File size: 16,289 Bytes
1b4b6d2 | 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 | """
End-to-end villa-QTO extraction pipeline.
Inputs:
--str path/to/STR.pdf
--arch path/to/ARCH.pdf
--apikey <AI API key> (optional; falls back to GOOGLE_API_KEY env)
Pipeline:
1. Classify every page of both PDFs.
2. Extract STR schedules (footings, tie beams, columns, slab beams).
3. Extract ARCH base data (plot/areas/perimeters/heights).
4. Extract opening schedules (doors + windows).
5. Estimate wall lengths (AI; user can refine in Streamlit).
6. Run sanity checks; print warnings.
7. Build a single _project_data.json that the BOQ engine consumes.
Output:
_project_data.json β single source of truth, replaces hardcoded `base` dict.
Usage:
py _master_extractor.py --str "B:/path/STR.pdf" --arch "B:/path/ARCH.pdf"
"""
import sys, os, json, argparse, subprocess
from typing import Optional
sys.stdout.reconfigure(encoding='utf-8')
sys.path.insert(0, os.path.dirname(__file__))
from _pdf_utils import pdf_page_count
from _page_classifier_ai import classify_and_save, OUT_JSON as CLASS_JSON
from _read_arch_ai import extract_arch, OUT_JSON as ARCH_JSON
from _read_openings_ai import extract_openings, OUT_JSON as OPEN_JSON
from _wall_length_extractor import extract_walls_all_floors, OUT_JSON as WALL_JSON
from _qto_config import (
SCHEDULE_JSON, PROJECT_JSON, SANITY, derive_villa_type, get_logger,
)
log = get_logger("master_extractor")
# ββ Sanity checks [W11] βββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def _in_range(val, key: str) -> bool:
"""Check val against SANITY[key]=(lo,hi)."""
if val is None:
return True
lo, hi = SANITY[key]
return lo <= val <= hi
def sanity_check(data: dict) -> list[str]:
"""Return list of warnings; empty list = all OK."""
warns = []
# ββ Base dimensions βββββββββββββββββββββββββββββββββββββββββββββββββββββββ
for k in ("plot_area", "gf_area", "external_perimeter", "total_villa_height"):
v = data.get(k)
if v is not None and not _in_range(v, k):
lo, hi = SANITY[k]
warns.append(f"{k}={v} outside typical range ({lo}-{hi})")
# ββ Bbox vs gf_area cross-check βββββββββββββββββββββββββββββββββββββββββββ
ll = data.get("longest_length")
lw = data.get("longest_width")
ga = data.get("gf_area")
if ll and lw and ga:
ratio = ga / (ll * lw)
if ratio < 0.5 or ratio > 1.05:
warns.append(
f"gf_area/bbox ratio = {ratio:.2f} (expected 0.5-1.05) β "
f"likely a mis-read of one of: gf_area, longest_length, longest_width"
)
# ββ Foundations ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
footings = data.get("schedules", {}).get("foundation", {}).get("footings", [])
if not footings:
warns.append("No footings extracted β check foundation schedule")
else:
total_v = sum(
(f.get("long_mm", 0) or 0) / 1000
* (f.get("short_mm", 0) or 0) / 1000
* (f.get("depth_mm", 0) or 0) / 1000
* (f.get("count", 0) or 0)
for f in footings
)
if not _in_range(total_v, "foundation_total_m3"):
lo, hi = SANITY["foundation_total_m3"]
warns.append(f"foundation total volume {total_v:.1f} mΒ³ outside {lo}-{hi}")
for f in footings:
d = f.get("depth_mm")
l = f.get("long_mm")
if d and not _in_range(d, "footing_depth_mm"):
lo, hi = SANITY["footing_depth_mm"]
warns.append(f"Footing {f.get('type')}: depth={d}mm outside {lo}-{hi}")
if l and not _in_range(l, "footing_long_mm"):
lo, hi = SANITY["footing_long_mm"]
warns.append(f"Footing {f.get('type')}: long={l}mm outside {lo}-{hi}")
if not f.get("count"):
warns.append(f"Footing {f.get('type')}: count missing or zero")
# ββ Tie beams βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
tbs = data.get("schedules", {}).get("tie_beam", {}).get("tie_beams", [])
for t in tbs:
if t.get("width_mm") and not _in_range(t["width_mm"], "tie_beam_width_mm"):
lo, hi = SANITY["tie_beam_width_mm"]
warns.append(f"TB {t.get('type')}: width={t['width_mm']}mm outside {lo}-{hi}")
if t.get("depth_mm") and not _in_range(t["depth_mm"], "tie_beam_depth_mm"):
lo, hi = SANITY["tie_beam_depth_mm"]
warns.append(f"TB {t.get('type')}: depth={t['depth_mm']}mm outside {lo}-{hi}")
# ββ Columns βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
cols = data.get("schedules", {}).get("column_schedule", {}).get("columns", [])
for c in cols:
if c.get("width_mm") and not _in_range(c["width_mm"], "column_width_mm"):
lo, hi = SANITY["column_width_mm"]
warns.append(f"Column {c.get('mark')}: width={c['width_mm']}mm outside {lo}-{hi}")
if c.get("depth_mm") and not _in_range(c["depth_mm"], "column_depth_mm"):
lo, hi = SANITY["column_depth_mm"]
warns.append(f"Column {c.get('mark')}: depth={c['depth_mm']}mm outside {lo}-{hi}")
# ββ Slab beams (all floors) βββββββββββββββββββββββββββββββββββββββββββββββ
for slab_key in ("slab_1f", "slab_2f", "slab_roof"):
slab = data.get("schedules", {}).get(slab_key, {})
st = slab.get("slab_thickness_mm")
if st and not _in_range(st, "slab_thickness_mm"):
lo, hi = SANITY["slab_thickness_mm"]
warns.append(f"{slab_key}: slab_thickness={st}mm outside {lo}-{hi}")
for b in slab.get("beams", []):
if b.get("width_mm") and not _in_range(b["width_mm"], "beam_width_mm"):
lo, hi = SANITY["beam_width_mm"]
warns.append(f"{slab_key} {b.get('mark')}: width={b['width_mm']}mm outside {lo}-{hi}")
if b.get("depth_mm") and not _in_range(b["depth_mm"], "beam_depth_mm"):
lo, hi = SANITY["beam_depth_mm"]
warns.append(f"{slab_key} {b.get('mark')}: depth={b['depth_mm']}mm outside {lo}-{hi}")
# ββ Openings ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
op_tot = data.get("openings", {}).get("totals", {})
if op_tot.get("door_count", 0) == 0:
warns.append("No doors extracted")
if op_tot.get("window_count", 0) == 0:
warns.append("No windows extracted")
# ββ Walls vs floor areas ββββββββββββββββββββββββββββββββββββββββββββββββββ
walls = data.get("walls", {})
floors = data.get("floors", {})
for fk, w in walls.items():
intl = w.get("internal_total_m")
fa = (floors.get(fk) or {}).get("area")
if intl and fa:
ratio = intl / fa
lo, hi = SANITY["wall_to_area_ratio"]
if not (lo <= ratio <= hi):
warns.append(
f"{fk}: internal_walls/area = {ratio:.2f} m/mΒ² outside {lo}-{hi}"
)
return warns
# ββ Stage runners βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def run_str_schedules(str_pdf: str, api_key: str | None):
"""Run the (already-existing) STR schedule extractor as a subprocess."""
env = os.environ.copy()
env["STR_PDF_PATH"] = str_pdf
if api_key:
env["GOOGLE_API_KEY"] = api_key
cmd = [sys.executable,
os.path.join(os.path.dirname(__file__), "_read_schedules_ai.py")]
if api_key:
cmd.append(api_key)
print(f"\n β invoking _read_schedules_ai.py")
subprocess.run(cmd, env=env, check=False)
# ββ Main pipeline βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def run_pipeline(str_pdf: str, arch_pdf: str, api_key: str | None = None,
skip_schedule: bool = False, skip_walls: bool = False,
skip_classify: bool = False) -> dict:
print("=" * 78)
print(" THE QS HUB MASTER EXTRACTOR")
print("=" * 78)
print(f" STR : {str_pdf} ({pdf_page_count(str_pdf)} pages)")
print(f" ARCH : {arch_pdf} ({pdf_page_count(arch_pdf)} pages)")
# ββ Step 1: classify pages ββββββββββββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 1: Page Classification ββββββββββββββββββββββββββββββββββ")
if skip_classify and os.path.exists(CLASS_JSON):
# Use the classification the user already reviewed/corrected in the UI.
with open(CLASS_JSON, encoding="utf-8") as f:
classification = json.load(f)
print(" (using existing / user-confirmed page classification)")
else:
classification = classify_and_save([str_pdf, arch_pdf], api_key=api_key)
# ββ Step 2: STR schedules βββββββββββββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 2: STR Schedule Extraction ββββββββββββββββββββββββββββββ")
if not skip_schedule:
run_str_schedules(str_pdf, api_key)
schedule_data = {}
if os.path.exists(SCHEDULE_JSON):
with open(SCHEDULE_JSON, encoding='utf-8') as f:
schedule_data = json.load(f)
# ββ Step 3: ARCH base data ββββββββββββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 3: ARCH Base Data βββββββββββββββββββββββββββββββββββββββ")
arch_data = extract_arch(arch_pdf, classification, api_key=api_key)
with open(ARCH_JSON, 'w', encoding='utf-8') as f:
json.dump(arch_data, f, indent=2, ensure_ascii=False)
# ββ Step 4: Openings ββββββββββββββββββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 4: Door & Window Schedules ββββββββββββββββββββββββββββββ")
open_data = extract_openings(arch_pdf, classification, api_key=api_key)
with open(OPEN_JSON, 'w', encoding='utf-8') as f:
json.dump(open_data, f, indent=2, ensure_ascii=False)
# ββ Step 5: Walls βββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
wall_data = {}
if not skip_walls:
print("\nββ STEP 5: Wall Length Estimation βββββββββββββββββββββββββββ")
# Pass floor areas so wall extractor can sanity-check its output [W7]
floor_areas = {
fk: f.get("area") for fk, f in arch_data.get("floors", {}).items()
if f.get("area")
}
wall_data = extract_walls_all_floors(
arch_pdf, classification, api_key=api_key, floor_areas=floor_areas
)
with open(WALL_JSON, 'w', encoding='utf-8') as f:
json.dump(wall_data, f, indent=2, ensure_ascii=False)
# ββ Step 6: Build _project_data.json ββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 6: Build _project_data.json βββββββββββββββββββββββββββββ")
villa_type = derive_villa_type(arch_data.get("structural_levels"))
project = {
# core dimensions
"villa_type": villa_type,
"plot_area": arch_data.get("plot_area"),
"gf_area": arch_data.get("gf_area"),
"external_perimeter": arch_data.get("external_perimeter"),
"ext_perimeter": arch_data.get("external_perimeter"),
"roof_slab_area": arch_data.get("roof_slab_area"),
"roof_perimeter": arch_data.get("roof_perimeter"),
"longest_length": arch_data.get("longest_length"),
"longest_width": arch_data.get("longest_width"),
"total_villa_height": arch_data.get("total_villa_height"),
"structural_levels": arch_data.get("structural_levels"),
"compound_length": arch_data.get("compound_length"),
"parapet_height": arch_data.get("parapet_height") or 1.0,
# default substructure (overrideable in UI)
"excavation_depth": 1.25,
# per-floor block
"floors": arch_data.get("floors", {}),
# source schedules
"schedules": schedule_data,
# openings
"openings": open_data,
# walls
"walls": wall_data,
# metadata
"_meta": {
"str_pdf": str_pdf,
"arch_pdf": arch_pdf,
},
}
# Merge wall data into per-floor block for engine consumption
for fk, w in wall_data.items():
if fk in project["floors"]:
project["floors"][fk]["wall_internal"] = w.get("internal_total_m")
project["floors"][fk]["wall_external"] = w.get("external_total_m")
# ββ Step 7: Sanity check ββββββββββββββββββββββββββββββββββββββββββββββββββ
print("\nββ STEP 7: Sanity Check βββββββββββββββββββββββββββββββββββββββββ")
warns = sanity_check(project)
project["_sanity_warnings"] = warns
if not warns:
print(" β All values within expected ranges.")
else:
print(f" {len(warns)} warnings:")
for w in warns:
print(f" - {w}")
with open(PROJECT_JSON, 'w', encoding='utf-8') as f:
json.dump(project, f, indent=2, ensure_ascii=False)
print(f"\n β Saved β {PROJECT_JSON}")
return project
def cli():
ap = argparse.ArgumentParser(description="THE QS HUB master extractor")
ap.add_argument("--str", required=True, help="path to STR PDF")
ap.add_argument("--arch", required=True, help="path to ARCH PDF")
ap.add_argument("--apikey", help="AI API key (optional)")
ap.add_argument("--skip-schedule", action="store_true",
help="skip the STR schedule pass (use existing _schedule_data.json)")
ap.add_argument("--skip-walls", action="store_true",
help="skip the wall-length AI pass")
args = ap.parse_args()
run_pipeline(args.str, args.arch,
api_key=args.apikey,
skip_schedule=args.skip_schedule,
skip_walls=args.skip_walls)
if __name__ == "__main__":
cli()
|