[add] migration script

api/migrations/versions/ef0787b85c35_202602061233.py

@@ -0,0 +1,32 @@
+"""202602061233
+
+Revision ID: ef0787b85c35
+Revises: 9b28b66cf8e8
+Create Date: 2026-02-06 12:33:26.114673
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision: str = 'ef0787b85c35'
+down_revision: Union[str, None] = '9b28b66cf8e8'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('files', sa.Column('file_url', sa.String(), nullable=True, comment='file comes from a website url'))
+    op.create_index(op.f('ix_files_file_url'), 'files', ['file_url'], unique=False)
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_index(op.f('ix_files_file_url'), table_name='files')
+    op.drop_column('files', 'file_url')
+    # ### end Alembic commands ###

api/uv.lock

@@ -3224,6 +3224,7 @@ dependencies = [
     { name = "hanziconv" },
     { name = "html5lib" },
     { name = "httptools" },
+    { name = "httpx" },
     { name = "huggingface-hub" },
     { name = "idna" },
     { name = "jieba" },
@@ -3237,6 +3238,7 @@ dependencies = [
     { name = "langchain-ollama" },
     { name = "langchain-openai" },
     { name = "langfuse" },
+    { name = "lxml" },
     { name = "mako" },
     { name = "mammoth" },
     { name = "markdown" },
@@ -3361,6 +3363,7 @@ requires-dist = [
     { name = "hanziconv", specifier = "==0.3.2" },
     { name = "html5lib", specifier = "==1.1" },
     { name = "httptools", specifier = "==0.7.1" },
+    { name = "httpx", specifier = ">=0.28.0" },
     { name = "huggingface-hub", specifier = "==0.25.2" },
     { name = "idna", specifier = "==3.11" },
     { name = "jieba", specifier = ">=0.42.1" },
@@ -3375,6 +3378,7 @@ requires-dist = [
     { name = "langchain-ollama" },
     { name = "langchain-openai", specifier = ">=1.0.2" },
     { name = "langfuse", specifier = ">=3.10.0" },
+    { name = "lxml", specifier = ">=4.9.0" },
     { name = "mako", specifier = "==1.3.10" },
     { name = "mammoth", specifier = "==1.11.0" },
     { name = "markdown", specifier = "==3.8" },