fix: honor parser params and image VLM system_prompt in Go ingestion (#17334)

## Summary
Fix the Go ingestion pipeline so that several parser setup switches and
the
image VLM prompt are actually honored end-to-end (previously the DSL
fields
existed but the Go code never read them).

- **DOCX** (`docx_parser.go`, `docx_postprocess.go`): read `remove_toc`
and
  `remove_header_footer`; apply to both JSON and markdown output paths
  (outline-based TOC removal with a text-heuristic fallback, plus
  header/footer section filtering).
- **HTML** (`html_parser.go`, `html_postprocess.go`, `text_toc.go`):
read
`remove_header_footer` (pre-parse strip of `<header>`/`<footer>` and
ARIA
`banner`/`contentinfo`) and `remove_toc` (post-parse
`remove_contents_table`
  heuristic).
- **Markdown** (`markdown_parser.go`): read `flatten_media_to_text` and
force
  media blocks to text when enabled.
- **Image VLM** (`media_dispatch.go`): read `system_prompt` instead of
`prompt`
  so the user-configured image VLM prompt is no longer silently dropped
  (`prompt` remains the video family key).

All flags are wired through `ConfigureFromSetup`, which the dispatch
layer
already invokes for every family, so the behavior is live rather than
dead code.

## Test plan
- New unit tests: `docx_postprocess_test.go`, `html_parser_test.go`,
`text_toc_test.go`, `markdown_parser_test.go`, `media_dispatch_test.go`.
- `bash build.sh --test ./internal/parser/parser/...
./internal/ingestion/component/...`

## Notes
- The `File` component is excluded from this migration scope.
- Relates to the Python→Go parity diff (Parser 1.8–1.11, 1.15).

internal/parser/parser/text_toc.go

@@ -0,0 +1,107 @@
+package parser
+
+import (
+	"regexp"
+	"strings"
+)
+
+// tocHeadingPattern matches TOC heading text. Mirrors Python
+// rag/nlp/__init__.py:945 — case-insensitive match against
+// contents/目录/目次/table of contents/致谢/acknowledge after
+// stripping all spaces (ASCII + fullwidth).
+var tocHeadingPattern = regexp.MustCompile(`(?i)^(contents|目录|目次|table of contents|致谢|acknowledge)$`)
+
+// whitespacePattern matches ASCII spaces and fullwidth space (U+3000)
+// for normalization before heading match.
+var whitespacePattern = regexp.MustCompile("[ \t\u3000]+")
+
+// itemText extracts the text field from a parser item map. Falls
+// back to empty string when the field is missing or not a string.
+func itemText(item map[string]any) string {
+	if s, ok := item["text"].(string); ok {
+		return strings.TrimSpace(s)
+	}
+	return ""
+}
+
+// removeContentsTable drops the table-of-contents entries from a list
+// of parser items. Mirrors Python rag/nlp/__init__.py:937-965
+// (remove_contents_table).
+//
+// Algorithm:
+//  1. Scan for a TOC heading (contents/目录/目次/...).
+//  2. Drop the heading.
+//  3. Take the next entry's prefix (first 3 chars for CJK, first 2
+//     words for English) as the TOC entry pattern.
+//  4. Drop that entry, then drop up to 128 following entries that
+//     start with the prefix.
+//  5. Non-matching entries are kept.
+func removeContentsTable(items []map[string]any, eng bool) []map[string]any {
+	i := 0
+	for i < len(items) {
+		text := itemText(items[i])
+		// Strip @@ suffix and whitespace before matching.
+		normalized := whitespacePattern.ReplaceAllString(strings.SplitN(text, "@@", 2)[0], "")
+		if !tocHeadingPattern.MatchString(normalized) {
+			i++
+			continue
+		}
+		// Drop the TOC heading.
+		items = append(items[:i], items[i+1:]...)
+		if i >= len(items) {
+			break
+		}
+		// Determine the prefix from the entry right after the heading.
+		prefix := tocPrefix(itemText(items[i]), eng)
+		for prefix == "" {
+			items = append(items[:i], items[i+1:]...)
+			if i >= len(items) {
+				break
+			}
+			prefix = tocPrefix(itemText(items[i]), eng)
+		}
+		if i >= len(items) || prefix == "" {
+			break
+		}
+		// Drop the first TOC entry (the one that supplied the prefix).
+		items = append(items[:i], items[i+1:]...)
+		if i >= len(items) {
+			break
+		}
+		// Drop up to 128 following entries that start with prefix.
+		limit := i + 128
+		if limit > len(items) {
+			limit = len(items)
+		}
+		for j := i; j < limit; j++ {
+			if !strings.HasPrefix(itemText(items[j]), prefix) {
+				continue
+			}
+			// Drop entries [i, j).
+			items = append(items[:i], items[j:]...)
+			break
+		}
+	}
+	return items
+}
+
+// tocPrefix returns the TOC-entry prefix: first 3 chars for CJK,
+// first 2 whitespace-separated words for English. Mirrors Python
+// remove_contents_table line 951.
+func tocPrefix(text string, eng bool) string {
+	if eng {
+		words := strings.Fields(text)
+		if len(words) >= 2 {
+			return words[0] + " " + words[1]
+		}
+		if len(words) == 1 {
+			return words[0]
+		}
+		return ""
+	}
+	runes := []rune(text)
+	if len(runes) < 3 {
+		return string(runes)
+	}
+	return string(runes[:3])
+}