Spaces:
Sleeping
Sleeping
Update main.py
Browse files
main.py
CHANGED
|
@@ -1,30 +1,155 @@
|
|
| 1 |
-
from fastapi import FastAPI, File, UploadFile, Form, HTTPException
|
| 2 |
-
from io import BytesIO
|
| 3 |
-
import pandas as pd
|
| 4 |
-
from fastapi.responses import StreamingResponse
|
| 5 |
-
from extract import extract_data_from_image
|
| 6 |
-
from parsing import process_extracted_text
|
| 7 |
import csv
|
|
|
|
| 8 |
import json
|
| 9 |
import math
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
from acc_processor import AccessoryProcessor
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
app = FastAPI()
|
| 13 |
|
| 14 |
-
|
| 15 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
try:
|
| 17 |
file_content = await file.read()
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
processed_data = process_extracted_text(extracted_text)
|
| 20 |
return {"diamonds": processed_data["Diamonds"]}
|
| 21 |
except Exception as e:
|
| 22 |
raise HTTPException(status_code=400, detail=str(e))
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
@app.post("/process-image/")
|
| 25 |
async def process_image(
|
| 26 |
-
file: UploadFile = File(
|
| 27 |
-
|
|
|
|
| 28 |
finding_14kt: float = Form(...),
|
| 29 |
finding_18kt: float = Form(...),
|
| 30 |
finding_plt: float = Form(...),
|
|
@@ -36,20 +161,30 @@ async def process_image(
|
|
| 36 |
diamond_markup_type: str = Form(...),
|
| 37 |
diamond_markup_value: float = Form(...),
|
| 38 |
finding_markup_type: str = Form(...),
|
| 39 |
-
finding_markup_value: float = Form(...)
|
| 40 |
):
|
|
|
|
| 41 |
try:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
processor = AccessoryProcessor(accessory_type)
|
| 43 |
-
|
| 44 |
diamond_rates_dict = json.loads(diamond_rates)
|
| 45 |
-
file_content = await file.read()
|
| 46 |
-
extracted_text = extract_data_from_image(BytesIO(file_content))
|
| 47 |
-
processed_data = process_extracted_text(extracted_text)
|
| 48 |
metal_data = processed_data["Metal Details"]
|
| 49 |
diamonds = processed_data["Diamonds"]
|
| 50 |
|
| 51 |
-
|
| 52 |
diamond_values_sum = 0
|
|
|
|
|
|
|
| 53 |
for idx, diamond in enumerate(diamonds):
|
| 54 |
rate = float(diamond_rates_dict.get(str(idx), 0))
|
| 55 |
diamond["Diamond Rate"] = rate
|
|
@@ -59,6 +194,18 @@ async def process_image(
|
|
| 59 |
diamond_values_sum += diamond_value
|
| 60 |
|
| 61 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
if finding_markup_type == "percentage":
|
| 63 |
finding_markup_factor = 1 + (finding_markup_value / 100)
|
| 64 |
finding_14kt = round(finding_14kt * finding_markup_factor, 2)
|
|
@@ -73,115 +220,209 @@ async def process_image(
|
|
| 73 |
finding_18kt = round(finding_18kt * finding_markup_value, 2)
|
| 74 |
finding_plt = round(finding_plt * finding_markup_value, 2)
|
| 75 |
|
| 76 |
-
|
| 77 |
metal_data = processor.calculate_labour(metal_data)
|
| 78 |
|
| 79 |
-
|
| 80 |
finding_key_map = {
|
| 81 |
"pendant": {
|
| 82 |
"14kt": "Finding For Gold 14KT",
|
| 83 |
"18kt": "Finding For Gold 18KT",
|
| 84 |
-
"plt": "Finding for Plt."
|
| 85 |
},
|
| 86 |
"ringband": {
|
| 87 |
"14kt": "CFP for Gold 14KT",
|
| 88 |
"18kt": "CFP for Gold 18KT",
|
| 89 |
-
"plt": "CFP for Plt."
|
| 90 |
},
|
| 91 |
"earring": {
|
| 92 |
"14kt": "Finding & CFP For Gold 14KT",
|
| 93 |
"18kt": "Finding For Gold 18KT",
|
| 94 |
-
"plt": "Finding for Plt."
|
| 95 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
}[accessory_type]
|
| 97 |
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 119 |
if gold_markup_type == "percentage":
|
| 120 |
gold_markup_factor = 1 + (gold_markup_value / 100)
|
| 121 |
-
metal_data["Gold Value 14K"] = round(
|
| 122 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
elif gold_markup_type == "fixed":
|
| 124 |
-
metal_data["Gold Value 14K"] = round(
|
| 125 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 126 |
elif gold_markup_type == "times":
|
| 127 |
-
metal_data["Gold Value 14K"] = round(
|
| 128 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
|
| 130 |
if platinum_markup_type == "percentage":
|
| 131 |
platinum_markup_factor = 1 + (platinum_markup_value / 100)
|
| 132 |
-
metal_data["Plat Value"] = round(
|
|
|
|
|
|
|
| 133 |
elif platinum_markup_type == "fixed":
|
| 134 |
-
|
| 135 |
-
|
|
|
|
| 136 |
elif platinum_markup_type == "times":
|
| 137 |
-
metal_data["Plat Value"] = round(
|
| 138 |
-
|
|
|
|
| 139 |
|
| 140 |
if diamond_markup_type == "percentage":
|
| 141 |
diamond_markup_factor = 1 + (diamond_markup_value / 100)
|
| 142 |
for diamond in diamonds:
|
| 143 |
-
diamond["Diamond Value"] = round(
|
|
|
|
|
|
|
| 144 |
elif diamond_markup_type == "fixed":
|
| 145 |
for diamond in diamonds:
|
| 146 |
-
diamond["Diamond Value"] = round(
|
|
|
|
|
|
|
| 147 |
elif diamond_markup_type == "times":
|
| 148 |
for diamond in diamonds:
|
| 149 |
-
diamond["Diamond Value"] = round(
|
|
|
|
|
|
|
| 150 |
|
| 151 |
diamond_values_sum = sum(diamond["Diamond Value"] for diamond in diamonds)
|
| 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 |
gbp_14k = usd_14k * 0.78
|
| 181 |
gbp_18k = usd_18k * 0.78
|
| 182 |
gbp_pt = usd_pt * 0.78
|
| 183 |
|
| 184 |
-
duty_14k = gbp_14k * 1.025
|
| 185 |
duty_18k = gbp_18k * 1.025
|
| 186 |
duty_pt = gbp_pt * 1.025
|
| 187 |
|
|
@@ -191,98 +432,132 @@ async def process_image(
|
|
| 191 |
landing_18k = duty_18k + freight
|
| 192 |
landing_pt = duty_pt + freight
|
| 193 |
|
| 194 |
-
metal_data.update(
|
| 195 |
-
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
|
| 199 |
-
|
| 200 |
-
|
| 201 |
-
|
| 202 |
-
|
| 203 |
-
|
| 204 |
-
|
| 205 |
-
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
|
| 209 |
-
|
| 210 |
-
|
| 211 |
-
|
| 212 |
-
|
| 213 |
-
|
| 214 |
-
|
|
|
|
|
|
|
| 215 |
|
| 216 |
if not diamonds:
|
| 217 |
final_df = pd.DataFrame([metal_data])
|
| 218 |
else:
|
| 219 |
columns_config = processor.get_column_config()
|
| 220 |
-
|
| 221 |
-
gold_lock_col = next(
|
| 222 |
-
|
| 223 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 224 |
metal_columns = [
|
| 225 |
-
col(metal_data) if callable(col) else col
|
| 226 |
for col in columns_config["metal_columns"]
|
| 227 |
]
|
| 228 |
-
|
| 229 |
-
metal_data_basic = {
|
| 230 |
-
|
| 231 |
-
|
|
|
|
| 232 |
findings_data = {}
|
| 233 |
for col in columns_config["finding_columns"]:
|
| 234 |
-
findings_data[col] = metal_data.get(col, 0)
|
| 235 |
|
| 236 |
findings_df = pd.DataFrame([findings_data])
|
| 237 |
-
|
| 238 |
-
labour_data = {
|
| 239 |
-
|
| 240 |
-
|
| 241 |
-
|
| 242 |
-
|
| 243 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 244 |
metal_df = pd.DataFrame([metal_data_basic])
|
| 245 |
labour_df = pd.DataFrame([labour_data])
|
| 246 |
price_df = pd.DataFrame([price_data])
|
| 247 |
diamonds_df = pd.DataFrame(diamonds)
|
| 248 |
-
|
| 249 |
-
empty_metal = pd.DataFrame(
|
| 250 |
-
|
| 251 |
-
|
| 252 |
-
|
| 253 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 254 |
combined_metal = pd.concat([metal_df, empty_metal], ignore_index=True)
|
| 255 |
-
combined_findings = pd.concat(
|
|
|
|
|
|
|
| 256 |
combined_labour = pd.concat([labour_df, empty_labour], ignore_index=True)
|
| 257 |
combined_prices = pd.concat([price_df, empty_prices], ignore_index=True)
|
| 258 |
-
|
| 259 |
-
final_df = pd.concat(
|
| 260 |
-
|
| 261 |
-
|
| 262 |
-
|
| 263 |
-
|
| 264 |
-
|
| 265 |
-
|
|
|
|
|
|
|
|
|
|
| 266 |
|
| 267 |
for col in columns_config["price_columns"]:
|
| 268 |
-
if col in final_df.columns:
|
| 269 |
final_df[col] = final_df[col].astype(str)
|
| 270 |
-
final_df[col] = final_df[col].apply(
|
|
|
|
|
|
|
| 271 |
if "GBP" in col:
|
| 272 |
-
final_df[col] = final_df[col].apply(
|
| 273 |
-
|
|
|
|
| 274 |
|
| 275 |
print("Finding columns config:", columns_config["finding_columns"])
|
| 276 |
-
print(
|
|
|
|
|
|
|
| 277 |
print("Findings data:", findings_data)
|
| 278 |
|
| 279 |
output = BytesIO()
|
| 280 |
final_df.to_csv(
|
| 281 |
-
output,
|
| 282 |
-
index=False,
|
| 283 |
-
encoding=
|
| 284 |
quoting=csv.QUOTE_MINIMAL,
|
| 285 |
-
escapechar=
|
| 286 |
)
|
| 287 |
output.seek(0)
|
| 288 |
|
|
@@ -291,8 +566,40 @@ async def process_image(
|
|
| 291 |
media_type="text/csv; charset=utf-8",
|
| 292 |
headers={
|
| 293 |
"Content-Disposition": "attachment; filename=output.csv",
|
| 294 |
-
"Content-Type": "text/csv; charset=utf-8"
|
| 295 |
-
}
|
| 296 |
)
|
| 297 |
except Exception as e:
|
| 298 |
-
raise HTTPException(status_code=400, detail=str(e))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import csv
|
| 2 |
+
import hashlib
|
| 3 |
import json
|
| 4 |
import math
|
| 5 |
+
from io import BytesIO
|
| 6 |
+
|
| 7 |
+
import pandas as pd
|
| 8 |
+
from fastapi import (FastAPI, File, Form, HTTPException,
|
| 9 |
+
UploadFile)
|
| 10 |
+
from fastapi.responses import StreamingResponse
|
| 11 |
+
|
| 12 |
from acc_processor import AccessoryProcessor
|
| 13 |
+
from cache import textract_cache
|
| 14 |
+
from extract import extract_data_from_image
|
| 15 |
+
from parsing import map_accessory_type, process_extracted_text
|
| 16 |
|
| 17 |
app = FastAPI()
|
| 18 |
|
| 19 |
+
|
| 20 |
+
def get_file_hash(file_content: bytes) -> str:
|
| 21 |
+
"""Generate a unique hash for a file to use as cache key"""
|
| 22 |
+
return hashlib.md5(file_content).hexdigest()
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
async def get_extracted_text(file: UploadFile = None, extraction_id: str = None) -> str:
|
| 26 |
+
"""Get extracted text from cache or Textract based on file or extraction ID"""
|
| 27 |
+
if extraction_id:
|
| 28 |
+
# Try to get from cache using extraction ID
|
| 29 |
+
cached_text = textract_cache.get(extraction_id)
|
| 30 |
+
if cached_text is not None:
|
| 31 |
+
return cached_text
|
| 32 |
+
raise HTTPException(
|
| 33 |
+
status_code=404,
|
| 34 |
+
detail=f"Extraction ID {extraction_id} not found or expired",
|
| 35 |
+
)
|
| 36 |
+
|
| 37 |
+
if not file:
|
| 38 |
+
raise HTTPException(
|
| 39 |
+
status_code=400, detail="Either file or extraction_id must be provided"
|
| 40 |
+
)
|
| 41 |
+
|
| 42 |
+
file_content = await file.read()
|
| 43 |
+
file_hash = get_file_hash(file_content)
|
| 44 |
+
|
| 45 |
+
# Try to get from cache using file hash
|
| 46 |
+
cached_text = textract_cache.get(file_hash)
|
| 47 |
+
if cached_text is not None:
|
| 48 |
+
# Ensure file.file is repositioned for any subsequent reads
|
| 49 |
+
await file.seek(0)
|
| 50 |
+
return cached_text
|
| 51 |
+
|
| 52 |
+
# Not in cache, extract using Textract
|
| 53 |
+
extracted_text = extract_data_from_image(BytesIO(file_content))
|
| 54 |
+
|
| 55 |
+
# Store in cache (no need to return the ID here)
|
| 56 |
+
textract_cache.set(file_hash, extracted_text)
|
| 57 |
+
|
| 58 |
+
# Ensure file.file is repositioned for any subsequent reads
|
| 59 |
+
await file.seek(0)
|
| 60 |
+
|
| 61 |
+
return extracted_text
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
@app.post("/extract-raw-data/")
|
| 65 |
+
async def extract_raw_data(file: UploadFile = File(...)):
|
| 66 |
+
"""
|
| 67 |
+
Extract raw text data from an image and cache it for later use.
|
| 68 |
+
Returns an extraction_id that can be used in other endpoints.
|
| 69 |
+
"""
|
| 70 |
try:
|
| 71 |
file_content = await file.read()
|
| 72 |
+
file_hash = get_file_hash(file_content)
|
| 73 |
+
|
| 74 |
+
# Try to get from cache first
|
| 75 |
+
cached_text = textract_cache.get(file_hash)
|
| 76 |
+
|
| 77 |
+
if cached_text is None:
|
| 78 |
+
# Not in cache, extract using Textract
|
| 79 |
+
extracted_text = extract_data_from_image(BytesIO(file_content))
|
| 80 |
+
else:
|
| 81 |
+
extracted_text = cached_text
|
| 82 |
+
|
| 83 |
+
# Store in cache and get extraction ID
|
| 84 |
+
extraction_id = textract_cache.set(file_hash, extracted_text)
|
| 85 |
+
|
| 86 |
+
# Return a short sample of the extracted text for verification
|
| 87 |
+
text_sample = (
|
| 88 |
+
extracted_text[:200] + "..."
|
| 89 |
+
if len(extracted_text) > 200
|
| 90 |
+
else extracted_text
|
| 91 |
+
)
|
| 92 |
+
|
| 93 |
+
return {
|
| 94 |
+
"extraction_id": extraction_id,
|
| 95 |
+
"text_sample": text_sample,
|
| 96 |
+
"message": "Extraction successful. Use this extraction_id for subsequent processing calls.",
|
| 97 |
+
}
|
| 98 |
+
except Exception as e:
|
| 99 |
+
raise HTTPException(status_code=500, detail=f"Extraction failed: {str(e)}")
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
@app.post("/extract-diamonds/")
|
| 103 |
+
async def extract_diamonds(
|
| 104 |
+
file: UploadFile = File(None), extraction_id: str = Form(None)
|
| 105 |
+
):
|
| 106 |
+
"""Extract diamonds data using either a file or an extraction_id"""
|
| 107 |
+
try:
|
| 108 |
+
if not file and not extraction_id:
|
| 109 |
+
raise HTTPException(
|
| 110 |
+
status_code=400, detail="Either file or extraction_id must be provided"
|
| 111 |
+
)
|
| 112 |
+
|
| 113 |
+
extracted_text = await get_extracted_text(file, extraction_id)
|
| 114 |
processed_data = process_extracted_text(extracted_text)
|
| 115 |
return {"diamonds": processed_data["Diamonds"]}
|
| 116 |
except Exception as e:
|
| 117 |
raise HTTPException(status_code=400, detail=str(e))
|
| 118 |
|
| 119 |
+
|
| 120 |
+
@app.post("/detect-accessory-type/")
|
| 121 |
+
async def detect_accessory_type(
|
| 122 |
+
file: UploadFile = File(None), extraction_id: str = Form(None)
|
| 123 |
+
):
|
| 124 |
+
"""Detect accessory type using either a file or an extraction_id"""
|
| 125 |
+
try:
|
| 126 |
+
if not file and not extraction_id:
|
| 127 |
+
raise HTTPException(
|
| 128 |
+
status_code=400, detail="Either file or extraction_id must be provided"
|
| 129 |
+
)
|
| 130 |
+
|
| 131 |
+
extracted_text = await get_extracted_text(file, extraction_id)
|
| 132 |
+
processed_data = process_extracted_text(extracted_text)
|
| 133 |
+
|
| 134 |
+
detected_type = processed_data.get("Accessory Type", "pendant")
|
| 135 |
+
|
| 136 |
+
return {
|
| 137 |
+
"detected_accessory_type": detected_type,
|
| 138 |
+
"raw_text_sample": (
|
| 139 |
+
extracted_text[:500] + "..."
|
| 140 |
+
if len(extracted_text) > 500
|
| 141 |
+
else extracted_text
|
| 142 |
+
),
|
| 143 |
+
}
|
| 144 |
+
except Exception as e:
|
| 145 |
+
raise HTTPException(status_code=400, detail=str(e))
|
| 146 |
+
|
| 147 |
+
|
| 148 |
@app.post("/process-image/")
|
| 149 |
async def process_image(
|
| 150 |
+
file: UploadFile = File(None),
|
| 151 |
+
extraction_id: str = Form(None),
|
| 152 |
+
accessory_type: str = Form(None),
|
| 153 |
finding_14kt: float = Form(...),
|
| 154 |
finding_18kt: float = Form(...),
|
| 155 |
finding_plt: float = Form(...),
|
|
|
|
| 161 |
diamond_markup_type: str = Form(...),
|
| 162 |
diamond_markup_value: float = Form(...),
|
| 163 |
finding_markup_type: str = Form(...),
|
| 164 |
+
finding_markup_value: float = Form(...),
|
| 165 |
):
|
| 166 |
+
"""Process image data using either a file or an extraction_id"""
|
| 167 |
try:
|
| 168 |
+
if not file and not extraction_id:
|
| 169 |
+
raise HTTPException(
|
| 170 |
+
status_code=400, detail="Either file or extraction_id must be provided"
|
| 171 |
+
)
|
| 172 |
+
|
| 173 |
+
extracted_text = await get_extracted_text(file, extraction_id)
|
| 174 |
+
processed_data = process_extracted_text(extracted_text)
|
| 175 |
+
|
| 176 |
+
detected_accessory_type = processed_data.get("Accessory Type", "pendant")
|
| 177 |
+
accessory_type = accessory_type or detected_accessory_type
|
| 178 |
+
|
| 179 |
processor = AccessoryProcessor(accessory_type)
|
| 180 |
+
|
| 181 |
diamond_rates_dict = json.loads(diamond_rates)
|
|
|
|
|
|
|
|
|
|
| 182 |
metal_data = processed_data["Metal Details"]
|
| 183 |
diamonds = processed_data["Diamonds"]
|
| 184 |
|
|
|
|
| 185 |
diamond_values_sum = 0
|
| 186 |
+
settings_sum = 0
|
| 187 |
+
|
| 188 |
for idx, diamond in enumerate(diamonds):
|
| 189 |
rate = float(diamond_rates_dict.get(str(idx), 0))
|
| 190 |
diamond["Diamond Rate"] = rate
|
|
|
|
| 194 |
diamond_values_sum += diamond_value
|
| 195 |
|
| 196 |
|
| 197 |
+
if accessory_type in ["bracelet", "necklace"]:
|
| 198 |
+
dia_qty = float(diamond.get("Dia Qty", 0))
|
| 199 |
+
diamond["Setting"] = round(dia_qty * 0.25, 2)
|
| 200 |
+
else:
|
| 201 |
+
diamond["Setting"] = 0
|
| 202 |
+
|
| 203 |
+
settings_sum = (
|
| 204 |
+
sum(diamond["Setting"] for diamond in diamonds)
|
| 205 |
+
if accessory_type in ["bracelet", "necklace"]
|
| 206 |
+
else 0
|
| 207 |
+
)
|
| 208 |
+
|
| 209 |
if finding_markup_type == "percentage":
|
| 210 |
finding_markup_factor = 1 + (finding_markup_value / 100)
|
| 211 |
finding_14kt = round(finding_14kt * finding_markup_factor, 2)
|
|
|
|
| 220 |
finding_18kt = round(finding_18kt * finding_markup_value, 2)
|
| 221 |
finding_plt = round(finding_plt * finding_markup_value, 2)
|
| 222 |
|
| 223 |
+
|
| 224 |
metal_data = processor.calculate_labour(metal_data)
|
| 225 |
|
| 226 |
+
|
| 227 |
finding_key_map = {
|
| 228 |
"pendant": {
|
| 229 |
"14kt": "Finding For Gold 14KT",
|
| 230 |
"18kt": "Finding For Gold 18KT",
|
| 231 |
+
"plt": "Finding for Plt.",
|
| 232 |
},
|
| 233 |
"ringband": {
|
| 234 |
"14kt": "CFP for Gold 14KT",
|
| 235 |
"18kt": "CFP for Gold 18KT",
|
| 236 |
+
"plt": "CFP for Plt.",
|
| 237 |
},
|
| 238 |
"earring": {
|
| 239 |
"14kt": "Finding & CFP For Gold 14KT",
|
| 240 |
"18kt": "Finding For Gold 18KT",
|
| 241 |
+
"plt": "Finding for Plt.",
|
| 242 |
+
},
|
| 243 |
+
"bracelet": {
|
| 244 |
+
"14kt": "CFP For Gold 14KT",
|
| 245 |
+
"18kt": "CFP For Gold 18KT",
|
| 246 |
+
"plt": "CFP For Gold PLT",
|
| 247 |
+
},
|
| 248 |
+
"necklace": {
|
| 249 |
+
"14kt": "CFP For Gold 14KT",
|
| 250 |
+
"18kt": "CFP For Gold 18KT",
|
| 251 |
+
"plt": "CFP For Gold PLT",
|
| 252 |
+
},
|
| 253 |
}[accessory_type]
|
| 254 |
|
| 255 |
+
if accessory_type in ["bracelet", "necklace"]:
|
| 256 |
+
metal_data.update(
|
| 257 |
+
{
|
| 258 |
+
"CFP For Gold 14KT": finding_14kt,
|
| 259 |
+
"CFP For Gold 18KT": finding_18kt,
|
| 260 |
+
"CFP For Gold PLT": finding_plt,
|
| 261 |
+
"Finding Markup Type": finding_markup_type,
|
| 262 |
+
"Finding Markup Value": finding_markup_value,
|
| 263 |
+
"Setting": settings_sum,
|
| 264 |
+
}
|
| 265 |
+
)
|
| 266 |
+
elif accessory_type == "ringband":
|
| 267 |
+
metal_data.update(
|
| 268 |
+
{
|
| 269 |
+
"CFP for Gold 14KT": finding_14kt,
|
| 270 |
+
"CFP for Gold 18KT": finding_18kt,
|
| 271 |
+
"CFP for Plt.": finding_plt,
|
| 272 |
+
"Finding Markup Type": finding_markup_type,
|
| 273 |
+
"Finding Markup Value": finding_markup_value,
|
| 274 |
+
}
|
| 275 |
+
)
|
| 276 |
+
elif accessory_type == "earring":
|
| 277 |
+
metal_data.update(
|
| 278 |
+
{
|
| 279 |
+
"Finding & CFP For Gold 14KT": finding_14kt,
|
| 280 |
+
"Finding For Gold 18KT": finding_18kt,
|
| 281 |
+
"Finding for Plt.": finding_plt,
|
| 282 |
+
"Finding Markup Type": finding_markup_type,
|
| 283 |
+
"Finding Markup Value": finding_markup_value,
|
| 284 |
+
}
|
| 285 |
+
)
|
| 286 |
+
else: # pendant case
|
| 287 |
+
metal_data.update(
|
| 288 |
+
{
|
| 289 |
+
"Finding For Gold 14KT": finding_14kt,
|
| 290 |
+
"Finding For Gold 18KT": finding_18kt,
|
| 291 |
+
"Finding for Plt.": finding_plt,
|
| 292 |
+
"Finding Markup Type": finding_markup_type,
|
| 293 |
+
"Finding Markup Value": finding_markup_value,
|
| 294 |
+
}
|
| 295 |
+
)
|
| 296 |
+
|
| 297 |
+
|
| 298 |
+
metal_data.update(
|
| 299 |
+
{
|
| 300 |
+
"Gold Lock": round(
|
| 301 |
+
metal_data.get("Wt. in gram 14K", 0) * 0.15
|
| 302 |
+
+ metal_data.get("Wt. in gram 18K", 0) * 0.15,
|
| 303 |
+
2,
|
| 304 |
+
),
|
| 305 |
+
"Platinum Lock": round(metal_data.get("Wt. in gram PLT", 0) * 0.25, 2),
|
| 306 |
+
}
|
| 307 |
+
)
|
| 308 |
+
|
| 309 |
+
|
| 310 |
if gold_markup_type == "percentage":
|
| 311 |
gold_markup_factor = 1 + (gold_markup_value / 100)
|
| 312 |
+
metal_data["Gold Value 14K"] = round(
|
| 313 |
+
metal_data["Gold Value 14K"] * gold_markup_factor, 2
|
| 314 |
+
)
|
| 315 |
+
metal_data["Gold Value 18K"] = round(
|
| 316 |
+
metal_data["Gold Value 18K"] * gold_markup_factor, 2
|
| 317 |
+
)
|
| 318 |
elif gold_markup_type == "fixed":
|
| 319 |
+
metal_data["Gold Value 14K"] = round(
|
| 320 |
+
metal_data["Gold Value 14K"] + gold_markup_value, 2
|
| 321 |
+
)
|
| 322 |
+
metal_data["Gold Value 18K"] = round(
|
| 323 |
+
metal_data["Gold Value 18K"] + gold_markup_value, 2
|
| 324 |
+
)
|
| 325 |
elif gold_markup_type == "times":
|
| 326 |
+
metal_data["Gold Value 14K"] = round(
|
| 327 |
+
metal_data["Gold Value 14K"] * gold_markup_value, 2
|
| 328 |
+
)
|
| 329 |
+
metal_data["Gold Value 18K"] = round(
|
| 330 |
+
metal_data["Gold Value 18K"] * gold_markup_value, 2
|
| 331 |
+
)
|
| 332 |
|
| 333 |
if platinum_markup_type == "percentage":
|
| 334 |
platinum_markup_factor = 1 + (platinum_markup_value / 100)
|
| 335 |
+
metal_data["Plat Value"] = round(
|
| 336 |
+
metal_data["Plat Value"] * platinum_markup_factor, 2
|
| 337 |
+
)
|
| 338 |
elif platinum_markup_type == "fixed":
|
| 339 |
+
metal_data["Plat Value"] = round(
|
| 340 |
+
metal_data["Plat Value"] + platinum_markup_value, 2
|
| 341 |
+
)
|
| 342 |
elif platinum_markup_type == "times":
|
| 343 |
+
metal_data["Plat Value"] = round(
|
| 344 |
+
metal_data["Plat Value"] * platinum_markup_value, 2
|
| 345 |
+
)
|
| 346 |
|
| 347 |
if diamond_markup_type == "percentage":
|
| 348 |
diamond_markup_factor = 1 + (diamond_markup_value / 100)
|
| 349 |
for diamond in diamonds:
|
| 350 |
+
diamond["Diamond Value"] = round(
|
| 351 |
+
diamond["Diamond Value"] * diamond_markup_factor, 2
|
| 352 |
+
)
|
| 353 |
elif diamond_markup_type == "fixed":
|
| 354 |
for diamond in diamonds:
|
| 355 |
+
diamond["Diamond Value"] = round(
|
| 356 |
+
diamond["Diamond Value"] + diamond_markup_value, 2
|
| 357 |
+
)
|
| 358 |
elif diamond_markup_type == "times":
|
| 359 |
for diamond in diamonds:
|
| 360 |
+
diamond["Diamond Value"] = round(
|
| 361 |
+
diamond["Diamond Value"] * diamond_markup_value, 2
|
| 362 |
+
)
|
| 363 |
|
| 364 |
diamond_values_sum = sum(diamond["Diamond Value"] for diamond in diamonds)
|
| 365 |
+
|
| 366 |
+
if accessory_type in ["bracelet", "necklace"]:
|
| 367 |
+
usd_14k = (
|
| 368 |
+
metal_data["Gold Value 14K"]
|
| 369 |
+
+ finding_14kt
|
| 370 |
+
+ diamond_values_sum
|
| 371 |
+
+ settings_sum
|
| 372 |
+
+ metal_data["Rhodium"]
|
| 373 |
+
)
|
| 374 |
+
usd_14k = math.ceil(usd_14k) + 1
|
| 375 |
+
|
| 376 |
+
usd_18k = (
|
| 377 |
+
metal_data["Gold Value 18K"]
|
| 378 |
+
+ finding_18kt
|
| 379 |
+
+ diamond_values_sum
|
| 380 |
+
+ settings_sum
|
| 381 |
+
+ metal_data["Rhodium"]
|
| 382 |
+
)
|
| 383 |
+
usd_18k = math.ceil(usd_18k) + 1
|
| 384 |
+
|
| 385 |
+
usd_pt = (
|
| 386 |
+
metal_data["Plat Value"]
|
| 387 |
+
+ finding_plt
|
| 388 |
+
+ diamond_values_sum
|
| 389 |
+
+ settings_sum
|
| 390 |
+
+ metal_data["Rhodium"]
|
| 391 |
+
)
|
| 392 |
+
usd_pt = math.ceil(usd_pt) + 1
|
| 393 |
+
else:
|
| 394 |
+
usd_14k = (
|
| 395 |
+
metal_data["Gold Value 14K"]
|
| 396 |
+
+ finding_14kt
|
| 397 |
+
+ diamond_values_sum
|
| 398 |
+
+ metal_data["Labour 14K"]
|
| 399 |
+
+ metal_data["Rhodium"]
|
| 400 |
+
)
|
| 401 |
+
usd_14k = math.ceil(usd_14k) + 1
|
| 402 |
+
|
| 403 |
+
usd_18k = (
|
| 404 |
+
metal_data["Gold Value 18K"]
|
| 405 |
+
+ finding_18kt
|
| 406 |
+
+ diamond_values_sum
|
| 407 |
+
+ metal_data["Labour 18K"]
|
| 408 |
+
+ metal_data["Rhodium"]
|
| 409 |
+
)
|
| 410 |
+
usd_18k = math.ceil(usd_18k) + 1
|
| 411 |
+
|
| 412 |
+
usd_pt = (
|
| 413 |
+
metal_data["Plat Value"]
|
| 414 |
+
+ finding_plt
|
| 415 |
+
+ diamond_values_sum
|
| 416 |
+
+ metal_data["Labour PT"]
|
| 417 |
+
+ metal_data["Rhodium"]
|
| 418 |
+
)
|
| 419 |
+
usd_pt = math.ceil(usd_pt) + 1
|
| 420 |
|
| 421 |
gbp_14k = usd_14k * 0.78
|
| 422 |
gbp_18k = usd_18k * 0.78
|
| 423 |
gbp_pt = usd_pt * 0.78
|
| 424 |
|
| 425 |
+
duty_14k = gbp_14k * 1.025
|
| 426 |
duty_18k = gbp_18k * 1.025
|
| 427 |
duty_pt = gbp_pt * 1.025
|
| 428 |
|
|
|
|
| 432 |
landing_18k = duty_18k + freight
|
| 433 |
landing_pt = duty_pt + freight
|
| 434 |
|
| 435 |
+
metal_data.update(
|
| 436 |
+
{
|
| 437 |
+
"Gold Markup Type": gold_markup_type,
|
| 438 |
+
"Gold Markup Value": gold_markup_value,
|
| 439 |
+
"Platinum Markup Type": platinum_markup_type,
|
| 440 |
+
"Platinum Markup Value": platinum_markup_value,
|
| 441 |
+
"Diamond Markup Type": diamond_markup_type,
|
| 442 |
+
"Diamond Markup Value": diamond_markup_value,
|
| 443 |
+
"Price of USD 14K": f"$ {usd_14k:.2f}",
|
| 444 |
+
"Price of USD 18K": f"$ {usd_18k:.2f}",
|
| 445 |
+
"Price of USD PT": f"$ {usd_pt:.2f}",
|
| 446 |
+
"Price of GBP 14K @ 1/0.78": f"£ {gbp_14k:.2f}",
|
| 447 |
+
"Price of GBP 18K @ 1/0.78": f"£ {gbp_18k:.2f}",
|
| 448 |
+
"Price of GBP PT @ 1/0.78": f"£ {gbp_pt:.2f}",
|
| 449 |
+
"14K Duty @ 2.5% Including": f"£ {duty_14k:.2f}",
|
| 450 |
+
"18K Duty @ 2.5% Including": f"£ {duty_18k:.2f}",
|
| 451 |
+
"PT950 Duty @ 2.5% Including": f"£ {duty_pt:.2f}",
|
| 452 |
+
"Freight": f"£ {freight:.2f}",
|
| 453 |
+
"14K Landing Cost (Duty + Freight)": f"£ {landing_14k:.2f}",
|
| 454 |
+
"18K Landing Cost (Duty + Freight)": f"£ {landing_18k:.2f}",
|
| 455 |
+
"PT950 Landing Cost (Duty + Freight)": f"£ {landing_pt:.2f}",
|
| 456 |
+
}
|
| 457 |
+
)
|
| 458 |
|
| 459 |
if not diamonds:
|
| 460 |
final_df = pd.DataFrame([metal_data])
|
| 461 |
else:
|
| 462 |
columns_config = processor.get_column_config()
|
| 463 |
+
|
| 464 |
+
gold_lock_col = next(
|
| 465 |
+
col for col in metal_data.keys() if "Gold Lock @" in col
|
| 466 |
+
)
|
| 467 |
+
platinum_lock_col = next(
|
| 468 |
+
col for col in metal_data.keys() if "Platinum Lock" in col
|
| 469 |
+
)
|
| 470 |
+
|
| 471 |
metal_columns = [
|
| 472 |
+
col(metal_data) if callable(col) else col
|
| 473 |
for col in columns_config["metal_columns"]
|
| 474 |
]
|
| 475 |
+
|
| 476 |
+
metal_data_basic = {
|
| 477 |
+
k: v for k, v in metal_data.items() if k in metal_columns
|
| 478 |
+
}
|
| 479 |
+
|
| 480 |
findings_data = {}
|
| 481 |
for col in columns_config["finding_columns"]:
|
| 482 |
+
findings_data[col] = metal_data.get(col, 0) # Use 0 as default value
|
| 483 |
|
| 484 |
findings_df = pd.DataFrame([findings_data])
|
| 485 |
+
|
| 486 |
+
labour_data = {
|
| 487 |
+
k: v
|
| 488 |
+
for k, v in metal_data.items()
|
| 489 |
+
if k in columns_config["labour_columns"]
|
| 490 |
+
}
|
| 491 |
+
|
| 492 |
+
price_data = {
|
| 493 |
+
k: v
|
| 494 |
+
for k, v in metal_data.items()
|
| 495 |
+
if k in columns_config["price_columns"]
|
| 496 |
+
}
|
| 497 |
+
|
| 498 |
metal_df = pd.DataFrame([metal_data_basic])
|
| 499 |
labour_df = pd.DataFrame([labour_data])
|
| 500 |
price_df = pd.DataFrame([price_data])
|
| 501 |
diamonds_df = pd.DataFrame(diamonds)
|
| 502 |
+
|
| 503 |
+
empty_metal = pd.DataFrame(
|
| 504 |
+
[{k: "" for k in metal_columns}] * (len(diamonds_df) - 1)
|
| 505 |
+
)
|
| 506 |
+
empty_findings = pd.DataFrame(
|
| 507 |
+
[{k: "" for k in columns_config["finding_columns"]}]
|
| 508 |
+
* (len(diamonds_df) - 1)
|
| 509 |
+
)
|
| 510 |
+
empty_labour = pd.DataFrame(
|
| 511 |
+
[{k: "" for k in columns_config["labour_columns"]}]
|
| 512 |
+
* (len(diamonds_df) - 1)
|
| 513 |
+
)
|
| 514 |
+
empty_prices = pd.DataFrame(
|
| 515 |
+
[{k: "" for k in columns_config["price_columns"]}]
|
| 516 |
+
* (len(diamonds_df) - 1)
|
| 517 |
+
)
|
| 518 |
+
|
| 519 |
combined_metal = pd.concat([metal_df, empty_metal], ignore_index=True)
|
| 520 |
+
combined_findings = pd.concat(
|
| 521 |
+
[findings_df, empty_findings], ignore_index=True
|
| 522 |
+
)
|
| 523 |
combined_labour = pd.concat([labour_df, empty_labour], ignore_index=True)
|
| 524 |
combined_prices = pd.concat([price_df, empty_prices], ignore_index=True)
|
| 525 |
+
|
| 526 |
+
final_df = pd.concat(
|
| 527 |
+
[
|
| 528 |
+
combined_metal,
|
| 529 |
+
diamonds_df,
|
| 530 |
+
combined_findings,
|
| 531 |
+
combined_labour,
|
| 532 |
+
combined_prices,
|
| 533 |
+
],
|
| 534 |
+
axis=1,
|
| 535 |
+
)
|
| 536 |
|
| 537 |
for col in columns_config["price_columns"]:
|
| 538 |
+
if col in final_df.columns: # Check if column exists
|
| 539 |
final_df[col] = final_df[col].astype(str)
|
| 540 |
+
final_df[col] = final_df[col].apply(
|
| 541 |
+
lambda x: x.strip('"').strip("'") if x else ""
|
| 542 |
+
)
|
| 543 |
if "GBP" in col:
|
| 544 |
+
final_df[col] = final_df[col].apply(
|
| 545 |
+
lambda x: x.replace("£", "£") if x else ""
|
| 546 |
+
)
|
| 547 |
|
| 548 |
print("Finding columns config:", columns_config["finding_columns"])
|
| 549 |
+
print(
|
| 550 |
+
"Metal data finding keys:", [k for k in metal_data.keys() if "Finding" in k]
|
| 551 |
+
)
|
| 552 |
print("Findings data:", findings_data)
|
| 553 |
|
| 554 |
output = BytesIO()
|
| 555 |
final_df.to_csv(
|
| 556 |
+
output,
|
| 557 |
+
index=False,
|
| 558 |
+
encoding="utf-8-sig",
|
| 559 |
quoting=csv.QUOTE_MINIMAL,
|
| 560 |
+
escapechar="\\",
|
| 561 |
)
|
| 562 |
output.seek(0)
|
| 563 |
|
|
|
|
| 566 |
media_type="text/csv; charset=utf-8",
|
| 567 |
headers={
|
| 568 |
"Content-Disposition": "attachment; filename=output.csv",
|
| 569 |
+
"Content-Type": "text/csv; charset=utf-8",
|
| 570 |
+
},
|
| 571 |
)
|
| 572 |
except Exception as e:
|
| 573 |
+
raise HTTPException(status_code=400, detail=str(e))
|
| 574 |
+
|
| 575 |
+
@app.post("/test-accessory-mapping/")
|
| 576 |
+
async def test_accessory_mapping(description: str = Form(...)):
|
| 577 |
+
"""Test endpoint to see how the LLM maps different accessory descriptions."""
|
| 578 |
+
try:
|
| 579 |
+
mapped_type = map_accessory_type(description)
|
| 580 |
+
|
| 581 |
+
return {"original_description": description, "mapped_type": mapped_type}
|
| 582 |
+
except Exception as e:
|
| 583 |
+
raise HTTPException(status_code=400, detail=str(e))
|
| 584 |
+
|
| 585 |
+
|
| 586 |
+
@app.post("/admin/clear-cache/")
|
| 587 |
+
async def clear_cache(key: str = None):
|
| 588 |
+
"""Admin endpoint to clear the cache"""
|
| 589 |
+
textract_cache.clear(key)
|
| 590 |
+
return {"status": "success", "message": "Cache cleared"}
|
| 591 |
+
|
| 592 |
+
|
| 593 |
+
@app.get("/check-extraction/")
|
| 594 |
+
async def check_extraction(extraction_id: str):
|
| 595 |
+
"""Check if an extraction ID is valid and not expired"""
|
| 596 |
+
is_valid = textract_cache.is_valid_id(extraction_id)
|
| 597 |
+
return {
|
| 598 |
+
"extraction_id": extraction_id,
|
| 599 |
+
"valid": is_valid,
|
| 600 |
+
"message": (
|
| 601 |
+
"Extraction ID is valid and can be used"
|
| 602 |
+
if is_valid
|
| 603 |
+
else "Extraction ID not found or expired"
|
| 604 |
+
),
|
| 605 |
+
}
|