annotate bin/sort_date.py @ 102:e606c609f813

reinstate better check to start queuing, check for missed match leading to potential runaway
author Henry S. Thompson <ht@inf.ed.ac.uk>
date Tue, 12 Sep 2023 12:14:04 +0100
parents 135a8c56dcc2
children fc9a045c872b
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
1 #!/usr/bin/python3
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
2 '''Process output of lmh_warc [original 2-column version]
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
3 Usage: <(uz ....warc.gz | fgrep $'\t'|sed "/GMT$/s/\([^ ]\)GMT$/\1 GMT/")
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
4 '''
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
5
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
6 # Assumes you have used grep -v $'\t' on input for speed
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
7 # Recommended to also sed '/GMT$/s/\([^ ]\)GMT$/\1 GMT/'
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
8 # to fix a common 'bad' timestamp (~ .2% of inputs)
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
9
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
10 import email.utils
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
11 import sys
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
12 from urllib.parse import urlsplit, quote, unquote
83
d92bd8527718 use surt instead of trying to create index term by hand
Henry Thompson <ht@markup.co.uk>
parents: 82
diff changeset
13 from surt import surt
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
14
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
15 import re, codecs
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
16
88
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
17 WPAT = re.compile('(,www\\d*)+\\)')
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
18
77
bf09a1d80d7b make CC's own sorting explicit
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 76
diff changeset
19 # Thanks to https://stackoverflow.com/a/8776871
bf09a1d80d7b make CC's own sorting explicit
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 76
diff changeset
20 import locale
bf09a1d80d7b make CC's own sorting explicit
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 76
diff changeset
21 from functools import cmp_to_key
76
eeef811f734d handle corner cases with final . and initial www..+
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 75
diff changeset
22
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
23 def percent_encode(ude):
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
24 #print(ude.object,ude.object[ude.start:ude.end])
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
25 return (''.join('%%%X'%c for c in ude.object[ude.start:ude.end]),
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
26 ude.end)
75
177f7df2bf46 handle %-encoded utf-8 as idna
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 73
diff changeset
27
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
28 codecs.register_error('percent',percent_encode)
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
29
86
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
30 # From RFC-3986:
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
31 # gen-delims = ":" / "/" / "?" / "#" / "[" / "]" / "@"
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
32 # sub-delims = "!" / "$" / "&" / "'" / "(" / ")"
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
33 # / "*" / "+" / "," / ";" / "="
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
34 # But # _is_ escaped in Java surt results
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
35 # and additionally " \ : < = > ? \ ^ _ ` { | } are not
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
36
88
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
37 # Note also that although quote already does _not_ quote - . / _ ~
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
38 # they are included below as that's what we find in surt.surt 0.3.1
88
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
39
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
40 # Also, Java surt strips _all_ leading 'www\d*.',
86
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
41 # where python3 surt only strips the first one.
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
42
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
43 def cdx_key(uristring):
86
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
44 _surt = quote(unquote(surt(uristring),
3a2ae6057242 handle double .www, more keep-me chars
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 85
diff changeset
45 errors='percent'),
88
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
46 safe='''!"$&'()*+,-./:;<=>?@[\]^_`{|}~''' # '
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
47 ).lower()
49faf679d7df final keystroke fixes, recurse and decimal www stripping
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 86
diff changeset
48 return WPAT.sub(')',_surt)
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
49
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
50 def keyed(l):
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
51 uri, cc_stamp, dateTime = l.split('\t',2)
77
bf09a1d80d7b make CC's own sorting explicit
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 76
diff changeset
52 #print('ul',uri,file=sys.stderr)
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
53 try:
76
eeef811f734d handle corner cases with final . and initial www..+
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 75
diff changeset
54 try:
eeef811f734d handle corner cases with final . and initial www..+
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 75
diff changeset
55 epoch = email.utils.parsedate_to_datetime(dateTime).timestamp()
eeef811f734d handle corner cases with final . and initial www..+
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 75
diff changeset
56 except OverflowError:
eeef811f734d handle corner cases with final . and initial www..+
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 75
diff changeset
57 epoch = 32535215999.0
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
58 return ((cdx_key(uri), cc_stamp, uri), epoch)
75
177f7df2bf46 handle %-encoded utf-8 as idna
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 73
diff changeset
59 except (TypeError,IndexError,ValueError) as e:
73
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
60 print(dateTime.rstrip(),e,sep='\t',file=sys.stderr)
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
61 return
e8c667bf8965 compute timestamps, key and sort lmh lines
Henry S. Thompson <ht@inf.ed.ac.uk>
parents:
diff changeset
62
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
63 fstr = sys.argv[1]
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
64
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
65 with open(fstr,"r") as ff:
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
66 # crucial that the following is done _after_ the file is opened
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
67 # with the default (utf-8) locale!
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
68 locale.setlocale(locale.LC_ALL, "C")
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
69 ctk=cmp_to_key(locale.strcoll)
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
70 for key, ts in sorted((kk for l in ff if (kk:=keyed(l)) is not None),
85
1daa8e444cfe work-around for weird handling of %-encoding in Java impl. of SURT
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 84
diff changeset
71 key=lambda x:x[0]):
97
135a8c56dcc2 include full URI in output
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 88
diff changeset
72 print(key[0],key[1],key[2],ts,sep='\t')
77
bf09a1d80d7b make CC's own sorting explicit
Henry S. Thompson <ht@inf.ed.ac.uk>
parents: 76
diff changeset
73