cd /home/daytona/workspace && pandoc --track-changes=all "/home/daytona/workspac…

ls /home/daytona/workspace/attachments/be8e8ecc-c01e-472d-9ffb-7ddcf45c14df/

ls /home/daytona/skills/docx/scripts/office/ 2>/dev/null && ls /home/daytona/ski…

mkdir -p /home/daytona/workspace/surgery-output && python /home/daytona/skills/d…

~/surgery-output/mark_answers.py

#!/usr/bin/env python3
"""
Mark correct answers in the Surgery.docx by adding green highlight to the correct answer paragraphs.
Strategy: parse document.xml, identify question blocks, find the correct answer paragraph, 
add w:highlight w:val="green" to each run in that paragraph.
"""

import re
import shutil
import subprocess
import os

# Correct answers (1-indexed, matching the answer text fragments)
# Each entry: question number -> substring to match in the correct answer
correct_answers = {
    1: "penetration",
    2: "pyloroduodenal stenosis",
    3: "deep ulcerative niche",
    4: "pneumoperitoneum",
    5: "deep ulcerative niche",
    6: "malignancy of the ulcer",
    7: "hemorrhagic shock",
    8: "pneumoperitoneum",
    9: "perforated ulcer",
    10: "Kloiber",
    11: "atheroma",
    12: "acute pancreatitis",
    13: "pancreatogenic shock",
    14: "omental bursa abscess",
    15: "acute appendicitis",
    16: "appendicular infiltrate",
    17: "abscess formation",
    18: "subungual panaritium",
    19: "paronychia",
    20: "postoperative ventral hernia",
    21: "without ulcers and inflammation",
    22: "Ischemia 2B",
    23: "acute thrombophlebitis of superficial veins",
    24: "chronic external hemorrhoids",
    25: "acute subcutaneous paraproctitis",
    26: "crack of the anus",
    27: "rectal prolapse",
    28: "chronic paraproctitis",
    29: "rectal prolapse, stage III",
    30: "pinched inguinal hernia on the right",
    31: "pinched femoral hernia on the left",
    32: "postoperative ventral hernia",
    33: "unrecoverable umbilical hernia",
    34: "dropsy of the testicle",
    35: "direct inguinal hernia on the right",
    36: "fracture of the radius in a typical location",
    37: "phlegmon of the foot",
    38: "furuncle",
    39: "late adhesive intestinal obstruction",
    40: "retinal hemorrhage",
    41: "FGDS with biopsy",
    42: "FGDS",
    43: "overview radiography of the abdominal cavity",
    44: "fibrogastroduodenoscopy",
    45: "FGDS",
    46: "fibrogastroduodenoscopy",
    47: "call of the JV team, referral to a surgical hospital in an emergency",
    48: "referral to a planned operation in a hospital",
    49: "transfer to the surgical department",
    50: "call of the JV team, referral to the emergency surgery department",
    51: "JV call and referral to the hospital in an emergency",
    52: "Ultrasound of the abdominal cavity",
    53: "Ultrasound of the abdominal cavity",
    54: "JV call and referral to the hospital in an emergency",
    55: "referral for surgical treatment to a hospital as planned",
    56: "referral to emergency surgery",
    57: "overview radiography of the abdominal cavity",
    58: "overview radiography of the abdominal cavity",
    59: "immediate referral to a surgical hospital by ambulance",
    60: "immediate referral to a surgical hospital by ambulance",
    61: "remove the foreign body after 2 times local drip anesthesia with a spear",
    62: "removal of a foreign body after 2-fold local drip anesthesia with a wet swab",
    63: "instillation of an anesthetic solution, removal of a foreign body, washing with water",
    64: "copious washing, instillation of anesthetics",
    65: "apply an aseptic bandage, send it to the eye department",
    66: "urgently send for hospitalization in the eye department",
    67: "apply an aseptic eye patch, send for hospitalization in the eye department",
    68: "primary surgical treatment",
    69: "local instillation anesthesia, removal of a foreign body",
    70: "perform primary surgical treatment of the wound, instill disinfecting drops",
    71: "prescribe diacarb 250-500 mg, 1 ml of 2% promedol",
    72: "lipoma removal",
    73: "opening of a boil in a purulent dressing",
    74: "send for consultation to an oncologist, puncture with biopsy",
    75: "diprospan",
    76: "instrumental and laboratory research in a polyclinic",
    77: "diacarb 250-500 mg, 1 ml of 2% promedol solution once, conjunctivally",
    78: "ear washing with a 1:5000 furacilin solution from a janet syringe",
    79: "infusion of warm oil",
    80: "rinse the ear with a warm solution of furacilin",
}

# Read the document XML
doc_path = "/home/daytona/workspace/surgery-output/unpacked/word/document.xml"
with open(doc_path, "r", encoding="utf-8") as f:
    content = f.read()

# We'll use a Python XML parser approach
from lxml import etree

W = "http://schemas.openxmlformats.org/wordprocessingml/2006/main"

def w(tag):
    return f"{{{W}}}{tag}"

tree = etree.fromstring(content.encode("utf-8"))
body = tree.find(w("body"))

paragraphs = body.findall(".//" + w("p"))

def get_para_text(para):
    texts = []
    for r in para.findall(".//" + w("r")):
        for t in r.findall(w("t")):
            if t.text:
                texts.append(t.text)
    return "".join(texts)

# Build list of (para_index, text) for all paragraphs
para_texts = [(i, get_para_text(p)) for i, p in enumerate(paragraphs)]

# Identify question start paragraphs: paragraphs containing "#N"
import re

question_starts = []
for i, (pi, text) in enumerate(para_texts):
    m = re.match(r'^#(\d+)\s*$', text.strip())
    if m:
        question_starts.append((int(m.group(1)), pi, i))  # (q_num, para_index, list_index)

print(f"Found {len(question_starts)} question headers")

def add_highlight_to_para(para, color="green"):
    """Add highlight to all runs in a paragraph."""
    for r in para.findall(w("r")):
        rpr = r.find(w("rPr"))
        if rpr is None:
            rpr = etree.SubElement(r, w("rPr"))
            r.insert(0, rpr)
        # Check if highlight already exists
        hl = rpr.find(w("highlight"))
        if hl is None:
            hl = etree.SubElement(rpr, w("highlight"))
        hl.set(w("val"), color)
    # Also add shading for better visibility
    ppr = para.find(w("pPr"))
    if ppr is None:
        ppr = etree.SubElement(para, w("pPr"))
        para.insert(0, ppr)
    shd = ppr.find(w("shd"))
    if shd is None:
        shd = etree.SubElement(ppr, w("shd"))
    shd.set(w("val"), "clear")
    shd.set(w("color"), "auto")
    shd.set(w("fill"), "92D050")  # green fill

marked = 0

for q_idx, (q_num, para_idx_start, list_idx_start) in enumerate(question_starts):
    if q_num not in correct_answers:
        continue
    correct_text = correct_answers[q_num].lower()
    
    # Find end of this question (start of next question or end of doc)
    if q_idx + 1 < len(question_starts):
        para_idx_end = question_starts[q_idx + 1][1]
        list_idx_end = question_starts[q_idx + 1][2]
    else:
        para_idx_end = len(paragraphs)
        list_idx_end = len(para_texts)
    
    # Among answer paragraphs for this question, find matching one
    # The structure is: #N / !question text / answer1 / answer2 / ...
    # We look for the answer paragraph whose text matches correct_text
    best_match_para = None
    best_score = 0
    
    for li in range(list_idx_start + 1, list_idx_end):
        pi, text = para_texts[li]
        if not text.strip():
            continue
        # Skip if this is the question text (starts with !)
        if text.strip().startswith("!"):
            continue
        # Check similarity
        text_lower = text.strip().lower()
        # Simple substring match
        if correct_text in text_lower:
            score = len(correct_text)
            if score > best_score:
                best_score = score
                best_match_para = pi
        # Partial match: check how many words of correct_text appear in text
        elif best_match_para is None:
            words = correct_text.split()
            matches = sum(1 for w_ in words if w_ in text_lower)
            if matches > best_score and matches >= max(1, len(words) // 2):
                best_score = matches
                best_match_para = pi
    
    if best_match_para is not None:
        add_highlight_to_para(paragraphs[best_match_para])
        marked += 1
        print(f"Q{q_num}: marked para {best_match_para} -> '{para_texts[best_match_para][1][:80]}'")
    else:
        print(f"Q{q_num}: NO MATCH for '{correct_text}'")

print(f"\nTotal marked: {marked}/{len(correct_answers)}")

# Write back
with open(doc_path, "wb") as f:
    f.write(etree.tostring(tree, xml_declaration=True, encoding="UTF-8", standalone=True))

print("Saved document.xml")

cd /home/daytona/workspace/surgery-output && python mark_answers.py 2>&1

python -c "import xml.etree.ElementTree; import re; print('stdlib ok')" && pytho…

python -c "import xml.etree.ElementTree as ET; print('ET ok')" && node -e "const…

~/surgery-output/mark_answers.py

#!/usr/bin/env python3
"""Mark correct answers in Surgery.docx using stdlib xml.etree.ElementTree"""

import xml.etree.ElementTree as ET
import re, os, shutil

# Correct answers - text fragments to match
correct_answers = {
    1: "penetration",
    2: "pyloroduodenal stenosis",
    3: "deep ulcerative niche",
    4: "pneumoperitoneum",
    5: "deep ulcerative niche",
    6: "malignancy of the ulcer",
    7: "hemorrhagic shock",
    8: "pneumoperitoneum",
    9: "perforated ulcer",
    10: "Kloiber",
    11: "atheroma",
    12: "acute pancreatitis",
    13: "pancreatogenic shock",
    14: "omental bursa abscess",
    15: "acute appendicitis",
    16: "appendicular infiltrate",
    17: "abscess formation",
    18: "subungual panaritium",
    19: "paronychia",
    20: "postoperative ventral hernia",
    21: "without ulcers and inflammation",
    22: "Ischemia 2B",
    23: "acute thrombophlebitis of superficial veins",
    24: "chronic external hemorrhoids",
    25: "acute subcutaneous paraproctitis",
    26: "crack of the anus",
    27: "rectal prolapse",
    28: "chronic paraproctitis",
    29: "rectal prolapse, stage III",
    30: "pinched inguinal hernia on the right",
    31: "pinched femoral hernia on the left",
    32: "postoperative ventral hernia",
    33: "unrecoverable umbilical hernia",
    34: "dropsy of the testicle",
    35: "direct inguinal hernia on the right",
    36: "fracture of the radius in a typical location",
    37: "phlegmon of the foot",
    38: "furuncle",
    39: "late adhesive intestinal obstruction",
    40: "retinal hemorrhage",
    41: "FGDS with biopsy",
    42: "FGDS",
    43: "overview radiography of the abdominal cavity",
    44: "fibrogastroduodenoscopy",
    45: "FGDS",
    46: "fibrogastroduodenoscopy",
    47: "call of the JV team",
    48: "referral to a planned operation in a hospital",
    49: "transfer to the surgical department",
    50: "call of the JV team, referral to the emergency surgery department",
    51: "JV call and referral to the hospital in an emergency",
    52: "Ultrasound of the abdominal cavity",
    53: "Ultrasound of the abdominal cavity",
    54: "JV call and referral to the hospital in an emergency",
    55: "referral for surgical treatment to a hospital as planned",
    56: "referral to emergency surgery",
    57: "overview radiography of the abdominal cavity",
    58: "overview radiography of the abdominal cavity",
    59: "immediate referral to a surgical hospital by ambulance",
    60: "immediate referral to a surgical hospital by ambulance",
    61: "remove the foreign body after 2 times local drip anesthesia with a spear",
    62: "removal of a foreign body after 2-fold local drip anesthesia with a wet swab",
    63: "instillation of an anesthetic solution, removal of a foreign body, washing with water",
    64: "copious washing, instillation of anesthetics",
    65: "apply an aseptic bandage, send it to the eye department",
    66: "urgently send for hospitalization in the eye department",
    67: "apply an aseptic eye patch, send for hospitalization in the eye department",
    68: "primary surgical treatment",
    69: "local instillation anesthesia, removal of a foreign body",
    70: "perform primary surgical treatment of the wound, instill disinfecting drops",
    71: "prescribe diacarb 250-500 mg, 1 ml of 2% promedol",
    72: "lipoma removal",
    73: "opening of a boil in a purulent dressing",
    74: "send for consultation to an oncologist, puncture with biopsy",
    75: "diprospan",
    76: "instrumental and laboratory research in a polyclinic",
    77: "diacarb 250-500 mg, 1 ml of 2% promedol solution once, conjunctivally",
    78: "ear washing with a 1:5000 furacilin solution from a janet syringe",
    79: "infusion of warm oil",
    80: "rinse the ear with a warm solution",
}

doc_path = "/home/daytona/workspace/surgery-output/unpacked/word/document.xml"

# Register namespaces to preserve them
namespaces = {
    'wpc': 'http://schemas.microsoft.com/office/word/2010/wordprocessingCanvas',
    'cx': 'http://schemas.microsoft.com/office/drawing/2014/chartex',
    'cx1': 'http://schemas.microsoft.com/office/drawing/2015/9/8/chartex',
    'cx2': 'http://schemas.microsoft.com/office/drawing/2015/10/21/chartex',
    'cx3': 'http://schemas.microsoft.com/office/drawing/2016/5/9/chartex',
    'cx4': 'http://schemas.microsoft.com/office/drawing/2016/5/10/chartex',
    'cx5': 'http://schemas.microsoft.com/office/drawing/2016/5/11/chartex',
    'cx6': 'http://schemas.microsoft.com/office/drawing/2016/5/12/chartex',
    'cx7': 'http://schemas.microsoft.com/office/drawing/2016/5/13/chartex',
    'cx8': 'http://schemas.microsoft.com/office/drawing/2016/5/14/chartex',
    'mc': 'http://schemas.openxmlformats.org/markup-compatibility/2006',
    'aink': 'http://schemas.microsoft.com/office/drawing/2016/ink',
    'am3d': 'http://schemas.microsoft.com/office/drawing/2017/model3d',
    'o': 'urn:schemas-microsoft-com:office:office',
    'r': 'http://schemas.openxmlformats.org/officeDocument/2006/relationships',
    'm': 'http://schemas.openxmlformats.org/officeDocument/2006/math',
    'v': 'urn:schemas-microsoft-com:vml',
    'wp14': 'http://schemas.microsoft.com/office/word/2010/wordprocessingDrawing',
    'wp': 'http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing',
    'w10': 'urn:schemas-microsoft-com:office:word',
    'w': 'http://schemas.openxmlformats.org/wordprocessingml/2006/main',
    'w14': 'http://schemas.microsoft.com/office/word/2010/wordml',
    'w15': 'http://schemas.microsoft.com/office/word/2012/wordml',
    'w16cex': 'http://schemas.microsoft.com/office/word/2018/wordml/cex',
    'w16cid': 'http://schemas.microsoft.com/office/word/2016/wordml/cid',
    'w16': 'http://schemas.microsoft.com/office/word/2018/wordml',
    'w16sdtdh': 'http://schemas.microsoft.com/office/word/2020/wordml/sdtdatahash',
    'w16se': 'http://schemas.microsoft.com/office/word/2015/wordml/symex',
    'wpg': 'http://schemas.microsoft.com/office/word/2010/wordprocessingGroup',
    'wpi': 'http://schemas.microsoft.com/office/word/2010/wordprocessingInk',
    'wne': 'http://schemas.microsoft.com/office/word/2006/wordml',
    'wps': 'http://schemas.microsoft.com/office/word/2010/wordprocessingShape',
}
for prefix, uri in namespaces.items():
    ET.register_namespace(prefix, uri)

W = "http://schemas.openxmlformats.org/wordprocessingml/2006/main"

def wtag(name):
    return f"{{{W}}}{name}"

with open(doc_path, "r", encoding="utf-8") as f:
    raw = f.read()

tree = ET.fromstring(raw.encode("utf-8"))
body = tree.find(wtag("body"))
paragraphs = list(body.iter(wtag("p")))

def get_para_text(para):
    parts = []
    for t in para.iter(wtag("t")):
        if t.text:
            parts.append(t.text)
    return "".join(parts)

para_texts = [get_para_text(p) for p in paragraphs]

# Find question starts (#N)
question_starts = []
for i, text in enumerate(para_texts):
    m = re.match(r'^#(\d+)\s*$', text.strip())
    if m:
        question_starts.append((int(m.group(1)), i))

print(f"Found {len(question_starts)} questions")

def highlight_para(para):
    """Add green shading to paragraph background via pPr/shd."""
    pPr = para.find(wtag("pPr"))
    if pPr is None:
        pPr = ET.Element(wtag("pPr"))
        para.insert(0, pPr)
    # Add/update shd
    shd = pPr.find(wtag("shd"))
    if shd is None:
        shd = ET.SubElement(pPr, wtag("shd"))
    shd.set(wtag("val"), "clear")
    shd.set(wtag("color"), "auto")
    shd.set(wtag("fill"), "00B050")
    # Also highlight each run
    for run in para.findall(wtag("r")):
        rPr = run.find(wtag("rPr"))
        if rPr is None:
            rPr = ET.Element(wtag("rPr"))
            run.insert(0, rPr)
        hl = rPr.find(wtag("highlight"))
        if hl is None:
            hl = ET.SubElement(rPr, wtag("highlight"))
        hl.set(wtag("val"), "green")

marked = 0
unmatched = []

for q_idx, (q_num, q_para_idx) in enumerate(question_starts):
    if q_num not in correct_answers:
        continue
    correct_fragment = correct_answers[q_num].lower()
    
    # range of paragraphs for this question
    if q_idx + 1 < len(question_starts):
        end_idx = question_starts[q_idx + 1][1]
    else:
        end_idx = len(paragraphs)
    
    best_para_idx = None
    best_score = -1
    
    for pi in range(q_para_idx + 1, end_idx):
        text = para_texts[pi].strip()
        if not text:
            continue
        if text.startswith("!"):
            continue  # skip question stem
        text_lower = text.lower()
        
        if correct_fragment in text_lower:
            score = len(correct_fragment)
            if score > best_score:
                best_score = score
                best_para_idx = pi
    
    if best_para_idx is None:
        # try word-by-word
        words = correct_fragment.split()
        for pi in range(q_para_idx + 1, end_idx):
            text = para_texts[pi].strip().lower()
            if not text or text.startswith("!"):
                continue
            score = sum(1 for w_ in words if w_ in text)
            if score > best_score:
                best_score = score
                best_para_idx = pi
    
    if best_para_idx is not None and best_score > 0:
        highlight_para(paragraphs[best_para_idx])
        marked += 1
        print(f"  Q{q_num:2d}: '{para_texts[best_para_idx][:80]}'")
    else:
        unmatched.append(q_num)
        print(f"  Q{q_num:2d}: NO MATCH for '{correct_fragment}'")

print(f"\nMarked {marked}, Unmatched: {unmatched}")

# Serialize back
ET.indent(tree, space="  ")
out_xml = ET.tostring(tree, encoding="unicode", xml_declaration=False)
out_xml = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>\n' + out_xml

with open(doc_path, "w", encoding="utf-8") as f:
    f.write(out_xml)

print("Saved document.xml")

~/surgery-output/mark_answers.py

Error: String not found in file: '#!/usr/bin/env python3
"""Mark correct answers in Surgery.docx using stdlib xml.etree.ElementTree"""'

cat > /home/daytona/workspace/surgery-output/mark2.py << 'PYEOF' #!/usr/bin/env …