fix: deduping by url
This commit is contained in:
@@ -121,9 +121,12 @@ def get_invoice_links(page: Page) -> list[dict]:
|
|||||||
seen = set()
|
seen = set()
|
||||||
|
|
||||||
def _add_link(url: str, label: str, **meta) -> None:
|
def _add_link(url: str, label: str, **meta) -> None:
|
||||||
if not url or url in seen:
|
if not url:
|
||||||
return
|
return
|
||||||
seen.add(url)
|
dedupe_key = meta.get("reference_id") or url
|
||||||
|
if dedupe_key in seen:
|
||||||
|
return
|
||||||
|
seen.add(dedupe_key)
|
||||||
entry = {"url": url, "label": label}
|
entry = {"url": url, "label": label}
|
||||||
entry.update(meta)
|
entry.update(meta)
|
||||||
links.append(entry)
|
links.append(entry)
|
||||||
|
|||||||
Reference in New Issue
Block a user