From e56691a1c57a0341fde08ff284986074299534c4 Mon Sep 17 00:00:00 2001
From: LukeZekes <41444778+LukeZekes@users.noreply.github.com>
Date: Wed, 5 Mar 2025 09:56:02 -0600
Subject: [PATCH] add filename option for text documents (#47)

---
 core/api.py                       | 2 ++
 core/models/request.py            | 1 +
 core/services/document_service.py | 3 ++-
 sdks/python/databridge/models.py  | 1 +
 sdks/python/databridge/sync.py    | 2 ++
 5 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/core/api.py b/core/api.py
index c5df9fe..ab5b161 100644
--- a/core/api.py
+++ b/core/api.py
@@ -261,6 +261,7 @@ async def ingest_text(
     Args:
         request: IngestTextRequest containing:
             - content: Text content to ingest
+            - filename: Optional filename to help determine content type
             - metadata: Optional metadata dictionary
             - rules: Optional list of rules. Each rule should be either:
                    - MetadataExtractionRule: {"type": "metadata_extraction", "schema": {...}}
@@ -283,6 +284,7 @@ async def ingest_text(
         ):
             return await document_service.ingest_text(
                 content=request.content,
+                filename=request.filename,
                 metadata=request.metadata,
                 rules=request.rules,
                 use_colpali=request.use_colpali,
diff --git a/core/models/request.py b/core/models/request.py
index 97facc2..203c5a6 100644
--- a/core/models/request.py
+++ b/core/models/request.py
@@ -24,6 +24,7 @@ class IngestTextRequest(BaseModel):
     """Request model for ingesting text content"""
 
     content: str
+    filename: Optional[str] = None
     metadata: Dict[str, Any] = Field(default_factory=dict)
     rules: List[Dict[str, Any]] = Field(default_factory=list)
     use_colpali: Optional[bool] = None
diff --git a/core/services/document_service.py b/core/services/document_service.py
index 483f3f2..8a6e5d6 100644
--- a/core/services/document_service.py
+++ b/core/services/document_service.py
@@ -36,7 +36,6 @@ import os
 logger = logging.getLogger(__name__)
 IMAGE = {im.mime for im in IMAGE}
 
-
 class DocumentService:
     def __init__(
         self,
@@ -185,6 +184,7 @@ class DocumentService:
     async def ingest_text(
         self,
         content: str,
+        filename: Optional[str] = None,
         metadata: Optional[Dict[str, Any]] = None,
         auth: AuthContext = None,
         rules: Optional[List[str]] = None,
@@ -197,6 +197,7 @@ class DocumentService:
 
         doc = Document(
             content_type="text/plain",
+            filename=filename,
             metadata=metadata or {},
             owner={"type": auth.entity_type, "id": auth.entity_id},
             access_control={
diff --git a/sdks/python/databridge/models.py b/sdks/python/databridge/models.py
index cbcd8d8..6f62f96 100644
--- a/sdks/python/databridge/models.py
+++ b/sdks/python/databridge/models.py
@@ -70,6 +70,7 @@ class IngestTextRequest(BaseModel):
     """Request model for ingesting text content"""
 
     content: str
+    filename: Optional[str] = None
     metadata: Dict[str, Any] = Field(default_factory=dict)
     rules: List[Dict[str, Any]] = Field(default_factory=list)
     use_colpali: bool = Field(default=False)
diff --git a/sdks/python/databridge/sync.py b/sdks/python/databridge/sync.py
index 405726f..0fd751e 100644
--- a/sdks/python/databridge/sync.py
+++ b/sdks/python/databridge/sync.py
@@ -150,6 +150,7 @@ class DataBridge:
     def ingest_text(
         self,
         content: str,
+        filename: Optional[str] = None,
         metadata: Optional[Dict[str, Any]] = None,
         rules: Optional[List[RuleOrDict]] = None,
         use_colpali: bool = True,
@@ -191,6 +192,7 @@ class DataBridge:
         """
         request = IngestTextRequest(
             content=content,
+            filename=filename,
             metadata=metadata or {},
             rules=[self._convert_rule(r) for r in (rules or [])],
             use_colpali=use_colpali,