You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 28153fc
Browse filesBrowse the repository at this point in the historyBrowse files
docs(storage): document GCS migration and recovery workflow
Document bucket and authentication setup, provider configuration, verified transfer and URL cutover, validation results, rollback, and delayed source cleanup.
@@ -120,7 +126,7 @@ Response on duplicate (`200`):
120
126
}
121
127
```
122
128
123
-
After creation, if the paper has a `pdf_url`, a background task auto-downloads the PDF to Supabase Storage. If the library has `auto_note_enabled`, AI notes are generated in the background.
129
+
After creation, if the paper has a `pdf_url`, a background task auto-downloads the PDF to the configured PDF storage provider. If the library has `auto_note_enabled`, AI notes are generated in the background.
124
130
125
131
---
126
132
@@ -204,19 +210,19 @@ Upload a PDF file for a paper.
204
210
205
211
Request: `multipart/form-data` with field `file` (must be `application/pdf` or `application/octet-stream`).
206
212
207
-
Stores the PDF in Supabase Storage at `{paper_id}.pdf`, updates `papers.pdf_url` to the public storage URL, then queues AI note generation as a background task.
213
+
Stores the PDF in the configured storage provider at `{paper_id}.pdf`, updates `papers.pdf_url` to the public storage URL, invalidates cached extracted text, then queues AI note generation as a background task.
208
214
209
215
Returns the updated Paper object.
210
216
211
217
### DELETE /api/papers/{id}/pdf
212
218
213
-
Remove the paper's PDF from Supabase Storage, clear `pdf_url`, and delete any cached extracted text.
219
+
Remove the paper's PDF from its recorded managed storage location, clear `pdf_url`, and delete any cached extracted text. Storage failures preserve the database reference for retry.
214
220
215
221
Returns `204 No Content`.
216
222
217
223
### POST /api/papers/{id}/pdf/fetch
218
224
219
-
Download the paper's existing external `pdf_url` into Supabase Storage.
225
+
Download the paper's existing external `pdf_url` into the configured storage provider. PDFs already in either configured managed bucket are rejected as already stored.
220
226
221
227
Use this for papers imported via DOI/arXiv where the PDF URL was set but not yet downloaded. Returns `422` if the paper has no `pdf_url` or the PDF is already in storage.
0 commit comments