feat: Rename ocr-text asset type to text-content across the application

backend/migrations/202511300000_rename_ocr_text_asset/down.sql

@@ -0,0 +1,3 @@
+UPDATE tenant.document_assets
+SET asset_type = 'ocr-text'
+WHERE asset_type = 'text-content';

backend/migrations/202511300000_rename_ocr_text_asset/up.sql

@@ -0,0 +1,3 @@
+UPDATE tenant.document_assets
+SET asset_type = 'text-content'
+WHERE asset_type = 'ocr-text';

backend/src/workers/analyze.rs

@@ -16,7 +16,7 @@ use super::{
     index::IndexDocumentTask,
     issued_at::DetermineIssuedAtTask,
     job_execution_from_task_error,
-    ocr::{GenerateOcrTask, OCR_TEXT_ASSET_TYPE},
+    ocr::{GenerateOcrTask, TEXT_CONTENT_ASSET_TYPE},
     taskflow::{
         document::DocumentVersionTaskContext, BoxedTask, Task, TaskError, TaskExecutor,
         TaskPlanner, TaskResult,
@@ -126,7 +126,7 @@ impl TaskPlanner<DocumentVersionTaskContext> for AnalyzePlanner {
             tasks.push(Box::new(GenerateThumbnailsTask::new(self.force)));
         }
 
-        let existing_ocr = ctx.asset(OCR_TEXT_ASSET_TYPE).await?.is_some();
+        let existing_ocr = ctx.asset(TEXT_CONTENT_ASSET_TYPE).await?.is_some();
         let mut should_index = existing_ocr;
 
         if document_supports_ocr(&document) {

backend/src/workers/index.rs

@@ -6,7 +6,7 @@ use reqwest::Client;
 use crate::documents::search::{build_quickwit_ingest_record, quickwit_ingest};
 
 use super::{
-    ocr::OCR_TEXT_ASSET_TYPE,
+    ocr::TEXT_CONTENT_ASSET_TYPE,
     taskflow::{document::DocumentVersionTaskContext, Task, TaskError, TaskResult},
 };
 
@@ -45,7 +45,7 @@ impl Task<DocumentVersionTaskContext> for IndexDocumentTask {
             .ok_or_else(|| TaskError::fail("tenant quickwit index not configured"))?;
 
         let asset = ctx
-            .asset(OCR_TEXT_ASSET_TYPE)
+            .asset(TEXT_CONTENT_ASSET_TYPE)
             .await?
             .ok_or_else(|| TaskError::fail("missing OCR text asset"))?;
 

backend/src/workers/issued_at.rs

@@ -12,7 +12,7 @@ use uuid::Uuid;
 
 use crate::issued_at::{DateOrder, IssuedAtSettings};
 use crate::schema::documents::dsl as documents_dsl;
-use crate::workers::ocr::OCR_TEXT_ASSET_TYPE;
+use crate::workers::ocr::TEXT_CONTENT_ASSET_TYPE;
 use crate::workers::taskflow::document::DocumentVersionTaskContext;
 use crate::workers::taskflow::{Task, TaskContext, TaskError, TaskResult};
 
@@ -427,10 +427,8 @@ fn month_name_to_number(value: &str, settings: &IssuedAtSettings) -> Option<u32>
 
 async fn load_document_text(ctx: &mut DocumentVersionTaskContext) -> TaskResult<Option<String>> {
     let object_key = {
-        let assets = ctx.assets().await?;
-        assets
-            .get(OCR_TEXT_ASSET_TYPE)
-            .map(|asset| asset.asset.s3_key.clone())
+        let asset = ctx.asset(TEXT_CONTENT_ASSET_TYPE).await?;
+        asset.map(|a| a.asset.s3_key.clone())
     };
 
     let Some(key) = object_key else {

backend/src/workers/ocr.rs

@@ -29,7 +29,7 @@ use super::taskflow::{
     document::DocumentVersionTaskContext, Task, TaskContext, TaskError, TaskResult,
 };
 
-pub const OCR_TEXT_ASSET_TYPE: &str = "ocr-text";
+pub const TEXT_CONTENT_ASSET_TYPE: &str = "text-content";
 const MIN_TEXT_LENGTH: usize = 50;
 
 pub struct GenerateOcrTask {
@@ -83,7 +83,7 @@ impl Task<DocumentVersionTaskContext> for GenerateOcrTask {
         let s3_key = document_asset_key(
             context.document.id,
             context.version.version_number,
-            OCR_TEXT_ASSET_TYPE,
+            TEXT_CONTENT_ASSET_TYPE,
             asset_id,
         );
 
@@ -129,7 +129,7 @@ async fn build_ocr_context(
 ) -> TaskResult<OcrContext> {
     let document = ctx.document().await?.clone();
     let version = ctx.version().await?.clone();
-    let asset = ctx.asset(OCR_TEXT_ASSET_TYPE).await?;
+    let asset = ctx.asset(TEXT_CONTENT_ASSET_TYPE).await?;
 
     let existing_asset = asset.map(|asset| asset.asset.clone());
 
@@ -226,7 +226,7 @@ fn persist_ocr_metadata(
     let new_asset = NewDocumentAsset {
         id: asset_id,
         document_version_id,
-        asset_type: OCR_TEXT_ASSET_TYPE.to_string(),
+        asset_type: TEXT_CONTENT_ASSET_TYPE.to_string(),
         mime_type: "text/plain".to_string(),
         metadata,
         s3_key: s3_key.to_string(),