Debian Codegraph @

tests/adaptdocx/extract_docx.py

raw
1 #! /usr/bin/python3
2
3 """
4 Extract text from a .xml in a .docx file
5 Reads the file from stdin
6 """
7
8 import lxml.etree as ET
9 import io
10 import re
11 import os, sys, argparse
12
13
14 class docx():
15
16 def __init__(self, linies):
17 self.parser = ET.XMLParser(remove_blank_text=True)
18 self.tot = ET.fromstring("".join(lin.strip() for lin in linies).encode("utf8"), self.parser)
19 self.pars = [paragraf(p) for p in docx.iter(self.tot, "w:p")]
20
21 @staticmethod
22 def iter(element, tag):
23 ns, nom = tag.split(":")
24 for e in element.iter("{%s}%s" % (docx.ns[ns], nom)):
25 yield e
26
27 ns = {
28 "r": "http://schemas.openxmlformats.org/officeDocument/2006/relationships",
29 "wp": "http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing",
30 "w": "http://schemas.openxmlformats.org/wordprocessingml/2006/main",
31 "w10": "urn:schemas-microsoft-com:office:word",
32 "w14": "http://schemas.microsoft.com/office/word/2010/wordml",
33 "wps": "http://schemas.microsoft.com/office/word/2010/wordprocessingShape",
34 "wpg": "http://schemas.microsoft.com/office/word/2010/wordprocessingGroup",
35 "mc": "http://schemas.openxmlformats.org/markup-compatibility/2006",
36 "v": "urn:schemas-microsoft-com:vml",
37 "o": "urn:schemas-microsoft-com:office:office",
38 "m": "http://schemas.openxmlformats.org/officeDocument/2006/math",
39 "xml": "http://www.w3.org/XML/1998/namespace"
40 }
41
42
43
44 class paragraf():
45
46 def __init__(self, element):
47 self.element = element
48 self.rr = [r for r in docx.iter(element, "w:r")]
49 self.tt = [t for r in self.rr for t in docx.iter(r, "w:t")]
50 texts = []
51 for t in self.tt:
52 if self.get_atrib(t, "xml:space") == "preserve":
53 texts.append(t.text)
54 else:
55 texts.append(t.text.strip())
56 self.txt = "|".join(texts)
57 self.mides = [len(t) for t in texts]
58
59 def get_atrib(self, element, nom, defecte = None):
60 pn = nom.split(":", 1)
61 pn = "{%s}%s" % (docx.ns[pn[0]], pn[1])
62 valor = element.get(pn, defecte)
63 return valor
64
65
66 linia = ""
67 while True:
68 c = sys.stdin.read(1)
69 if not c or c == '\0':
70 break
71 else:
72 linia += c
73 linies = [ linia ]
74 #sys.stdout.write("#lĂ­nies: %d" % (len(linies)))
75 doc = docx(linies)
76 for p in doc.pars:
77 print(p.txt)
78 sys.stdout.write('\0')
79 sys.stdout.flush()
80