# render_capture URL: /kida/api/render_capture/ Section: api Description: Kida RenderCapture — opt-in content capture during template rendering. This module provides render-time data capture for block content, context snapshots, and content hashing. Enables search indexing, block-level caching, and semantic diffing as byproducts of rendering. Zero overhead when disabled (get_capture() returns None). RFC: render-capture Example: from kida import Environment, FileSystemLoader from kida.render_capture import captured_render env = Environment(loader=FileSystemLoader("templates/"), enable_capture=True) template = env.get_template("page.html") # Normal render (no overhead) html = template.render(page=page) # Captured render (opt-in) with captured_render( capture_blocks=frozenset({"content", "sidebar"}), capture_context=frozenset({"page"}), ) as capture: html = template.render(page=page) print(capture.template_name) # "page.html" print(capture.blocks["content"].html) # "
...
" print(capture.blocks["content"].content_hash) # "a1b2c3d4e5f67890" print(capture.context_keys) # {"page": } --- > For a complete page index, fetch /kida/llms.txt. Open LLM text (/kida/api/render_capture/index.txt) Share with AI Ask Claude (https://claude.ai/new?q=Please%20help%20me%20understand%20this%20documentation%3A%20%2Fkida%2Fapi%2Frender_capture%2Findex.txt) Ask ChatGPT (https://chatgpt.com/?q=Please%20help%20me%20understand%20this%20documentation%3A%20%2Fkida%2Fapi%2Frender_capture%2Findex.txt) Ask Gemini (https://gemini.google.com/app?q=Please%20help%20me%20understand%20this%20documentation%3A%20%2Fkida%2Fapi%2Frender_capture%2Findex.txt) Ask Copilot (https://copilot.microsoft.com/?q=Please%20help%20me%20understand%20this%20documentation%3A%20%2Fkida%2Fapi%2Frender_capture%2Findex.txt) Module # `render_capture` Kida RenderCapture — opt-in content capture during template rendering. This module provides render-time data capture for block content, context snapshots, and content hashing. Enables search indexing, block-level caching, and semantic diffing as byproducts of rendering. Zero overhead when disabled (get_capture() returns None). RFC: render-capture Example: ``` from kida import Environment, FileSystemLoader from kida.render_capture import captured_render env = Environment(loader=FileSystemLoader("templates/"), enable_capture=True) template = env.get_template("page.html") # Normal render (no overhead) html = template.render(page=page) # Captured render (opt-in) with captured_render( capture_blocks=frozenset({"content", "sidebar"}), capture_context=frozenset({"page"}), ) as capture: html = template.render(page=page) print(capture.template_name) # "page.html" print(capture.blocks["content"].html) # "
...
" print(capture.blocks["content"].content_hash) # "a1b2c3d4e5f67890" print(capture.context_keys) # {"page": } ``` 2Classes3Functions ## Classes `Fragment` 6 ▼ A single block's rendered output with semantic metadata. A single block's rendered output with semantic metadata. #### Attributes Name Type Description `name` `str` Block name (e.g., "content", "sidebar") `role` `str` Semantic role from BlockMetadata.inferred_role (e.g., "content", "navigation", "unknown") `html` `str` Raw rendered HTML of this block `content_hash` `str` SHA256[:16] hex digest of the rendered HTML `depends_on` `frozenset[str]` Context paths this block reads (from BlockMetadata) `cache_scope` `str` Caching granularity from BlockMetadata.cache_scope ("site", "page", "none", "unknown") `RenderCapture` 8 ▼ Collects fragments and context during a single render() call. Populated by compiler-injected hooks… Collects fragments and context during a single render() call. Populated by compiler-injected hooks when an active capture exists. Use`captured_render`() to create and activate a capture. #### Attributes Name Type Description `template_name` `str` Name of the rendered template `blocks` `dict[str, Fragment]` Block name -> Fragment mapping for captured blocks `context_keys` `dict[str, Any]` Snapshotted context values (top-level keys only) `_capture_blocks` `frozenset[str] | None` — `_capture_context` `frozenset[str] | None` — `_block_metadata` `dict[str, Any] | None` — #### Methods `changed_from` 1 `dict[str, tuple[Fragment…` ▼ Blocks whose content_hash differs between two captures. Returns a dict mapping… `def changed_from(self, other: RenderCapture) -> dict[str, tuple[Fragment, Fragment]]` Blocks whose content_hash differs between two captures. Returns a dict mapping block name to (old_fragment, new_fragment) for blocks present in both captures with different content. ##### Parameters Name Type Description `other` `—` The earlier capture to compare against. ##### Returns `dict[str, tuple[Fragment, Fragment]]` Dict of block name -> (other's fragment, self's fragment) for changed blocks. Internal Methods 1 ▼ `_record` 2 ▼ Record a block's rendered output. Called by compiler-injected hooks at block c… `def _record(self, name: str, html: str) -> None` Record a block's rendered output. Called by compiler-injected hooks at block call sites. Skips recording if the block is not in the capture set. Pulls role and depends_on from _block_metadata when available. ##### Parameters Name Type Description `name` `—` Block name `html` `—` Rendered HTML output of the block ## Functions `_compute_content_hash` 1 `str` ▼ Compute content hash for a rendered block. Uses SHA256 truncated to 16 hex cha… `def _compute_content_hash(html: str) -> str` Compute content hash for a rendered block. Uses SHA256 truncated to 16 hex chars, matching the block_hash precedent in kida.analysis.analyzer. ##### Parameters Name Type Description `html` `str` ##### Returns `str` `get_capture` 0 `RenderCapture | None` ▼ Get current capture (None if capture not active). `def get_capture() -> RenderCapture | None` ##### Returns `RenderCapture | None` `captured_render` 2 `Iterator[RenderCapture]` ▼ Context manager for captured rendering. Creates a RenderCapture and makes it a… `def captured_render(capture_blocks: frozenset[str] | None = None, capture_context: frozenset[str] | None = None) -> Iterator[RenderCapture]` Context manager for captured rendering. Creates a RenderCapture and makes it available via get_capture() for the duration of the with block. Compiler-injected hooks record block content and context snapshots into the capture. ##### Parameters Name Type Description `capture_blocks` `frozenset[str] | None` Set of block names to capture. None captures all blocks. Default:`None` `capture_context` `frozenset[str] | None` Set of top-level context keys to snapshot. None captures no context (explicit opt-in for context). Default:`None` ##### Returns `Iterator[RenderCapture]`