Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • 14:12
    Blackclaws commented #726
  • 14:10
    Blackclaws commented #2885
  • 08:58

    pombredanne on improve-gradle

    Improve Gradle script handling … (compare)

  • 04:36
    rohitcoder starred nexB/scancode-toolkit
  • 01:11
    JonoYang synchronize #2974
  • 01:11

    JonoYang on 2972-summary-consider-copyrights

    Do not use party members as dec… Get canonical version of holder… Use detected holders from packa… (compare)

  • May 23 21:48
    martonilles starred nexB/scancode-toolkit
  • May 23 21:37
    johnmhoran labeled #2976
  • May 23 21:37
    johnmhoran labeled #2976
  • May 23 21:37
    johnmhoran opened #2976
  • May 23 21:06
    pombredanne commented #2975
  • May 23 21:05
    rspier commented #2975
  • May 23 21:03
    pombredanne commented #2878
  • May 23 21:02
    pombredanne opened #2975
  • May 23 19:59
    johnmhoran synchronize #2958
  • May 23 19:59

    johnmhoran on 2945-file-cat

    Include directories in codebase… (compare)

  • May 23 17:36
    JonoYang closed #2967
  • May 23 17:36
    JonoYang commented #2967
  • May 23 15:46
    johnmhoran synchronize #2958
  • May 23 15:46

    johnmhoran on 2945-file-cat

    Add initial JSON-based test Re… (compare)

Philippe Ombredanne
@pombredanne
@adii21-Ux good catch :) the doc needs some significant love alright (much more than just a few typos)
Aditya Sangave
@adii21-Ux
should I fix this and open a PR?
Philippe Ombredanne
@pombredanne
@adii21-Ux sure thing, and having something that touches code and not just doc is always welcomed too
Aditya Sangave
@adii21-Ux
ok, I'll make sure if I can do some other changes
Mike Rombout
@mrombout

I am having a bit of trouble understanding Query.unknown_by_pos. As far as I can tell the query tests/licensedcode/data/datadriven/external/fossology-tests/BSD/lz4.license.txt matches the rule src/licensedcode/data/rules/bsd-simplified_and_gpl-2.0_1.RULE exactly (apart from everything after line 7). Yet in the refine_matches phase (second iteration) it reports the following unknowns_by_pos = defaultdict(<class 'int'>, {43: 0, 41: 0, 15: 0, 21: 0, 20: 0}), I am particularly surprised by 15, 20 and 21. And this is throwing off #2637.

I was under the impression that token will be considered unknown if token not in query.idx.dictionary? But it is not that simple?

20 replies
Mike Rombout
@mrombout

I'm afraid I have another case that I'm not able to work out. Another one where the ispan is too inclusive: https://github.com/softsense/scancode-toolkit/blob/issue-2637-allow-license-rules-to-require-the-presence-of-certain-defining-keywords/tests/licensedcode/test_match.py#L325

The ispan of the match containsSpan(2,22), but I feel it should be Span(2,4)|Span(7...) so that it does not include the key phrase of Span(2,8)

Philippe Ombredanne
@pombredanne
@mrombout hey :wave: ...let me check.
Philippe Ombredanne
@pombredanne
@mrombout https://github.com/softsense/scancode-toolkit/pull/1/files#r758425422 you are being misled by the weirdness that may exist in very small indexes
and may be the actual nature of what is in an ispan?
Mike Rombout
@mrombout

Ah, so as far as the ispan is concerned all the words are matching, it is not concerned about the extra words in there, that'd be for the qspan. So what is still missing from the key_phrase_filter is checking if key_phrase_span is uninterrupted in the qspan.

If I would use the zip(qspan, ispan) to create a query_key_phrase_span = Span(qpos, qpos + len(key_phrase_span)) if ipos in key_phrase_span to create a Span offset by where it matches. And then check if that is in query_key_phrase_span in qspan. Would that reasoning be in the right direction?

Philippe Ombredanne
@pombredanne
@mrombout hum... let me think a bit... this is a dense question! :D
Mike Rombout
@mrombout
This message was deleted
What I'm trying right now is:
for qpos, ipos in zip(match.qspan, match.ispan):            
    if ipos in key_phrase_span:
        query_key_phrase_span = Span(qpos, qpos + len(key_phrase_span))
        if query_key_phrase_span not in match.qspan:
            has_key_phrases = False
            break
Philippe Ombredanne
@pombredanne
For reference, we have stopwords and unknown words:
  • unknown words DO NOT exist anywhere in any RULE or LICENSE. They can be seen only in the Query.unknowns_by_pos where we only track how many unknown words exist after a known word position. They are not present in the ispan nor the qspan
  • stopwords exist in RULEs and LICENSEs are short, too common words to be useful. They are skipped both on the index and query side. They are not present in the ispan nor the qspan. They can be seen only in the Query.stopwords_by_pos where we only track how many stopwords exist after a known word position.

by construction, key_phrase_span should be :

  • containing no stopwords
  • containing no unknowns

Therefore, it should be possible to do key_phrase_span in match.qspan.
I reckon the code snippet above is for the next step.

Your snippet looks fine to me :+1:
paraphrasing it, I read it this way:
Philippe Ombredanne
@pombredanne
if a matched rule word position is present in a key phrase (i.e. is the first position of a key phrase), then create a query-side span of positions as long as the key phrase, and check if it exists entirely in the matched query
@mrombout is key_phrase_span for a single key phrase or all the key_phrases of a rule?
Mike Rombout
@mrombout

Yes, that's what I was trying to communicate. So that under the {{Creative Commons Attribution 4.0 International License}} (the "License"); won't match under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

The key_phrase_span is a single key phrase.

Philippe Ombredanne
@pombredanne
BTW do you need to keep track of a complete span for a keyphrase? or rather may be you need only its start position and its length? and may be a Rule.keyphrases could be a mapping of {start_position: length} ?

Yes, that's what I was trying to communicate. So that under the {{Creative Commons Attribution 4.0 International License}} (the "License"); won't match under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

yes, remember my comment on your PR... if you have a only one rule in your test index, the univers of unknown words is very large :D

Mike Rombout
@mrombout

That was just a bad attempt to make my example easier to run, the same problem persists when running as a datadriven test (they use the full index right?). Hence why I am taking another look.

https://dev.azure.com/nexB/scancode-toolkit/_build/results?buildId=5135&view=logs&jobId=58aae7ae-0fb4-5f3e-3a0c-bb9f2080987e&j=58aae7ae-0fb4-5f3e-3a0c-bb9f2080987e&t=8cf9f4a1-3731-59c7-8acb-d9b49562ab2c

Philippe Ombredanne
@pombredanne
checking the test and matched rule texts...
You should expect a single cc-by-nc-sa-4.0 match IMHO
Philippe Ombredanne
@pombredanne
BTW this other thing from Amazon is totally fubar https://github.com/awsdocs/aws-net-developer-guide
it is both under cc-by-sa and cc-by-nc-sa which are completely different puppies
@mrombout In you failing test above you are likely missing a rule
Your new filtering code ... does filter! and it exposes that we are missing some rules there
(which makes me think of a new refinement.... IMHO you might want to avoid filtering later not only when there is only one match... but when there is only one match in a given region possibly a query run? ... just a thought)
Philippe Ombredanne
@pombredanne
@mrombout why have you closed your PR at softsense/scancode-toolkit#1 ? also do not forget to merge in softsense/scancode-toolkit#2 IMHO
Mike Rombout
@mrombout
I interactively rebased so softsense/scancode-toolkit#2 couldn't merge so I just cherry picked it into my branch and created nexB/scancode-toolkit#2773, I am hoping to be done with this soon as I'm running over budget ;)
Philippe Ombredanne
@pombredanne
ah you moved it there ...... nexB/scancode-toolkit#2773 :+1: and merged @dd-jy PR .. :+1:
excellent
Aditya Sangave
@adii21-Ux
Hello @pombredanne I am trying to work on nexB/scancode-toolkit#2767 correct me if I am wrong I have to imporve the license rule for given license which is lgpl2.0 right?
Philippe Ombredanne
@pombredanne
@adii21-Ux hey, yes, the resolution could go a few different ways!
I added a few comments in https://github.com/nexB/scancode-toolkit/issues/2767#issuecomment-987735258
Aditya Sangave
@adii21-Ux
hello @pombredanne added a comment under nexB/scancode-toolkit#2767 please checkout
Salt
@salt:sal.td
[m]
Yay, finally, the correct place to ask ScanCode questions. I'm pretty excited about this project and am incorporating it into two research projects that are dealing with license detection. However, I'm running into memory issues and it keeps getting chomped by OOM killer. Could someone suggest a process to identify where the issue is taking place? I've already tried setting --max-in-memory to use disk-caching...
Salt
@salt:sal.td
[m]
Or perhaps there's some way to detected when it is killed and re-start the scan from that point forward?
Salt
@salt:sal.td
[m]
bleh, the reaper keeps coming. Most scancode runs are ~15% but then it spikes to 95+ and gets killed.
Philippe Ombredanne
@pombredanne
@salt:sal.td hey :wave:
How much ram do you have on hand?
it uses roughly one GB per process (and this is mostly static usage for the index, but unfortunately not memory-mapped hence not shared between processes)
and it then needs RAM to assemble the final output
This part may be the most memory hungry
which output format do you use?
the jsonlines has been designed for a smaller footprint as things do not nee to be all loaded in memory to create the output
Can you paste your scan cli args details?
Philippe Ombredanne
@pombredanne
@salt:sal.td I am intrigued by your research projects too! tell me more :)