Ignore invalid URLs that cause urljoin to raise.

html2text/__init__.py

@@ -489,7 +489,12 @@ def handle_tag(
             self.quote = not self.quote
 
         def link_url(self: HTML2Text, link: str, title: str = "") -> None:
-            url = urlparse.urljoin(self.baseurl, link)
+            url = link
+            try:
+                url = urlparse.urljoin(self.baseurl, link)
+            except ValueError:
+                # Ignore malformed URLs.
+                pass
             title = ' "{}"'.format(title) if title.strip() else ""
             self.o("]({url}{title})".format(url=escape_md(url), title=title))
 
@@ -584,9 +589,12 @@ def link_url(self: HTML2Text, link: str, title: str = "") -> None:
                     self.o("![" + escape_md(alt) + "]")
                     if self.inline_links:
                         href = attrs.get("href") or ""
-                        self.o(
-                            "(" + escape_md(urlparse.urljoin(self.baseurl, href)) + ")"
-                        )
+                        try:
+                            href = urlparse.urljoin(self.baseurl, href)
+                        except ValueError:
+                            # Ignore malformed URLs.
+                            pass
+                        self.o("(" + escape_md(href) + ")")
                     else:
                         i = self.previousIndex(attrs)
                         if i is not None:
@@ -820,12 +828,13 @@ def o(
                 newa = []
                 for link in self.a:
                     if self.outcount > link.outcount:
-                        self.out(
-                            "   ["
-                            + str(link.count)
-                            + "]: "
-                            + urlparse.urljoin(self.baseurl, link.attrs["href"])
-                        )
+                        href = link.attrs["href"]
+                        try:
+                            href = urlparse.urljoin(self.baseurl, href)
+                        except ValueError:
+                            # Ignore malformed URLs, and also calm mypy.
+                            assert href is not None
+                        self.out("   [" + str(link.count) + "]: " + href)
                         if "title" in link.attrs:
                             assert link.attrs["title"] is not None
                             self.out(" (" + link.attrs["title"] + ")")

test/invalid_link.html

@@ -0,0 +1 @@
+<a href="http://example.com]path">The typo in the link here causes urljoin to raise.</a>

test/invalid_link.md

@@ -0,0 +1,2 @@
+[The typo in the link here causes urljoin to raise.](http://example.com\]path)
+