Spaces:
Sleeping
Sleeping
File size: 47,290 Bytes
31c9e69 05ce027 bacf0f5 17b84ad 845ffeb bacf0f5 31c9e69 2e5e8b0 bacf0f5 31c9e69 d135896 31c9e69 d135896 31c9e69 2e5e8b0 d135896 bacf0f5 d135896 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 d135896 d7dc6ab f94f352 d135896 bacf0f5 d135896 f94f352 8e377a8 f94f352 e660d9c 2e5e8b0 e660d9c bacf0f5 e660d9c bacf0f5 2e5e8b0 31c9e69 2e5e8b0 e660d9c 2e5e8b0 bacf0f5 31c9e69 bacf0f5 31c9e69 bacf0f5 e660d9c bacf0f5 2e5e8b0 bacf0f5 31c9e69 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 31c9e69 bacf0f5 2e5e8b0 bacf0f5 e660d9c bacf0f5 31c9e69 bacf0f5 e660d9c bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 1281878 2e5e8b0 1281878 bacf0f5 2e5e8b0 bacf0f5 d7dc6ab a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 a024892 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 bacf0f5 2e5e8b0 deaa144 2e5e8b0 4448cdf 845ffeb 2e5e8b0 845ffeb 2e5e8b0 845ffeb 2e5e8b0 845ffeb 2e5e8b0 845ffeb 2e5e8b0 845ffeb 2e5e8b0 845ffeb bccbf11 68c1fb8 2e5e8b0 acd267e 2e5e8b0 bccbf11 2e5e8b0 bccbf11 acd267e 2e5e8b0 acd267e 845ffeb 2e5e8b0 8e377a8 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 4448cdf 2e5e8b0 6cef15c 2e5e8b0 e660d9c e979a7b 2e5e8b0 e979a7b 2e5e8b0 a024892 2e5e8b0 e660d9c e979a7b 2e5e8b0 e979a7b 2e5e8b0 e979a7b e660d9c 2e5e8b0 e660d9c 2e5e8b0 e660d9c 2e5e8b0 e660d9c 2e5e8b0 8386401 2e5e8b0 e660d9c 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 e660d9c 2e5e8b0 e660d9c 2e5e8b0 17b84ad 2e5e8b0 17b84ad 2e5e8b0 e660d9c 2e5e8b0 e660d9c 17b84ad 31c9e69 b12c580 6cef15c 46ae896 2932016 672a5e4 6cef15c 672a5e4 2e5e8b0 46ae896 2e5e8b0 845ffeb 2e5e8b0 68c1fb8 845ffeb 2e5e8b0 46ae896 7353ed1 1bc28b6 46ae896 acd267e 46ae896 1bc28b6 46ae896 2e5e8b0 1bc28b6 2e5e8b0 46ae896 68c1fb8 46ae896 6cef15c 2e5e8b0 46ae896 6cef15c 2e5e8b0 6cef15c 2932016 2e5e8b0 2932016 2e5e8b0 6639ebf 1bc28b6 2e5e8b0 46ae896 2e5e8b0 68c1fb8 2e5e8b0 2932016 2e5e8b0 2932016 2e5e8b0 b12c580 68c1fb8 2e5e8b0 46ae896 1bc28b6 46ae896 1bc28b6 46ae896 2e5e8b0 46ae896 b12c580 2e5e8b0 2932016 46ae896 6cef15c 672a5e4 6cef15c 672a5e4 6cef15c 2e5e8b0 46ae896 1bc28b6 2e5e8b0 68c1fb8 2e5e8b0 68c1fb8 2e5e8b0 46ae896 1bc28b6 2e5e8b0 49c6001 46ae896 672a5e4 46ae896 2e5e8b0 1bc28b6 2e5e8b0 3f649d8 2e5e8b0 d135896 31c9e69 68c1fb8 2e5e8b0 31c9e69 2e5e8b0 31c9e69 d135896 2e5e8b0 d135896 2e5e8b0 bacf0f5 984af9c bacf0f5 1bc28b6 6cef15c 31c9e69 2e5e8b0 572315f 2e5e8b0 ad18b49 2e5e8b0 ad18b49 2e5e8b0 4448cdf 2e5e8b0 7c678e5 4448cdf 2e5e8b0 8386401 2e5e8b0 7c678e5 4448cdf 2e5e8b0 826de52 2e5e8b0 31c9e69 2e5e8b0 f94f352 2e5e8b0 2932016 672a5e4 2932016 2e5e8b0 f94f352 4448cdf f94f352 1bc28b6 2e5e8b0 672a5e4 1bc28b6 2e5e8b0 68c1fb8 1bc28b6 46ae896 2e5e8b0 6cef15c 1bc28b6 2e5e8b0 6cef15c 2e5e8b0 6cef15c 2e5e8b0 31c9e69 2e5e8b0 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 |
import os
import gradio as gr
import geopandas as gpd
import folium
import requests
# osmnx opsiyonel (fallback için), yoksa sorun değil
try:
import osmnx as ox
except ImportError:
ox = None
from huggingface_hub import InferenceClient, login
# ==========================================
# HF TOKEN & MODEL
# ==========================================
HF_TOKEN = (os.environ.get("HUGGINGFACE_HUB_TOKEN", "") or "").strip()
if HF_TOKEN:
login(token=HF_TOKEN)
else:
print("UYARI: HF_TOKEN / HUGGINGFACE_HUB_TOKEN bulunamadı, gated modellere erişilemeyebilir.")
#client = InferenceClient(
# model="abacusai/Dracarys-72B-Instruct",
# token=HF_TOKEN if HF_TOKEN else None,
#)
# ==========================================
# ÖNCEDEN HAZIRLANMIŞ OSM VERİSİ
# ==========================================
NEIGH_PATH = "data/neighborhoods.geojson"
POIS_PATH = "data/pois.geojson"
if not (os.path.exists(NEIGH_PATH) and os.path.exists(POIS_PATH)):
print("UYARI: data/ klasöründe neighborhoods.geojson / pois.geojson bulunamadı!")
neighborhoods_gdf = None
pois_gdf = None
else:
neighborhoods_gdf = gpd.read_file(NEIGH_PATH)
pois_gdf = gpd.read_file(POIS_PATH)
print(f"{len(neighborhoods_gdf)} mahalle/semt, {len(pois_gdf)} POI yüklendi.")
DEFAULT_TAGS = {
"amenity": ["school", "pharmacy", "hospital", "restaurant", "cafe", "bank"],
"leisure": ["park", "playground"],
"shop": True,
# >>> ULAŞIM ETİKETLERİ
"highway": ["bus_stop"], # otobüs durakları
"railway": ["station", "halt", "tram_stop"], # tren / metro / tramvay istasyonları
"public_transport": ["stop_position", "platform"], # toplu taşıma durak/istasyonları
}
# ==========================================
# OSM / CBS FONKSİYONLARI
# ==========================================
from shapely.geometry import Point
import math
def get_neighborhood_center(city, district, neighborhood):
"""
Verilen şehir/ilçe/mahalle için poligonu alır ve merkez (centroid) koordinatını döndürür.
(lat, lon) şeklinde WGS84 (EPSG:4326) döner.
"""
gdf = get_neighborhood_gdf(city, district, neighborhood)
if gdf is None or len(gdf) == 0:
return None
# Emin olmak için WGS84'e çevir
try:
gdf_ll = gdf.to_crs(epsg=4326)
except Exception:
gdf_ll = gdf # zaten 4326 ise
poly = gdf_ll.geometry.iloc[0]
c = poly.centroid
return c.y, c.x # (lat, lon)
def _build_nearest_poi_query(lat, lon, key, values, radius):
"""
Overpass için: verilen key + values (örn: amenity=university, shop=supermarket)
etrafında en yakın POI'yi arayan sorguyu üretir.
"""
if isinstance(values, str):
values = [values]
filters = []
for v in values:
filters.append(f' node["{key}"="{v}"](around:{radius},{lat},{lon});')
filters.append(f' way["{key}"="{v}"](around:{radius},{lat},{lon});')
filters.append(f' rel["{key}"="{v}"](around:{radius},{lat},{lon});')
filters_str = "\n".join(filters)
query = f"""
[out:json][timeout:25];
(
{filters_str}
);
out center 1;
"""
return query
def _haversine_m(lat1, lon1, lat2, lon2):
"""
İki nokta arasındaki yaklaşık mesafeyi (metre) hesaplar.
"""
R = 6371000 # Dünya yarıçapı (m)
phi1 = math.radians(lat1)
phi2 = math.radians(lat2)
dphi = math.radians(lat2 - lat1)
dlambda = math.radians(lon2 - lon1)
a = math.sin(dphi/2)**2 + math.cos(phi1)*math.cos(phi2)*math.sin(dlambda/2)**2
c = 2 * math.atan2(math.sqrt(a), math.sqrt(1 - a))
return R * c
def find_nearest_poi_overpass(city, district, neighborhood, key, values, radius=5000):
"""
Mahalle merkezinden belirli bir yarıçap içinde (m cinsinden)
verilen OSM tag'ine göre (örn: amenity, shop) en yakın POI'yi arar.
Örnek:
key="amenity", values=["university", "college"]
key="shop", values=["supermarket", "convenience"]
Dönüş:
None -> bulunamadı
dict -> {"name", "lat", "lon", "tag_key", "tag_value", "distance_m", "distance_km"}
"""
center = get_neighborhood_center(city, district, neighborhood)
if center is None:
return None
lat_c, lon_c = center
query = _build_nearest_poi_query(lat_c, lon_c, key, values, radius)
url = "https://overpass-api.de/api/interpreter"
try:
resp = requests.post(url, data={"data": query}, timeout=30)
resp.raise_for_status()
data = resp.json()
except Exception as e:
print("Overpass nearest POI isteği hatası:", e)
return None
elements = data.get("elements", [])
if not elements:
return None
# İlk element en yakın kabul ediliyor
e = elements[0]
tags = e.get("tags", {}) or {}
name = tags.get("name", "(isimsiz POI)")
# node ise lat/lon direk var; way/rel ise center altında
lat_p = e.get("lat") or (e.get("center") or {}).get("lat")
lon_p = e.get("lon") or (e.get("center") or {}).get("lon")
tag_value = None
if isinstance(values, list):
for v in values:
if tags.get(key) == v:
tag_value = v
break
if tag_value is None:
tag_value = tags.get(key)
else:
tag_value = tags.get(key)
dist_m = None
dist_km = None
if lat_p is not None and lon_p is not None:
dist_m = _haversine_m(lat_c, lon_c, lat_p, lon_p)
dist_km = dist_m / 1000.0
return {
"name": name,
"lat": lat_p,
"lon": lon_p,
"tag_key": key,
"tag_value": tag_value,
"distance_m": dist_m,
"distance_km": dist_km,
}
def answer_nearest_poi(city, district, neighborhood, key, values, human_label=None, radius=5000):
"""
Kullanıcıya gösterilecek generic cevap metni.
human_label: 'üniversite', 'market', 'hastane' gibi daha okunaklı isim.
"""
info = find_nearest_poi_overpass(city, district, neighborhood, key, values, radius=radius)
if info is None:
label = human_label or f"{key}={values}"
return f"{city} / {district} / {neighborhood} çevresinde belirli bir yarıçapta {label} bulunamadı."
name = info["name"]
lat = info["lat"]
lon = info["lon"]
dist_km = info["distance_km"]
label = human_label or info.get("tag_value") or f"{key}={values}"
if lat is not None and lon is not None and dist_km is not None:
return (
f"{city} / {district} / {neighborhood} mahallesi merkezine en yakın {label}: "
f"{name} (yaklaşık {dist_km:.2f} km, konum: {lat:.5f}, {lon:.5f})."
)
else:
return (
f"{city} / {district} / {neighborhood} mahallesi merkezine en yakın {label}: {name}."
)
def nearest_poi_wrapper(city, district, neighborhood, poi_type):
"""
Gradio dropdown'dan seçilen poi_type'ı gerçek tag'lere çeviren yardımcı.
"""
if poi_type == "Üniversite":
key = "amenity"
values = ["university", "college"]
label = "üniversite"
elif poi_type == "Market":
key = "shop"
values = ["supermarket", "convenience", "mall", "department_store"]
label = "market"
elif poi_type == "Hastane":
key = "amenity"
values = "hospital"
label = "hastane"
elif poi_type == "Eczane":
key = "amenity"
values = "pharmacy"
label = "eczane"
elif poi_type == "Park":
key = "leisure"
values = "park"
label = "park"
else:
# varsayılan: üniversite
key = "amenity"
values = ["university", "college"]
label = "üniversite"
return answer_nearest_poi(city, district, neighborhood, key, values, human_label=label)
def get_neighborhood_gdf(city: str, district: str, neighborhood: str):
"""
1) Önce önceden kaydedilmiş GeoJSON'dan arar (city + district + neighborhood).
2) Eğer orada yoksa ve osmnx mevcutsa, OSM'den canlı çeker (fallback).
"""
city = (city or "").strip()
district = (district or "").strip()
neighborhood = (neighborhood or "").strip()
# 1) Precomputed veri
if neighborhoods_gdf is not None:
mask = (
(neighborhoods_gdf["city"] == city)
& (neighborhoods_gdf["district"] == district)
& (neighborhoods_gdf["neighborhood"] == neighborhood)
)
gdf_local = neighborhoods_gdf[mask]
if gdf_local is not None and len(gdf_local) > 0:
return gdf_local
# 2) Fallback: canlı OSM çağrısı
if ox is None:
print("OSM fallback kullanılamıyor: osmnx yüklü değil.")
return None
# İlçe bilgisini de geocode sorgusuna ekliyoruz
query = f"{neighborhood}, {district}, {city}, Türkiye"
print(f"OSM fallback: {query}")
try:
gdf_osm = ox.geocode_to_gdf(query)
except Exception as e:
print("Mahalle geocode hatası (OSM fallback):", e)
return None
if gdf_osm is None or len(gdf_osm) == 0:
return None
gdf_osm = gdf_osm.copy()
gdf_osm["city"] = city
gdf_osm["district"] = district
gdf_osm["neighborhood"] = neighborhood
return gdf_osm
def get_pois_within(gdf, tags=None):
"""
1) Eğer GeoJSON'da bu mahalle için POI varsa, oradan döner.
2) Yoksa ve osmnx mevcutsa, poligon üzerinden OSM'den canlı çeker (fallback).
"""
if gdf is None:
return None
if tags is None:
tags = DEFAULT_TAGS
# 1) Precomputed POI verisi
if (
pois_gdf is not None
and all(col in gdf.columns for col in ["city", "district", "neighborhood"])
):
row = gdf.iloc[0]
city = row["city"]
district = row["district"]
neighborhood = row["neighborhood"]
mask = (
(pois_gdf["city"] == city)
& (pois_gdf["district"] == district)
& (pois_gdf["neighborhood"] == neighborhood)
)
pois_local = pois_gdf[mask]
if pois_local is not None and len(pois_local) > 0:
return pois_local
# 2) Fallback: OSM'den canlı POI çek
if ox is None:
print("OSM POI fallback kullanılamıyor: osmnx yüklü değil.")
return None
try:
polygon = gdf.geometry.iloc[0]
pois_osm = ox.features_from_polygon(polygon, tags)
return pois_osm
except Exception as e:
print("POI hatası (OSM fallback):", e)
return None
def summarize_pois(gdf, pois):
summary = {}
try:
area_m2 = gdf.to_crs(epsg=32636).geometry.iloc[0].area
summary["alan_m2"] = float(area_m2)
summary["alan_km2"] = float(area_m2 / 1_000_000)
except Exception as e:
print("Alan hesaplama hatası:", e)
summary["alan_m2"] = None
summary["alan_km2"] = None
if pois is None or len(pois) == 0:
summary["toplam_poi"] = 0
return summary
summary["toplam_poi"] = int(len(pois))
if "amenity" in pois.columns:
amenity_counts = pois["amenity"].value_counts().to_dict()
for k, v in amenity_counts.items():
summary[f"amenity_{k}"] = int(v)
if "leisure" in pois.columns:
leisure_counts = pois["leisure"].value_counts().to_dict()
for k, v in leisure_counts.items():
summary[f"leisure_{k}"] = int(v)
if "shop" in pois.columns:
shop_counts = pois["shop"].value_counts().to_dict()
for k, v in shop_counts.items():
summary[f"shop_{k}"] = int(v)
# >>> ULAŞIM: highway / railway / public_transport
if "highway" in pois.columns:
hw_counts = pois["highway"].value_counts().to_dict()
for k, v in hw_counts.items():
summary[f"highway_{k}"] = int(v)
if "railway" in pois.columns:
rw_counts = pois["railway"].value_counts().to_dict()
for k, v in rw_counts.items():
summary[f"railway_{k}"] = int(v)
if "public_transport" in pois.columns:
pt_counts = pois["public_transport"].value_counts().to_dict()
for k, v in pt_counts.items():
summary[f"public_transport_{k}"] = int(v)
return summary
def build_poi_names_text(pois, max_per_category=15) -> str:
"""
POI GeoDataFrame'inden okul, eczane, kafe, restoran, market vb.
için isim listeleri çıkarır. LLM bağlamında kullanılacak metni döndürür.
"""
if pois is None or len(pois) == 0:
return "Bu mahalle için isim verisi olan POI bulunamadı.\n"
if "name" not in pois.columns:
return "Bu mahallede POI'ler için 'name' alanı bulunamadı.\n"
lines = []
def add_category(title, mask):
sub = pois[mask]
if sub is None or len(sub) == 0:
return
names = (
sub["name"]
.dropna()
.astype(str)
.str.strip()
)
names = [n for n in names if n]
if not names:
return
unique_names = sorted(set(names))[:max_per_category]
lines.append(f"{title}:")
for n in unique_names:
lines.append(f" - {n}")
lines.append("") # kategori arası boş satır
# Kategoriler
if "amenity" in pois.columns:
amenity = pois["amenity"]
add_category("Okullar", amenity == "school")
add_category("Eczaneler", amenity == "pharmacy")
add_category("Kafeler", amenity == "cafe")
add_category("Restoranlar", amenity == "restaurant")
if "shop" in pois.columns:
shop = pois["shop"]
# İstersen buraya başka shop türleri de ekleyebilirsin
add_category("Marketler", shop.isin(["supermarket", "convenience", "mall", "department_store"]))
if not lines:
return "Bu mahallede adı bilinen POI listesi çıkarılamadı.\n"
return "\n".join(lines)
def build_stats_text(summary: dict) -> str:
if not summary:
return "Veri bulunamadı."
alan = summary.get("alan_km2", 0) or 0.0
toplam_poi = summary.get("toplam_poi", 0)
okul = summary.get("amenity_school", 0)
park = summary.get("leisure_park", 0)
eczane = summary.get("amenity_pharmacy", 0)
cafe = summary.get("amenity_cafe", 0)
restoran = summary.get("amenity_restaurant", 0)
# >>> ULAŞIM SAYILARI
otobus_duragi = summary.get("highway_bus_stop", 0)
tren_istasyonu = summary.get("railway_station", 0) + summary.get("railway_halt", 0)
tramvay_duragi = summary.get("railway_tram_stop", 0)
pt_platform = summary.get("public_transport_platform", 0)
lines = [
f"- Tahmini alan: {alan:.2f} km²",
f"- Toplam POI (ilgi noktası): {toplam_poi}",
f"- Okul sayısı: {okul}",
f"- Park sayısı: {park}",
f"- Eczane sayısı: {eczane}",
f"- Kafe sayısı: {cafe}",
f"- Restoran sayısı: {restoran}",
# >>> ULAŞIM SATIRLARI
f"- Otobüs durağı sayısı: {otobus_duragi}",
f"- Tren/metro istasyonu sayısı: {tren_istasyonu}",
f"- Tramvay durağı sayısı: {tramvay_duragi}",
f"- Toplu taşıma platform/istasyon öğesi: {pt_platform}",
]
return "\n".join(lines)
import requests
import json
import geopandas as gpd
from shapely.geometry import shape
from pathlib import Path
def is_osm_query(message: str) -> bool:
return isinstance(message, str) and message.strip().lower().startswith("/osm")
OSM_PROMPT_PATH = Path("osm_query_system.txt") # veya Path("data/osm_query_system.txt")
try:
with OSM_PROMPT_PATH.open("r", encoding="utf-8") as f:
OSM_QUERY_SYSTEM = f.read()
print(f"OSM system prompt '{OSM_PROMPT_PATH}' dosyasından yüklendi.")
except FileNotFoundError:
print(f"UYARI: {OSM_PROMPT_PATH} bulunamadı, kısa yedek prompt kullanılacak.")
OSM_QUERY_SYSTEM = """
You are an expert in OpenStreetMap and Overpass API.
Produce only valid Overpass QL.
No explanations. No markdown.
Use [out:json][timeout:25]; at top.
End with: out center;
"""
def generate_overpass_query_from_llm(prompt: str, model_name: str) -> str:
"""
Doğal dilde verilen prompt'u kullanarak Overpass QL sorgusu üretir.
Sadece geçerli Overpass QL döndürmeye çalışır, markdown vs. temizler.
"""
prompt = (prompt or "").strip()
if not prompt:
prompt = "Generate an Overpass QL query for my request."
client = InferenceClient(model=model_name, token=HF_TOKEN if HF_TOKEN else None)
messages = [
{"role": "system", "content": OSM_QUERY_SYSTEM},
{"role": "user", "content": prompt},
]
# Streaming'e gerek yok, tek seferde alalım
result = client.chat_completion(
messages=messages,
max_tokens=400,
temperature=0.2,
top_p=0.9,
stream=False,
)
# HF InferenceClient sonucu
query_text = result.choices[0].message.content
# Olası ```ql ``` bloklarını temizle
query_text = query_text.replace("```ql", "").replace("```QL", "").replace("```", "")
return query_text.strip()
def normalize_overpass_query(raw_query: str) -> str:
if not raw_query:
return ""
q = raw_query.strip()
if "{{geocodeArea:" in q:
# Kullanıcıyı uyarmak için
raise ValueError("Overpass API, {{geocodeArea:...}} makrosunu desteklemez. "
"Lütfen bunun yerine area[...] filtresi veya numeric area id kullanın.")
# (senin mevcut temizleme kodların)
for token in ("```ql", "```QL", "```", "<code>", "</code>", "<pre>", "</pre>"):
q = q.replace(token, "")
return q.strip()
# ==========================================
# MAHALLE KARŞILAŞTIRMA BAĞLAMI
# ==========================================
def prepare_comparison(city1, district1, neigh1,city2, district2, neigh2):
"""
Butona basıldığında:
- Her iki mahalle için OSM özetini hazırlar
- İki metin döndürür
- LLM için karşılaştırma bağlamını üretir
"""
city1 = (city1 or "").strip()
district1 = (district1 or "").strip()
neigh1 = (neigh1 or "").strip()
city2 = (city2 or "").strip()
district2 = (district2 or "").strip()
neigh2 = (neigh2 or "").strip()
if not city1 or not district1 or not city1 or not neigh1 or not district2 or not neigh2:
msg = "Şehir, ilçe ve iki mahalle de girilmelidir."
return (msg, msg, "", "<b>Harita için yeterli veri yok.</b>")
# Varsayılan olarak boş metinler
labels1 = "Bu mahalle için isim verisi çıkarılamadı."
labels2 = "Bu mahalle için isim verisi çıkarılamadı."
# Mahalle 1
gdf1 = get_neighborhood_gdf(city1, district1, neigh1)
if gdf1 is None or len(gdf1) == 0:
stats1 = f"{city1} / {district1} / {neigh1} için veri bulunamadı."
summary1 = None
else:
pois1 = get_pois_within(gdf1)
summary1 = summarize_pois(gdf1, pois1)
stats1 = build_stats_text(summary1)
# >>> POI isim metni
labels1 = build_poi_names_text(pois1)
# Mahalle 2
gdf2 = get_neighborhood_gdf(city2, district2, neigh2)
if gdf2 is None or len(gdf2) == 0:
stats2 = f"{city2} / {district2} / {neigh2} için veri bulunamadı."
summary2 = None
else:
pois2 = get_pois_within(gdf2)
summary2 = summarize_pois(gdf2, pois2)
stats2 = build_stats_text(summary2)
# >>> POI isim metni
labels2 = build_poi_names_text(pois2)
# LLM bağlamı
compare_context_parts = [
f"Şehir: {city1}",
"",
f"1. Mahalle: {neigh1} (İlçe: {district1})",
stats1,
"",
"1. mahalledeki önemli POI isimleri (okullar, eczaneler, marketler, kafeler vb.):",
labels1,
"",
f"Şehir: {city2}",
"",
f"2. Mahalle: {neigh2} (İlçe: {district2})",
stats2,
"",
"2. mahalledeki önemli POI isimleri (okullar, eczaneler, marketler, kafeler vb.):",
labels2,
"",
"Bu iki mahalleyi alan, toplam POI sayısı, park, okul, kafe, restoran, eczane sayıları"
" ve verilen POI isimleri açısından karşılaştır."
" Kullanıcı soru sorarsa, hem sayısal verilere hem de POI isimlerine dayanarak"
" açıklayıcı ve dengeli bir karşılaştırma yap."
]
compare_context = "\n".join(compare_context_parts)
# Harita HTML'i
map_html = create_comparison_map(gdf1, gdf2)
return stats1, stats2, compare_context, map_html,gdf1, gdf2
def add_poi_markers_to_map(pois, m, layer_prefix="POI"):
"""
POI GeoDataFrame'ini alır, amenity/leisure/shop/railway/highway/public_transport
sütunlarına göre kategorik katmanlar oluşturup haritaya ekler.
"""
if pois is None or len(pois) == 0:
return
# Gerekirse WGS84'e (lat/lon) çevir
try:
if pois.crs is not None and pois.crs.to_epsg() != 4326:
pois = pois.to_crs(epsg=4326)
except Exception:
# CRS yoksa veya hata olursa direkt devam
pass
# >>> ÖNEMLİ: layer_groups burada tanımlanmalı
layer_groups = {} # {kategori_ismi: folium.FeatureGroup}
for _, row in pois.iterrows():
geom = row.geometry
if geom is None or geom.is_empty:
continue
# Nokta değilse centroid al
try:
if geom.geom_type == "Point":
lat, lon = geom.y, geom.x
else:
c = geom.centroid
lat, lon = c.y, c.x
except Exception:
continue
amenity = row.get("amenity")
leisure = row.get("leisure")
shop = row.get("shop")
highway = row.get("highway")
railway = row.get("railway")
public_transport = row.get("public_transport")
# Kategori belirle (öncelik: amenity > leisure > shop > railway > highway > public_transport)
if isinstance(amenity, str):
cat = f"Amenity: {amenity}"
elif isinstance(leisure, str):
cat = f"Leisure: {leisure}"
elif isinstance(shop, str):
cat = f"Shop: {shop}"
elif isinstance(railway, str):
cat = f"Railway: {railway}"
elif isinstance(highway, str):
cat = f"Highway: {highway}"
elif isinstance(public_transport, str):
cat = f"PT: {public_transport}"
else:
cat = "Diğer"
layer_name = f"{layer_prefix} - {cat}"
if layer_name not in layer_groups:
fg = folium.FeatureGroup(name=layer_name, show=True)
fg.add_to(m)
layer_groups[layer_name] = fg
name = row.get("name")
popup_items = []
# Önce isim
if isinstance(name, str) and name.strip():
popup_items.append(name.strip())
if isinstance(amenity, str):
popup_items.append(f"amenity={amenity}")
if isinstance(leisure, str):
popup_items.append(f"leisure={leisure}")
if isinstance(shop, str):
popup_items.append(f"shop={shop}")
if isinstance(railway, str):
popup_items.append(f"railway={railway}")
if isinstance(highway, str):
popup_items.append(f"highway={highway}")
if isinstance(public_transport, str):
popup_items.append(f"public_transport={public_transport}")
popup_text = ", ".join(popup_items) if popup_items else "POI"
folium.CircleMarker(
location=[lat, lon],
radius=4,
popup=popup_text,
tooltip=name.strip() if isinstance(name, str) and name.strip() else None,
weight=1,
fill=True,
fill_opacity=0.7,
).add_to(layer_groups[layer_name])
def create_comparison_map(gdf1, gdf2):
"""
İki mahalle poligonunu tek bir Folium haritasında gösterir.
+ Her mahalle için POI'leri kategorik katmanlar hâlinde ekler.
"""
# Hiç veri yoksa
if (gdf1 is None or len(gdf1) == 0) and (gdf2 is None or len(gdf2) == 0):
return "<b>Harita için yeterli veri yok.</b>"
# Merkez olarak mevcut bir poligonun centroid'ini al
centroid = None
if gdf1 is not None and len(gdf1) > 0:
centroid = gdf1.geometry.iloc[0].centroid
elif gdf2 is not None and len(gdf2) > 0:
centroid = gdf2.geometry.iloc[0].centroid
if centroid is None:
return "<b>Geometri bulunamadı.</b>"
m = folium.Map(location=[centroid.y, centroid.x], zoom_start=14)
# Mahalle 1
pois1 = None
if gdf1 is not None and len(gdf1) > 0:
name1 = str(gdf1.get("neighborhood", ["Mahalle 1"]).iloc[0])
folium.GeoJson(
gdf1.geometry.__geo_interface__,
name=name1,
style_function=lambda feat: {
"color": "red",
"fill": False,
"weight": 3,
},
).add_to(m)
# >>> Mahalle 1 için POI'ler
pois1 = get_pois_within(gdf1)
add_poi_markers_to_map(pois1, m, layer_prefix=f"{name1} POI")
# Mahalle 2
pois2 = None
if gdf2 is not None and len(gdf2) > 0:
name2 = str(gdf2.get("neighborhood", ["Mahalle 2"]).iloc[0])
folium.GeoJson(
gdf2.geometry.__geo_interface__,
name=name2,
style_function=lambda feat: {
"color": "blue",
"fill": False,
"weight": 3,
},
).add_to(m)
# >>> Mahalle 2 için POI'ler
pois2 = get_pois_within(gdf2)
add_poi_markers_to_map(pois2, m, layer_prefix=f"{name2} POI")
folium.LayerControl().add_to(m)
return m._repr_html_()
def run_overpass_to_map(query: str,
previous_elements: list | None,
previous_spatial: str | None,
layer_color: str,
base_gdf1=None,
base_gdf2=None):
layer_name = "Spatial Query"
query = normalize_overpass_query(query) if 'normalize_overpass_query' in globals() else query
if not query or not query.strip():
return (
"<b>Overpass sorgusu boş.</b>",
previous_spatial or "Geçerli bir Overpass sorgusu sağlanmadı.",
previous_elements,
)
url = "https://overpass-api.de/api/interpreter"
try:
resp = requests.post(url, data={"data": query}, timeout=30)
resp.raise_for_status()
data = resp.json()
except Exception as e:
print("Overpass isteği hatası:", e)
try:
print("Overpass response text:", resp.text[:500])
except Exception:
pass
return (
f"<b>Overpass isteği hatası:</b> {e}",
previous_spatial or "Overpass isteğinde hata oluştu, veri yok.",
previous_elements,
)
new_elements = data.get("elements", [])
if not new_elements:
return (
"<b>Overpass sonucu: veri bulunamadı.</b>",
previous_spatial or "Overpass sonucu: hiç element bulunamadı.",
previous_elements,
)
# ---- Eski + yeni elementleri birleştir ----
if not previous_elements:
all_elements = new_elements
else:
all_elements = previous_elements + new_elements
# ---- Merkez bul (tüm elementler üzerinden) ----
center_lat, center_lon = None, None
for el in all_elements:
if "lat" in el and "lon" in el:
center_lat, center_lon = el["lat"], el["lon"]
break
if center_lat is None:
return (
"<b>Overpass sonucu: nokta verisi yok.</b>",
previous_spatial or "Overpass sonucu: nokta verisi bulunamadı.",
all_elements,
)
# ---- Haritayı çiz ----
# Önce merkez: mümkünse mahallelerden, yoksa Overpass noktalardan
if base_gdf1 is not None and len(base_gdf1) > 0:
c = base_gdf1.geometry.iloc[0].centroid
center_lat, center_lon = c.y, c.x
elif base_gdf2 is not None and len(base_gdf2) > 0:
c = base_gdf2.geometry.iloc[0].centroid
center_lat, center_lon = c.y, c.x
# (aksi halde yukarıda all_elements'tan zaten bulduk)
m = folium.Map(location=[center_lat, center_lon], zoom_start=14)
# 1. mahalleyi tekrar çiz
if base_gdf1 is not None and len(base_gdf1) > 0:
name1 = str(base_gdf1.get("neighborhood", ["Mahalle 1"]).iloc[0])
folium.GeoJson(
base_gdf1.geometry.__geo_interface__,
name=name1,
style_function=lambda feat: {
"color": "red",
"fill": False,
"weight": 3,
},
).add_to(m)
pois1 = get_pois_within(base_gdf1)
add_poi_markers_to_map(pois1, m, layer_prefix=f"{name1} POI")
# 2. mahalleyi tekrar çiz
if base_gdf2 is not None and len(base_gdf2) > 0:
name2 = str(base_gdf2.get("neighborhood", ["Mahalle 2"]).iloc[0])
folium.GeoJson(
base_gdf2.geometry.__geo_interface__,
name=name2,
style_function=lambda feat: {
"color": "blue",
"fill": False,
"weight": 3,
},
).add_to(m)
pois2 = get_pois_within(base_gdf2)
add_poi_markers_to_map(pois2, m, layer_prefix=f"{name2} POI")
# Burada tek bir "spatial" katman oluşturuyoruz
fg_nodes = folium.FeatureGroup(
name=f"{layer_name} - Noktalar", # katman adı
show=True
)
fg_nodes.add_to(m)
fg_ways = folium.FeatureGroup(
name=f"{layer_name} - Yollar",
show=True
)
fg_ways.add_to(m)
for el in all_elements:
etype = el.get("type")
tags = el.get("tags", {}) or {}
name = tags.get("name", "")
popup_items = []
if name:
popup_items.append(name)
for k, v in tags.items():
if k != "name":
popup_items.append(f"{k}={v}")
popup_text = "<br>".join(popup_items) if popup_items else etype
if etype == "node" and "lat" in el and "lon" in el:
folium.CircleMarker(
location=[el["lat"], el["lon"]],
radius=4,
popup=popup_text,
tooltip=name or None,
weight=1,
fill=True,
fill_opacity=0.7,
color=layer_color, # 👈 stroke rengi
fill_color=layer_color, # 👈 doldurma rengi
).add_to(fg_nodes)
elif etype == "way" and "geometry" in el:
coords = [(p["lat"], p["lon"]) for p in el["geometry"]]
if len(coords) >= 2:
folium.PolyLine(
locations=coords,
popup=popup_text,
weight=3,
color=layer_color, # 👈 yol rengi
).add_to(fg_ways)
folium.LayerControl().add_to(m)
map_html = m._repr_html_()
# ---- Yeni spatial özet ----
new_summary = summarize_overpass_data({"elements": new_elements})
if previous_spatial:
combined_spatial = previous_spatial + "\n\n--- Yeni Sorgu ---\n" + new_summary
else:
combined_spatial = new_summary
return map_html, combined_spatial, all_elements
def llm_overpass_to_map(natural_prompt: str,
model_name: str,
previous_elements: list | None,
previous_spatial: str | None,
layer_color: str,
base_gdf1=None,
base_gdf2=None):
if not natural_prompt or not natural_prompt.strip():
return (
"Doğal dil sorgu boş.",
"<b>Overpass sonucu: sorgu üretilemedi.</b>",
previous_spatial or "Overpass sonucu: sorgu üretilemedi.",
previous_elements,
)
try:
query = generate_overpass_query_from_llm(natural_prompt, model_name)
except Exception as e:
print("LLM Overpass üretim hatası:", e)
return (
f"LLM Overpass üretim hatası: {e}",
"<b>Overpass sonucu: LLM hatası.</b>",
previous_spatial or "Overpass sonucu: LLM hatası.",
previous_elements,
)
map_html, combined_spatial, all_elements = run_overpass_to_map(
query,
previous_elements,
previous_spatial,
layer_color,
base_gdf1,
base_gdf2,
)
# 4 output: üretilen query, harita, birikmiş spatial metin, birikmiş element listesi
return query, map_html, combined_spatial, all_elements
def summarize_overpass_data(data: dict, max_examples: int = 30) -> str:
"""
Overpass JSON sonucundan LLM'e verilecek metinsel bir özet üretir.
Çok büyük verilerde token patlamaması için sınırlı örnek verir.
"""
if not data:
return "Overpass sonucu boş veya geçersiz.\n"
elements = data.get("elements", [])
if not elements:
return "Overpass sonucu: hiç element bulunamadı.\n"
total_nodes = sum(1 for e in elements if e.get("type") == "node")
total_ways = sum(1 for e in elements if e.get("type") == "way")
total_rel = sum(1 for e in elements if e.get("type") == "relation")
# Basit tag istatistikleri
amenity_counts = {}
leisure_counts = {}
shop_counts = {}
highway_counts = {}
railway_counts = {}
pt_counts = {}
examples = []
for e in elements[:max_examples]:
etype = e.get("type")
tags = e.get("tags", {})
name = tags.get("name", "(isimsiz)")
amenity = tags.get("amenity")
leisure = tags.get("leisure")
shop = tags.get("shop")
highway = tags.get("highway")
railway = tags.get("railway")
pt = tags.get("public_transport")
opening_hours = tags.get("opening_hours")
wheelchair = tags.get("wheelchair")
if amenity:
amenity_counts[amenity] = amenity_counts.get(amenity, 0) + 1
if leisure:
leisure_counts[leisure] = leisure_counts.get(leisure, 0) + 1
if shop:
shop_counts[shop] = shop_counts.get(shop, 0) + 1
if highway:
highway_counts[highway] = highway_counts.get(highway, 0) + 1
if railway:
railway_counts[railway] = railway_counts.get(railway, 0) + 1
if pt:
pt_counts[pt] = pt_counts.get(pt, 0) + 1
# Örnek satır
tag_parts = []
for k in ["amenity", "leisure", "shop", "highway", "railway", "public_transport",
"opening_hours", "wheelchair"]:
v = tags.get(k)
if v:
tag_parts.append(f"{k}={v}")
tag_text = ", ".join(tag_parts) if tag_parts else "etiket yok"
examples.append(f"- {etype} | {name} | {tag_text}")
def dict_to_lines(title, d):
if not d:
return []
items = sorted(d.items(), key=lambda x: -x[1])
lines = [title]
for k, v in items:
lines.append(f" - {k}: {v}")
return lines
lines = [
f"Toplam node sayısı: {total_nodes}",
f"Toplam way sayısı: {total_ways}",
f"Toplam relation sayısı: {total_rel}",
"",
]
lines += dict_to_lines("Amenity türleri:", amenity_counts)
lines += dict_to_lines("Leisure türleri:", leisure_counts)
lines += dict_to_lines("Shop türleri:", shop_counts)
lines += dict_to_lines("Highway türleri:", highway_counts)
lines += dict_to_lines("Railway türleri:", railway_counts)
lines += dict_to_lines("Public transport türleri:", pt_counts)
if examples:
lines.append("")
lines.append(f"İlk {len(examples)} elementten bazı örnekler:")
lines.extend(examples)
return "\n".join(lines)
# ==========================================
# LLM SOHBET FONKSİYONU
# ==========================================
def respond(
message,
history,
model_name,
system_message,
max_tokens,
temperature,
top_p,
compare_context, # Mahalle karşılaştırma
spatial_context, # Overpass sonuçlar
):
# --------------- OSM SPATIAL QUERY MODU ---------------
# --- /osm ile başlayan mesajlar: sadece Overpass sorgusu üret ---
if is_osm_query(message):
user_text = message.lstrip()[4:].strip()
if not user_text:
user_text = "Generate an Overpass QL query for my request."
client = InferenceClient(model=model_name, token=HF_TOKEN if HF_TOKEN else None)
messages = [
{"role": "system", "content": OSM_QUERY_SYSTEM},
{"role": "user", "content": user_text},
]
query_text = ""
for chunk in client.chat_completion(
messages=messages,
max_tokens=400,
stream=True,
temperature=0.2,
top_p=0.9,
):
choices = chunk.choices
token_text = ""
if len(choices) and choices[0].delta.content:
token_text = choices[0].delta.content
query_text += token_text
# Kullanıcıya sadece sorguyu göster
yield f"Üretilen Overpass Sorgusu:\n```ql\n{query_text}\n```"
return
# --------------- NORMAL CHAT AKIŞI ---------------
"""
Streaming chat using Hugging Face Inference API.
history: list of {"role": "...", "content": "..."}
"""
# Temperature güvenli aralık (model max 2)
temperature = max(0.0, min(2.0, float(temperature)))
top_p = max(0.0, min(1.0, float(top_p)))
client = InferenceClient(model=model_name, token=HF_TOKEN if HF_TOKEN else None)
# System mesajına mahalle karşılaştırma bağlamını ekle
full_system = system_message
if compare_context:
full_system += (
"\n\nAşağıda aynı şehirdeki iki mahalleye ait sayısal özetler var.\n"
"Kullanıcı bu mahalleler hakkında soru sorarsa bu bağlama göre cevap ver:\n"
f"{compare_context}"
)
if spatial_context:
full_system += (
"\n\nAyrıca kullanıcı tarafından en son çalıştırılan bir Overpass (spatial) sorgusunun"
" özet sonuçları var. Kullanıcı bu sorgudan gelen veriler hakkında soru sorarsa,"
" bu özet bağlamına dayanarak cevap ver:\n"
f"{spatial_context}"
)
messages = [{"role": "system", "content": full_system}]
messages.extend(history)
messages.append({"role": "user", "content": message})
response = ""
for chunk in client.chat_completion(
messages=messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
choices = chunk.choices
token_text = ""
if len(choices) and choices[0].delta.content:
token_text = choices[0].delta.content
response += token_text
yield response + f"\n\n---\n**Model:** {model_name}"
# ==========================================
# GRADIO ARAYÜZÜ (SOL CHAT, SAĞ KARŞILAŞTIRMA PANELİ)
# ==========================================
with gr.Blocks() as demo:
gr.Markdown("## Mahalle Karşılaştırmalı Chat Botu")
compare_state = gr.State("")
spatial_state = gr.State("") # son Overpass özetleri
overpass_elements_state = gr.State([]) # 👈 tüm Overpass sonuçlarını biriktireceğimiz liste
gdf1_state = gr.State(None) # 1. mahalle geometri
gdf2_state = gr.State(None) # 2. mahalle geometri
with gr.Row():
# SOL SÜTUN: CHAT
with gr.Column(scale=2):
chatbox = gr.Chatbot(height=800, scale=1)
model_dropdown = gr.Dropdown(
choices=[
# Küçük modeller
"google/gemma-2-2b-it", # 2B
"meta-llama/Meta-Llama-3.1-8B-Instruct", # 8B
# Büyük modeller
"abacusai/Dracarys-72B-Instruct", # 72B
"Qwen/Qwen2.5-72B-Instruct", # 72B
# Çok büyük model
"openai/gpt-oss-120b", # 120Bdı
],
label="Model Seç (Bu listedeki modeller Hugging Face Inference API chat_completion ile uyumludur)",
value="google/gemma-2-2b-it"
)
system_box = gr.Textbox(
value="Sen şehir planlama ve mahalleler hakkında bilgi veren yardımsever bir asistansın.",
label="System message",
)
max_tokens_slider = gr.Slider(
minimum=1,
maximum=2048,
value=512,
step=1,
label="Max new tokens",
)
temperature_slider = gr.Slider(
minimum=0.0,
maximum=2.0, # model max 2
value=0.5,
step=0.1,
label="Temperature",
)
top_p_slider = gr.Slider(
minimum=0.1,
maximum=1.0,
value=0.5,
step=0.05,
label="Top-p (nucleus sampling)",
)
chatbot = gr.ChatInterface(
respond,
chatbot=chatbox,
type="messages",
title="Basit Chat Botu",
description="Küçük bir sohbet botu, HF Inference API ve OSM verisi ile çalışıyor.",
additional_inputs=[
model_dropdown,
system_box,
max_tokens_slider,
temperature_slider,
top_p_slider,
compare_state, # >>> mahalle karşılaştırma bağlamı
spatial_state, # >>> son Overpass sonucu özeti
],
)
# SAĞ SÜTUN: MAHALLE KARŞILAŞTIRMA PANELİ
with gr.Column(scale=1):
gr.Markdown("### Mahalle Karşılaştırma")
# 1. mahalle: ilçe + mahalle aynı satırda
with gr.Row():
city_in1 = gr.Textbox(
label="1. Şehir",
value="Ankara",
placeholder="Örn: Ankara",
)
district1_in = gr.Textbox(
label="1. İlçe",
value="Gölbaşı",
scale=1,
placeholder="Örn: Gölbaşı",
)
neigh1_in = gr.Textbox(
label="1. Mahalle",
value="İncek",
scale=2,
placeholder="Örn: İncek Mahallesi",
)
# 2. mahalle: ilçe + mahalle aynı satırda
with gr.Row():
city_in2 = gr.Textbox(
label="2. Şehir",
value="Ankara",
placeholder="Örn: Ankara",
)
district2_in = gr.Textbox(
label="2. İlçe",
value="Gölbaşı",
scale=1,
placeholder="Örn: Gölbaşı",
)
neigh2_in = gr.Textbox(
label="2. Mahalle",
value="Kızılcaşar",
scale=2,
placeholder="Örn: Kızılcaşar Mahallesi",
)
compare_btn = gr.Button("Karşılaştırmayı Hazırla")
with gr.Row():
stats1_box = gr.Textbox(
label="1. Mahalle Özeti",
lines=4,
)
stats2_box = gr.Textbox(
label="2. Mahalle Özeti",
lines=4,
)
map_html = gr.HTML(label="Mahalle Haritası")
gr.Markdown("### Spatial Query (Overpass)")
# 1) LLM'e doğal dil prompt'u
osm_nl_prompt = gr.Textbox(
label="LLM ile Overpass Sorgusu (Doğal Dil)",
lines=3,
placeholder="Örn: İncek ve Kızılcaşar çevresindeki tüm park ve okulları getir"
)
gen_and_run_btn = gr.Button("LLM ile Sorguyu Üret ve Çalıştır")
# 2) Üretilen veya manuel Overpass sorgusu
overpass_box = gr.Textbox(
label="Overpass Sorgusu",
lines=6,
placeholder="Buraya LLM'in ürettiği Overpass QL sorgusunu yapıştırın..."
)
layer_color_dd = gr.Dropdown(
choices=["red", "blue", "green", "orange", "purple"],
value="red",
label="Spatial katman rengi"
)
run_overpass_btn = gr.Button("Sorguyu Çalıştır ve Haritayı Güncelle")
# En yakın POI (generic)
poi_type_dropdown = gr.Dropdown(
choices=["Üniversite", "Market", "Hastane", "Eczane", "Park"],
value="Üniversite",
label="En yakın neyi bulmak istiyorsun?"
)
nearest_poi_btn = gr.Button("1. Mahalle için en yakın noktayı bul")
nearest_poi_box = gr.Textbox(
label="En Yakın POI",
lines=2,
)
nearest_poi_btn.click(
fn=nearest_poi_wrapper,
inputs=[city_in1, district1_in, neigh1_in, poi_type_dropdown],
outputs=[nearest_poi_box],
)
# LLM ile üret + çalıştır (birikimli)
gen_and_run_btn.click(
fn=llm_overpass_to_map,
inputs=[osm_nl_prompt, model_dropdown, overpass_elements_state,
spatial_state, layer_color_dd, gdf1_state, gdf2_state],
outputs=[overpass_box, map_html, spatial_state, overpass_elements_state],
)
# Manuel Overpass çalıştırma (birikimli)
run_overpass_btn.click(
fn=run_overpass_to_map,
inputs=[overpass_box, overpass_elements_state, spatial_state,
layer_color_dd, gdf1_state, gdf2_state],
outputs=[map_html, spatial_state, overpass_elements_state],
)
compare_btn.click(
fn=prepare_comparison,
inputs=[city_in1, district1_in, neigh1_in, city_in1, district2_in, neigh2_in],
outputs=[stats1_box, stats2_box, compare_state, map_html, gdf1_state, gdf2_state],
)
if __name__ == "__main__":
demo.launch()
|