# Rebuild the kitchen ingredient table from USDA FoodData Central (SR Legacy, 2018-04). # Every macro and every fat-class fraction is computed from real FDC nutrient rows. import csv, json, os, urllib.request, zipfile # ---------------------------------------------------------------- USDA loader # Downloads and parses the SR Legacy release. No third-party imports, no other # project files: running `python build-ingredients.py` reproduces the table. SR_URL = ('https://fdc.nal.usda.gov/fdc-datasets/' 'FoodData_Central_sr_legacy_food_csv_2018-04.zip') SR_DIR = 'FoodData_Central_sr_legacy_food_csv_2018-04' # nutrient_id -> short name. NOTE 1261/1262 are caprylic/capric (C8/C10); # 1263/1264 are lauric/myristic and must NOT be used for the MCT class. NUTRIENTS = { '1008': 'kcal', '1003': 'protein', '1004': 'fat', '1005': 'carb', '1079': 'fiber', '1258': 'sfa', '1292': 'mufa', '1293': 'pufa', '1278': 'epa', '1272': 'dha', '1280': 'dpa', '1404': 'ala', '1270': 'ala_gen', '1316': 'la', '1269': 'la_gen', '1406': 'aa', '1271': 'aa_gen', '1321': 'gla', '1261': 'c8', '1262': 'c10', } def load_usda(): if not os.path.isdir(SR_DIR): zp = 'sr_legacy.zip' if not os.path.exists(zp): print('downloading', SR_URL) req = urllib.request.Request(SR_URL, headers={'User-Agent': 'Mozilla/5.0'}) try: with urllib.request.urlopen(req, timeout=300) as r, open(zp, 'wb') as f: f.write(r.read()) except Exception as exc: raise SystemExit( 'Could not download the USDA release (%s: %s).\n' 'Download it manually from:\n %s\n' 'then save it next to this script as "%s" and run again.' % (type(exc).__name__, exc, SR_URL, zp)) with zipfile.ZipFile(zp) as z: z.extractall('.') foods, vals, portions = {}, {}, {} with open(os.path.join(SR_DIR, 'food.csv'), encoding='utf-8') as fh: for row in csv.DictReader(fh): foods[row['fdc_id']] = row['description'] with open(os.path.join(SR_DIR, 'food_nutrient.csv'), encoding='utf-8') as fh: for row in csv.DictReader(fh): name = NUTRIENTS.get(row['nutrient_id']) if name: try: vals.setdefault(row['fdc_id'], {})[name] = float(row['amount']) except ValueError: pass with open(os.path.join(SR_DIR, 'food_portion.csv'), encoding='utf-8') as fh: for row in csv.DictReader(fh): try: grams = float(row['gram_weight']) except ValueError: continue portions.setdefault(row['fdc_id'], []).append( (row.get('portion_description') or row.get('modifier', ''), row.get('amount', ''), grams)) return {'foods': foods, 'vals': vals, 'portions': portions} DB = load_usda() U = lambda fdc, **kw: dict(fdc=str(fdc), src='usda', **kw) # my_id -> mapping. g_per: grams per piece for unit 'ct' (USDA portion where available). MAP = { # ---------------- seafood ---------------- 'salmon': U(175167, n='Salmon fillet (farmed Atlantic)', c='sea', u='g'), 'smokedsalmon':U(173687, n='Smoked salmon (chinook)', c='sea', u='g'), 'sardines': U(175139, n='Sardines (92 g tin, in oil, drained)', c='sea', u='ct', g_per=92, pw='USDA portion: 1 can (3.75 oz)', one='tin of sardines', pl='tins of sardines'), # SR Legacy lists only a 361 g can, so the retail tin size is an assumption. 'mackereltin': U(175121, n='Mackerel (125 g tin, jack, drained)', c='sea', u='ct', g_per=125, pw='ASSUMED: typical 125 g retail tin (USDA lists only a 361 g can)', one='tin of mackerel', pl='tins of mackerel'), 'anchovy': U(174183, n='Anchovies (45 g tin, in oil, drained)', c='sea', u='ct', g_per=45, pw='USDA portion: 1 can (2 oz)', one='tin of anchovies', pl='tins of anchovies'), 'shrimp': U(175179, n='Shrimp, peeled, raw', c='sea', u='g'), 'cod': U(171955, n='Cod, Atlantic, raw', c='sea', u='g'), 'mussels': U(174216, n='Mussels, blue, raw', c='sea', u='g'), 'oyster': U(171978, n='Oyster (eastern, wild)', c='sea', u='ct', g_per=14, pw='USDA portion: 6 medium = 84 g', one='oyster', pl='oysters'), # ---------------- meat, poultry & eggs ---------------- 'eggs': U(171287, n='Eggs', c='meat', u='ct', g_per=50, pw='USDA portion: 1 large', one='egg', pl='eggs'), 'chickenthigh':U(172385, n='Chicken thighs (skin-on), raw', c='meat', u='g'), 'chickenbreast':U(171077, n='Chicken breast, skinless, raw', c='meat', u='g'), 'groundbeef': U(174036, n='Ground beef (80/20), raw', c='meat', u='g'), 'steak': U(168726, n='Beef top sirloin steak, raw', c='meat', u='g'), 'liver': U(169451, n='Beef liver, raw', c='meat', u='g'), 'lamb': U(172500, n='Lamb shoulder, raw', c='meat', u='g'), 'porkchop': U(167839, n='Pork loin chop, raw', c='meat', u='g'), 'porkbelly': U(167812, n='Pork belly, raw', c='meat', u='g'), 'bacon': U(168277, n='Bacon, cured, unprepared', c='meat', u='g'), # ---------------- produce ---------------- 'spinach': U(168462, n='Spinach', c='prod', u='g'), 'kale': U(168421, n='Kale', c='prod', u='g'), 'arugula': U(169387, n='Arugula', c='prod', u='g'), 'romaine': U(169247, n='Romaine lettuce', c='prod', u='g'), 'butterlettuce':U(168429, n='Butter lettuce', c='prod', u='g'), 'endive': U(168412, n='Endive', c='prod', u='g'), 'broccoli': U(170379, n='Broccoli', c='prod', u='g'), 'cauliflower': U(169986, n='Cauliflower', c='prod', u='g'), 'cabbage': U(169975, n='Cabbage', c='prod', u='g'), 'brussels': U(170383, n='Brussels sprouts', c='prod', u='g'), 'zucchini': U(169291, n='Zucchini', c='prod', u='g'), 'asparagus': U(168389, n='Asparagus', c='prod', u='g'), 'bellpepper': U(170108, n='Bell pepper, red', c='prod', u='g'), 'mushrooms': U(168434, n='Mushrooms (cremini)', c='prod', u='g'), 'cherrytom': U(170457, n='Tomatoes, red, ripe', c='prod', u='g'), 'cucumber': U(168409, n='Cucumber, with peel', c='prod', u='g'), 'avocado': U(171706, n='Avocado (Hass / California)', c='prod', u='ct', g_per=136, pw='USDA portion: 1 fruit, without skin and seed', one='avocado', pl='avocados'), 'celery': U(169988, n='Celery', c='prod', u='g'), 'greenbeans': U(169961, n='Green beans', c='prod', u='g'), 'radish': U(169276, n='Radishes', c='prod', u='g'), 'scallion': U(170005, n='Scallion', c='prod', u='ct', g_per=15, pw='USDA portion: 1 medium', one='scallion', pl='scallions'), 'onion': U(170000, n='Onion', c='prod', u='ct', g_per=110, pw='USDA portion: 1 medium', one='onion', pl='onions'), 'garlic': U(169230, n='Garlic clove', c='prod', u='ct', g_per=3, pw='USDA portion: 1 clove', one='garlic clove', pl='garlic cloves'), 'ginger': U(169231, n='Fresh ginger', c='prod', u='g'), 'lemon': U(167746, n='Lemon', c='prod', u='ct', g_per=58, pw='USDA portion: 1 fruit (2-1/8 in dia)', one='lemon', pl='lemons'), 'lime': U(168155, n='Lime', c='prod', u='ct', g_per=67, pw='USDA portion: 1 fruit (2 in dia)', one='lime', pl='limes'), 'raspberries': U(167755, n='Raspberries', c='prod', u='g'), 'blackberries':U(173946, n='Blackberries', c='prod', u='g'), 'strawberries':U(167762, n='Strawberries', c='prod', u='g'), 'blueberries': U(171711, n='Blueberries', c='prod', u='g'), 'parsley': U(170416, n='Parsley (bunch)', c='prod', u='ct', g_per=60, pw='ASSUMED bunch = 60 g (no USDA bunch; matches its 1 cup chopped)', one='bunch of parsley', pl='bunches of parsley'), 'cilantro': U(169997, n='Cilantro (bunch)', c='prod', u='ct', g_per=40, pw='ASSUMED bunch = 40 g (no USDA bunch; its 9 sprigs = 20 g)', one='bunch of cilantro', pl='bunches of cilantro'), 'basil': U(172232, n='Fresh basil', c='prod', u='g'), 'rosemary': U(173473, n='Rosemary sprig', c='prod', u='ct', g_per=1.7, pw='USDA portion: 1 tbsp, used as a sprig equivalent', one='rosemary sprig', pl='rosemary sprigs'), 'thyme': U(173470, n='Thyme sprig', c='prod', u='ct', g_per=0.8, pw='USDA portion: 1 tsp, used as a sprig equivalent', one='thyme sprig', pl='thyme sprigs'), 'dill': U(172233, n='Dill (bunch)', c='prod', u='ct', g_per=20, pw='ASSUMED bunch = 20 g (was 9 g = USDA 1 cup of loose sprigs, too light for a retail bunch)', one='bunch of dill', pl='bunches of dill'), # ---------------- dairy ---------------- 'butter': U(173410, n='Butter, salted', c='dairy', u='g'), # FDC's ghee entry reports only SFA and PUFA totals (no MUFA, no individual acids), # which would render ghee as ~100% saturated. Ghee is clarified butter, so take the # fat profile from butter (173410) and keep ghee's own macros. 'ghee': dict(fdc='171314', fp_from='173410', src='composite', n='Ghee (clarified butter)', c='dairy', u='g'), 'cream': U(170859, n='Heavy cream', c='dairy', u='ml', density=1.008, pw='density from USDA portion 1 fl oz = 29.8 g'), 'feta': U(173420, n='Feta', c='dairy', u='g'), 'parmesan': U(170848, n='Parmesan, hard', c='dairy', u='g'), 'cheddar': U(173414, n='Aged cheddar', c='dairy', u='g'), 'goatcheese': U(173433, n='Goat cheese, semisoft', c='dairy', u='g'), 'yogurt': U(171304, n='Greek yogurt (whole milk, plain)', c='dairy', u='g'), 'kefir': U(170904, n='Kefir (lowfat, plain)', c='dairy', u='ml', density=1.027, pw='density from USDA portion 1 cup = 243 g'), 'creamcheese': U(173418, n='Cream cheese', c='dairy', u='g'), # ---------------- nuts & seeds ---------------- 'macadamia': U(170178, n='Macadamias, raw', c='nuts', u='g'), 'walnut': U(170187, n='Walnuts, English', c='nuts', u='g'), 'pecan': U(170182, n='Pecans', c='nuts', u='g'), 'almond': U(170567, n='Almonds', c='nuts', u='g'), 'hazelnut': U(170581, n='Hazelnuts', c='nuts', u='g'), 'brazil': U(170569, n='Brazil nut', c='nuts', u='ct', g_per=5, pw='USDA portion: 1 kernel', one='Brazil nut', pl='Brazil nuts'), 'pumpkinseeds':U(170557, n='Pumpkin seeds, roasted', c='nuts', u='g'), 'hemphearts': U(170148, n='Hemp hearts (hulled)', c='nuts', u='g'), 'chia': U(170554, n='Chia seeds', c='nuts', u='g'), 'flax': U(169414, n='Ground flaxseed', c='nuts', u='g'), 'almondflour': U(170568, n='Almond flour (blanched, ground)', c='nuts', u='g'), 'tahini': U(170189, n='Tahini', c='nuts', u='g'), 'pinenuts': U(170591, n='Pine nuts, dried', c='nuts', u='g'), # ---------------- oils & pantry ---------------- 'evoo': U(171413, n='Extra-virgin olive oil', c='pantry', u='ml', density=0.913, st=1, pw='density from USDA portion 1 cup = 216 g'), 'avooil': U(173573, n='Avocado oil', c='pantry', u='ml', density=0.921, st=1, pw='density from USDA portion 1 cup = 218 g'), 'coconutoil': U(171412, n='Coconut oil', c='pantry', u='g', st=1), 'coconutmilk': U(170173, n='Coconut milk (canned, full-fat)', c='pantry', u='ml', density=0.955, pw='density from USDA portion 1 cup = 226 g'), 'redwinevin': U(172240, n='Red wine vinegar', c='pantry', u='ml', density=1.010, st=1, pw='density from USDA portion 1 cup = 239 g'), # Bones are simmered and discarded, so they carry no "as eaten" macros. They exist # only so the shopping list tells you what to buy; the broth's own nutrition comes # from the finished-broth values on the bonebroth row (see the base recipe's `out`). 'beefbones': dict(src='label', n='Beef marrow / knuckle bones', c='meat', u='g', per100=[0, 0, 0, 0], fp=None, shoponly=1), 'chickencarcass': dict(src='label', n='Chicken carcass or wings (for broth)', c='meat', u='g', per100=[0, 0, 0, 0], fp=None, shoponly=1), 'acv': U(173469, n='Apple cider vinegar', c='pantry', u='ml', density=1.010, st=1, pw='density from USDA portion 1 cup = 239 g'), 'dijon': U(172234, n='Dijon mustard', c='pantry', u='g', st=1), 'darkchoc': U(170273, n='Dark chocolate (70–85 % cacao)', c='pantry', u='g'), # composite: mayonnaise macros from FDC 171009, fat profile from avocado oil FDC 173573 'mayo': dict(fdc='171009', fp_from='173573', src='composite', n='Mayonnaise (avocado oil)', c='pantry', u='g'), # no SR Legacy equivalent -> typical manufacturer label values, flagged as such 'mctoil': dict(src='label', n='MCT oil', c='pantry', u='ml', st=1, per100=[830, 0, 0, 93], fp={'mct': 1.0}), 'aminos': dict(src='label', n='Coconut aminos', c='pantry', u='ml', st=1, per100=[100, 20, 0, 0], fp=None), 'bonebroth': dict(src='label', n='Bone broth', c='pantry', u='ml', per100=[18, 0, 4.0, 0.2], fp=None), # ---------------- fermented & jarred ---------------- 'sauerkraut': U(169279, n='Sauerkraut', c='ferm', u='g'), 'kimchi': U(170392, n='Kimchi', c='ferm', u='g'), 'olives': U(169096, n='Olives, green', c='ferm', u='g'), 'capers': U(172238, n='Capers', c='ferm', u='g'), # ---------------- herbs, spices & salt ---------------- 'salt': U(173468, n='Sea salt', c='spice', u='g', st=1), 'pepper': U(170931, n='Black pepper', c='spice', u='g', st=1), 'paprika': U(171329, n='Paprika', c='spice', u='g', st=1), 'chiliflakes': U(170932, n='Chili flakes', c='spice', u='g', st=1), 'oregano': U(171328, n='Dried oregano', c='spice', u='g', st=1), 'cumin': U(170923, n='Ground cumin', c='spice', u='g', st=1), 'turmeric': U(172231, n='Ground turmeric', c='spice', u='g', st=1), 'cinnamon': U(171320, n='Cinnamon', c='spice', u='g', st=1), } def g(v, *keys): """first present nutrient among keys, else 0""" for k in keys: if k in v: return v[k] return 0.0 def fat_profile(v, total_fat): """Fat-class fractions computed from real USDA fatty-acid rows. mct = C8+C10 ; s = total SFA - mct ; m = total MUFA ; p3m = EPA+DHA+DPA ; p3p = ALA ; p6 = LA + AA + GLA.""" if total_fat < 0.5: return None mct = g(v,'c8') + g(v,'c10') s = max(0.0, g(v,'sfa') - mct) m = g(v,'mufa') p3m = g(v,'epa') + g(v,'dha') + g(v,'dpa') p3p = g(v,'ala','ala_gen') p6 = g(v,'la','la_gen') + g(v,'aa','aa_gen') + g(v,'gla') out = {} for k, grams in (('m',m),('p3m',p3m),('p3p',p3p),('p6',p6),('s',s),('mct',mct)): frac = grams / total_fat if frac >= 0.005: out[k] = round(frac, 3) return out or None rows, report, missing = [], [], [] for key, mp in MAP.items(): if mp['src'] == 'label': kcal, netc, prot, fat = mp['per100'] fp = mp.get('fp'); fdc = None; desc = '(manufacturer label, typical)' else: fdc = mp['fdc'] v = DB['vals'].get(fdc) if not v: missing.append((key, fdc)); continue desc = DB['foods'][fdc] kcal = g(v,'kcal'); prot = g(v,'protein'); fat = g(v,'fat') netc = max(0.0, g(v,'carb') - g(v,'fiber')) src_v = DB['vals'].get(mp['fp_from'], v) if mp.get('fp_from') else v fat_for_profile = g(src_v,'fat') if mp.get('fp_from') else fat fp = fat_profile(src_v, fat_for_profile) dens = mp.get('density', 1.0) scale = (mp['g_per'] if mp['u'] == 'ct' else 100.0 * (dens if mp['u'] == 'ml' else 1.0)) / 100.0 m4 = [round(kcal*scale,1), round(netc*scale,2), round(prot*scale,2), round(fat*scale,2)] rows.append((key, mp, m4, fp, fdc, desc)) report.append(dict(key=key, fdc=fdc, desc=desc, unit=mp['u'], portion_note=mp.get('pw'), per=('per %g g piece' % mp['g_per']) if mp['u']=='ct' else ('per 100 %s' % mp['u']), macros=m4, fp=fp, src=mp['src'], shoponly=mp.get('shoponly'))) print("mapped:", len(rows), " missing:", missing) json.dump(report, open('ing_report.json','w'), indent=1) # ---- emit JS ---- def js_num(x): s = ('%g' % x) return s lines = [] CATNAME = {'sea':'seafood','meat':'meat, poultry & eggs','prod':'produce','dairy':'dairy', 'nuts':'nuts & seeds','pantry':'oils & pantry','ferm':'fermented & jarred','spice':'herbs, spices & salt'} cur = None for key, mp, m4, fp, fdc, desc in rows: if mp['c'] != cur: cur = mp['c']; lines.append(' /* %s */' % CATNAME[cur]) parts = ["n: %s" % json.dumps(mp['n']), "c: %s" % json.dumps(mp['c']), "u: %s" % json.dumps(mp['u'])] parts.append("m: [%s]" % ', '.join(js_num(x) for x in m4)) if fp: parts.append("fp: { %s }" % ', '.join('%s: %s' % (k, js_num(v)) for k, v in fp.items())) if mp.get('st'): parts.append("st: 1") if mp.get('shoponly'): parts.append("shoponly: 1") if mp['u'] == 'ct': parts.append("one: %s" % json.dumps(mp['one'])); parts.append("pl: %s" % json.dumps(mp['pl'])) if fdc: parts.append("fdc: %s" % json.dumps(fdc)) else: parts.append("fdc: null") lines.append(" %-14s { %s }," % (key + ':', ', '.join(parts))) open('ing_block.js','w',encoding='utf-8').write('\n'.join(lines)) print("wrote ing_block.js (%d lines)" % len(lines))