# -*- coding: utf-8 -*- ''' Created : 2015-03-12 @author: Eric Lapouyade ''' __version__ = '0.1.9' from lxml import etree from docx import Document from jinja2 import Template from cgi import escape import re import six class DocxTemplate(object): """ Class for managing docx files as they were jinja2 templates """ def __init__(self, docx): self.docx = Document(docx) def __getattr__(self, name): return getattr(self.docx, name) def get_docx(self): return self.docx def get_xml(self): # Be careful : pretty_print MUST be set to False, otherwise patch_xml() won't work properly return etree.tostring(self.docx._element.body, encoding='unicode', pretty_print=False) def write_xml(self,filename): with open(filename,'w') as fh: fh.write(self.get_xml()) def patch_xml(self,src_xml): # strip all xml tags inside {% %} and {{ }} that MS word can insert into xml source src_xml = re.sub(r'(?<={)(<[^>]*>)+(?=[\{%])|(?<=[%\}])(<[^>]*>)+(?=\})','',src_xml,flags=re.DOTALL) def striptags(m): return re.sub('.*?(|]*>)','',m.group(0),flags=re.DOTALL) src_xml = re.sub(r'{%(?:(?!%}).)*|{{(?:(?!}}).)*',striptags,src_xml,flags=re.DOTALL) # manage table cell background color def cellbg(m): cell_xml = m.group(1) + m.group(3) cell_xml = re.sub(r'](?:(?!]).)*.*?','',cell_xml,flags=re.DOTALL) cell_xml = re.sub(r'','', cell_xml, count=1) return re.sub(r'(]*>)',r'\1' % m.group(2), cell_xml) src_xml = re.sub(r'(](?:(?!]).)*){%\s*cellbg\s+([^%]*)\s*%}(.*?)',cellbg,src_xml,flags=re.DOTALL) for y in ['tr', 'p', 'r']: # replace into xml code the row/paragraph/run containing {%y xxx %} or {{y xxx}} template tag # by {% xxx %} or {{ xx }} without any surronding xml tags : # This is mandatory to have jinja2 generating correct xml code pat = r'](?:(?!]).)*({%%|{{)%(y)s ([^}%%]*(?:%%}|}})).*?' % {'y':y} src_xml = re.sub(pat, r'\1 \2',src_xml,flags=re.DOTALL) src_xml = src_xml.replace(r"‘","'") return src_xml def render_xml(self,src_xml,context,jinja_env=None): if jinja_env: template = jinja_env.from_string(src_xml) else: template = Template(src_xml) dst_xml = template.render(context) dst_xml = dst_xml.replace('{_{','{{').replace('}_}','}}').replace('{_%','{%').replace('%_}','%}') return dst_xml def build_xml(self,context,jinja_env=None): xml = self.get_xml() xml = self.patch_xml(xml) xml = self.render_xml(xml, context, jinja_env) return xml def map_xml(self,xml): root = self.docx._element body = root.body root.replace(body,etree.fromstring(xml)) def render(self,context,jinja_env=None): xml = self.build_xml(context,jinja_env) self.map_xml(xml) def new_subdoc(self): return Subdoc(self) class Subdoc(object): """ Class for subdocument to insert into master document """ def __init__(self, tpl): self.tpl = tpl self.docx = tpl.get_docx() self.subdocx = Document() self.subdocx._part = self.docx._part def __getattr__(self, name): return getattr(self.subdocx, name) def _get_xml(self): xml = re.sub(r']*>','',etree.tostring(self.subdocx._element.body, encoding='unicode', pretty_print=False)) return xml def __unicode__(self): return self._get_xml() def __str__(self): return self._get_xml() class RichText(object): """ class to generate Rich Text when using templates variables This is much faster than using Subdoc class, but this only for texts INSIDE an existing paragraph. """ def __init__(self, text=None, **text_prop): self.xml = '' if text: self.add(text, **text_prop) def add(self, text, style=None, color=None, highlight=None, size=None, bold=False, italic=False, underline=False, strike=False): if not isinstance(text, six.text_type): text = text.decode('utf-8',errors='ignore') text = escape(text).replace('\n','') prop = u'' if style: prop += u'' % style if color: if color[0] == '#': color = color[1:] prop += u'' % color if highlight: if highlight[0] == '#': highlight = highlight[1:] prop += u'' % highlight if size: prop += u'' % size prop += u'' % size if bold: prop += u'' if italic: prop += u'' if underline: if underline not in ['single','double']: underline = 'single' prop += u'' % underline if strike: prop += u'' self.xml += u'' if prop: self.xml += u'%s' % prop self.xml += u'%s\n' % text def __unicode__(self): return self.xml def __str__(self): return self.xml