two version of R2R are here HEAD master

1 files changed, 25 insertions, 0 deletions

diff --git a/R2R/r2r/parsers/structured/csv_parser.py b/R2R/r2r/parsers/structured/csv_parser.py
new file mode 100755
index 00000000..343d9fbf
--- /dev/null
+++ b/R2R/r2r/parsers/structured/csv_parser.py
@@ -0,0 +1,25 @@
+from typing import AsyncGenerator, Union
+
+from r2r.base.abstractions.document import DataType
+from r2r.base.parsers.base_parser import AsyncParser
+
+
+class CSVParser(AsyncParser[DataType]):
+    """A parser for CSV data."""
+
+    def __init__(self):
+        import csv
+        from io import StringIO
+
+        self.csv = csv
+        self.StringIO = StringIO
+
+    async def ingest(
+        self, data: Union[str, bytes]
+    ) -> AsyncGenerator[str, None]:
+        """Ingest CSV data and yield text from each row."""
+        if isinstance(data, bytes):
+            data = data.decode("utf-8")
+        csv_reader = self.csv.reader(self.StringIO(data))
+        for row in csv_reader:
+            yield ", ".join(row)