def pisaParser(src, context, default_css="", xhtml=False, encoding=None, xml_output=None):
"""
- Parse HTML and get miniDOM
- Extract CSS informations, add default CSS, parse CSS
- Handle the document DOM itself and build reportlab story
- Return Context object
"""
global CSSAttrCache
CSSAttrCache = {}
if xhtml:
#TODO: XHTMLParser doesn't see to exist...
parser = html5lib.XHTMLParser(tree=treebuilders.getTreeBuilder("dom"))
else:
parser = html5lib.HTMLParser(tree=treebuilders.getTreeBuilder("dom"))
if type(src) in StringTypes:
if type(src) is unicode:
# If an encoding was provided, do not change it.
if not encoding:
encoding = "utf-8"
src = src.encode(encoding)
src = pisaTempFile(src, capacity=context.capacity)
# Test for the restrictions of html5lib
if encoding:
# Workaround for html5lib<0.11.1
if hasattr(inputstream, "isValidEncoding"):
if encoding.strip().lower() == "utf8":
encoding = "utf-8"
if not inputstream.isValidEncoding(encoding):
log.error("%r is not a valid encoding e.g. 'utf8' is not valid but 'utf-8' is!", encoding)
else:
if inputstream.codecName(encoding) is None:
log.error("%r is not a valid encoding", encoding)
document = parser.parse(
src,
encoding=encoding)
if xml_output:
if encoding:
xml_output.write(document.toprettyxml(encoding=encoding))
else:
xml_output.write(document.toprettyxml(encoding="utf8"))
if default_css:
context.addDefaultCSS(default_css)
pisaPreLoop(document, context)
#try:
context.parseCSS()
#except:
# context.cssText = DEFAULT_CSS
# context.parseCSS()
# context.debug(9, pprint.pformat(context.css))
pisaLoop(document, context)
return context
def pisaParser(src, context, default_css="", xhtml=False, encoding=None, xml_output=None):
"""
- Parse HTML and get miniDOM
- Extract CSS informations, add default CSS, parse CSS
- Handle the document DOM itself and build reportlab story
- Return Context object
"""
global CSSAttrCache
CSSAttrCache = {}
if xhtml:
#TODO: XHTMLParser doesn't see to exist...
parser = html5lib.XHTMLParser(tree=treebuilders.getTreeBuilder("dom"))
else:
parser = html5lib.HTMLParser(tree=treebuilders.getTreeBuilder("dom"))
if type(src) in StringTypes:
if type(src) is unicode:
# If an encoding was provided, do not change it.
if not encoding:
encoding = "utf-8"
src = src.encode(encoding)
src = pisaTempFile(src, capacity=context.capacity)
# Test for the restrictions of html5lib
if encoding:
# Workaround for html5lib<0.11.1
if hasattr(inputstream, "isValidEncoding"):
if encoding.strip().lower() == "utf8":
encoding = "utf-8"
if not inputstream.isValidEncoding(encoding):
log.error("%r is not a valid encoding e.g. 'utf8' is not valid but 'utf-8' is!", encoding)
else:
if inputstream.codecName(encoding) is None:
log.error("%r is not a valid encoding", encoding)
document = parser.parse(
src,
encoding=encoding)
if xml_output:
if encoding:
xml_output.write(document.toprettyxml(encoding=encoding))
else:
xml_output.write(document.toprettyxml(encoding="utf8"))
if default_css:
context.addDefaultCSS(default_css)
pisaPreLoop(document, context)
#try:
context.parseCSS()
#except:
# context.cssText = DEFAULT_CSS
# context.parseCSS()
# context.debug(9, pprint.pformat(context.css))
pisaLoop(document, context)
return context
def test_unicode(self):
"""Asserts bytes generated by reportlab are returned"""
src = pisaTempFile()
value = b'%PDF-1.4\r\n%\x93\x8c\x8b\x9e ReportLab Generated PDF document http://www.reportlab.com'
try:
src.write(value)
except UnicodeDecodeError as error:
self.fail(error)
def test_unicode(self):
"""Asserts bytes generated by reportlab are returned"""
src = pisaTempFile()
value = b'%PDF-1.4\r\n%\x93\x8c\x8b\x9e ReportLab Generated PDF document http://www.reportlab.com'
try:
src.write(value)
except UnicodeDecodeError as error:
self.fail(error)
def pisaErrorDocument(dest, c):
out = pisaTempFile(capacity=c.capacity)
out.write("<p style='background-color:red;'><strong>%d error(s) occured:</strong><p>" % c.err)
for mode, line, msg, _ in c.log:
if mode=="error":
out.write("<pre>%s in line %d: %s</pre>" % (mode, line, cgi.escape(msg)))
out.write("<p><strong>%d warning(s) occured:</strong><p>" % c.warn)
for mode, line, msg, _ in c.log:
if mode=="warning":
out.write("<p>%s in line %d: %s</p>" % (mode, line, cgi.escape(msg)))
return pisaDocument(out.getvalue(), dest, raise_exception=False)
def join(self, file=None):
output = PyPDF2.PdfFileWriter()
for pdffile in self.files:
input = PyPDF2.PdfFileReader(pdffile)
for pageNumber in xrange(input.getNumPages()):
output.addPage(input.getPage(pageNumber))
if file is not None:
output.write(file)
return file
out = pisaTempFile(capacity=self.capacity)
output.write(out)
return out.getvalue()
def join(self, file=None):
import pyPdf # TODO: Why is this in the middle of everything?
if pyPdf:
output = pyPdf.PdfFileWriter()
for pdffile in self.files:
input = pyPdf.PdfFileReader(pdffile)
for pageNumber in xrange(input.getNumPages()):
output.addPage(input.getPage(pageNumber))
if file is not None:
output.write(file)
return file
out = pisaTempFile(capacity=self.capacity)
output.write(out)
return out.getvalue()
def join(self, file=None):
import pyPdf # TODO: Why is this in the middle of everything?
if pyPdf:
output = pyPdf.PdfFileWriter()
for pdffile in self.files:
input = pyPdf.PdfFileReader(pdffile)
for pageNumber in xrange(input.getNumPages()):
output.addPage(input.getPage(pageNumber))
if file is not None:
output.write(file)
return file
out = pisaTempFile(capacity=self.capacity)
output.write(out)
return out.getvalue()
def pisaDocument(src,
dest=None,
path=None,
link_callback=None,
debug=0,
default_css=None,
xhtml=False,
encoding=None,
xml_output=None,
raise_exception=True,
capacity=100 * 1024,
**kw):
log.debug(
"pisaDocument options:\n src = %r\n dest = %r\n path = %r\n link_callback = %r\n xhtml = %r",
src, dest, path, link_callback, xhtml)
# Prepare simple context
context = pisaContext(path, debug=debug, capacity=capacity)
context.pathCallback = link_callback
# Build story
context = pisaStory(src,
path,
link_callback,
debug,
default_css,
xhtml,
encoding,
context=context,
xml_output=xml_output)
# Buffer PDF into memory
out = pisaTempFile(capacity=context.capacity)
doc = PmlBaseDoc(out,
pagesize=context.pageSize,
author=context.meta["author"].strip(),
subject=context.meta["subject"].strip(),
keywords=[
x.strip()
for x in context.meta["keywords"].strip().split(",")
if x
],
title=context.meta["title"].strip(),
showBoundary=0,
allowSplitting=1)
# Prepare templates and their frames
if "body" in context.templateList:
body = context.templateList["body"]
del context.templateList["body"]
else:
x, y, w, h = getBox("1cm 1cm -1cm -1cm", context.pageSize)
body = PmlPageTemplate(id="body",
frames=[
Frame(x,
y,
w,
h,
id="body",
leftPadding=0,
rightPadding=0,
bottomPadding=0,
topPadding=0)
],
pagesize=context.pageSize)
doc.addPageTemplates([body] + context.templateList.values())
# Use multibuild e.g. if a TOC has to be created
if context.multiBuild:
doc.multiBuild(context.story)
else:
doc.build(context.story)
# Add watermarks
if pyPdf:
for bgouter in context.pisaBackgroundList:
# If we have at least one background, then lets do it
if bgouter:
istream = out
output = pyPdf.PdfFileWriter()
input1 = pyPdf.PdfFileReader(istream)
ctr = 0
# TODO: Why do we loop over the same list again?
# see bgouter at line 137
for bg in context.pisaBackgroundList:
page = input1.getPage(ctr)
if (bg and not bg.notFound()
and (bg.mimetype == "application/pdf")):
bginput = pyPdf.PdfFileReader(bg.getFile())
pagebg = bginput.getPage(0)
pagebg.mergePage(page)
page = pagebg
else:
log.warn(
context.warning("Background PDF %s doesn't exist.",
bg))
output.addPage(page)
ctr += 1
out = pisaTempFile(capacity=context.capacity)
output.write(out)
# data = sout.getvalue()
# Found a background? So leave loop after first occurence
break
else:
log.warn(context.warning("pyPDF not installed!"))
# Get the resulting PDF and write it to the file object
# passed from the caller
if dest is None:
# No output file was passed - Let's use a pisaTempFile
dest = pisaTempFile(capacity=context.capacity)
context.dest = dest
data = out.getvalue(
) # TODO: That load all the tempfile in RAM - Why bother with a swapping tempfile then?
context.dest.write(data) # TODO: context.dest is a tempfile as well...
return context
def pisaDocument(src, dest=None, path=None, link_callback=None, debug=0,
default_css=None, xhtml=False, encoding=None, xml_output=None,
raise_exception=True, capacity=100*1024, **kw):
log.debug("pisaDocument options:\n src = %r\n dest = %r\n path = %r\n link_callback = %r\n xhtml = %r",
src,
dest,
path,
link_callback,
xhtml)
# Prepare simple context
context = pisaContext(path, debug=debug, capacity=capacity)
context.pathCallback = link_callback
# Build story
context = pisaStory(src, path, link_callback, debug, default_css, xhtml,
encoding, context=context, xml_output=xml_output)
# Buffer PDF into memory
out = pisaTempFile(capacity=context.capacity)
doc = PmlBaseDoc(
out,
pagesize=context.pageSize,
author=context.meta["author"].strip(),
subject=context.meta["subject"].strip(),
keywords=[x.strip() for x in
context.meta["keywords"].strip().split(",") if x],
title=context.meta["title"].strip(),
showBoundary=0,
allowSplitting=1)
# Prepare templates and their frames
if "body" in context.templateList:
body = context.templateList["body"]
del context.templateList["body"]
else:
x, y, w, h = getBox("1cm 1cm -1cm -1cm", context.pageSize)
body = PmlPageTemplate(
id="body",
frames=[
Frame(x, y, w, h,
id="body",
leftPadding=0,
rightPadding=0,
bottomPadding=0,
topPadding=0)],
pagesize = context.pageSize)
doc.addPageTemplates([body] + context.templateList.values())
# Use multibuild e.g. if a TOC has to be created
if context.multiBuild:
doc.multiBuild(context.story)
else:
doc.build(context.story)
# Add watermarks
if pyPdf:
for bgouter in context.pisaBackgroundList:
# If we have at least one background, then lets do it
if bgouter:
istream = out
output = pyPdf.PdfFileWriter()
input1 = pyPdf.PdfFileReader(istream)
ctr = 0
# TODO: Why do we loop over the same list again?
# see bgouter at line 137
for bg in context.pisaBackgroundList:
page = input1.getPage(ctr)
if (bg and not bg.notFound()
and (bg.mimetype=="application/pdf")):
bginput = pyPdf.PdfFileReader(bg.getFile())
pagebg = bginput.getPage(0)
pagebg.mergePage(page)
page = pagebg
else:
log.warn(context.warning(
"Background PDF %s doesn't exist.", bg))
output.addPage(page)
ctr += 1
out = pisaTempFile(capacity=context.capacity)
output.write(out)
# data = sout.getvalue()
# Found a background? So leave loop after first occurence
break
else:
log.warn(context.warning("pyPDF not installed!"))
# Get the resulting PDF and write it to the file object
# passed from the caller
if dest is None:
# No output file was passed - Let's use a pisaTempFile
dest = pisaTempFile(capacity=context.capacity)
context.dest = dest
data = out.getvalue() # TODO: That load all the tempfile in RAM - Why bother with a swapping tempfile then?
context.dest.write(data) # TODO: context.dest is a tempfile as well...
return context
def addFromString(self, data):
self.files.append(pisaTempFile(data, capacity=self.capacity))
def addFromString(self, data):
self.files.append(pisaTempFile(data, capacity=self.capacity))
def pisaDocument(src,
dest=None,
path=None,
link_callback=None,
debug=0,
default_css=None,
xhtml=False,
encoding=None,
xml_output=None,
raise_exception=True,
capacity=100 * 1024,
context_meta=None,
**kw):
log.debug(
"pisaDocument options:\n src = %r\n dest = %r\n path = %r\n link_callback = %r\n xhtml = %r\n context_meta = %r",
src, dest, path, link_callback, xhtml, context_meta)
# Prepare simple context
context = pisaContext(path, debug=debug, capacity=capacity)
if context_meta is not None:
context.meta.update(context_meta)
context.pathCallback = link_callback
# Build story
context = pisaStory(src,
path,
link_callback,
debug,
default_css,
xhtml,
encoding,
context=context,
xml_output=xml_output)
# Buffer PDF into memory
out = io.BytesIO()
doc = PmlBaseDoc(out,
pagesize=context.pageSize,
author=context.meta["author"].strip(),
subject=context.meta["subject"].strip(),
keywords=[
x.strip()
for x in context.meta["keywords"].strip().split(",")
if x
],
title=context.meta["title"].strip(),
showBoundary=0,
allowSplitting=1)
# Prepare templates and their frames
multi_template_list = False
if "body" in context.templateList:
body = context.templateList["body"]
del context.templateList["body"]
else:
x, y, w, h = getBox("1cm 1cm -1cm -1cm", context.pageSize)
body = PmlPageTemplate(id="body",
frames=[
Frame(x,
y,
w,
h,
id="body",
leftPadding=0,
rightPadding=0,
bottomPadding=0,
topPadding=0)
],
pagesize=context.pageSize)
ptl = build_grid_templates(doc, context)
if ptl == []:
doc.addPageTemplates([body] + list(context.templateList.values()))
if ptl != []:
if out_grid == []:
doc.addPageTemplates(ptl)
else:
doc.addPageTemplates([body] + ptl)
# Use multibuild e.g. if a TOC has to be created
if context.multiBuild:
doc.multiBuild(context.story)
else:
doc.build(context.story)
# Add watermarks
if PyPDF2:
file_handler = None
for bgouter in context.pisaBackgroundList:
# If we have at least one background, then lets do it
if bgouter:
istream = out
output = PyPDF2.PdfFileWriter()
input1 = PyPDF2.PdfFileReader(istream)
ctr = 0
# TODO: Why do we loop over the same list again?
# see bgouter at line 137
for bg in context.pisaBackgroundList:
page = input1.getPage(ctr)
if (bg and not bg.notFound()
and (bg.mimetype == "application/pdf")):
file_handler = open(bg.uri, 'rb')
bginput = PyPDF2.PdfFileReader(file_handler)
pagebg = bginput.getPage(0)
pagebg.mergePage(page)
page = pagebg
# Todo: the else-statement doesn't make a lot of sense to me; it's just throwing warnings
# on unittesting \tests. Probably we have to rewrite the whole "background-image" stuff
# to deal with cases like:
# Page1 .jpg background
# Page1 .pdf background
# Page1 .jpg background, Page2 no background
# Page1 .pdf background, Page2 no background
# Page1 .jpg background, Page2 .pdf background
# Page1 .pdf background, Page2 .jpg background
# etc.
# Right now it's kind of confusing. (fbernhart)
# else:
# log.warning(context.warning(
# "Background PDF %s doesn't exist.", bg))
output.addPage(page)
ctr += 1
out = pisaTempFile(capacity=context.capacity)
output.write(out)
if file_handler:
file_handler.close()
# data = sout.getvalue()
# Found a background? So leave loop after first occurence
break
else:
log.warning(context.warning("PyPDF2 not installed!"))
# Get the resulting PDF and write it to the file object
# passed from the caller
if dest is None:
# No output file was passed - Let's use a pisaTempFile
dest = io.BytesIO()
context.dest = dest
data = out.getvalue()
context.dest.write(data) # TODO: context.dest is a tempfile as well...
return context
def pisaParser(src,
context,
default_css="",
xhtml=False,
encoding=None,
xml_output=None):
"""
- Parse HTML and get miniDOM
- Extract CSS informations, add default CSS, parse CSS
- Handle the document DOM itself and build reportlab story
- Return Context object
"""
global CSSAttrCache
CSSAttrCache = {}
if xhtml:
# TODO: XHTMLParser doesn't seem to exist...
parser = html5lib.XHTMLParser(tree=treebuilders.getTreeBuilder("dom"))
else:
parser = html5lib.HTMLParser(tree=treebuilders.getTreeBuilder("dom"))
parser_kwargs = {}
if isinstance(src, six.text_type):
# If an encoding was provided, do not change it.
if not encoding:
encoding = "utf-8"
src = src.encode(encoding)
src = pisaTempFile(src, capacity=context.capacity)
# To pass the encoding used to convert the text_type src to binary_type
# on to html5lib's parser to ensure proper decoding
parser_kwargs['transport_encoding'] = encoding
# # Test for the restrictions of html5lib
# if encoding:
# # Workaround for html5lib<0.11.1
# if hasattr(inputstream, "isValidEncoding"):
# if encoding.strip().lower() == "utf8":
# encoding = "utf-8"
# if not inputstream.isValidEncoding(encoding):
# log.error("%r is not a valid encoding e.g. 'utf8' is not valid but 'utf-8' is!", encoding)
# else:
# if inputstream.codecName(encoding) is None:
# log.error("%r is not a valid encoding", encoding)
document = parser.parse(src, **parser_kwargs) # encoding=encoding)
if xml_output:
if encoding:
xml_output.write(document.toprettyxml(encoding=encoding))
else:
xml_output.write(document.toprettyxml(encoding="utf8"))
if default_css:
context.addDefaultCSS(default_css)
pisaPreLoop(document, context)
# try:
context.parseCSS()
# except:
# context.cssText = DEFAULT_CSS
# context.parseCSS()
# context.debug(9, pprint.pformat(context.css))
pisaLoop(document, context)
return context
def pisaDocument(
src,
dest = None,
path = None,
link_callback = None,
debug = 0,
show_error_as_pdf = False,
default_css = None,
xhtml = False,
encoding = None,
xml_output = None,
raise_exception = True,
capacity = 100 * 1024, # -1,
**kw):
c = None
if show_error_as_pdf:
raise_exception = False
try:
log.debug("pisaDocument options:\n src = %r\n dest = %r\n path = %r\n link_callback = %r\n xhtml = %r",
src,
dest,
path,
link_callback,
xhtml)
# Prepare simple context
c = pisaContext(path, debug=debug, capacity=capacity)
c.pathCallback = link_callback
if dest is None:
dest = pisaTempFile(capacity=c.capacity)
c.dest = dest
# Build story
c = pisaStory(src, path, link_callback, debug, default_css, xhtml, encoding, c=c, xml_output=xml_output)
# Buffer PDF into memory
out = pisaTempFile(capacity=c.capacity)
doc = PmlBaseDoc(
out,
pagesize = c.pageSize,
author = c.meta["author"].strip(),
subject = c.meta["subject"].strip(),
keywords = [x.strip() for x in c.meta["keywords"].strip().split(",") if x],
title = c.meta["title"].strip(),
showBoundary = 0,
allowSplitting = 1)
# XXX It is not possible to access PDF info, because it is private in canvas
# doc.info.producer = "pisa <http://www.holtwick.it>"
# Prepare templates and their frames
if c.templateList.has_key("body"):
body = c.templateList["body"]
del c.templateList["body"]
else:
x, y, w, h = getBox("1cm 1cm -1cm -1cm", c.pageSize)
body = PmlPageTemplate(
id="body",
frames=[
Frame(x, y, w, h,
id = "body",
leftPadding = 0,
rightPadding = 0,
bottomPadding = 0,
topPadding = 0)],
pagesize = c.pageSize)
# print body.frames
# print [body] + c.templateList.values()
doc.addPageTemplates([body] + c.templateList.values())
doc._pisa_page_counter = 0
def _page_counter(page_no):
doc._pisa_page_counter += 1
doc.setPageCallBack(_page_counter)
# Use multibuild e.g. if a TOC has to be created
if c.multiBuild:
doc.multiBuild(c.story)
else:
doc.build(c.story)
c._pisa_page_counter = doc._pisa_page_counter
# Add watermarks
if pyPdf:
for bgouter in c.pisaBackgroundList:
# If we have at least one background, then lets do it
if bgouter:
istream = out
try:
output = pyPdf.PdfFileWriter()
input1 = pyPdf.PdfFileReader(istream)
ctr = 0
for bg in c.pisaBackgroundList:
page = input1.getPage(ctr)
if bg and not bg.notFound() and (bg.mimetype=="application/pdf"):
bginput = pyPdf.PdfFileReader(bg.getFile())
pagebg = bginput.getPage(0)
pagebg.mergePage(page)
page = pagebg
else:
log.warn(c.warning("Background PDF %s doesn't exist.", bg))
output.addPage(page)
ctr += 1
out = pisaTempFile(capacity=c.capacity)
output.write(out)
# data = sout.getvalue()
except Exception:
log.exception(c.error("pyPDF error"))
if raise_exception:
raise
# Found a background? So leave loop after first occurence
break
else:
log.warn(c.warning("pyPDF not installed!"))
# In web frameworks for debugging purposes maybe an output of
# errors in a PDF is preferred
if show_error_as_pdf and c and c.err:
return pisaErrorDocument(c.dest, c)
# Get the resulting PDF and write it to the file object
# passed from the caller
data = out.getvalue()
c.dest.write(data)
c.dest.close()
except: # TODO: Kill catch-all!
# log.exception(c.error("Document error"))
log.exception("Document error")
c.err += 1
if raise_exception:
raise
if raise_exception and c.err:
raise Exception("Errors occured, please see log files for more informations")
return c