Mercurial > repos > artbio > facturation_ibps
changeset 1:927553056183 draft
planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/master/tools/facturation_ibps commit 52ed46f7aa2a7b7c1dc478e02b5b53fd0445e155
| author | artbio |
|---|---|
| date | Tue, 23 Oct 2018 11:13:24 -0400 |
| parents | 32b0db39551c |
| children | 3bbcf49ad677 |
| files | facturation.py facturation.xml newtests/test.xlsx |
| diffstat | 3 files changed, 8 insertions(+), 7 deletions(-) [+] |
line wrap: on
line diff
--- a/facturation.py Mon Oct 22 17:17:11 2018 -0400 +++ b/facturation.py Tue Oct 23 11:13:24 2018 -0400 @@ -1,3 +1,4 @@ +#!/usr/bin/env python2 # -*- coding: utf-8 -*- @@ -25,7 +26,10 @@ # ouverture fichier input with open(input_file, 'r') as file_object: facture_html = file_object.read() - + # convert to unicode utf-8, remove   and € + facture_html = facture_html.decode('utf-8') + facture_html = facture_html.replace(r' ', r' ') + facture_html = facture_html.replace(u' \u20ac', '') # parsing de la date et de la période de facturation date = re.search(r'Paris le (.*?)</p>'.decode('utf-8'), facture_html).group(1) @@ -40,11 +44,8 @@ flavor='bs4') # remove 'Adresse de l'appel à facturation : ' (\xa0:\xa0) adresse = facture_parsed[0].replace( - r"Adresse de l'appel \xe0 facturation\xa0:\xa0", r'', regex=True) - - # supression des symboles € (ça fait planter les calculs dans excel sinon) - # ' € ' == \xa0\u20ac - elements = facture_parsed[1].replace(r"\xa0\u20ac", r'', regex=True) + r"Adresse de l\'appel \xe0 facturation : ", r'', regex=True) + elements = facture_parsed[1] # conversion des noms de colonnes elements_col = elements.iloc[0]
--- a/facturation.xml Mon Oct 22 17:17:11 2018 -0400 +++ b/facturation.xml Tue Oct 23 11:13:24 2018 -0400 @@ -1,4 +1,4 @@ -<tool id="facturation_ibps" name="IBPS facturation parser" version="0.2.1"> +<tool id="facturation_ibps" name="IBPS facturation parser" version="0.3.0"> <description /> <requirements> <requirement type="package" version="4.6.3=py27_0">beautifulsoup4</requirement>