# Changelog of Japan New Companies & Corporate Changes API (`japan_data_tools/japan-corporate-events-feed`) Actor

- **URL**: https://apify.com/japan_data_tools/japan-corporate-events-feed/changelog.md
- **Full Actor documentation**: https://apify.com/japan_data_tools/japan-corporate-events-feed.md

## Changelog

### 0.5

- Added `feed_date` to separate the NTA differential-file date from the underlying `event_date`.
- Added `feedDate` to the run SUMMARY record.
- Surfaced `source` and `source_url` in the default Dataset overview/export view.
- Renamed visible dates to `Change date`, `Feed date`, and `NTA update date`.
- Repositioned Store copy around **newly published corporate events** instead of ambiguous “today changed” wording.
- Added explicit date-semantics documentation.

### 0.4

- Fixed the NTA differential download flow: the site uses a tokenized POST (`event=download`, `selDlFileNo`) rather than direct ZIP hrefs.
- Extracts the official request token and Unicode CSV `doDownload(fileNo)` value from the NTA page.
- Validates the ZIP signature before parsing, producing a clear error instead of `zipfile.BadZipFile` when an HTML response is returned.
- Added a regression test for the NTA POST download-page structure.
- Updated the request User-Agent to a normal browser signature for compatibility.

### 0.3

- Updated Python SDK 3.x PPE call to `charged_event_name=`.
- Stops cleanly when the user's maximum charge limit is reached.
- Counts only actually delivered/charged dataset items in PPE mode.
- Added 128–512 MB Actor memory limits to control platform cost.

### v0.2 — Corporate Events Feed repositioning

- Renamed Store-facing product to **Japan New Companies & Corporate Changes API**.
- Repositioned from registry/data wrapper to **Japan Corporate Events Feed**.
- Renamed `NEW` output signal to `NEW_COMPANY` (`NEW` remains accepted as an input alias).
- Added deterministic `event_id` for downstream deduplication.
- Added PPE-aware result delivery using the `company-event` charge event.
- Added event counts and pricing status to `SUMMARY`.
- Updated input/output/dataset schemas around the four productized signals.
- Added monetization setup and 30-day validation rules.
- Kept corrections excluded by default.