WEBDIENSTE/theHarvester
1
0
Fork 0
mirror of https://github.com/laramies/theHarvester.git synced 2026-08-17 19:35:40 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix: move Shodan discovery into adapter (#2535)

Browse Source 
* fix: move Shodan discovery into adapter (#283)

* docs: clarify Shodan transport and persistence

* fix: query every Shodan-resolved IPv4

* fix: call Shodan Host API directly

* fix: serve HarvestView assets locally

* Revert "fix: serve HarvestView assets locally"

This reverts commit 413e8b25ab.

* docs: clarify Shodan proxy transport

* docs: clarify Shodan changelog entry

* feat: persist structured Shodan host evidence

* feat: expand Shodan discovery with TLS search
This commit is contained in:
Matt
2026-08-14 21:11:12 -04:00
committed by GitHub
parent 259455f1c1
commit b8a8f7c7ca
27 changed files with 2054 additions and 443 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CHANGELOG.md
+1
View File
@@ -35,6 +35,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- Added root contributor and security policies, structured issue forms, repository agent guidance, discovery terminology, and an operator-focused documentation wiki ([d090a29a](https://github.com/laramies/theHarvester/commit/d090a29a), [7c491ef5](https://github.com/laramies/theHarvester/commit/7c491ef5), [8b9d420b](https://github.com/laramies/theHarvester/commit/8b9d420b)).
### Changed
- Replaced Shodan's synchronous Python SDK with cancellable async Host API requests that honor configured proxies, query every unique resolved IPv4, paginate target-bound hostname and TLS-certificate searches without an adapter-specific result cap, retain successful partial results, and add no source-local deadline. Shodan now stores one canonical `shodan-host` result per IP with every normalized TCP or UDP service and scoped certificate CN/SAN metadata in native JSONL, SQLite, API, and HarvestView details instead of an escaped JSON value.
- Removed the transport-wide delay before reading ready HTTP responses, bounded Wayback Archive to 30 seconds and Common Crawl to 120 seconds, kept both sources within the requested result limit, and made long-source progress visible in verbose mode. Common Crawl now requests one 50-record page at a time instead of bursting page batches.
- Made Baidu, crt.sh, HackerTarget, Have I Been Pwned, Mojeek, OTX, and Robtex report blocked, malformed, or transport failures truthfully. Also fixed HackerTarget CSV parsing and Robtex AAAA results.
- Hardened BufferOver, ProjectDiscovery, DNSDumpster, ONYPHE, and URLScan parsing and result attribution, including scoped typed results and bounded URLScan pagination.