1
0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-29 09:02:27 +01:00

8 Commits

Author SHA1 Message Date
Ilya Kreymer
9da5bd1083
Decoding and Recorder Fixes (#313)
* redisindex: use decode_resposes=True for redisindex
* recorder: close_file(): return true if closed, close_key() return filename if closed
* logging: if debug=True, log warc load failures
* appveyor build fix: add pypiwin32 as dependency for windows build
2018-03-29 13:42:00 -07:00
Ilya Kreymer
459cd706d3 include the collection in Memento Link outputs: (#259)
* include the collection in Memento Link outputs:
- add new cdx 'source-coll' field, storing only the collection
- ensure rel="collection" property included in the TimeMap and Link header
- tests: update all tests to include the 'source-coll' property
- docs: add 'collection provenance' to auto-all collection configuration docs
2017-10-23 15:33:23 -07:00
Ilya Kreymer
54b265aaa8 s3 and zipnum fixes: (#253)
* s3 and zipnum fixes:
- update s3 to use boto3
- ensure zipnum indexes (.idx, .summary) are picked up automatically via DirectoryAggregator
- ensure showNumPages query always return a json object, ignoring output=
- add tests for auto-configured zipnum indexes

* reqs: add boto3 dependency, init boto Config only if avail

* s3 loader: first try with credentials, then with no-cred config
archive paths: don't add anything if path is fully qualified (contains '://')

* s3 loader: on first load, if credentialed load fails, try uncredentialed
fix typo
tests: add zinum auto collection tests

* zipnum page count query: don't add 'source' field to page count query (if 'url' key not present in dict)

* s3 loader: fix no-range load, add test, update skip check to boto3

* fix spacing

* boto -> boto3 rename error message, cleanup comments
2017-10-11 15:33:57 -07:00
Ilya Kreymer
5791980132 warcserver: DirectoryAggregator:
- support naming directory aggregator such that source is reflected as '<name>:<path/to/index>' if optional name is present
- for default WarcServer use colls dir as name, defaulting to 'collections:<coll/indexes/index.cdxj>' for 'source' entries
- tests: update tests to use name with directory aggregator for more consistent source names
2017-09-28 01:52:07 -07:00
Ilya Kreymer
bcb5bef39d Windows Build Fixes/Appveyor CI (#225)
windows build fixes: all tests should pass, ci with appveyor
- add appveyor.yml
- path fixes for windows, use os.path.join
- templates_dir: use '/' always for jinja2 paths
- auto colls: ensure chdir before deleting dir
- recorder: ensure warc writer is always closed
- recorder: disable locking in warcwriter on windows for now (read access not avail, shared
lock seems to not be working)
- zipnum: ensure block is closed after read!
- cached dir test: wait before adding file
- tests: adjust timeout tests to allow more leeway in timing
2017-08-05 17:12:16 -07:00
Ilya Kreymer
d12f715d81 refactor: split warcserver.utils into utils package:
- utils.io for stream/compression related utils
- utils.format for string formatting
- utils.memento for memento
- load_config -> utils.loaders.load_overlay_config
- also: use warcio.utils.to_native_str instead of utils.loaders.to_native_str
2017-06-05 17:43:46 -07:00
Ilya Kreymer
dbc56b864b Merge branch 'aggregator-improvements' into refactor2 2017-06-02 21:33:23 -07:00
Ilya Kreymer
ad33dc6728 refactor: webagg -> warcserver rename
- ResAggApp -> BaseWarcServer
- AutoApp -> WarcServer
- move index related files to warcserver.index package, tests to warcserver.index.test
- move resource loading related files to warcserver.resource package, tests to warcserver.resource.test
- pywb.cdx -> pywb.warcserver.index
- split pywb.warc -> pywb.warcserver.resource or pywb.indexer (for cdx generation)
- bump to 0.51.0 for now!
- tests for pywb.warcserver should be working
2017-05-23 09:21:43 -07:00