return img_url
                if part.get("type") == "text":
                    text = part.get("text", "")
                    found = self._extract_from_text(text)
                    if found:
                        return found
            raise RuntimeError(f"multimodal parts 里没找到图片：{str(content)[:300]}")

        if isinstance(content, str):
            found = self._extract_from_text(content)
            if found:
                return found
            raise RuntimeError(f"text content 里没找到图片：{content[:300]}")

        raise RuntimeError(f"未知 content 类型：{type(content)}")

    def _extract_from_text(self, text: str) -> Optional[str]:
        # 1) data:image/xxx;base64,YYY
        m = re.search(r"data:image/[\w]+;base64,[A-Za-z0-9+/=]+", text)
        if m:
            return m.group(0)
        # 2) markdown 图片 ![](url)
        m = re.search(r"!\[.*?\]\((https?://[^\)\s]+)\)", text)
        if m:
            return m.group(1)
        # 3) 裸图片 URL
        m = re.search(r"(https?://[^\s\)]+\.(?:png|jpg|jpeg|webp|gif))", text, re.IGNORECASE)
        if m:
            return m.group(1)
        # 4) 任意 http 链接（最后兜底）
        m = re.search(r"(https?://[^\s\)]+)", text)
        if m:
            return m.group(1)
        return None

    # ---------- 端点 1: /v1/chat/completions ----------

    def _request_via_chat(self, prompt: str, size: str, reference_image_path: Optional[str] = None) -> str:
        """流式请求 chat completions，从增量文本中拼接出图片 URL / b64。

        中转站（如聚灵）把图片模型挂在 chat completions 上时，通常要求 stream=True。
        响应里会先推送进度文本（"> 进度：25%"），最后吐 markdown 图片
        （"![image](http://...png)"）或 base64。

        reference_image_path 不为空时，把该图片作为多模态 input 一并塞进 messages，
        让 gpt-image-2 按它的视觉风格出新图（高保真 / 模板克隆模式）。
        """
        url = f"{self.base_url}/v1/chat/completions"
        # 用比例描述而不是具体像素 ---- gpt-image 类模型更听自然语言 "宽屏 16:9"，
        # 写具体像素值反而被忽略。
        if self.aspect_ratio == "9:16":
            aspect_hint = (
                "\n\n【画面比例 -- 强制】严格按 9:16 竖版手机屏幕生成 "
                "(portrait, vertical 9:16, height much taller than width). "
                "绝对不要方图。"
            )
        elif self.aspect_ratio == "1:1":
            aspect_hint = "\n\n【画面比例 -- 强制】1:1 方图。"
        else:
            aspect_hint = (
                "\n\n【画面比例 -- 强制要求】生成图片必须是 16:9 横版宽屏 "
                "(landscape orientation, widescreen 16:9 aspect ratio, "
                "ultrawide horizontal banner format). "
                "宽度必须明显大于高度，宽高比约 1.78:1。"
                "绝对不要生成方图(square)、近方图(near-square)或竖图(portrait)。"
                "Output MUST be 16:9 landscape widescreen, NEVER square or portrait."
            )

        full_prompt = f"{prompt}{aspect_hint}"
        if reference_image_path and os.path.exists(reference_image_path):
            with open(reference_image_path, "rb") as f:
                ref_b64 = base64.b64encode(f.read()).decode("ascii")
            ref_data_url = f"data:image/png;base64,{ref_b64}"
            user_content: Any = [
                {"type": "image_url", "image_url": {"url": ref_data_url}},
                {"type": "text", "text": (
                    "请以上面这张图作为视觉风格参考（配色 / 字体 / 装饰元素 / 布局氛围），"
                    "按下方新内容生成一张全新的 PPT 页，不要复制原图的文字内容。\n\n"
                    + full_prompt
                )},
            ]
        else:
            user_content = full_prompt

        payload = {
            "model": self.model_name,
            "messages": [
                {"role": "user", "content": user_content}
            ],
            "stream": True,
            "temperature": 0.7,
        }
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json",
            "Accept": "text/event-stream",
        }
        print(f"🔗 POST {url}  size={size}  stream=True")
        resp = requests.post(
            url, headers=headers, json=payload,
            stream=True, timeout=REQUEST_TIMEOUT_SECS,
        )
        print(f"📥 status={resp.status_code}")
        if resp.status_code != 200:
            raise RuntimeError(
                f"chat 调用失败 (status={resp.status_code}): {resp.text[:500]}"
            )

        full_text = []
        for line in resp.iter_lines(decode_unicode=True):
            if not line or not line.startswith("data:"):
                continue
            data_str = line[5:].strip()
            if data_str == "[DONE]":
                break
            try:
                import json as _json
                chunk = _json.loads(data_str)
            except Exception:
                continue
            choices = chunk.get("choices") or []
            if not choices:
                continue
            delta = choices[0].get("delta") or {}
            content = delta.get("content")
            if content:
                full_text.append(content)
                # 实时打印进度（截短）
                snippet = content.replace("\n", " ").strip()
                if snippet:
                    print(f"  ↳ {snippet[:80]}")

        merged = "".join(full_text)
        found = self._extract_from_text(merged)
        if not found:
            raise RuntimeError(
                f"流式响应里没找到图片 URL/base64。完整文本：{merged[:500]}"
            )
        return found

    # ---------- 端点 2: /v1/images/generations ----------

    def _request_via_images(self, prompt: str, size: str) -> str:
        url = f"{self.base_url}/v1/images/generations"
        payload = {
            "model": self.model_name,
            "prompt": prompt,
            "n": 1,
            "size": size,
            "quality": self.quality,
            "response_format": "b64_json",
        }
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json",
        }
        print(f"🔗 POST {url}  size={size}  quality={self.quality}")
        resp = requests.post(url, headers=headers, json=payload, timeout=REQUEST_TIMEOUT_SECS)
        print(f"📥 status={resp.status_code}")

        if resp.status_code != 200:
            raise RuntimeError(
                f"images 调用失败 (status={resp.status_code}): {resp.text[:500]}"
            )

        result = resp.json()
        data = result.get("data") or []
        if not data:
            raise RuntimeError(f"响应没有 data: {str(result)[:300]}")
        first = data[0]
        b64 = first.get("b64_json")
        url_field = first.get("url")
        if b64:
            return f"data:image/png;base64,{b64}"
        if url_field:
            return url_field
        raise RuntimeError(f"data[0] 既没 b64_json 也没 url: {str(first)[:300]}")

    # ---------- 对外入口 ----------

    def generate_scene_image(
        self,
        scene_data: Dict[str, Any],
        output_path: str,
        size: str = "auto",
        reference_image_path: Optional[str] = None,
    ) -> str:
        scene_index = scene_data.get("index", 0)
        prompt = scene_data.get("image_prompt", "")
        if not prompt:
            raise ValueError(f"场景 {scene_index} 缺少 image_prompt")

        out_dir = os.path.dirname(output_path)
        if out_dir:
            os.makedirs(out_dir, exist_ok=True)

        target_size = self.default_size if size == "auto" else size
        print(f"📝 prompt[:100]: {prompt[:100].replace(chr(10), ' ')}{'...' if len(prompt) > 100 else ''}")

        import time as _time

        # 外层：比例不合格重生（最多 MAX_ASPECT_RETRIES 次）
        for ratio_attempt in range(MAX_ASPECT_RETRIES + 1):
            last_err = None
            # 内层：瞬态错误重试（524 / 超时等）
            for attempt in range(1, MAX_RETRIES + 1):
                try:
                    if self.endpoint == "images":
                        # images endpoint 暂不支持 reference image，自动 fallback 到 chat
                        if reference_image_path:
                            payload = self._request_via_chat(prompt, target_size, reference_image_path)
                        else:
                            payload = self._request_via_images(prompt, target_size)
                    elif self.endpoint == "chat":
                        payload = self._request_via_chat(prompt, target_size, reference_image_path)
                    else:  # auto
                        try:
                            if reference_image_path:
                                payload = self._request_via_chat(prompt, target_size, reference_image_path)
                            else:
                                payload = self._request_via_images(prompt, target_size)
                        except Exception as e:
                            print(f"(!) images 失败，回退到 chat: {str(e)[:120]}")
                            payload = self._request_via_chat(prompt, target_size, reference_image_path)

                    self._save_payload(payload, output_path)
                    break  # 成功落盘，跳出瞬态重试循环
                except Exception as e:
                    last_err = e
                    msg = str(e)[:200]
                    transient = any(s in msg for s in ("524", "502", "503", "504", "timeout", "Read timed out",
                                                        "Connection aborted", "RemoteDisconnected"))
                    if attempt < MAX_RETRIES and transient:
                        print(f"(!) [scene {scene_index}] 第 {attempt} 次失败({msg})，{RETRY_DELAY_SECS}s 后重试")
                        _time.sleep(RETRY_DELAY_SECS)
                        continue
                    raise
            else:
                raise RuntimeError(f"重试 {MAX_RETRIES} 次仍失败: {last_err}")

            # 比例校验
            w, h = read_png_dimensions(output_path)
            if aspect_acceptable(w, h, self.aspect_ratio):
                print(f"[OK] 已保存: {output_path}  ({w}x{h}, 比例 {w/h:.3f}, 目标 {self.aspect_ratio})")
                return output_path

            # 比例偏离
            actual_ratio = w / h if h else 0
            expected = ASPECT_RATIO_VALUES.get(self.aspect_ratio, 16/9)
            dev_pct = abs(actual_ratio - expected) / expected * 100 if expected else 0
            if ratio_attempt < MAX_ASPECT_RETRIES:
                print(f"📐 [scene {scene_index}] 尺寸 {w}x{h} (比例 {actual_ratio:.3f}) "
                      f"偏离目标 {self.aspect_ratio} {dev_pct:.0f}%，重生 ({ratio_attempt+1}/{MAX_ASPECT_RETRIES})")
                try:
                    os.remove(output_path)
                except OSError:
                    pass
                continue
            else:
                print(f"(!) [scene {scene_index}] 尺寸 {w}x{h} 仍偏离 {dev_pct:.0f}%，已达比例重试上限，保留")
                return output_path

        return output_path


if __name__ == "__main__":
    import sys
    gen = GptImage2Generator(aspect_ratio="16:9")
    gen.generate_scene_image(
        {"index": 0, "image_prompt": "A clean blue gradient background with the bold white text 'gpt-image-2 Test'"},
        "test_output.png",
    )
    print(f"自检完成: {Path('test_output.png').resolve()}")
    sys.exit(0)