Delete app.py

app.py

@@ -1,124 +0,0 @@
-
-import fitz
-import re
-import gradio as gr
-import pandas as pd
-import tempfile
-import pytesseract
-from PIL import Image, ImageEnhance, ImageFilter
-import io
-
-def classificar(nome, valor):
-    faixas = {
-        "HB": (12, 17), "HT": (36, 50), "GLI": (70, 99), "UREIA": (10, 50),
-        "CR": (0.6, 1.3), "K+": (3.5, 5.5), "NA+": (135, 145), "TGO": (0, 40),
-        "TGP": (0, 40), "ALB": (3.5, 5.0), "INR": (0.8, 1.2), "TAP": (10, 14),
-        "TTP": (25, 35), "LAC": (0.5, 2.2), "PLT": (150, 450), "LEUCO": (4, 11),
-        "CKMB": (0, 24), "CPK": (0, 190), "TROPO": (0, 0.04), "AMIL": (28, 100),
-        "LIP": (0, 60), "PCR": (0, 1), "ÁC UR": (3.5, 7.2), "FAL": (44, 147),
-        "GGT": (0, 38), "FÓS": (2.5, 4.5), "MG++": (1.6, 2.6), "CA TOTAL": (8.6, 10.2),
-        "CAI": (1.1, 1.35), "BT": (0.2, 1.2), "BD": (0, 0.4), "BI": (0.1, 0.8), "CL-": (96, 106)
-    }
-    try:
-        val = float(valor.replace("K", "").replace(">", "").replace("<", "").strip())
-        if nome in faixas:
-            min_v, max_v = faixas[nome]
-            if val < min_v:
-                return f"{valor} ↓"
-            elif val > max_v:
-                return f"{valor} ↑"
-        return valor
-    except:
-        return valor
-
-def melhorar_imagem(img):
-    img = img.convert('L')
-    img = ImageEnhance.Contrast(img).enhance(2)
-    img = img.filter(ImageFilter.SHARPEN)
-    return img
-
-def limpar_texto(texto):
-    texto = re.sub(r'\b([A-Z])\s+([A-Z])\b', r'\1\2', texto)  # Junta siglas com espaços (ex: "C P K" → "CPK")
-    return re.sub(r'\s+', ' ', texto)
-
-def extrair_texto_pdf(pdf_file):
-    texto_fitz = ""
-    imagens_ocr = []
-    with fitz.open(pdf_file.name) as doc:
-        for page in doc:
-            texto_fitz += page.get_text()
-            pix = page.get_pixmap(dpi=400)
-            img = Image.open(io.BytesIO(pix.tobytes("png")))
-            imagens_ocr.append(melhorar_imagem(img))
-    texto_fitz = limpar_texto(texto_fitz)
-    texto_ocr = limpar_texto(" ".join(pytesseract.image_to_string(im) for im in imagens_ocr))
-    return texto_fitz, texto_ocr
-
-def buscar_exame(textos, padrao):
-    for texto in textos:
-        match = re.search(padrao, texto, re.IGNORECASE)
-        if match:
-            return match.group(1).replace(",", ".").strip()
-    return None
-
-def extrair_exames_formatado(pdf_file):
-    if not pdf_file:
-        return "Nenhum arquivo enviado.", None
-
-    texto_fitz, texto_ocr = extrair_texto_pdf(pdf_file)
-    textos = [texto_fitz, texto_ocr]  # Sempre considerar ambos
-
-    exames = {
-        "AMIL": r"amilase[^\d]{0,10}([\d.,]+)",
-        "ÁC UR": r"ácido[\s]?úrico[^\d]{0,10}([\d.,]+)",
-        "BT": r"bilirrubina total|bt[^\d]{0,10}([\d.,]+)",
-        "BD": r"bilirrubina direta|bd[^\d]{0,10}([\d.,]+)",
-        "BI": r"bilirrubina indireta|bi[^\d]{0,10}([\d.,]+)",
-        "CAI": r"cálcio ionizável|cai[^\d]{0,10}([\d.,]+)",
-        "CA TOTAL": r"cálcio total[^\d]{0,10}([\d.,]+)",
-        "CL-": r"cloro[^\d]{0,10}([\d.,]+)",
-        "CR": r"creatinina[^\d]{0,10}([\d.,]+)",
-        "FAL": r"fosfatase alcalina|fal[^\d]{0,10}([\d.,]+)",
-        "FÓS": r"f[óo]sforo[^\d]{0,10}([\d.,]+)",
-        "GGT": r"gama.*?gt|ggt[^\d]{0,10}([\d.,]+)",
-        "GLI": r"glicose[^\d]{0,10}([\d.,]+)",
-        "LIP": r"lipase[^\d]{0,10}([\d.,]+)",
-        "MG++": r"magn[ée]sio[^\d]{0,10}([\d.,]+)",
-        "PCR": r"pcr[^\d]{0,10}([\d.,]+)",
-        "K+": r"pot[áa]ssio[^\d]{0,10}([\d.,]+)",
-        "PTN": r"proteínas totais[^\d]{0,10}([\d.,]+)",
-        "ALB": r"albumina[^\d]{0,10}([\d.,]+)",
-        "GLOB": r"globulina[^\d]{0,10}([\d.,]+)",
-        "RELAÇÃO": r"relação.*?a/g[^\d]{0,10}([\d.,]+)",
-        "NA+": r"s[óo]dio[^\d]{0,10}([\d.,]+)",
-        "TGO": r"tgo[^\d]{0,10}([\d.,]+)",
-        "TGP": r"tgp[^\d]{0,10}([\d.,]+)",
-        "TAP": r"tap[^\d]{0,10}([\d.,]+)",
-        "INR": r"inr[^\d]{0,10}([\d.,]+)",
-        "TTP": r"ttpa[^\d]{0,10}([\d.,]+)",
-        "UREIA": r"ureia[^\d]{0,10}([\d.,]+)",
-        "LAC": r"lactato[^\d]{0,10}([\d.,]+)",
-        "LEUCO": r"leuc[óo]citos[^\d]{0,10}([\d.,]+)",
-        "HB": r"hemoglobina[^\d]{0,10}([\d.,]+)",
-        "HT": r"hemat[óo]crito[^\d]{0,10}([\d.,]+)",
-        "PLT": r"plaquetas[^\d]{0,10}([\d.,]+)",
-        "CPK": r"(?:cpk|creatinofosfoquinase)[^\d]{0,10}([\d.,]+)",
-        "CKMB": r"(?:ck[- ]?mb|ckmb massa)[^\d]{0,10}([\d.,]+)",
-        "TROPO": r"(?:troponina)[^\d]{0,10}([<>]?[\d.,]+)"
-    }
-
-    resultados = [(exame, classificar(exame, buscar_exame(textos, padrao) or "—")) for exame, padrao in exames.items()]
-    df = pd.DataFrame(resultados, columns=["Exame", "Valor"])
-    temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".csv")
-    df.to_csv(temp_file.name, index=False)
-    texto_final = "\n".join(f"{e}: {v}" for e, v in resultados)
-    return texto_final, temp_file.name
-
-with gr.Blocks() as demo:
-    pdf_file = gr.File(label="📄 PDF Exames")
-    extract_button = gr.Button("🔍 Extrair")
-    output_text = gr.Textbox(label="📋 Resultados", lines=25)
-    download_button = gr.File(label="📥 CSV")
-    extract_button.click(extrair_exames_formatado, pdf_file, [output_text, download_button])
-
-demo.launch()