Improve handling of base64 images

Also makes images links absolute in the rendering

CHANGELOG

@@ -3,8 +3,10 @@
 ## 2.0-beta2 - unreleased
 Changes since beta1
 - "--sync" can optionnaly take some lists as arguments, in order to make for specific sync
+- improves handling of base64 images
 - fixes gophermap being considered as gemtext files 
 - fixes opening mailto links
+- attempt at hiding XMLparsedAsHTMLWarning from BS4 library
 
 ## 2.0-beta1 - September 05th 2023
 This is an an experimental release. Bug reports and feedbacks are welcome on the offpunk-devel list.

ansicat.py

@@ -13,6 +13,7 @@ import fnmatch
 import netcache
 import offthemes
 from offutils import run,term_width,is_local,looks_like_base64
+import base64
 from offutils import _DATA_DIR
 try:
     from readability import Document
@@ -982,6 +983,7 @@ class HtmlRenderer(AbstractRenderer):
                     #4 followings line are there to translate the URL into cache path
                     img = netcache.get_cache_path(imgurl)
                     if imgdata:
+                        os.makedirs(os.path.dirname(img), exist_ok=True)
                         with open(img,"wb") as cached:
                             cached.write(base64.b64decode(imgdata))
                             cached.close()
@@ -1116,10 +1118,10 @@ class HtmlRenderer(AbstractRenderer):
                 else:
                     text += "[IMG]"
                 if src:
-                    links.append(src+" "+text)
                     if not mode in self.images:
                         self.images[mode] = []
-                    abs_url = urllib.parse.urljoin(self.url, src)
+                    abs_url,data = looks_like_base64(src,self.url)
+                    links.append(abs_url+" "+text)
                     self.images[mode].append(abs_url)
                     link_id = " [%s]"%(len(links)+startlinks)
                     r.add_block(ansi_img)

offutils.py

@@ -136,7 +136,11 @@ def looks_like_base64(src,baseurl):
     if src and src.startswith("data:image/"):
         if ";base64," in src:
             splitted = src.split(";base64,")
-            extension = splitted[0].strip("data:image/")[:3]
+            #splitted[0] is something like data:image/jpg
+            if "/" in splitted[0]:
+                extension = splitted[0].split("/")[1]
+            else:
+                extension = "data"
             imgdata = splitted[1]
             imgname = imgdata[:20] + "." + extension
             imgurl = urllib.parse.urljoin(baseurl, imgname)