Merge pull request #113 from ZumZoom/master

add baseurl functionality to inplace urls Thanks Mikhail Melnik @ZumZoom
Alir3z4 · Jan 31, 2016 · 54ea451 · 54ea451
2 parents d546297 + fbe1b76
commit 54ea451
Show file tree

Hide file tree

Showing 6 changed files with 25 additions and 3 deletions.
diff --git a/AUTHORS.rst b/AUTHORS.rst
@@ -19,6 +19,7 @@ The AUTHORS/Contributors are (and/or have been):
 * Albert Berger <gh: nbdsp>
 * Etienne Millon <[email protected]>
 * John C F <gh: critiqjo>
+* Mikhail Melnik <[email protected]>
 
 
 Maintainer:

diff --git a/ChangeLog.rst b/ChangeLog.rst
@@ -3,6 +3,7 @@
 ----
 
 * Fix #89: Python 3.5 support.
+* Fix #113: inplace baseurl substitution for <a> and <img> tags
 
 
 2016.1.8

diff --git a/html2text/__init__.py b/html2text/__init__.py
@@ -420,9 +420,9 @@ def handle_tag(self, tag, attrs, start):
                             try:
                                 title = escape_md(a['title'])
                             except KeyError:
-                                self.o("](" + escape_md(a['href']) + ")")
+                                self.o("](" + escape_md(urlparse.urljoin(self.baseurl, a['href'])) + ")")
                             else:
-                                self.o("](" + escape_md(a['href'])
+                                self.o("](" + escape_md(urlparse.urljoin(self.baseurl, a['href']))
                                        + ' "' + title + '" )')
                         else:
                             i = self.previousIndex(a)
@@ -476,7 +476,7 @@ def handle_tag(self, tag, attrs, start):
                     self.o("![" + escape_md(alt) + "]")
                     if self.inline_links:
                         href = attrs.get('href') or ''
-                        self.o("(" + escape_md(href) + ")")
+                        self.o("(" + escape_md(urlparse.urljoin(self.baseurl, href)) + ")")
                     else:
                         i = self.previousIndex(attrs)
                         if i is not None:

diff --git a/test/inplace_baseurl_substitution.html b/test/inplace_baseurl_substitution.html
@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<head></head>
+<body>
+<p>
+<img src="/uploads/2012/01/read2textheader.jpg" alt="read2text header image" width="650" height="165"/>
+</p>
+<p>
+<a href="/">BrettTerpstra.com</a>
+</p>
+</body>
+</html>
diff --git a/test/inplace_baseurl_substitution.md b/test/inplace_baseurl_substitution.md
@@ -0,0 +1,3 @@
+![read2text header image](http://brettterpstra.com/uploads/2012/01/read2textheader.jpg)
+
+[BrettTerpstra.com](http://brettterpstra.com/)
diff --git a/test/test_html2text.py b/test/test_html2text.py
@@ -173,6 +173,12 @@ def test_func(self):
     if base_fn not in ['bodywidth_newline.html', 'abbr_tag.html']:
         test_func = None
 
+    if base_fn == 'inplace_baseurl_substitution.html':
+        module_args['baseurl'] = 'http://brettterpstra.com'
+        module_args['body_width'] = 0
+        # there is no way to specify baseurl in cli :(
+        test_cmd = None
+
     return test_mod, test_cmd, test_func
 
 # Originally from http://stackoverflow.com/questions/32899/\
-Original file line number
+Diff line change
@@ Expand Up / @@ -3,6 +3,7 @@ @@
     ----
     * Fix #89: Python 3.5 support.
+    * Fix #113: inplace baseurl substitution for <a> and <img> tags
 .1.8
@@ Expand Down @@
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,3 @@
		![read2text header image](http://brettterpstra.com/uploads/2012/01/read2textheader.jpg)

		[BrettTerpstra.com](http://brettterpstra.com/)