Hi Sandro, Martin, Sorry for not replying sooner, I am just drowning in work currently.
I have looked at the pisa package and tried to patch it the way I mentionned earlier, but it turned out to not be sufficient, so I have made some more modifications. You can find a first (somewhat dirty) draft of patch in attachment. With this patch, xhtml2pdf seemed to work correctly on several test pages (e.g. this one[0]). I'll try to improve/test this patch further. I'd not upload it in its current state. Could anybody have a look at it ? Getting rid of pisa for buster would be a good idea indeed. Thanks ! Cheers, Hugo [0] http://noir.liw.fi/ -- Hugo Lefeuvre (hle) | www.owl.eu.com 4096/ ACB7 B67F 197F 9B32 1533 431C AC90 AC3E C524 065E
--- a/sx/pisa3/pisa_parser.py 2017-02-08 15:38:26.127678118 +0100
+++ b/sx/pisa3/pisa_parser.py 2017-02-08 15:38:49.755588784 +0100
@@ -26,7 +26,8 @@
import os.path
import html5lib
-from html5lib import treebuilders, serializer, treewalkers, inputstream
+from html5lib import treebuilders, treewalkers
+from html5lib import _inputstream as inputstream
from xml.dom import Node
import xml.dom.minidom
@@ -611,9 +612,7 @@
if inputstream.codecName(encoding) is None:
log.error("%r is not a valid encoding", encoding)
- document = parser.parse(
- src,
- encoding=encoding)
+ document = parser.parse(src)
if xml_output:
xml_output.write(document.toprettyxml(encoding="utf8"))
--- a/sx/pisa3/pisa_context.py 2017-02-08 15:38:26.127678118 +0100
+++ b/sx/pisa3/pisa_context.py 2017-02-08 15:38:26.123678133 +0100
@@ -40,8 +40,6 @@
from sx.w3c import css, cssDOMElementInterface
-from html5lib.sanitizer import *
-
import logging
log = logging.getLogger("ho.pisa")
signature.asc
Description: PGP signature
