filter_scored_matches.py

1 files changed, 110 insertions, 0 deletions

diff --git a/python/filter_scored_matches.py b/python/filter_scored_matches.py
new file mode 100755
index 0000000..819e13c
--- /dev/null
+++ b/python/filter_scored_matches.py
@@ -0,0 +1,110 @@
+#!/usr/bin/env python3
+"""
+Filters an input stream of sorted "matchcrossref" scalding job, and outputs
+"good enough" matches to be inserted to fatcat.
+
+Currently works on DOI numbers. Filters for a high enough string match (doesn't
+re-do title match), and checks author lists. Filters out slugs with too many
+matches, and outputs one-line-per-sha1 (aka, file).
+
+No dependencies (only python3 stdlib)
+"""
+
+import sys
+import json
+
+# out of 1000
+score_threshold = 900
+
+max_slug_lines = 10
+
+require_authors = 1
+
+
+def tokenize(s, remove_whitespace=False):
+
+    s.replace(''', "'")
+    # Remove non-alphanumeric characters
+    s = ''.join([c for c in s.lower() if c.isalnum() or c.isspace()])
+
+    if remove_whitespace:
+        s = ''.join(s.split())
+
+    # Encode as dumb ASCII (TODO: this is horrible)
+    return s.encode('ascii', 'replace').replace(b'?', b'')
+
+def check_authors(left, right):
+    if len(left) == 0:
+        return False
+    if len(left) > len(right):
+        return False
+    right_all = tokenize(" ".join(right))
+    for i in range(len(left)):
+        l = left[i].lower().replace('jr.', '').split()
+        if len(l) == 0:
+            return False
+        l = tokenize(l[-1])
+        if len(l) <= 1:
+            # weird author name (single char)
+            return False
+        if not l in right_all:
+            #print("MISSING: {} from {}".format(l.decode('utf8'), right_all.decode('utf8')))
+            return False
+    return True
+
+def test_check_authors():
+    assert False == check_authors([], [])
+    assert False == check_authors([], ['one'])
+    assert True == check_authors(['one'], ['one'])
+    assert True == check_authors(['one two'], ['One Two'])
+    assert True == check_authors(['two'], ['One Two'])
+    assert True == check_authors(['two'], ['two, one'])
+    assert True == check_authors(['Mr. Magoo'], ['mago'])
+    assert True == check_authors(['one', 'two', 'three'], ['one', 'tw', 'thr'])
+
+# Rows are (score, grobid, crossref)
+def process_group(rows):
+    if len(rows) > max_slug_lines:
+        return
+    keepers = dict()
+    for row in rows:
+        score = int(row[0])
+        if score < score_threshold:
+            continue
+        grobid = json.loads(row[1])
+        crossref = json.loads(row[2])
+        if not check_authors(crossref['authors'], grobid['authors']):
+            #print("NO (crossref/grobid): {} {}".format(crossref['authors'], grobid['authors']))
+            continue
+        else:
+            #print("YES: {} {}".format(crossref['authors'], grobid['authors']))
+            pass
+        sha1 = grobid['sha1']
+        doi = crossref['doi'].lower()
+        l = keepers.get(sha1, list())
+        l.append(doi)
+        keepers[sha1] = l
+    for key, value in keepers.items():
+        print("{}\t{}".format(sha1, json.dumps(value)))
+
+def run():
+
+    last_slug = None
+    lines = []
+
+    # group lines by slug, and process in batches
+    for line in sys.stdin:
+        line = line.strip().split('\t')
+        assert(len(line) == 4)
+        slug = line[0]
+        if last_slug and slug != last_slug and len(lines) > 0:
+            process_group(lines)
+            lines = []
+        last_slug = slug
+        lines.append(line[1:])
+
+    if len(lines) > 0:
+        process_group(lines)
+
+if __name__=='__main__':
+    run()