two version of R2R are here HEAD master

1 files changed, 32 insertions, 0 deletions

diff --git a/.venv/lib/python3.12/site-packages/core/parsers/text/html_parser.py b/.venv/lib/python3.12/site-packages/core/parsers/text/html_parser.py
new file mode 100644
index 00000000..a04331e0
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/core/parsers/text/html_parser.py
@@ -0,0 +1,32 @@
+# type: ignore
+from typing import AsyncGenerator
+
+from bs4 import BeautifulSoup
+
+from core.base.parsers.base_parser import AsyncParser
+from core.base.providers import (
+    CompletionProvider,
+    DatabaseProvider,
+    IngestionConfig,
+)
+
+
+class HTMLParser(AsyncParser[str | bytes]):
+    """A parser for HTML data."""
+
+    def __init__(
+        self,
+        config: IngestionConfig,
+        database_provider: DatabaseProvider,
+        llm_provider: CompletionProvider,
+    ):
+        self.database_provider = database_provider
+        self.llm_provider = llm_provider
+        self.config = config
+
+    async def ingest(
+        self, data: str | bytes, *args, **kwargs
+    ) -> AsyncGenerator[str, None]:
+        """Ingest HTML data and yield text."""
+        soup = BeautifulSoup(data, "html.parser")
+        yield soup.get_text()