diff options
| -rwxr-xr-x | mapreduce/backfill_hbase_from_cdx.py | 11 | ||||
| -rw-r--r-- | mapreduce/common.py | 2 | ||||
| -rwxr-xr-x | mapreduce/extraction_cdx_grobid.py | 2 | ||||
| -rw-r--r-- | mapreduce/tests/test_backfill_hbase_from_cdx.py | 19 | ||||
| -rw-r--r-- | mapreduce/tests/test_common.py | 10 | ||||
| -rw-r--r-- | mapreduce/tests/test_extraction_cdx_grobid.py | 1 |
6 files changed, 39 insertions, 6 deletions
diff --git a/mapreduce/backfill_hbase_from_cdx.py b/mapreduce/backfill_hbase_from_cdx.py index 8a28ec1..57c18e4 100755 --- a/mapreduce/backfill_hbase_from_cdx.py +++ b/mapreduce/backfill_hbase_from_cdx.py @@ -75,16 +75,19 @@ class MRCDXBackfillHBase(MRJob): # Skip line # XXX: tests don't cover this path; need coverage! self.increment_counter('lines', 'invalid') - return _, dict(status="invalid") + yield _, dict(status="invalid", reason="line prefix") + return info = parse_cdx_line(raw_cdx) if info is None: self.increment_counter('lines', 'invalid') - return _, dict(status="invalid") + yield _, dict(status="invalid") + return if info['file:mime'] not in self.mime_filter: self.increment_counter('lines', 'skip') - return _, dict(status="skip") + yield _, dict(status="skip", reason="unwanted mimetype") + return key = info.pop('key') info['f:c'] = json.dumps(info['f:c'], sort_keys=True, indent=None) @@ -96,6 +99,6 @@ class MRCDXBackfillHBase(MRJob): yield _, dict(status="success") -if __name__ == '__main__': +if __name__ == '__main__': # pragma: no cover MRCDXBackfillHBase.run() diff --git a/mapreduce/common.py b/mapreduce/common.py index 65b2744..1b8e572 100644 --- a/mapreduce/common.py +++ b/mapreduce/common.py @@ -29,6 +29,8 @@ def test_normalize_mime(): assert normalize_mime("Application/PDF") == "application/pdf" assert normalize_mime("application/p") == None assert normalize_mime("application/xml+stuff") == "text/xml" + assert normalize_mime("application/x-pdf") == "application/pdf" + assert normalize_mime("application/x-html") == None def parse_cdx_line(raw_cdx): diff --git a/mapreduce/extraction_cdx_grobid.py b/mapreduce/extraction_cdx_grobid.py index c102a59..ea36e6e 100755 --- a/mapreduce/extraction_cdx_grobid.py +++ b/mapreduce/extraction_cdx_grobid.py @@ -207,6 +207,6 @@ class MRExtractCdxGrobid(MRJob): yield _, dict(status="success", grobid_status=grobid_status) -if __name__ == '__main__': +if __name__ == '__main__': # pragma: no cover MRExtractCdxGrobid.run() diff --git a/mapreduce/tests/test_backfill_hbase_from_cdx.py b/mapreduce/tests/test_backfill_hbase_from_cdx.py index 1a13e5b..2dbbc25 100644 --- a/mapreduce/tests/test_backfill_hbase_from_cdx.py +++ b/mapreduce/tests/test_backfill_hbase_from_cdx.py @@ -52,3 +52,22 @@ com,pbworks,educ333b)/robots.txt 20170705063311 http://educ333b.pbworks.com/robo f_c = json.loads(row[b'f:c'].decode('utf-8')) assert f_c['u'] == "http://cadmus.eui.eu/bitstream/handle/1814/36635/RSCAS_2015_03.pdf%3Bjsessionid%3D761393014319A39F40D32AE3EB3A853F?sequence%3D1" assert b'i' not in f_c + +def test_parse_cdx_skip(job): + + job.mapper_init() + + print("CDX prefix") + raw = " com,sagepub,cep)/content/28/9/960.full.pdf 20170705062200 http://cep.sagepub.com/content/28/9/960.full.pdf application/pdf 200 3I42H3S6NNFQ2MSVX7XZKYAYSCX5QBYJ - - 401 313356621 CITESEERX-CRAWL-2017-06-20-20170705062052659-00043-31209~wbgrp-svc284.us.archive.org~8443.warc.gz" + info, status = job.mapper(None, raw).__next__() + assert info is None + assert status['status'] == "invalid" + assert 'prefix' in status['reason'] + + print("mimetype") + raw = "com,sagepub,cep)/content/28/9/960.full.pdf 20170705062200 http://cep.sagepub.com/content/28/9/960.full.pdf text/html 200 3I42H3S6NNFQ2MSVX7XZKYAYSCX5QBYJ - - 401 313356621 CITESEERX-CRAWL-2017-06-20-20170705061647307-00039-00048-wbgrp-svc284/CITESEERX-CRAWL-2017-06-20-20170705062052659-00043-31209~wbgrp-svc284.us.archive.org~8443.warc.gz" + info, status = job.mapper(None, raw).__next__() + assert info is None + assert status['status'] == "skip" + assert 'mimetype' in status['reason'] + diff --git a/mapreduce/tests/test_common.py b/mapreduce/tests/test_common.py index e2f96bb..34d50ed 100644 --- a/mapreduce/tests/test_common.py +++ b/mapreduce/tests/test_common.py @@ -28,3 +28,13 @@ def test_parse_cdx_line(): assert parse_cdx_line(raw) == correct assert parse_cdx_line(raw + "\n") == correct assert parse_cdx_line(raw + " extra_field") == correct + +def test_invalid_cdx(): + + print("missing warc") + raw = "edu,upenn,ldc)/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf 20170828233154 https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf application/pdf 200 WL3FEA62TEU4F52Y5DOVQ62VET4QJW7G - - 210251 931661233 -" + assert parse_cdx_line(raw) == None + + print("bad datetime") + raw = "edu,upenn,ldc)/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf 2070828233154 https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf application/pdf 200 WL3FEA62TEU4F52Y5DOVQ62VET4QJW7G - - 210251 931661233i SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828231135742-00000-00009-wbgrp-svc284/SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828232253025-00005-3480~wbgrp-svc284.us.archive.org~8443.warc.gz" + assert parse_cdx_line(raw) == None diff --git a/mapreduce/tests/test_extraction_cdx_grobid.py b/mapreduce/tests/test_extraction_cdx_grobid.py index 514a830..1d32c9f 100644 --- a/mapreduce/tests/test_extraction_cdx_grobid.py +++ b/mapreduce/tests/test_extraction_cdx_grobid.py @@ -121,7 +121,6 @@ def test_parse_cdx_skip(job): job.mapper_init() - print("warc format") raw = "com,sagepub,cep)/content/28/9/960.full.pdf 20170705062200 http://cep.sagepub.com/content/28/9/960.full.pdf application/pdf 200 3I42H3S6NNFQ2MSVX7XZKYAYSCX5QBYJ - - 401 313356621 CITESEERX-CRAWL-2017-06-20-20170705062052659-00043-31209~wbgrp-svc284.us.archive.org~8443.warc.gz" info, status = job.mapper(None, raw).__next__() |