.

help_processor.py

@@ -22,6 +22,7 @@ help_processor.py
 import os
 import re
 import sys
+import html as html_lib
 import json
 import hashlib
 import logging
@@ -164,7 +165,7 @@ class ProcessedSection:
     keywords: str           # "кл1, кл2, кл3"
     text: str
     images_json: str = "[]"        # JSON масив с относителни пътища
-    html_text: str   = ""          # rich HTML (само за HTML-source файлове)
+    html_text: str   = ""          # rich HTML (HTML + DOCX; bold/цветове/картинки)
     char_count: int = 0
 
     def __post_init__(self):
@@ -556,6 +557,12 @@ _HTML_BLOCK_TAGS  = ["h1", "h2", "h3", "h4", "h5", "h6",
 _HTML_PLAIN_NL_TAGS = frozenset({"ul", "ol", "table", "dl", "pre", "blockquote"})
 _HTML_DROP_ATTRS  = ("class", "style", "id", "lang", "dir", "align",
                      "valign", "width", "height", "bgcolor", "border")
+# От style пазим само inline форматиране, нужно за viewer (bold/italic/color).
+_HTML_STYLE_KEEP_RES = (
+    ("color", re.compile(r"(?:^|;)\s*color\s*:\s*([^;]+)", re.I)),
+    ("font-weight", re.compile(r"(?:^|;)\s*font-weight\s*:\s*([^;]+)", re.I)),
+    ("font-style", re.compile(r"(?:^|;)\s*font-style\s*:\s*([^;]+)", re.I)),
+)
 _HTML_HEADING_MAP = {"h1": 1, "h2": 2, "h3": 3, "h4": 3, "h5": 3, "h6": 3}
 _HEADING_TOKEN_RE = re.compile(
     r"^(heading|title|subtitle|заглавие|подзаглавие|наименование|überschrift|msoheading)"
@@ -786,12 +793,42 @@ def _html_block_plain_text(el) -> str:
     return el.get_text(sep, strip=True)
 
 
+def _kept_inline_style(style: str) -> str:
+    """Извлича color / bold / italic от CSS style; останалото се маха."""
+    if not style:
+        return ""
+    parts: list[str] = []
+    for prop, rx in _HTML_STYLE_KEEP_RES:
+        m = rx.search(style)
+        if not m:
+            continue
+        val = m.group(1).strip()
+        if not val:
+            continue
+        low = val.lower()
+        if prop == "font-weight" and low not in (
+            "bold", "bolder", "600", "700", "800", "900"
+        ):
+            continue
+        if prop == "font-style" and "italic" not in low and "oblique" not in low:
+            continue
+        parts.append(f"{prop}:{val}")
+    return ";".join(parts)
+
+
 def _strip_attrs(el):
-    """Премахва decorative атрибути (class, style, on*, data-*)."""
+    """Премахва decorative атрибути; пази color/bold/italic от style и font color=."""
     for t in el.find_all(True):
+        kept_style = ""
+        raw_style = t.attrs.get("style") if t.attrs else None
+        if isinstance(raw_style, str):
+            kept_style = _kept_inline_style(raw_style)
         for a in list(t.attrs):
             if a in _HTML_DROP_ATTRS or a.startswith("on") or a.startswith("data-"):
                 del t[a]
+        # <font color="..."> остава (color не е в DROP)
+        if kept_style:
+            t["style"] = kept_style
 
 
 def _swap_imgs_in_block(el, base_dir: Path, sec_images: list, img_counter: list) -> None:
@@ -975,6 +1012,69 @@ def _extract_docx_paragraph_images(para, doc) -> list[ImageRef]:
     return imgs
 
 
+def _docx_run_color_css(run) -> Optional[str]:
+    """RGB от run.font.color → '#RRGGBB'; theme/auto цветове се пропускат."""
+    try:
+        color = run.font.color
+        if color is None or color.rgb is None:
+            return None
+        return f"#{color.rgb}"
+    except Exception:
+        return None
+
+
+def _docx_run_to_html(run) -> str:
+    """Един Word run → HTML с <b>/<i>/<u> и color span. Спец. символи се escape-ват."""
+    text = run.text or ""
+    if not text:
+        return ""
+    s = html_lib.escape(text, quote=False).replace("\n", "<br>")
+    if run.bold:
+        s = f"<b>{s}</b>"
+    if run.italic:
+        s = f"<i>{s}</i>"
+    if run.underline:
+        s = f"<u>{s}</u>"
+    css_color = _docx_run_color_css(run)
+    if css_color:
+        s = f'<span style="color:{css_color}">{s}</span>'
+    return s
+
+
+def _docx_para_to_html(para) -> str:
+    """Параграф → <p>…</p> с inline форматиране от runs."""
+    parts = [_docx_run_to_html(r) for r in para.runs]
+    inner = "".join(parts)
+    if not inner.strip():
+        return ""
+    return f"<p>{inner}</p>"
+
+
+def _table_to_html(table) -> str:
+    """Таблица → прост HTML <table> (plain клетки, без вложен rich text)."""
+    rows_html: list[str] = []
+    try:
+        rows = table.rows
+    except Exception:
+        return ""
+    for row in rows:
+        cells_html: list[str] = []
+        try:
+            cells = row.cells
+        except Exception:
+            continue
+        for cell in cells:
+            cell_text = " ".join((cell.text or "").split())
+            if not cell_text:
+                continue
+            cells_html.append(f"<td>{html_lib.escape(cell_text, quote=False)}</td>")
+        if cells_html:
+            rows_html.append("<tr>" + "".join(cells_html) + "</tr>")
+    if not rows_html:
+        return ""
+    return "<table>" + "".join(rows_html) + "</table>"
+
+
 def _iter_docx_blocks(doc):
     """Параграфи и таблици в document-order (python-docx .paragraphs пропуска таблиците)."""
     from docx.oxml.ns import qn
@@ -1011,6 +1111,7 @@ def parse_docx(path: Path) -> list[Section]:
     sections: list[Section] = []
     current_title, current_level = "", 1
     buf: list[str] = []
+    buf_html: list[str] = []
     sec_images: list[ImageRef] = []
     img_counter = [0]
     toc_state = _TocSplitState()
@@ -1019,21 +1120,30 @@ def parse_docx(path: Path) -> list[Section]:
         if current_title or buf or sec_images:
             sec = Section(current_title, "\n".join(buf), current_level)
             sec.images = list(sec_images)
+            sec.html_text = "\n".join(buf_html) if buf_html else None
             sections.append(sec)
 
-    def append_para(text: str, para_imgs: list[ImageRef]):
+    def append_para(text: str, para_imgs: list[ImageRef], html_frag: str = ""):
         if text:
             buf.append(text)
+        if html_frag:
+            buf_html.append(html_frag)
         for im in para_imgs:
             img_counter[0] += 1
             im.placeholder = f"img_{img_counter[0]:02d}"
             sec_images.append(im)
-            buf.append(f"[IMG: {im.placeholder}]")
+            ph = f"[IMG: {im.placeholder}]"
+            buf.append(ph)
+            buf_html.append(f"<p>{ph}</p>")
+
+    def append_from_para(para, text: str, para_imgs: list[ImageRef]):
+        html_frag = _docx_para_to_html(para) if text else ""
+        append_para(text, para_imgs, html_frag)
 
     def start_section(title: str, level: int = 1):
-        nonlocal current_title, current_level, buf, sec_images
+        nonlocal current_title, current_level, buf, buf_html, sec_images
         flush()
-        buf, sec_images = [], []
+        buf, buf_html, sec_images = [], [], []
         current_title = title
         current_level = level
 
@@ -1043,6 +1153,9 @@ def parse_docx(path: Path) -> list[Section]:
                 toc_state.leave_phase()
             for line in _table_lines(block):
                 buf.append(line)
+            th = _table_to_html(block)
+            if th:
+                buf_html.append(th)
             continue
 
         para = block
@@ -1071,17 +1184,17 @@ def parse_docx(path: Path) -> list[Section]:
             else:
                 if toc_state.phase:
                     toc_state.leave_phase()
-                append_para(text, para_imgs)
+                append_from_para(para, text, para_imgs)
             continue
 
         if text and _is_toc_heading(text):
             toc_state.note_toc_heading()
-            append_para(text, para_imgs)
+            append_from_para(para, text, para_imgs)
             continue
 
         numbered_action = toc_state.handle_numbered(text) if text else "no"
         if numbered_action == "toc":
-            append_para(text, para_imgs)
+            append_from_para(para, text, para_imgs)
             continue
         if numbered_action == "split":
             start_section(text, level or 1)
@@ -1091,7 +1204,7 @@ def parse_docx(path: Path) -> list[Section]:
         if (level or 0) >= 2 or is_bold_heading:
             if toc_state.phase:
                 toc_state.leave_phase()
-            append_para(text, para_imgs)
+            append_from_para(para, text, para_imgs)
             continue
 
         if level == 1:
@@ -1102,7 +1215,7 @@ def parse_docx(path: Path) -> list[Section]:
 
         if toc_state.phase and text and not _is_numbered_chapter_heading(text):
             toc_state.leave_phase()
-        append_para(text, para_imgs)
+        append_from_para(para, text, para_imgs)
 
     flush()
 
@@ -1521,8 +1634,13 @@ def merge_preamble_sections(sections: list[Section]) -> list[Section]:
 
 
 def clean_text(text: str) -> str:
-    """Collapse spaces/tabs but keep newlines (lists, paragraphs)."""
+    """Collapse spaces/tabs but keep newlines (lists, paragraphs).
+
+    Маха опасни C0 контроли (NUL и др.); пази Unicode символи (•, →, NBSP→space).
+    """
     text = text.replace("\r\n", "\n").replace("\r", "\n")
+    text = re.sub(r"[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]", "", text)
+    text = text.replace("\u00a0", " ")
     text = re.sub(r"[^\S\n]+", " ", text)
     text = re.sub(r"\n{3,}", "\n\n", text)
     text = "\n".join(line.strip() for line in text.split("\n"))