Merge pull request aaronsw#55 from mgontav/fix-img-links

<a> tags with no text or <img> tags no longer discarded.

Thanks @mgontav

Author: Alir3z4

AUTHORS.rst

@@ -11,6 +11,7 @@ The AUTHORS/Contributors are (and/or have been):
 * Ivan Gromov <[email protected]>
 * Jocelyn Delalande <[email protected]>
 * Matt Dorn <[email protected]>
+* Miguel Tavares <[email protected]>
 
 Maintainer:
 

ChangeLog.rst

@@ -1,3 +1,10 @@
+2015.02.18
+==========
+----
+
+* Fix #38: Anchor tags with empty text or with `<img>` tags inside are no longer stripped.
+
+
 2014.12.29
 ==========
 ----

html2text/__init__.py

@@ -84,6 +84,7 @@ def __init__(self, out=None, baseurl='', bodywidth=config.BODY_WIDTH):
         self.a = []
         self.astack = []
         self.maybe_automatic_link = None
+        self.empty_link = False
         self.absolute_url_matcher = re.compile(r'^[a-zA-Z+]+://')
         self.acount = 0
         self.list = []
@@ -370,16 +371,21 @@ def handle_tag(self, tag, attrs, start):
                                  attrs['href'].startswith('#')):
                     self.astack.append(attrs)
                     self.maybe_automatic_link = attrs['href']
+                    self.empty_link = True
                     if self.protect_links:
                         attrs['href'] = '<'+attrs['href']+'>'
                 else:
                     self.astack.append(None)
             else:
                 if self.astack:
                     a = self.astack.pop()
-                    if self.maybe_automatic_link:
+                    if self.maybe_automatic_link and not self.empty_link:
                         self.maybe_automatic_link = None
                     elif a:
+                        if self.empty_link:
+                            self.o("[")
+                            self.empty_link = False
+                            self.maybe_automatic_link = None
                         if self.inline_links:
                             self.o("](" + escape_md(a['href']) + ")")
                         else:
@@ -399,6 +405,19 @@ def handle_tag(self, tag, attrs, start):
                     attrs['href'] = attrs['src']
                 alt = attrs.get('alt') or ''
 
+                # If we have a link to create, output the start
+                if not self.maybe_automatic_link is None:
+                    href = self.maybe_automatic_link
+                    if self.images_to_alt and escape_md(alt) == href and \
+                            self.absolute_url_matcher.match(href):
+                        self.o("<" + escape_md(alt) + ">")
+                        self.empty_link = False
+                        return
+                    else:
+                        self.o("[")
+                        self.maybe_automatic_link = None
+                        self.empty_link = False                   
+
                 # If we have images_to_alt, we discard the image itself,
                 # considering only the alt text.
                 if self.images_to_alt:
@@ -637,10 +656,12 @@ def handle_data(self, data):
             href = self.maybe_automatic_link
             if href == data and self.absolute_url_matcher.match(href):
                 self.o("<" + data + ">")
+                self.empty_link = False
                 return
             else:
                 self.o("[")
                 self.maybe_automatic_link = None
+                self.empty_link = False
 
         if not self.code and not self.pre:
             data = escape_md_section(data, snob=self.escape_snob)

setup.py

@@ -34,7 +34,7 @@ def run(self):
 
 setup(
     name="html2text",
-    version="2014.12.29",
+    version="2015.02.18",
     description="Turn HTML into equivalent Markdown-structured text.",
     author="Aaron Swartz",
     author_email="[email protected]",

test/empty-link.html

@@ -0,0 +1,6 @@
+<h1>Processing empty hyperlinks</h1>
+
+<p>This test checks wheter empty hyperlinks still appear in the markdown result.</p>
+
+<a href="http://some.link"></a>
+<a href="http://some.link"><p></p></a>

test/empty-link.md

@@ -0,0 +1,8 @@
+# Processing empty hyperlinks
+
+This test checks wheter empty hyperlinks still appear in the markdown result.
+
+[](http://some.link)
+
+[](http://some.link)
+

test/images_to_alt.html

@@ -1,3 +1,7 @@
 <a href="http://example.com">
 <img src="http://example.com/img.png" alt="ALT TEXT" />
 </a>
+<br>
+<a href="http://example.com"><img src="http://example.com/img.png" alt="ALT TEXT" /></a>
+<br>
+<a href="http://example.com"><img src="http://example.com/img.png" alt="http://example.com" /></a>

test/images_to_alt.md

@@ -1,2 +1,4 @@
-[ ALT TEXT ](http://example.com)
+[ ALT TEXT ](http://example.com)  
+[ALT TEXT](http://example.com)  
+<http://example.com>
 

test/img-tag-with-link.html

@@ -0,0 +1,9 @@
+<h1>Processing images with links</h1>
+
+<p>This test checks images with associated links.</p>
+
+<a href="http://some.link"><img src="http://placehold.it/350x150#(banana)" width="350" height="150" alt="(banana)"></a>
+<a href="http://some.link"><img src="http://placehold.it/350x150#[banana]" width="350" height="150" alt="[banana]"></a>
+<a href="http://some.link"><img src="http://placehold.it/350x150#{banana}" width="350" height="150" alt="{banana}"></a>
+<a href="http://some.link"><img src="http://placehold.it/350x150#([{}])" width="350" height="150" alt="([{}])"></a>
+<a href="http://some.link"><img src="http://placehold.it/350x150#([{}])" width="350" height="150" alt></a>

test/img-tag-with-link.md

@@ -0,0 +1,10 @@
+# Processing images with links
+
+This test checks images with associated links.
+
+[![\(banana\)](http://placehold.it/350x150#\(banana\))](http://some.link)
+[![\[banana\]](http://placehold.it/350x150#\[banana\])](http://some.link)
+[![{banana}](http://placehold.it/350x150#{banana})](http://some.link)
+[![\(\[{}\]\)](http://placehold.it/350x150#\(\[{}\]\))](http://some.link)
+[![](http://placehold.it/350x150#\(\[{}\]\))](http://some.link)
+