Add XLIFF uploading capabilities

backend/alembic/versions/dcd734dfe9c4_convert_documentrecordhistorychangetype_enum.py

@@ -0,0 +1,63 @@
+"""Convert DocumentRecordHistoryChangeType enum to string
+
+Revision ID: dcd734dfe9c4
+Revises: 71a5eef94341
+Create Date: 2026-02-17 02:32:03.411573
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# pylint: disable=E1101
+
+# revision identifiers, used by Alembic.
+revision: str = 'dcd734dfe9c4'
+down_revision: Union[str, None] = '71a5eef94341'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+change_type_enum = sa.Enum(
+    'initial_import',
+    'machine_translation',
+    'tm_substitution',
+    'glossary_substitution',
+    'repetition',
+    'manual_edit',
+    name='documentrecordhistorychangetype'
+)
+
+
+def upgrade() -> None:
+    # Convert enum column to text/string for flexibility
+    op.alter_column(
+        'document_record_history',
+        'change_type',
+        type_=sa.Text(),
+    )
+    change_type_enum.drop(op.get_bind(), checkfirst=True)
+
+
+def downgrade() -> None:
+    # Convert any translation_update values to manual_edit before recreating enum
+    op.execute(
+        sa.text("""
+            UPDATE document_record_history
+            SET change_type = 'manual_edit'
+            WHERE change_type = 'translation_update';
+        """)
+    )
+
+    # Recreate enum type (without translation_update)
+    change_type_enum.create(op.get_bind(), checkfirst=True)
+
+    # Convert text column back to enum
+    op.execute(
+        sa.text("""
+            ALTER TABLE document_record_history
+            ALTER COLUMN change_type TYPE documentrecordhistorychangetype
+            USING change_type::text::documentrecordhistorychangetype;
+        """)
+    )

backend/app/documents/models.py

@@ -28,6 +28,7 @@ class DocumentRecordHistoryChangeType(Enum):
     glossary_substitution = "glossary_substitution"
     repetition = "repetition"
     manual_edit = "manual_edit"
+    translation_update = "translation_update"
 
 
 class TmMode(Enum):
@@ -203,7 +204,9 @@ class DocumentRecordHistory(Base):
     diff: Mapped[str] = mapped_column()
     author_id: Mapped[int | None] = mapped_column(ForeignKey("user.id"), nullable=True)
     timestamp: Mapped[datetime] = mapped_column(default=utc_time)
-    change_type: Mapped[DocumentRecordHistoryChangeType] = mapped_column()
+    change_type: Mapped[DocumentRecordHistoryChangeType] = mapped_column(
+        SqlEnum(DocumentRecordHistoryChangeType, native_enum=False)
+    )
 
     record: Mapped["DocumentRecord"] = relationship(back_populates="history")
     author: Mapped["User"] = relationship()

backend/app/documents/schema.py

@@ -124,3 +124,10 @@ class DocumentUpdateResponse(BaseModel):
     project_id: int | None
 
     model_config = ConfigDict(from_attributes=True)
+
+
+class XliffUploadOptions(BaseModel):
+    update_approved: bool = Field(
+        default=False,
+        description="If True, forcefully update approved records. If False (default), skip approved records.",
+    )

backend/app/formats/xliff.py

@@ -211,6 +211,8 @@ def extract_xliff_content(content: bytes) -> XliffData:
     root: etree._Element = etree.fromstring(
         content, parser=etree.XMLParser(recover=True)
     )
+    if root is None:
+        raise RuntimeError("Error: Invalid XML file")
 
     version = root.attrib.get("version")
     if not version or version != "1.2":

backend/app/routers/document.py

@@ -1,6 +1,15 @@
 from typing import Annotated
 
-from fastapi import APIRouter, Depends, File, HTTPException, Query, UploadFile, status
+from fastapi import (
+    APIRouter,
+    Depends,
+    File,
+    Form,
+    HTTPException,
+    Query,
+    UploadFile,
+    status,
+)
 from fastapi.responses import StreamingResponse
 from sqlalchemy.orm import Session
 
@@ -260,3 +269,28 @@ def update_document(
             status_code=status.HTTP_403_FORBIDDEN,
             detail=str(e),
         )
+
+
+@router.post("/upload_xliff")
+async def upload_xliff(
+    service: Annotated[DocumentService, Depends(get_service)],
+    current_user: Annotated[int, Depends(get_current_user_id)],
+    file: Annotated[UploadFile, File()],
+    update_approved: Annotated[bool, Form()] = False,
+) -> models.StatusMessage:
+    try:
+        return await service.upload_xliff(
+            file,
+            doc_schema.XliffUploadOptions(update_approved=update_approved),
+            current_user,
+        )
+    except EntityNotFound as e:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(e),
+        )
+    except BusinessLogicError as e:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=str(e),
+        )

backend/app/services/document_service.py

@@ -13,13 +13,17 @@
 from app.documents import schema as doc_schema
 from app.documents.models import (
     Document,
+    DocumentRecord,
+    DocumentRecordHistoryChangeType,
     DocumentType,
     TmMode,
     XliffRecord,
 )
 from app.documents.query import (
+    DocumentRecordHistoryQuery,
     GenericDocsQuery,
 )
+from app.documents.utils import compute_diff
 from app.formats.txt import extract_txt_content
 from app.formats.xliff import (
     SegmentState,
@@ -54,6 +58,7 @@ def __init__(self, db: Session):
         self.__query = GenericDocsQuery(db)
         self.__glossary_query = GlossaryQuery(db)
         self.__tm_query = TranslationMemoryQuery(db)
+        self.__history_query = DocumentRecordHistoryQuery(db)
 
     def get_document(self, doc_id: int) -> doc_schema.DocumentWithRecordsCount:
         """
@@ -653,3 +658,101 @@ def encode_to_latin_1(self, original: str):
         for c in original:
             output += c if (c.isalnum() or c in "'().[] -") else "_"
         return output
+
+    async def upload_xliff(
+        self,
+        file: UploadFile,
+        options: doc_schema.XliffUploadOptions,
+        current_user: int,
+    ) -> models.StatusMessage:
+        """
+        Upload XLIFF file and update document records.
+
+        Args:
+            file: Uploaded XLIFF file
+            options: Upload options including update_approved flag
+            current_user: ID of user performing the upload
+
+        Returns:
+            StatusMessage indicating success
+
+        Raises:
+            EntityNotFound: If document not found
+        """
+        # Read file content
+        file_data = await file.read()
+        original_document = file_data.decode("utf-8")
+
+        # Parse XLIFF
+        try:
+            xliff_data = extract_xliff_content(original_document.encode("utf-8"))
+        except RuntimeError:
+            raise BusinessLogicError("Invalid XLIFF format")
+
+        # Extract document ID from first file element
+        file_element = xliff_data.xliff_file.find(
+            ".//{urn:oasis:names:tc:xliff:document:1.2}file",
+            namespaces=xliff_data.xliff_file.nsmap,
+        )
+        if file_element is None:
+            raise BusinessLogicError("Invalid XLIFF format: no file element found")
+
+        doc_id_str = file_element.get("original")
+        if not doc_id_str:
+            raise BusinessLogicError(
+                "Invalid XLIFF format: file element missing original attribute"
+            )
+
+        try:
+            doc_id = int(doc_id_str)
+        except ValueError:
+            raise BusinessLogicError(f"Invalid document ID in XLIFF: {doc_id_str}")
+
+        # Validate document exists
+        self._get_document_by_id(doc_id)
+
+        # Prepare history entries for bulk creation
+        history_entries = []
+
+        # Update records
+        for segment in xliff_data.segments:
+            record = (
+                self.__db.query(DocumentRecord)
+                .filter_by(document_id=doc_id, id=segment.id_)
+                .first()
+            )
+
+            if not record:
+                continue
+
+            # Check if we should update this record
+            should_update = options.update_approved or not record.approved
+
+            if not should_update:
+                continue
+
+            old_target = record.target
+            new_target = segment.translation or ""
+            # Only update if the new translation is different and not empty
+            if old_target != new_target and new_target:
+                record.target = new_target
+                record.approved = segment.approved
+
+                # Prepare history entry
+                history_entries.append(
+                    (record.id, compute_diff(old_target, new_target))
+                )
+
+        # Bulk create history entries
+        if history_entries:
+            self.__history_query.bulk_create_history_entry(
+                history_entries,
+                current_user,
+                DocumentRecordHistoryChangeType.translation_update,
+            )
+
+        self.__db.commit()
+        updated_count = len(history_entries)
+        return models.StatusMessage(
+            message=f"Successfully updated {updated_count} record(s)"
+        )

backend/tests/fixtures/upload_test.xliff

@@ -0,0 +1,32 @@
+<?xml version="1.0" encoding="utf-8"?>
+<xliff version="1.2"
+    xmlns:sc="SmartcatXliff"
+    xmlns="urn:oasis:names:tc:xliff:document:1.2">
+    <file datatype="plaintext" date="2023-11-07T20:56:05.0017382Z" original="1" source-language="en" target-language="ru">
+        <header>
+            <tool tool-name="Smartcat.ai" tool-id="40c7d5b2-da26-4b36-84f1-8305b3aadb03" tool-version="1.0.11.0" />
+        </header>
+        <body>
+            <trans-unit id="1" approved="no" sc:locked="false" sc:last-modified-date="2023-10-19T18:29:22.711Z" sc:last-modified-user="Person 1">
+                <source xml:space="preserve">Regional Effects</source>
+                <target state="translated" xml:space="preserve">Региональные эффекты</target>
+            </trans-unit>
+            <trans-unit id="2" approved="no" sc:locked="false" sc:last-modified-date="2023-10-19T18:29:22.911Z" sc:last-modified-user="Person 2">
+                <source xml:space="preserve">User Interface</source>
+                <target state="needs-translation" xml:space="preserve">Пользовательский интерфейс</target>
+            </trans-unit>
+            <trans-unit id="3" approved="yes" sc:locked="false" sc:last-modified-date="2023-10-19T18:29:22.967Z" sc:last-modified-user="Person 3">
+                <source xml:space="preserve">Approved Segment</source>
+                <target state="final" xml:space="preserve">Утвержденный сегмент</target>
+            </trans-unit>
+            <trans-unit id="4" approved="no" sc:locked="false" sc:last-modified-date="2023-10-19T18:29:22.968Z" sc:last-modified-user="Person 4">
+                <source xml:space="preserve">123456789</source>
+                <target state="needs-translation" xml:space="preserve"></target>
+            </trans-unit>
+            <trans-unit id="5" approved="no" sc:locked="false" sc:last-modified-date="2023-10-19T18:29:23.968Z" sc:last-modified-user="Person 5">
+                <source xml:space="preserve">Something else</source>
+                <target state="translated" xml:space="preserve">Что-то еще</target>
+            </trans-unit>
+        </body>
+    </file>
+</xliff>