1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
|
#!/usr/bin/python2.6
# coding: utf-8
import os
import sys
import pyPdf
from whoosh.index import create_in, open_dir
import whoosh.fields as fields
import time
from cStringIO import StringIO
from Queue import Queue, Empty
import multiprocessing as mp
schema_book = fields.Schema(
pagenumber=fields.NUMERIC(stored=True),
metadata_docnum=fields.NUMERIC(stored=True),
content=fields.TEXT(stored=True),
)
schema_metadata = fields.Schema(
title = fields.TEXT(stored=True),
path=fields.ID(stored=True,unique=True),
createtime=fields.NUMERIC(stored=True) )
if not os.path.exists(u"index"):
create_index = True
os.mkdir(u"index")
index_book = create_in(u"index", schema_book, u"book")
index_metadata = create_in(u"index", schema_metadata, u"metadata")
else:
create_index = False
index_book = open_dir(u"index", u"book")
index_metadata = open_dir(u"index", u"metadata")
filepaths = []
directory = unicode(sys.argv[1], "utf8")
searcher_book = index_book.searcher()
searcher_metadata = index_metadata.searcher()
print u"Walking {0}".format(directory)
filecount = 0
skipped = 0
for path, directories, files in os.walk(directory):
for filename in files:
if filename.endswith(u".pdf"):
filepath = os.path.join(path, filename)
docnum = create_index or searcher_metadata.document_number(path=filepath)
if not docnum:
skipped += 1
else:
filepaths.append(filepath)
filecount += 1
print u"\r{0} files found {1} skipped".format(filecount+skipped, skipped),
print ""
if not create_index: #update index for deleted files
writer_book = index_book.writer()
writer_metadata = index_metadata.writer()
deleted = 0
processed = 0
for fields in searcher_metadata.all_stored_fields():
path = fields['path']
processed += 1
if not os.path.exists(path):
writer_book.delete_by_term(u'metadata_docnum', searcher_metadata.document_number(path=path))
writer_metadata.delete_by_term('path', path)
deleted += 1
print u"\r{0} pages processed. {1} deleted".format(processed, deleted),
print ""
writer_book.commit()
writer_metadata.commit()
searcher_book.close()
searcher_metadata.close()
def process_file(filepath):
try:
print u"{0} processing {1}".format(os.getpid(), filepath)
inputfile = pyPdf.PdfFileReader(file(filepath, 'r'))
title = inputfile.getDocumentInfo().title
writer_metadata = index_metadata.writer()
writer_metadata.add_document(title=title, path=filepath, createtime=time.time())
writer_metadata.commit()
searcher_metadata = index_metadata.searcher()
metadata_docnum = searcher_metadata.document_number(path=filepath)
searcher_metadata.close()
pagenumber = 1
for page in inputfile.pages:
print u"{0} processing {1} Page {2}".format(os.getpid(), filepath, pagenumber)
content = page.extractText()
writer_book = index_book.writer()
writer_book.add_document(pagenumber=pagenumber,
metadata_docnum=metadata_docnum,
content=content)
writer_book.commit()
pagenumber += 1
except KeyboardInterrupt:
return 'KeyboardInterrupt'
except Exception,e:
print u"{0} failed at {1}: {2}".format(os.getpid(), filepath, e)
try:
pool = mp.Pool()
jobs = []
for filepath in filepaths:
jobs.append( pool.apply_async( process_file, (filepath,) ) )
pool.close()
pool.join()
except KeyboardInterrupt:
pool.terminate()
except ImportError:
for filepath in filepaths:
process_file(filepath)
print u"optimize indexes"
index_metadata.optimize()
index_metadata.close()
index_book.optimize()
index_book.close()
|
