Parameterized image fetcher function. Fixed handling for anchor with no href attribute.

Author: johnjor

README.md

@@ -1,18 +1,23 @@
 # html3docx
-A fork of https://github.com/pqzx/html2docx.  This version will focus on expedient changes for our particular use case.
+A fork of https://github.com/pqzx/html2docx.  This version will focus on expedient changes for our particular use case,
+and thus will receive infrequent updates.
 
 Dependencies: `python-docx` & `bs4`
 
 ### To install
 
 `pip install html3docx`
 
+PyPI: https://pypi.org/project/html3docx/
+
 ### Improvements
 
 - Fix for KeyError when handling an img tag without a src attribute.
 - Images with a width attribute will be scaled according to that width.
 - Fix for AttributeError when handling a leading br tag, either at the top of the HTML snippet, or within a td or th cell.
 - Fix for IndexError when a table has more cells in latter rows than in the first row.
+- Parameterized image fetcher function.
+- Fix for KeyError when handling an anchor with no href attribute.
 
 ## Original README
 

htmldocx/h2d.py

@@ -17,10 +17,11 @@
 import urllib.request
 from urllib.parse import urlparse
 from html.parser import HTMLParser
+from typing import Callable
 
 import docx, docx.table
 from docx import Document
-from docx.shared import RGBColor, Pt, Inches
+from docx.shared import RGBColor, Inches
 from docx.enum.text import WD_COLOR, WD_ALIGN_PARAGRAPH
 from docx.image.image import Image
 from docx.oxml import OxmlElement
@@ -196,7 +197,8 @@ def __init__(self,
                  ul_style="List Bullet",
                  ol_style="List Number",
                  table_style=DEFAULT_TABLE_STYLE,
-                 paragraph_style=DEFAULT_PARAGRAPH_STYLE):
+                 paragraph_style=DEFAULT_PARAGRAPH_STYLE,
+                 custom_image_fetcher: Callable = fetch_image):
         super().__init__()
         self.options = {
             'fix-html': True,
@@ -214,6 +216,7 @@ def __init__(self,
         self.paragraph_style = paragraph_style
         self.ul_style = ul_style
         self.ol_style = ol_style
+        self.image_fetcher = custom_image_fetcher
 
     def set_initial_attrs(self, document=None):
         self.tags = {
@@ -224,9 +227,9 @@ def set_initial_attrs(self, document=None):
             self.doc = document
         else:
             self.doc = Document()
-        self.bs = self.options['fix-html'] # whether or not to clean with BeautifulSoup
+        self.bs = self.options['fix-html']  # whether or not to clean with BeautifulSoup
         self.document = self.doc
-        self.include_tables = True #TODO add this option back in?
+        self.include_tables = True  # TODO add this option back in?
         self.include_images = self.options['images']
         self.include_styles = self.options['styles']
         self.paragraph = None
@@ -240,6 +243,7 @@ def copy_settings_from(self, other):
         self.paragraph_style = other.paragraph_style
         self.ul_style = other.ul_style
         self.ol_style = other.ol_style
+        self.image_fetcher = other.image_fetcher
 
     def get_cell_html(self, soup):
         # Returns string of td element with opening and closing <td> tags removed
@@ -351,7 +355,7 @@ def handle_img(self, current_attrs):
         src_is_url = is_url(src)
         if src_is_url:
             try:
-                image = fetch_image(src)
+                image = self.image_fetcher(src)
             except urllib.error.URLError:
                 image = None
         else:
@@ -598,7 +602,7 @@ def handle_data(self, data):
         # https://html.spec.whatwg.org/#interactive-content
         link = self.tags.get('a')
         if link:
-            self.handle_link(link['href'], data)
+            self.handle_link(link.get("href", ""), data)
         else:
             # If there's a link, dont put the data directly in the run
             self.run = self.paragraph.add_run(data)

tests/test.py

@@ -1,4 +1,6 @@
 import os
+import io
+import urllib.request
 from pathlib import Path
 import unittest
 from docx import Document
@@ -257,7 +259,6 @@ def test_leading_br(self):
         self.document.add_heading('Test: Leading BR', level=1)
         self.parser.add_html_to_document("<br /><p>Hello</p>", self.document)
 
-
     def test_unbalanced_table(self):
         # A table with more td elements in latter rows than in the first
         self.document.add_heading(
@@ -292,6 +293,30 @@ def test_custom_ul(self):
         ]:
             custom_parser.add_html_to_document(snippet, self.document)
 
+    def test_custom_image_fetcher(self):
+
+        def custom_fetcher(url):
+            url.replace("oops.githubusercontent.com", "raw.githubusercontent.com")
+            try:
+                with urllib.request.urlopen(url) as response:
+                    # security flaw?
+                    return io.BytesIO(response.read())
+            except urllib.error.URLError:
+                return None
+
+        custom_parser = HtmlToDocx(custom_image_fetcher=custom_fetcher)
+
+        self.document.add_heading(
+            'Test: Handling Images with Custom Fetch Function',
+            level=1
+        )
+        custom_parser.add_html_to_document(
+            "<img src='https://oops.githubusercontent.com/pqzx/h2d/master/testimg.png' />", self.document)
+        custom_parser.add_html_to_document(
+            "<table><tbody><tr><td>"
+            "<img src='https://oops.githubusercontent.com/pqzx/h2d/master/testimg.png' />"
+            "</td></tr></tbody></table>", self.document)
+
 
 if __name__ == '__main__':
     unittest.main()

tests/text1.html

@@ -43,6 +43,9 @@ <h1>heading 1</h1>
 <p style="text-align: left;">asdfsa</p>
 <p style="text-align: left;"><a class="fr-green fr-strong" href="abc://def.ghi" rel="noopener noreferrer"
         target="_blank">link</a></p>
+<p style="text-align: left;"><a class="fr-green fr-strong nohref" rel="noopener noreferrer"
+        target="_blank">link</a></p>
+
 <ol>
     <li>Ordered list</li>
     <ul>