File size: 13,030 Bytes
bef3741
1e6c6f2
9fd65e9
bef3741
 
64322bd
 
813f6fc
64322bd
1e6c6f2
813f6fc
 
1e6c6f2
0550b9b
ceffbde
1e6c6f2
813f6fc
 
1e6c6f2
0550b9b
ceffbde
813f6fc
 
 
 
 
 
ceffbde
813f6fc
ceffbde
0550b9b
 
 
 
 
 
 
1e6c6f2
0550b9b
ec75df1
0550b9b
 
1e6c6f2
0550b9b
813f6fc
0550b9b
813f6fc
 
81a32a1
 
0550b9b
813f6fc
 
 
 
81a32a1
 
813f6fc
81a32a1
813f6fc
81a32a1
9fd65e9
813f6fc
 
 
 
 
 
 
 
 
ec75df1
813f6fc
81a32a1
9fd65e9
 
81a32a1
9fd65e9
 
 
 
 
81a32a1
9fd65e9
 
 
 
81a32a1
 
9fd65e9
81a32a1
 
 
9fd65e9
 
 
81a32a1
9fd65e9
 
81a32a1
9fd65e9
 
 
 
 
 
813f6fc
 
 
81a32a1
0550b9b
1e6c6f2
 
ec75df1
ceffbde
 
813f6fc
ceffbde
 
 
 
bef3741
64322bd
bef3741
64322bd
81a32a1
813f6fc
 
 
 
 
 
81a32a1
813f6fc
 
 
 
 
 
 
 
 
 
81a32a1
ec75df1
813f6fc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ec75df1
64322bd
 
0550b9b
ceffbde
64322bd
1e6c6f2
81a32a1
1e6c6f2
64322bd
813f6fc
64322bd
ceffbde
64322bd
9fd65e9
 
 
0550b9b
81a32a1
 
 
64322bd
1e6c6f2
0550b9b
ceffbde
64322bd
ceffbde
 
 
 
 
 
64322bd
bef3741
64322bd
bef3741
64322bd
ec75df1
 
813f6fc
 
ec75df1
 
64322bd
ceffbde
 
64322bd
ec75df1
9fd65e9
0550b9b
64322bd
ceffbde
0f91f56
ceffbde
64322bd
 
9fd65e9
64322bd
1e6c6f2
64322bd
9ec2497
64322bd
 
 
 
9ec2497
64322bd
0550b9b
64322bd
 
813f6fc
ec75df1
813f6fc
64322bd
0550b9b
64322bd
9ec2497
64322bd
bef3741
 
 
 
0550b9b
64322bd
0550b9b
 
 
 
ec75df1
0550b9b
 
 
 
 
ec75df1
0550b9b
 
 
64322bd
0550b9b
ec75df1
0550b9b
64322bd
0f91f56
64322bd
1e6c6f2
813f6fc
 
ec75df1
 
 
0550b9b
ec75df1
 
 
813f6fc
0550b9b
 
 
 
 
1e6c6f2
0550b9b
 
 
813f6fc
0550b9b
 
 
 
 
 
 
 
 
813f6fc
0550b9b
 
 
 
 
64322bd
 
1e6c6f2
9ec2497
813f6fc
64322bd
ec75df1
9ec2497
1e6c6f2
64322bd
 
 
0550b9b
9ec2497
81a32a1
 
 
 
1e6c6f2
64322bd
813f6fc
 
1e6c6f2
64322bd
 
0550b9b
813f6fc
0550b9b
813f6fc
81a32a1
9fd65e9
 
81a32a1
9fd65e9
 
81a32a1
813f6fc
9fd65e9
ec75df1
 
05bf4a2
1e6c6f2
64322bd
0550b9b
813f6fc
64322bd
813f6fc
 
 
64322bd
0550b9b
813f6fc
64322bd
 
 
 
 
1e6c6f2
bef3741
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
import os
import json
import re
import gradio as gr
from openai import OpenAI

##############################################################################
# 1. 读取外部文件
##############################################################################
try:
    with open("furry_species.json", "r", encoding="utf-8") as ff:
        FURRY_DATA = json.load(ff)
except:
    FURRY_DATA = {}

try:
    with open("gender_rules.json", "r", encoding="utf-8") as gf:
        GENDER_RULES = json.load(gf)
except:
    GENDER_RULES = {}

try:
    with open("transform_rules.json", "r", encoding="utf-8") as tf:
        TRANSFORM_DICT = json.load(tf)
except:
    TRANSFORM_DICT = {}

##############################################################################
# 2. 多级菜单函数
##############################################################################
def get_top_categories(furry_data):
    return sorted(list(furry_data.keys()))

def get_sub_categories(furry_data, top_category):
    if top_category in furry_data:
        return sorted(list(furry_data[top_category].keys()))
    return []

def get_species_list(furry_data, top_category, sub_category):
    if top_category in furry_data and sub_category in furry_data[top_category]:
        return sorted(furry_data[top_category][sub_category])
    return []

##############################################################################
# 3. 合并规则文本
##############################################################################
def merge_transform_rules_into_prompt(rules_json):
    """
    将 transform_rules.json 中的相关字段转为统一文本,
    供 system_prompt 里参考。
    """
    if not rules_json:
        return "(No transform rules loaded)"
    
    gt = rules_json.get("gender_transform", {})
    sp = rules_json.get("shared_preferences", {})
    td = rules_json.get("table_details", {})

    text_parts = []
    text_parts.append("==== GENDER TRANSFORM RULES ====")
    text_parts.append(str(gt))

    text_parts.append("==== SHARED PREFERENCES ====")
    text_parts.append(str(sp))

    text_parts.append("==== TABLE DETAILS (PRO ACTIONS) ====")
    text_parts.append(str(td))

    return "\n".join(text_parts)

RULES_TEXT_FULL = merge_transform_rules_into_prompt(TRANSFORM_DICT)

##############################################################################
# 4. 强制替换:根据 override_conflicting_descriptors
##############################################################################

# 前端选项 -> transform_rules.json 里的 override_conflicting_descriptors key
transform_map = {
    "Trans_to_Male": "female_to_male",
    "Trans_to_Female": "male_to_female",
    "Trans_to_Mannequin": "any_to_genderless",
    "Trans_to_Intersex": "any_to_intersex",
    "Trans_to_Furry": "trans_to_furry"
}

def forced_replace(prompt, direction):
    """
    读取 transform_rules.json["override_conflicting_descriptors"][direction] 的键值,
    用正则整词替换 prompt 中出现的 old->new。
    """
    if not TRANSFORM_DICT:
        return prompt

    override_section = TRANSFORM_DICT.get("override_conflicting_descriptors", {})
    replacements = override_section.get(direction, {})
    if not replacements:
        return prompt  # 该方向没有映射表,直接返回

    for old, new in replacements.items():
        # (?i)不分大小写, \b表示单词边界
        pattern = r"(?i)\b" + re.escape(old) + r"\b"
        prompt = re.sub(pattern, new, prompt)
    return prompt

##############################################################################
# 5. 核心 GPT/DeepSeek 调用
##############################################################################
def generate_transformed_output(prompt, gender_option, top_cat, sub_cat, species_item, api_mode, api_key):
    """
    最终在 GPT/DeepSeek 中生成 (tags)\n\n(description)。
    """
    if not api_key:
        return "Error: No API Key provided."

    if api_mode == "GPT":
        base_url = None
        model_name = "gpt-3.5-turbo"
    else:
        base_url = "https://api.deepseek.com"
        model_name = "deepseek-chat"

    client = OpenAI(api_key=api_key)
    if base_url:
        client.base_url = base_url

    # 如果用户选 Furry, 记录一下当前选到的物种路径
    if gender_option == "Trans_to_Furry":
        furry_path = f"{top_cat} > {sub_cat} > {species_item}" if (top_cat and sub_cat and species_item) else "unknown"
        extra_line = f"\nFurry chosen: {furry_path}\n"
    else:
        extra_line = ""

    # 根据 user 选择加载 gender_rules.json 里的东西
    gender_specific_rule = ""
    if gender_option == "Trans_to_Male":
        gender_specific_rule = GENDER_RULES.get("male", "")
    elif gender_option == "Trans_to_Female":
        gender_specific_rule = GENDER_RULES.get("female", "")
    elif gender_option == "Trans_to_Mannequin":
        gender_specific_rule = GENDER_RULES.get("genderless", "")
    elif gender_option == "Trans_to_Intersex":
        gender_specific_rule = GENDER_RULES.get("intersex", "")

    # 组装 System Prompt
    system_prompt = f"""
You are a creative assistant that transforms the user's base prompt 
to reflect correct gender/furry transformations. Follow these references:

1) Detailed Transform Rules (transform_rules.json):
{RULES_TEXT_FULL}

2) Additional short gender rules (gender_rules.json):
{gender_specific_rule}

{extra_line}
Instructions:
- Original prompt tags: {prompt}
- Convert them into NEW combined tags, removing or replacing conflicting ones.
- Only output two parts:
  1) One line of final tags in parentheses, e.g. (male, short hair, dynamic pose, ...)
  2) A blank line.
  3) Then 3~6 sentences of imaginative scene description in English.
- No extra lines, no headings, no 'gender:' or 'base_prompt:'.
- End of instructions.
""".strip()

    try:
        resp = client.chat.completions.create(
            model=model_name,
            messages=[
                {"role": "system", "content": system_prompt},
                {"role": "user",   "content": "Generate final tags and description now."}
            ],
        )
        return resp.choices[0].message.content.strip()
    except Exception as e:
        return f"{api_mode} generation failed. Error: {e}"

##############################################################################
# 6. 翻译函数
##############################################################################
def translate_text(content, lang, api_mode, api_key):
    """
    后期翻译, 不更改上方生成逻辑.
    """
    if not api_key:
        return "Error: No API Key provided."
    if not content.strip():
        return ""

    if api_mode == "GPT":
        base_url = None
        model_name = "gpt-3.5-turbo"
    else:
        base_url = "https://api.deepseek.com"
        model_name = "deepseek-chat"

    client = OpenAI(api_key=api_key)
    if base_url:
        client.base_url = base_url

    translate_system_prompt = f"""
You are a translator. Translate the following text to {lang}, 
keeping parentheses line and blank line if present. 
No extra headings.
""".strip()

    try:
        resp = client.chat.completions.create(
            model=model_name,
            messages=[
                {"role": "system", "content": translate_system_prompt},
                {"role": "user",   "content": content}
            ],
        )
        return resp.choices[0].message.content.strip()
    except Exception as e:
        return f"{api_mode} translation failed. Error: {e}"

##############################################################################
# 7. Gradio 界面
##############################################################################
def build_interface():
    with gr.Blocks() as demo:
        gr.Markdown("## Prompt Trans-Tool - 提示词物种性别转换器")

        with gr.Row():
            with gr.Column():
                api_mode = gr.Radio(
                    label="Select API 选择API厂商 (GPT/DeepSeek)",
                    choices=["GPT", "DeepSeek"],
                    value="GPT"
                )
                api_key = gr.Textbox(
                    label="API Key",
                    type="password",
                    placeholder="Input your GPT or DeepSeek Key"
                )

                gender_option = gr.Radio(
                    label="Trans-Option 选择转换目标",
                    choices=[
                        "Trans_to_Male",
                        "Trans_to_Female",
                        "Trans_to_Mannequin",
                        "Trans_to_Intersex",
                        "Trans_to_Furry"
                    ],
                    value="Trans_to_Male"
                )

                top_cat_dd = gr.Dropdown(
                    label="Furry: Top Category",
                    choices=get_top_categories(FURRY_DATA),
                    value=None,
                    visible=False
                )
                sub_cat_dd = gr.Dropdown(
                    label="Furry: Sub Category",
                    choices=[],
                    value=None,
                    visible=False
                )
                species_dd = gr.Dropdown(
                    label="Furry: Species",
                    choices=[],
                    value=None,
                    visible=False
                )

                def show_furry_options(opt):
                    if opt == "Trans_to_Furry":
                        return (gr.update(visible=True),
                                gr.update(visible=True),
                                gr.update(visible=True))
                    else:
                        return (gr.update(visible=False),
                                gr.update(visible=False),
                                gr.update(visible=False))

                gender_option.change(
                    fn=show_furry_options,
                    inputs=[gender_option],
                    outputs=[top_cat_dd, sub_cat_dd, species_dd]
                )

                def on_top_cat_select(selected):
                    subs = get_sub_categories(FURRY_DATA, selected)
                    return gr.update(choices=subs, value=None)

                top_cat_dd.change(
                    fn=on_top_cat_select,
                    inputs=[top_cat_dd],
                    outputs=[sub_cat_dd]
                )

                def on_sub_cat_select(top_c, sub_c):
                    sp = get_species_list(FURRY_DATA, top_c, sub_c)
                    return gr.update(choices=sp, value=None)

                sub_cat_dd.change(
                    fn=on_sub_cat_select,
                    inputs=[top_cat_dd, sub_cat_dd],
                    outputs=[species_dd]
                )

            with gr.Column():
                user_prompt = gr.Textbox(
                    label="Original Prompt 原始提示词 (e.g. 1girl, butterfly, solo, ...)",
                    lines=5
                )
                final_output = gr.Textbox(
                    label="Transformed Output 翻译结果 (tags + description)",
                    lines=10
                )

        with gr.Row():
            translate_lang = gr.Dropdown(
                label="Translate to Language 翻译语言",
                choices=[
                    "English", "Chinese", "Japanese", "French", "German", 
                    "Italian", "Spanish", "Russian", "Dutch", "Persian", "Arabic", "Thai"
                ],
                value="English"
            )
            translated_text = gr.Textbox(
                label="Translated Result",
                lines=10
            )

        ######################################################################
        # 生成
        ######################################################################
        def on_generate(prompt, gender, tc, sc, spc, mode, key, lang):
            # 1) 找到 override_conflicting_descriptors 的方向
            direction = transform_map.get(gender, None)
            if direction:
                # 先做强制替换
                prompt = forced_replace(prompt, direction)

            # 2) 再执行原先逻辑
            merged = generate_transformed_output(prompt, gender, tc, sc, spc, mode, key)
            # 3) 翻译
            trans = translate_text(merged, lang, mode, key)
            return merged, trans

        user_prompt.submit(
            fn=on_generate,
            inputs=[user_prompt, gender_option, top_cat_dd, sub_cat_dd, species_dd, api_mode, api_key, translate_lang],
            outputs=[final_output, translated_text]
        )

        gen_btn = gr.Button("Generate")
        gen_btn.click(
            fn=on_generate,
            inputs=[user_prompt, gender_option, top_cat_dd, sub_cat_dd, species_dd, api_mode, api_key, translate_lang],
            outputs=[final_output, translated_text]
        )

    return demo

if __name__ == "__main__":
    demo = build_interface()
    demo.launch()