Initial commit: svg backend

backend/app/services/svg_display_processor.py

@@ -0,0 +1,187 @@
+from __future__ import annotations
+
+import logging
+import re
+from typing import Any
+
+from lxml import etree
+
+from app.core.config import settings
+
+
+logger = logging.getLogger(__name__)
+
+ALLOWED_MODES = {"passthrough", "optimized"}
+
+
+def _parse_length(value: str | None) -> float | None:
+    if not value:
+        return None
+    cleaned = value.strip().replace("px", "")
+    try:
+        return float(cleaned)
+    except ValueError:
+        return None
+
+
+def _local_name(tag: str) -> str:
+    if "}" in tag:
+        return tag.split("}", 1)[1]
+    return tag
+
+
+def _is_hidden(node: etree._Element) -> bool:
+    display = (node.attrib.get("display") or "").strip().lower()
+    visibility = (node.attrib.get("visibility") or "").strip().lower()
+    style = (node.attrib.get("style") or "").replace(" ", "").lower()
+    return (
+        display == "none"
+        or visibility == "hidden"
+        or "display:none" in style
+        or "visibility:hidden" in style
+    )
+
+
+def _is_seat_related(node: etree._Element) -> bool:
+    probe = " ".join(
+        [
+            node.attrib.get("id", ""),
+            node.attrib.get("class", ""),
+            node.attrib.get("data-seat-id", ""),
+            node.attrib.get("data-sector-id", ""),
+            node.attrib.get("data-group-id", ""),
+        ]
+    ).lower()
+    return any(token in probe for token in ["seat", "sector", "group", "place"])
+
+
+def _font_size(node: etree._Element) -> float | None:
+    direct = _parse_length(node.attrib.get("font-size"))
+    if direct is not None:
+        return direct
+
+    style = node.attrib.get("style") or ""
+    match = re.search(r"font-size\s*:\s*([0-9.]+)", style, flags=re.IGNORECASE)
+    if match:
+        return _parse_length(match.group(1))
+    return None
+
+
+def _is_technical_text(node: etree._Element) -> bool:
+    patterns = [
+        item.strip().lower()
+        for item in settings.svg_display_technical_text_patterns.split(",")
+        if item.strip()
+    ]
+    haystack = " ".join(
+        [
+            node.attrib.get("id", ""),
+            node.attrib.get("class", ""),
+            "".join(node.itertext()),
+        ]
+    ).lower()
+    return any(pattern in haystack for pattern in patterns)
+
+
+def _force_viewbox(root: etree._Element) -> None:
+    if not settings.svg_display_force_viewbox:
+        return
+    if root.attrib.get("viewBox"):
+        return
+
+    width = _parse_length(root.attrib.get("width"))
+    height = _parse_length(root.attrib.get("height"))
+    if width and height:
+        w = int(width) if width.is_integer() else width
+        h = int(height) if height.is_integer() else height
+        root.attrib["viewBox"] = f"0 0 {w} {h}"
+
+
+def _extract_meta(root: etree._Element) -> dict[str, Any]:
+    return {
+        "view_box": root.attrib.get("viewBox"),
+        "width": root.attrib.get("width"),
+        "height": root.attrib.get("height"),
+    }
+
+
+def generate_display_svg(content: bytes, mode: str) -> tuple[bytes, dict[str, Any]]:
+    if mode not in ALLOWED_MODES:
+        raise ValueError(f"Unsupported display mode: {mode}")
+
+    parser = etree.XMLParser(
+        resolve_entities=False,
+        remove_blank_text=False,
+        remove_comments=False,
+        no_network=True,
+        recover=False,
+        huge_tree=True,
+    )
+    root = etree.fromstring(content, parser=parser)
+
+    defs_count = len(root.xpath("//*[local-name()='defs']"))
+    use_count = len(root.xpath("//*[local-name()='use']"))
+    style_count = len(root.xpath("//*[local-name()='style']"))
+    clip_count = len(root.xpath("//*[local-name()='clipPath']"))
+
+    logger.info(
+        "display_svg.generate mode=%s size_bytes=%s has_style=%s defs=%s use=%s clipPath=%s",
+        mode,
+        len(content),
+        bool(style_count),
+        defs_count,
+        use_count,
+        clip_count,
+    )
+
+    removed_hidden_count = 0
+    removed_small_text_count = 0
+    removed_technical_text_count = 0
+
+    if mode == "optimized":
+        for node in list(root.iter()):
+            tag_name = _local_name(node.tag)
+
+            if settings.svg_display_remove_hidden_elements and not _is_seat_related(node) and _is_hidden(node):
+                parent = node.getparent()
+                if parent is not None:
+                    parent.remove(node)
+                    removed_hidden_count += 1
+                continue
+
+            if tag_name in {"text", "tspan"}:
+                if settings.svg_display_hide_small_text and not _is_seat_related(node):
+                    size = _font_size(node)
+                    if size is not None and size < settings.svg_display_min_text_font_size:
+                        parent = node.getparent()
+                        if parent is not None:
+                            parent.remove(node)
+                            removed_small_text_count += 1
+                        continue
+
+                if settings.svg_display_hide_technical_text and not _is_seat_related(node) and _is_technical_text(node):
+                    parent = node.getparent()
+                    if parent is not None:
+                        parent.remove(node)
+                        removed_technical_text_count += 1
+                    continue
+
+    _force_viewbox(root)
+
+    output = etree.tostring(
+        root,
+        encoding="utf-8",
+        xml_declaration=True,
+        pretty_print=False,
+    )
+
+    meta = _extract_meta(root)
+    meta.update(
+        {
+            "mode": mode,
+            "removed_hidden_count": removed_hidden_count,
+            "removed_small_text_count": removed_small_text_count,
+            "removed_technical_text_count": removed_technical_text_count,
+        }
+    )
+    return output, meta