File size: 8,343 Bytes
1c05cb6 e110644 1c05cb6 075ff48 cf1935e 1c05cb6 cf1935e 1c05cb6 075ff48 e110644 1c05cb6 e110644 1c05cb6 075ff48 cf1935e e110644 cf1935e e110644 1c05cb6 e110644 1c05cb6 |
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 |
import json
import base64
import os, io
from PIL import Image
import gradio as gr
from gradio import processing_utils, utils
def import_history(history, file):
if os.path.getsize(file.name) > 100e6:
raise ValueError("History larger than 100 MB")
with open(file.name, mode="rb") as f:
content = f.read().decode('utf-8', 'replace')
import_data = json.loads(content)
# Handle different import formats
if 'messages' in import_data:
# New OpenAI-style format
messages = import_data['messages']
system_prompt_value = ''
chat_history = []
msg_num = 1
for msg in messages:
if msg['role'] == 'system':
system_prompt_value = msg['content']
continue
if msg['role'] == 'user':
content = msg['content']
if isinstance(content, list):
for item in content:
if item.get('type', '') == 'image_url':
data_uri = item['image_url']['url']
img_bytes = base64.b64decode(data_uri.split(',')[1])
fname = f"img{msg_num}.webp"
cache_path = processing_utils.save_bytes_to_cache(img_bytes, fname, utils.get_upload_folder())
chat_history.append({
"role": msg['role'],
"content": {"path": cache_path}
})
elif item.get('type', '') == 'file':
fname = os.path.basename(item['file'].get('name', f'download{msg_num}'))
file_data = base64.b64decode(item['file']['url'].split(',')[1])
if (len(file_data) > 15e6):
raise ValueError(f"file content `{fname}` larger than 15 MB")
cache_path = processing_utils.save_bytes_to_cache(file_data, fname, utils.get_upload_folder())
chat_history.append({
"role": msg['role'],
"content": {"path": cache_path}
})
else:
chat_history.append(msg)
else:
chat_history.append(msg)
elif msg['role'] == 'assistant':
chat_history.append(msg)
msg_num = msg_num + 1
else:
# Legacy format handling
if 'history' in import_data:
legacy_history = import_data['history']
system_prompt_value = import_data.get('system_prompt', '')
else:
legacy_history = import_data
system_prompt_value = ''
chat_history = []
# Convert tuple/pair format to messages format
for pair in legacy_history:
if pair[0]: # User message
if isinstance(pair[0], dict) and 'file' in pair[0]:
if 'data' in pair[0]['file']:
# Legacy format with embedded data
file_data = pair[0]['file']['data']
mime_type = file_data.split(';')[0].split(':')[1]
if mime_type.startswith('image/'):
image_bytes = base64.b64decode(file_data.split(',')[1])
fname = 'legacy_img.webp'
cache_path = processing_utils.save_bytes_to_cache(image_bytes, fname, utils.get_upload_folder())
chat_history.append({
"role": "user",
"content": {"path": cache_path}
})
else:
fname = pair[0]['file'].get('name', 'download')
file_bytes = base64.b64decode(file_data.split(',')[1])
cache_path = processing_utils.save_bytes_to_cache(file_bytes, fname, utils.get_upload_folder())
chat_history.append({
"role": "user",
"content": {"path": cache_path}
})
else:
# Keep as-is but convert to message format
chat_history.append({
"role": "user",
"content": pair[0]
})
else:
chat_history.append({
"role": "user",
"content": pair[0]
})
if pair[1]: # Assistant message
chat_history.append({
"role": "assistant",
"content": pair[1]
})
return chat_history, system_prompt_value
def get_export_js():
return """
async (chat_history, system_prompt) => {
let messages = [];
if (system_prompt) {
messages.push({
"role": "system",
"content": system_prompt
});
}
async function processFile(file_url) {
const response = await fetch(file_url);
const blob = await response.blob();
return new Promise((resolve) => {
const reader = new FileReader();
reader.onloadend = () => resolve({
data: reader.result,
type: blob.type
});
reader.onerror = (error) => resolve(null);
reader.readAsDataURL(blob);
});
}
for (let message of chat_history) {
if (!message.role || !message.content) continue;
if (message.content && typeof message.content === 'object') {
if (message.content.file) {
try {
const file_data = await processFile(message.content.file.url);
if (!file_data) continue;
if (file_data.type.startsWith('image/')) {
messages.push({
"role": message.role,
"content": [{
"type": "image_url",
"image_url": {
"url": file_data.data
}
}]
});
} else {
const fileLink = document.querySelector(`a[data-testid="chatbot-file"][download][href*="${message.content.file.url.split('/').pop()}"]`);
const fileName = fileLink ? fileLink.getAttribute('download') : (message.content.file.name || "download");
messages.push({
"role": message.role,
"content": [{
"type": "file",
"file": {
"url": file_data.data,
"name": fileName,
"mime_type": file_data.type
}
}]
});
}
} catch (error) {}
}
} else {
messages.push({
"role": message.role,
"content": message.content
});
}
}
const export_data = { messages };
const blob = new Blob([JSON.stringify(export_data)], {type: 'application/json'});
const url = URL.createObjectURL(blob);
const a = document.createElement('a');
a.href = url;
a.download = 'chat_history.json';
document.body.appendChild(a);
a.click();
document.body.removeChild(a);
URL.revokeObjectURL(url);
}
""" |