New style APP (Tab Changes v1.2)

core/formatters.py

@@ -1,9 +1,9 @@
 """
 🇬🇧 Module: formatters.py
-Purpose: Converts LLM Markdown output (comparison/analysis) into styled HTML tables.
+Purpose: Cleanly format LLM Markdown portfolio comparison output into styled HTML.
 
 🇷🇺 Модуль: formatters.py
-Назначение: преобразует Markdown-таблицы и текст отчёта LLM в аккуратное HTML-представление.
+Назначение: форматирует Markdown-ответ LLM в аккуратную HTML-таблицу без дубликатов и пустых столбцов.
 """
 
 import re
@@ -11,66 +11,99 @@ from html import escape
 
 
 def highlight_metrics(text: str) -> str:
-    """Highlight numeric percentage values in green/red."""
+    """Color numeric values depending on sign."""
     def colorize(match):
         val = match.group(0)
         try:
             num = float(val.replace("%", ""))
-            color = "#10b981" if num > 0 else "#ef4444" if num < 0 else "#e6edf3"
+            if num > 0:
+                color = "#10b981"  # green
+            elif num < 0:
+                color = "#ef4444"  # red
+            else:
+                color = "#e6edf3"
             return f"<span style='color:{color};font-weight:600;'>{val}</span>"
         except ValueError:
             return val
+
     return re.sub(r"-?\d+\.?\d*%", colorize, text)
 
 
+def clean_table_lines(lines):
+    """Remove invalid / empty Markdown table lines."""
+    result = []
+    for ln in lines:
+        if not ln.strip():
+            continue
+        if ln.strip().startswith("|") and len(ln.split("|")) > 3:
+            result.append(ln)
+    return result
+
+
 def md_table_to_html(md_table: str) -> str:
-    """Convert a single Markdown table block into HTML."""
-    lines = [ln.strip() for ln in md_table.strip().splitlines() if "|" in ln]
+    """Convert one Markdown table to HTML safely."""
+    lines = clean_table_lines(md_table.strip().splitlines())
     if len(lines) < 2:
         return md_table
 
-    # remove the Markdown alignment line (---|---)
     header_line = lines[0]
-    body_lines = [ln for ln in lines[2:]]
+    data_lines = lines[2:]  # skip the ---|--- line
+    headers = [escape(c.strip()) for c in header_line.split("|")[1:-1] if c.strip()]
 
-    headers = [escape(c.strip()) for c in header_line.split("|")[1:-1]]
-    html = (
+    html = [
         "<table style='width:100%;border-collapse:collapse;"
-        "margin:10px 0;font-size:14px;font-family:Inter,sans-serif;'>"
+        "margin:10px 0;font-size:14px;font-family:Inter,sans-serif;'>",
         "<thead><tr>"
         + "".join(
-            f"<th style='border-bottom:1px solid #30363d;padding:6px;color:#c9d1d9;font-weight:600;text-align:center;'>"
-            f"{h}</th>" for h in headers
+            f"<th style='border-bottom:1px solid #30363d;padding:6px;color:#c9d1d9;"
+            f"font-weight:600;text-align:center;'>{h}</th>"
+            for h in headers
         )
         + "</tr></thead><tbody>"
-    )
-    for ln in body_lines:
-        cols = [highlight_metrics(escape(c.strip())) for c in ln.split("|")[1:-1]]
-        html += "<tr>" + "".join(
-            f"<td style='padding:6px 8px;border-bottom:1px solid #30363d;text-align:center;color:#e6edf3;'>"
-            f"{c}</td>" for c in cols
-        ) + "</tr>"
-    html += "</tbody></table>"
-    return html
+    ]
+
+    for ln in data_lines:
+        cols = [c.strip() for c in ln.split("|")[1:-1] if c.strip()]
+        if len(cols) != len(headers):
+            continue  # skip malformed rows
+        html.append(
+            "<tr>"
+            + "".join(
+                f"<td style='padding:6px 8px;border-bottom:1px solid #30363d;"
+                f"text-align:center;color:#e6edf3;'>"
+                f"{highlight_metrics(escape(c))}</td>"
+                for c in cols
+            )
+            + "</tr>"
+        )
+
+    html.append("</tbody></table>")
+    return "".join(html)
 
 
 def format_comparison_output(raw: str) -> str:
-    """Format LLM Markdown response into rich HTML."""
+    """Transform Markdown to HTML with correct table parsing."""
     if not raw:
         return "<i>No comparison data available.</i>"
 
     text = raw.strip()
 
-    # Section titles
+    # Section headers
     text = re.sub(r"###\s*1️⃣.*", "<h3 style='color:#93c5fd;margin-top:12px;'>1. Summary Table</h3>", text)
     text = re.sub(r"###\s*2️⃣.*", "<h3 style='color:#a78bfa;margin-top:12px;'>2. Comparative Analysis</h3>", text)
     text = re.sub(r"###\s*3️⃣.*", "<h3 style='color:#fbbf24;margin-top:12px;'>3. Recommendation</h3>", text)
 
-    # Extract and replace Markdown tables
-    tables = re.findall(r"(\|[^\n]+\|\n(\|[^\n]+\|\n)+)", text)
-    for tbl, _ in tables:
-        html_table = md_table_to_html(tbl)
-        text = text.replace(tbl, html_table)
+    # Extract Markdown tables only once
+    table_blocks = re.findall(r"((?:\|[^\n]+\n)+)", text)
+    used_blocks = set()
+
+    for tbl in table_blocks:
+        tbl_clean = tbl.strip()
+        if tbl_clean in used_blocks:
+            continue
+        used_blocks.add(tbl_clean)
+        html_table = md_table_to_html(tbl_clean)
+        text = text.replace(tbl_clean, html_table)
 
     text = highlight_metrics(text)
     text = text.replace("**", "")